Chemistry and Processing of Sugarbeet and Sugarcane Volume 9 PDF

SUGAR SCIENCES
Vol. 1. Standard Fabrication Practices for Cane Sugar Mills (Delden)

Vol. 2. Manufacture and Refining of Raw Cane Sugar (Baikow)
Vol. 3. By-Products of the Cane Sugar Industry (Paturau)
Vol. 4. Unit Operations in Cane Sugar Production (Payne)
Vol. 5. Noël Deerr: Classic Papers of a Sugar Cane Technologist (Payne,
Compiler)
Vol. 6. The Energy Cane Alternative (Alexander)
Vol.7. Handbook of Cane Sugar Engineering (Hugot, 3rd edition)
Vol. 8. Management Accounting for the Sugar Cane Industry (Fok Kam)
Vol. 9. Chemistry and Processing of Sugarbeet and Sugarcane (Clarke and
Godshall, Editors)
sugar series, 9
chemistry and
processing of
sugarbeet and
sugarcane
Proceedings of the Symposium on the Chemistry and Processing
of Sugarbeet, Denver, Colorado, April 6 , 1 9 8 7 and the Symposium
on the Chemistry and Processing of Sugarcane, New Orleans,
Louisiana, September 3 - 4 , 1 9 8 7
edited by
Margaret A. Clarke and Mary An Godshall

Sugar Processing Research, Inc., 1100 Robert E. Lee Boulevard, New Orleans,
Louisiana 70124, U.S.A.
Elsevier
Amsterdam — Oxford — New York — Tokyo 1988

ELSEVIER SCIENCE PUBLISHERS B.V.
Sara Burgerhartstraat 25
P.O. Box 211, 1000 AE Amsterdam, The Netherlands
Distributors for the United States and Canada:
ELSEVIER SCIENCE PUBLISHING COMPANY INC.

52, Vanderbilt Avenue
New York, NY 10017, U.S.A.
ISBN 0-444-43020-2(Vol. 9)
ISBN 0-444-41897-0(Series)
© Elsevier Science Publishers B.V., 1988
All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system or transmitted in any form or by any means, electronic, mechanical, photocopying,
recording or otherwise, without the prior written permission of the publisher, Elsevier
Science Publishers B.V./Physical Sciences & Engineering Division, P.O. Box 330, 1000 AH
Amsterdam, The Netherlands.
Special regulations for readers in the USA - This publication has been registered with the
Copyright Clearance Center Inc. (CCC), Salem, Massachusetts. Information can be obtained
from the CCC about conditions under which photocpies of parts of this publication may be
made in the USA. All other copyright questions, including photocopying outside of the USA,
should be referred to the publisher.
No responsibility is assumed by the Publisher for any injury and/or damage to persons or
property as a matter of products liability, negligence or otherwise, or from any use or
operation of any methods, products, instructions or ideas contained in the material herein.
Printed in The Netherlands

This book is dedicated to the memory of
Benjamin A. Oxnard
VII
PREFACE
In 1987, the American Chemical Society held its spring meeting in

Denver, Colorado, an historic center of the U.S. beet sugar industry, and
its fall meeting in New Orleans, a traditional center of the U.S. cane sugar
industry. The settings initiated the idea for symposia on the two sugar
crops and their processing industries. Papers from these symposia are
published in this book.
The world of sugar production has undergone massive changes in the last
ten years: increased production in the European countries has made an
additional three to four million tons of white sugar annually available on
world markets; replacement of cane and beet sugars in the U.S. by cheaper
corn syrups has reduced imports from sugarcane-growing tropical countries by
80Z or almost four million tons; increase in sugar consumption within these
same countries has increased the demand there for white sugar and created a
demand for systems to produce more high grade white sugars for small initial
investment.
The overall effect has been an apparent surplus of sugar, now

fortunately diminishing. The ample supply has had its negative effect on
price, and so processing costs have been tightly trimmed, or, in many
countries, cut to the bone. There is pressure on technologists to develop
more efficient and lower cost processes, and so technological changes have
emerged from the pressure of changing markets.
The purposes of this book, and of the symposia, are:
1. To present recent developments in beet and cane sugar manufacturing

as part of a picture of current sugar technology.
2. To present the chemistry of sugar processing and products
3. To consider trends and future possibilities in sugar production

systems and products.
To these ends, the papers herein, by various international authorities

in the sugar industries, have been collected. The book is in two sections,
beet and cane; each section begins with an overview of the crop and the
production systems that serves as a framework for the papers that follow.
Several papers—those on sucrose chemistry (Chapters 16, 18, and 23),

sucrose and health (Chapter 19) and Chapters 7, 8, 10 and 21—are relevant
to both sugarcane and sugarbeet.
The editors wish to express thanks to the Officers of the Agricultural

and Food Chemistry Division of the American Chemical Society, under whose
auspices the Symposia on the Chemistry and Processing of Sugarbeet and
Sugarcane were held. We express our thanks to the Board of Directors of
Sugar Processing Research, Inc., for their support for the Symposia and the
editing of this book, and to Southern Regional Research Center for the use
of facilities.
Vili
We offer our gratitude and appreciation to the contributors to this

book, and to their principals, for their time and efforts in participating
in the symposia and composing these papers.
We extend special thanks to our editorial assistant, Jacqueline E.

Smith, for her extensive and untiring efforts in assembling, typing and
formatting this material.
February, 1988. Margaret A. Clarke

New Orleans, Louisiana Mary An Godshall
IX
CONTRIBUTORS
VISTI ANDERSSON, DDS Engineering, P. 0. Box 2249, DK-1019, Copenhagen K,

Denmark.
SVEN BARFOED, DDS Engineering, P. 0. Box 2249, DK-1019, Copenhagen K,
Denmark.
STANLEY E. BICHSEL, American Crystal Sugar Company, 1700 N. 11th Street,
Moorhead, Minnesota USA 56560.
G. NORRIS BOLLENBACK, (The Sugar Association; retired), 4739 Berkeley
Terrace N.W., Washington, D.C. USA 20007.
NICHOLAS W. BR0UGHT0N, British Sugar plc, Research Center, Colney Lane,
Colney, Norwich, United Kingdom NR4 7UB.
B. W. BROWN, British Sugar plc, Research Center, Colney Lane, Colney,
Norwich, United Kingdom NR4 7UB.
MARGARET A. CLARKE, Sugar Processing Research, Inc., 1100 Robert E. Lee
Blvd., New Orleans, Louisiana USA 70124.
MICHAEL CLEARY, American Crystal Sugar Company, 1700 N. 11th Street,
Moorhead, Minnesota USA 56560.
MARY A. GODSHALL, Sugar Processing Research, Inc., 1100 Robert E. Lee Blvd.,
New Orleans, Louisiana USA 70124.
T. R. HANSSENS, Suiker Unie U.A., Postbus 1308, 4700 BH Roosendaal, The
Netherlands.
RHETT HARRIS, West India Rum Refinery, Brighton, St. Michael, Barbados.
ROBERT A. JOHNSON, Syrinx Research Institute Pty. Ltd., P. 0. Box 1583,
Bundaberg, Queensland 4670, Australia.
GRAHAM C. JONES, British Sugar plc, Research Center, Colney Lane, Colney,
Norwich, United Kingdom NR4 7UB.
HAYDN F. JONES, Philip Lyle Memorial Research Laboratory, Tate and Lyle plc,
Whiteknights, P. 0. Box 68, Reading, Berks., United Kingdom RG6 2BX.
RIAZ KHAN, Philip Lyle Memorial Research Laboratory, Tate and Lyle plc,
Whiteknights, P. 0. Box 68, Reading, Berks., United Kingdom RG6 2BX.
RICHARD A. KITCHEN, B.C. Sugar Refining Company, P. 0. Box 2150, Vancouver,
B.C., Canada V6B 3V2.
KEES KOERTS, Suiker Unie U.A., Postbus 1308, 4700 BH Roosendaal, The
Netherlands.
X
MICHEL S. LEFEBVRE, Syrinx Research Institute Pty. Ltd., MLC Centre,

Martin Place, Sydney, NSW 2000, Australia.
B. L. LEGENDRE, Agricultural Research Service, U.S.D.A., Sugarcane Research
Unit, Houma, Louisiana USA 70361.
LEIF H. RAMM-SCHMIDT, Finnish Sugar Company, Ltd., Porkkala Plant, 02460
Kantvik, Finland.
GEOFFREY N. RICHARDS, Wood Chemistry Laboratory, University of Montana,
Missoula, Montana USA 59812-1201.
JOHN A. RICHMOND, Holly Sugar Corporation, P. 0. Box 1052, Colorado Springs,
Colorado USA 80901.
RICHARD RIFFER, C and H Sugar Company, Crockett, California USA 94525.
CARLOS E. VAZ RUSSELL, Copersucar, Centro de Technologia, P. 0. Box 162,
13400 Piracicaba, S.P., Brazil.
DAVID SARGENT, British Sugar pic, Research Center, Colney Lane,
Colney, Norwich, United Kingdom NR4 7UB,
HUBERT SCHIWECK, Suddeutsche Zucker AG, Postfach 1127, 6718 Grünstadt I,
Federal Republic of Germany.
MERVYN SHORE, British Sugar pic, Research Center, Colney Lane,
Colney, Norwich, United Kingdom NR4 7UB.
GARRY A. SMITH, Crops Research Laboratory, U.S.D.A., 1701 Center Avenue,
Fort Collins, Colorado USA 80526.
GEORGE STEINLE, Suddeutsche Zucker AG, Postfach 1127, 6718 Grünstadt I,
Federal Republic of Germany.
ROBERT W. STRICKLAND, Holly Sugar Corporation, P. 0. Box 1052, Colorado
Springs, Colorado USA 80901.
JAN TJEBBES, Sockerbolaget, P. 0. Box 6, 232 00 Arlöv, Sweden.
ROBERT R. TROTT, Tate and Lyle Process Technology, 55 Liddon Road, Bromley,
Kent, United Kingdom BRI 2SR.
W.S. CHARLES TSANG, Sugar Processing Research, Inc., 1100 Robert E. Lee
Blvd., New Orleans, Louisiana USA 70124.
DAVID WEST, Sugar Technology Research Unit, Edgehill, St. Thomas,
Barbados, West Indies.
Chemistry and Processing of Sugarbeet and Sugarcane, edited by M.A. Clarke and M.A. Godshall
Elsevier Science Publishers B.V., Amsterdam, 1988 — Printed in the Netherlands
Chapter 1 1
AN OVERVIEW OF THE U.S. BEET SUGAR INDUSTRY
S. E. BICHSEL
INTRODUCTION
The United States beet sugar industry derives its historical origin
from the European beet sugar industry beginning in France and Germany in the
early 1800*s. At the present time sugarbeets are produced in 12 states
located in the western and central part of the United States as indicatd in
Figure 1. Since the inception of the U.S. beet industry in 1900, there has
been a dramatic increase in yield per acre and harvested acres of
sugarbeets, from a minimal amount in 1900 to approximately 1.4 million acres
(5.67 x 10 s ha) harvested at an average yield/acre of 19.0 tons (17.24
tonnes) (Figure 2). The increase in acres of land dedicated to sugarbeet
production and the increase in productivity per acre resulted primarily from
sugar price stability, advances in cultural practices, and the availability
of disease resistant monogerm hybrid sugarbeet varieties which offered
adequate resistance against common sugarbeet diseases and reduced labor
costs associated with thinning the original multigerm varieties. The
combination of stable sugar costs to the consumer, and technological
advances in sugarbeet production in the field and beet sugar production in
the factory, ensured the growth of the domestic beet sugar industry.
Figure 3 shows the increasing trend in refined sugar production/acre
from the mid 60's through the mid 80*s. During this period, continuous
gains were made in agronomic practices and hybrid sugarbeet variety
productivity in the field. Improvement in refined sugar production/acre
compensated for the decrease in harvested acres as indicated in a flat total
refined sugar production trend approximating 60 million cwt (2.7 million
tonnes) of domestic beet sugar produced from the mid 60*s through the mid
80*s (Figure 4).
The field production of sugarbeets followed by storage and subsequent
processing to produce refined sugar for sale are described in this paper.
2
LEGEND
Y///\ SU«.H Ikvl Counties
| | Sun.»r O I K· Counlio
"V IWvl Siifrir I'Uils
. Clin· Su>vw Mills
Fig. 1. The beet sugar industry in the United States. Grey areas represent
counties where sugarbeets are grown.
O1500
o
CO
LU
>
er ooooco^w
< OOJ^COCOCDCD
O>0>00>0>0>0>
x CROP YEAR
Fig. 2. Harvested acres and yield/acre vs years.

3
I I I I I I I I I I
1970 1975 1980 1985
CROP YEAR
Fig. 3. Pounds of refined sugar/acre vs crop years.
<80M
"dì 70M
S 60M
2
ft 50M
* 40MH
I I I | I I I I | I I I I | I I I I |
% 30MH
1965 1970 1975 1980 1985
CROP YEAR
Fig. 4. Crop year vs cwt refined sugar.
FIELD PRODUCTION
Sugarbeet monogerm seed is planted from March to May, using a precision
planter, as a row crop (generally 22 in rows). Seed spacing varies from 3-
5 in (7.6-12 cm) to obtain optimum plant population/acre. Seeds are planted
from 3/5 " to 1 1/4 in (1.5-4.4 cm) deep in a soil which has been prepared
to ensure a firm yet porous seed bed for optimum germination. Pre- or post-
emergence registered herbicides are utilized to minimize hand labor
k
associated with weed control. Sugarbeets are grown on irrigated land

(primarily lighter western soil) and in nonirrigated areas with heavier,
more organic soil in Minnesota, North Dakota, Ohio, and Michigan.
Beet seedlings are generally thinned at the six leaf stage to optimum
spacing for maximum yield and root quality. Approved chemical sprays may be
necessary at critical times during the growing season to minimize the
possibility of airborne fungal diseases which can damage the leaf and
significantly reduce yield and quality of the harvested root (ref. 1).
Sugarbeets in the field are defoliated using a multi-row implement to
remove all top growth from the crown of the beet. The top petiole and leaf
material are then incorporated into the soil as organic matter or fed to
livestock. After defoliation the beets are generally scalped to eliminate
low quality top crown tissue immediately before harvest.
Sugarbeets are generally harvested from September through the first
part of November, using multi-row lifting wheel harvesters. In areas of
central and northern California a designated percentage of the crop is
allowed to overwinter for harvest in the spring. Winter temperatures do not
generally reach below freezing in these areas. This approach to beet
storage in the ground extends total time of factory operation per year and
contributes to economic return on factory capital investment.
SUGARBEET LONG TERM STORAGE

Sugarbeets harvested in California are delivered direct to the
processing factory, with normal storage prior to slicing not exceeding 48
hours, with the exception of short term stockpiling in the case of fall or
spring rains, to maintain uninterrupted factory operation.
In the inland western, north central, and midwest sugarbeet growing
areas, harvesting and processing proceeds in a manner similar to the
California harvest until the first part of October. At this time, beets are
harvested at a rate in excess of factory processing capacity. Harvest is
generally completed by mid November to avoid irreversible freeze damage of
sugarbeets in the ground. The excess beets are stored in piles generally
180-200 ft (54.9-61.0 m) wide and 20-24 ft (6.1-7.3 m) high. These piles
are located on the processing plant grounds and in piling grounds remote
from the factory location. Beets are stored in the long term pile storage
from 80 to 160 days depending upon winter average ambient temperatures.
5
Sugar loss in stored sugarbeets, caused by live root respiration and

external and internal microorganism induced tissue rot, is specifically a
function of storage time and room temperature. Under ordinary storage
conditions a sugar loss value of 0.50 lb/ton/day (0.25 kg/tonne) may be
assumed. All means to induce natural circulation of cool air through the
beet piles are promoted. In some beet growing areas, outside air is forced
through the beet pile utilizing ventilation fans and air distribution
channels. The purpose is to lower root temperatures to several degrees
above freezing to reduce sugar loss from respiration to a practical minimum.
A 10°F (6°C) increase in root temperature more than doubles the root
respiration and subsequent sugar loss. In the north central area located in
North Dakota and Minnesota, an arctic-type winter temperature profile allows
deep freezing of sugarbeet roots which results in total cessation of sugar
loss due to respiration and internal and external microorganism attack.
SUGARBEET PROCESSING
Sugarbeets from the storage piles are introduced to the factory,
washed, and sliced into "V" shaped cossette sections measuring approximately
three sixteenths of an inch square (5 mm) and 2-3 in ( 5-7.5 cm) in length.
This shape of cossette ensures maximum surface area for sugar diffusion into
the hot aqueous extraction solvent. The cossettes are introduced into a
counter current diffuser which may be a horizontal drum, sloped vat, or
vertical cylindrical tower type (ref. 2).
All of the counter current continuous diffusion units involve a
mechanism to transport sugarbeet cossettes in a direction counterflow to the
hot water extractant. Sugarbeet cell walls are heat denatured to enhance
diffusion of sugar through the cell walls to the lower concentration of
sugar in the aqueous extractant. Approximately 98Z of the sugar is
extracted in the counter current diffusion operation.
Wet beet pulp discharged from the diffuser at approximately 92Z
moisture is pressed mechanically to reduce beet pulp moistures to
approximately 782 prior to drying in direct gas, oil, or coal fired rotary
drum dryers to a moisture content of approximately 10Z prior to sale as
animal feed. Diffusion juice from the diffuser, containing approximately
12Z sugar by weight and 21 soluble impurities in addition to soluble and
6
semi-soluble colloidal proteins, pectins, and saponins extracted from the

cossettes, is heated to 85°C prior to lime-carbon dioxide purification.
Purification of diffusion juice involves clarification by coagulation
and precipitation of the semi-soluble colloids as insoluble salts, and
precipitation of soluble impurities, i.e. potassium and sodium salts of
oxalic, malic, citric acids, phosphates, and sulfates, as insoluble calcium
salts (ref. 3). Continuous lime-carbon dioxide purification involves
addition of milk of lime to diffusion juice concurrent with carbon dioxide
addition to the carbonation tank, followed by a settling clarification for
separation of the precipitated insoluble impurities. The retention time in
the carbonation tank approximates 12 to 15 minutes under ideal conditions at
a temperature of 85° to 88°C.
Thickened underflow mud from the clarifier is generally filtered and
washed to minimize sugar loss on rotary vacuum drum filters. Lime and
carbon dioxide are produced from limerock at each factory in a vertical
shaft Belgian type lime kiln. Carbon dioxide is reclaimed from the kiln for
direct use in the lime-carbon dioxide first carbonation purification tank.
Approximately 2.00Z lime is used on sugarbeets for purification and
clarification.
Clarified overflow juice is then subjected to second carbonation with
carbon dioxide to reduce the concentration of residual lime salts. The
calcium carbonate precipitate is filtered from the juice using pressure
filters. Sulfur dioxide is added to the filtered juice to minimize color
formation during subsequent processing steps. The carbonation process is
described in detail in the chapter by Cleary. The chapter by Andersson and
Barfoed describes an alternate purification system.
The thin juice, with all the colloidal impurities and approximately 352
of the soluble impurities eliminated as precipitate, is concentrated from
about 14Z dissolved solids to approximately 60 to 65Z dissolved solids in
energy-efficient quintuple effect evaporators. Steam for evaporation at all
beet sugar factories is produced in relatively low pressure boilers, 400
psig, for cogeneration of steam turbine derived electricity for factory use,
and to supply steam for heating, evaporation, and crystallization of sugar
in the latter stages of the process in batch vacuum pan crystallizers.
Thick juice from the evaporators is concentrated, and sugar is
crystallized in a three stage crystallization process. In the traditional
7
process, the first vacuum pan crystallization produces white refined sugar
for direct sales, whereas the second and third vacuum pan crystallizers
produce lower purity raw sugars, which are redissolved after separation of
mother liquor using continuous conic basket centrifugals developed
specifically for the sugar processing industry. White sugar is produced
utilizing basket type batch centrifugals. A thick juice with a purity of
92Z (92Z of the dissolved solids are sugar, 8Z are nonsugars) is exhausted
to 60Z purity in the end molasses by-product. The white sugar is discharged
wet from the batch centrifuges, and dried in rotary hot air dryers to a
moisture content approximately 0.02Z moisture. The white refined sugar
produced is 99.92 pure as a result of the successive purifications by
diffusion, lime-carbon dioxide purification, and triple crystallization in
the batch vacuum crystallizer pans. In some factories additional
purification steps, employing ion exchange resins (see Chapter 5, this
volume) or carbon adsorbents, are part of process.
SUMMARY
As indicated in Figure 5, the objective of sugarbeet processing is to
separate pure sugar from insoluble pulp, soluble nonsugars, and water with
minimum production cost. A typical harvested sugarbeet contains 75.9Z
insoluble cellulose, hemicellulose, and pectin fraction remaining after
sugar extraction from cossettes. Figure 6 indicates the disposition of
white sugar from one ton of 16.00Z sugar by weight of beets. In
quantitative terms of white sugar, 266.0 lb sugar/ton (133 kg/tonne) is
produced for sale or 83.10Z of the total sugar is extracted. Loss of sugar
in molasses accounts for 12.5Z of the sugar introduced in conjunction with
26.7 lb (12.1 kg) of impurities, 17.2 lb (7.8 kg) water to produce 83.9 lb
(38.05 kg) of 80Z dry substance molasses at 60.OZ purity (Z of the total dry
substance which is sugar). In addition to the 83.4 lb (37.8 kg) of molasses
by-products produced, 11.0 lb (5 kg) of beet pulp are produced per ton of
fresh sugar beets processed. Other sugar losses include sugar remaining in
rotary vacuum drum filters; sugar lost due to bacterial action; sugar-
containing liquid spills; and sugar inverted and caramelized as a result of
exposure to hot heating surfaces.
WATER 75.97ο
^WWWW/ NONSUGARS 2.6%

SUGAR 16.0%
PULP 5.57ο
Fig. 5. Harvested sugarbeet composition percent on total weight.
WHITE SUGAR
2Θ6.0 LB SUGAR/TON
83.10% RECOVERY
FRESH SUGAR MOLASSES
BEET. 16.00% 40.0 LB SUGAR/TON
SUGAR BY 26.7 LB IMPURITIES
WEIGHT- 17.2 LB WATER
320 LB 83.9 LB MOLASSES
SUGAR/ 12.50% SUGAR LOSS
TON
PULP
110.0 LB/TON
OTHER SUGAR LOSSES
14.0 LB SUGAR/TON
4.40% SUGAR LOSS
F i g . 6. Sugar r e c o v e r e d , sugar l o s t and by-products produced from

one t o n of f r e s h s u g a r b e e t s .
REFERENCES
1 R. A. McGinnis, Beet Sugar Technology, published by the Beet Sugar

Development Foundation, Ft. Collins, Colorado, 1982.
2 K. Vukov, Physics and Chemistry of Sugar Beet in Sugar Manufacture,
Elsevier Scientific Publishing Company, New York, 1977.
3 P. M. Silin, Technology of Beet Sugar Production and Refining (Engl.
Transi, by L. Markin); National Science Foundation; Washington, D.C.,
1964; p. 47; Tekhnologiya sveklosakharnogo i rafinadnogo proizvodstva;
Pishchepromizdat; Moscow, USSR, 1958.
Chapter 2
EFFECTS OF PLANT BREEDING ON SUGARBEET COMPOSITION

G. A. SMITH
INTRODUCTION
Considerable attention in the sugarbeet industry has been focused on

sugarbeet quality and factors affecting it. Sugarbeet quality is determined
basically by two standards: sucrose and purity percentage. Sucrose
percentage is the percentage by weight of sucrose in the fresh beet root.
Purity is the ratio of sucrose to total soluble solids expressed as a
percentage. Purity of extracted sugarbeet juice is of vital importance to
the sugar processing industry because soluble nonsucrose constituents or
so-called impurity components impede crystallization and thereby lower
extraction of sucrose. Each pound of nonsucrose substance in the juice,
depending upon the particular operating conditions of the factory, inhibits
crystallization of 1.5 to 1.8 pounds (0.68-0.82 kg) of sucrose which,
therefore, is lost to molasses (ref. 1).
Many nonsucrose constituents may affect crystallization. Sugarbeets
metabolize nitrate not only into the pool of amino acids and amides required
for protein and enzyme synthesis, but also into many other compounds
including the organic base betaine. These N-containing compounds, some
essential to the plant, but all impurities as far as sugar extraction is
concerned, sometimes are produced in excess of amounts needed for optimum
sucrose production and storage. This has also been a problem in sugarbeet
variety development.
Nitrogeneous compounds are not the only nonsucrose constituents
affecting the solubility and crystallization rate of sucrose. Rorabaugh
and Norman (ref. 2) examined the effect of impurities on sucrose
crystallization and generalized that inorganic salts, particularly those of
carbonate and chloride, were especially deleterious, in addition to amino
acids, betaine, and non-nitrogenous organic acids. Furthermore, Carruthers
and Oldfield (ref. 3) pointed out that potassium and sodium salts, amino
acids, and betaine constituted about 70Z of the nonsugars in factory second
carbonation juice, implying that they are little removed by juice
purification procedures and remain in the purified juice to exert their
individual inhibitory effects on sucrose recovery. In our past research, we
10
have reported that Κ+, betaine and Na + followed by N03N (nitrate N ) , and
amino N are the most important factors affecting purity (ref. 4 ) .
Theoretically, the predominance of additive gene action controlling
non-sucrose beet components as reported by Smith et al. (ref. 5), should
result in significant component shift following selection. What effects
this selection might have on other components and on purity and sucrose
yield prompted this investigation.
Accordingly, the primary objective of this study was to test the
effects of selection for Na*, K + and amino N (previously identified by us as
three of the most important purity components (ref. 4)), on other chemical
characters known to affect purity and on purity and sucrose yield per se.
Before we can assume that more rapid progress can be made by breeding for
changes in the components of purity, we considered answers to the following
questions a prerequisite:
1. To what degree can population means for specific non-sucrose
purity components be shifted by selection?
2. What effect would such a mean shift have on other components?
MATERIALS AND METHODS

Laboratory analyses for variables discussed in this report were
determined as follows:
Percentage sucrose was determined on the beet pulp using a method
standard with many beet sugar companies and similar to the method outlined
in A.O.A.C, (ref. 6). Purity was determined by the method of Carruthers
and Oldfield (ref 3). Extractable sucrose and percentage extraction were
calculated according to the procedure of Dexter et al (ref. 17). In
addition, pressed juice was obtained from a laboratory root-pulp sample of
each plot. These samples were analyzed for Na* and K"·" by flame photometry,
for nitrate by specific ion electrode, for specific conductance, and for
chloride by coulometric titration with silver and amperometric endpoint-
detection. Determined spectrophotometrically were total N (including
nitrate) by direct Nesslerization of acid-digested samples, amino N using
ninhydrin, and betaine using ammonium reineckate (refs. 7, 8 ) .
The data base for the study was provided by two cycles of selection in
which individual plants (roots) were analyzed and selected to synthesize new
high/low selection populations. These populations along with the original
11
parental population were field evaluated in replicated tests for three

years.
RESULTS
Population means for selected characters. The means of the original
parental population for Na+, K"*" and amino N and the resulting populations
following one and two cycles of selection are presented in Table 1. Means
presented are the average of three years of replicated testing from
1983-1985. Although the magnitude of the character means differed in
response to years, the patterns shown by the means in Table 1 are
essentially identical to those shown by the data for each separate study
year. Population means for Na + , K + and amino N were shifted significantly
following the first cycle of high and low selection.
In all cases, mean shifts were greater following high selection than
for low selection. A second cycle of high selection for amino N increased
the already significant first cycle mean to a 98Z increase over the original
parental population mean. First cycle low selection reduced Na + , K"*" and
amino N by 21, 17.5 and 312 of the parental population, respectively.
TABLE 1
Character means and ranges (combined over test years) for parental (original)
population following low and high selection. "·"
Low selection High selection

Parental First Second First Second
Character population cycle cycle cycle cycle
Na 152.75* 120.69c 130.00c 200.73c **

(72.3-224.A) (42.8-196.5) (47.1-221.8) (86.5-292.1)
K 126.37x 104.23y 93.17z 169.08w

(97.7-163.3) (82.1-128.3) (66.0-123.3) (143.2-202.6)
Amino N 32.98o 22.70p 20.91p 50.00n 63.35mm

(28.2-41.7) (17.9-27.9) (16.5-26.0) (46.6-55.7) (53.1-85 1
-*- Values are in milligrams/100 ml of raw juice.
* Means within a row followed by the same letter are not significantly different
(P β 0.05) according to Duncan's multiple range test.
**Data not obtained.

12
Effect of selection for Na"*-. K+ and amino N on other characters. The

effects of selection for low Na"*" K"*· and amino N on each other and on other
characters including purity and extractable sucrose are presented in Figures 1
and 2. Data presented are percent increase or decrease from the unselected
parental population mean (P) for each character. The actual parental population
and selection cycle means for all characters are presented in Tables 2-4.
TABLE 2
Mean effects of low and high selection for Na*.
Parental First cycle selection Second cycle selection

Character** population Low High Low
Na 152.75b*** 120.69c 200.73a 130.00c

Purity X 88.66b 91.02a 84.89c 89.51b
Sucrose X 11.84b 12.31a 10.70c 11.99b
K 126.37a 113.81b 126.56a 12X.52ab
Chloride 22.34b 14.15c 32.37a 14.04c
Total N 113.63a 104.41ab 99.00b 108.72ab
Amino N 32.98a 26.09b 27.72b 29.90ab
Betaine 56.59a 51.66a 44.00b 52.34a
Extractable
sucrose X 9.07a 10.03a 7.31b 9.40a
Extraction X 76.6a 81.5a 68.3b 79.0a
* All values are the average of 3 years of replicated tests for all
characters except Cl, total N and betaine which were tested for one
year, two years and three years, respectively.
** Biochemical values are expressed as mg/100 ml. Extractable sucrose X and

extraction X calculated according to ref. 17, pp. 450-451.
***Means within a row followed by the same letter are not significantly
different (P < .05) according to Duncan's multiple range test.
Selection for low Na+ was accompanied by reductions in all other non-
sucrose chemical characters, especially chloride and a significant increase in
purity (Fig. 1A and Table 2). In contrast, selection for low K* resulted in no
significant changes in characters even after two cycles of selection for low K"*"
(Fig. IB and Table 3). Selection for low Na* increased extractable sucrose X and
X extraction after one or two cycles of selection. Extractable sugar per ton was
thus increased from 181.4 pounds per ton to 200.6 pounds (90.7 kg/tonne to 100.3
kg/tonne) after the first selection cycle. Selection for low K* had only slight
13
LOW Na SELECTION ^ 1
m Ht cycle
2i >
Py c β 2nd Cycle j
Ut
<
<-> IP
h Na
5* SS* i " SS ^
7 PO
Ss
TN ■
*
β
•4C
ί
LOW K SELECTION g
g H t cycle
2() r
f Na S 2nd cycle
tu
uI 1
<
*
K AN
-4C
LOW AMINO-N SELECTION Ç

80r
Cl
Λ δ H t cycle
60H
* « A 2nd cycle
ίÉ 1I
4
4
w Γ
?o
z r
<i Lr
r T
-?ol·
IF ? '■' I I
! -40H
-60»-
Fig. 1. The effects of selection for low Na"4·, K + or amino-N (AN) on purity
(Py), sucrose percentage (S), chloride (CL), total N (TN) and betaine (B).
Asterisks indicate significant difference between the character and the
parental population mean for that character at the 0.05 probability level.
1*
HIGH Na SELECTION * 1
Y Cl g u i cycle
40>[ H·
*
g
r B 5
h? *>l· B
< 1 m fi
X K
u
*
Γ fi
* ! s i e
* ,N
AN B
•2C)L
•4C
HIGH K SELECTION β
6
6( )r
- i
5 1st cycle CL fi
Ü
<
I £ =
Ξ
AN
*
B
B
I 2C Na B * B m
T fi
L)
s * ■ fi 5 5
* * B fi S ■ £
P fi fi m cz M er
Pv
-2C>L
AN HIGH A M I N O - N SELECTION Q
iOOr *
5 Ut cycle
BOl· 13 2nd cycle
h? 4 TN
, Il 1
B
**
< 1
7 K Ξ" œ
Pv
s* " 1
"a *
CL
-40L
Fig. 2. The effects of selection for high Na-, K- or amino-N (AN) on purity
(Py), sucrose percentage (S), chloride (CL), total N (TN) and betaine (B).
Asterisks indicate significant difference between the character and the
parental population mean for that character at the 0.05 probability level.
15
effects on X extraction. A different pattern was seen following selection for

low amino N. The significant reductions in amino N were accompanied by
reductions in K"1", total N and betaine and by significant increases in Na"·"
(21-342) and in Cl" (53-63Z) (Fig. 1C and Table 4 ) . These increases and
decreases following low amino N selections resulted in reductions in extractable
sucrose and X extraction. Percent extraction was reduced a significant 6X after
two cycles of selection compared to the parental population (76.6 vs 70.6).
Selection for high Na"·" was accompanied by a 45* increase in Cl~ and
reductions in all other characters except K~" (Figure 2A and Table 2 ) . Selection
for high K+ significantly reduced purity and significantly increased all other
characters except extractable sucrose and percentage extraction which were not
different from the parental population (Fig. 2B and Table 3). High amino N
selection which increased amino N up to 98Z also significantly increased total N
and betaine and significantly reduced all other chemical characters except K*
(Fig. 2C and Table 4 ) . Purity and sucrose were significantly reduced only after
two selection cycles for high amino N. Na"** and Cl~ were significantly reduced
after one but not after two cycles of high amino N selection. Significant
reductions in extractable sucrose X and X extraction occurred only after the
second cycle of high amino N selection.
Realized heritability estimates. Realized or true heritability estimates
based on high and low selection from the original parental population are
summarized in Table 5. Heritabilities calculated from high-low selection adjust
for unknown environmental effects which may affect a character more in one
direction than in another. Heritabilities were high, especially for amino N
(0.81) and K+ (0.66). As would be expected from such heritability estimates,
selection in either a high or low direction was very effective in shifting
population means.
DISCUSSION
This study was stimulated by the interest in and lack of information on the
effect of selection for reduced non-sucrose components present in the processing
juice from sugarbeet. Most plant breeding programs for improving beet quality
have concentrated on selection for high sucrose and on high purity. Few breeding
programs have consciously selected against the non- sucrose chemical impurity
components. In those few cases where selection against specific non-sucrose
juice constituents was practiced, interest was on the gross effect on yield,
16
TABLE 3
Mean effects of low and high selection for K*

Character** population Low High Low
K 126.37b*** 104.23c 169.08a 93.17d

Purity Z 88.66a 88.90a 86.19b 88.92a
Sucrose Z 11.84b 11.72b 12.74a 11.55b
Na 152.75b 151.35b 173.08a 157.75ab
Chloride 23.34b 22.78b 32.22a 20.79b
Total N 113.63b 103.78b 135.81a 101.38b
Amino N 32.98b 29.03b 44.15a 26.47b
Betaine 56.59b 49.09b 89.03a 49.91b
Extractable
sucrose Z 9.07a 9.03a 9.11a 8.91a
Extraction Z 76.6a 77.1a 71.5b 77.1a
All values are the average over 3 years of replicated tests for all
year, two years and three years respectively.
Biochemical values are expressed as mg/100 ml. Extractable sucrose Z and

extraction Z calculated according to ref. 17, pp. 450-451.
*Means within a row followed by the same letter are not significantly
sucrose or purity and not on the effects on other important non-sucrose

constituents (refs. 9-15).
The objectives of the present study were to determine to what degree
population means can be shifted for specific purity components and what effects
these shifts would have on other purity components.
Our previous genetic studies had indicated that selection for these
non-sucrose biochemical juice constituents should be effective due to the
predominance of additive genetic variance (ref. 5 ) . Results of the present
research confirmed this—shifts in populations means for Na + , K+ and amino N were
significant after only a single selection cycle. The magnitude.of selection
response was greatest following high selection for all three characters. These
results are indicative of greater genetic variance for high selection than for
low selection. This greater variance might have been expected, since past
selection for improved purity and sucrose has at the same time reduced the
genetic variance for downward selection of certain components associated with
juice quality. Although the effect of selection for low or high Na·*-, K + and
17
TABLE 4
Mean effects of low and high selection for Amino N*

Character** population High Low High Low
Amino N 32.98c*** 22.70d 50.00b 20.91d 63.35a

Purity Z 88.66ab 87.62bc 89.25a 85.95cd 86.30cd
Sucrose Z 11.84a 11.81a 11.90a 10.67b 10.43b
Na 152.75cd 185.50b 121.42c 205.21a 137.90de
K 126.37ab 124.38ab 125.73ab 115.81b 129.27a
Chloride 22.34b 34.23a 15.95c 36.49a 20.64bc
Total N 113.63b 96.00c 144.25a 92.44c 146.83a
Betaine 56.59DC 44.81c 70.16a 37.94d 67.28ab
Extractable
sucrose 9.07a 8.80ab 9.26a 7.54b 7.44b
Extraction Z 76.6a 74.4a 77.8a 70.6b 71.3b
All values are the average over 3 years of replicated tests for all
year, two years, three years respectively.
Biochemical values are expressed as mg/100 ml. Extractable sucrose Z and

extraction X calculated according to ref. 17, pp. 450-451.
*Means across rows followed by the same letter are not significantly
amino N may have been somewhat predictable, based on the known type of gene
action, the effects on the array of other non-sucrose chemical constituents
was certainly not predictable. Furthermore, significant increases in
extractable sucrose and extractability resulted after only one cycle of
selection for low Na*. On the other hand, selection for low K* had little
effect on sucrose Z, extractable sucroseZ or on sugar extractability.
(Fig. 1A and Table 2). Selection for low K*, while effectively reducing K*
(262), also reduced all other chemical characters except Na* which showed a
slight (3Z) increase after two selection cycles.
It is noteworthy that selection for low Na* increased Cl~ by 45Z (Figs.
1A, 2A). In all other cases, with the exception of low K* selection, the
response of Na* and Cl" was in the same direction. Doxtator et al (ref. 12)
-
reported that selection for low Cl also reduced Na*. These patterns
suggest a positive association between these two ions of which the
biological significance is not clear.
18
The research reported herein clearly demonstrated that selection for

low or high Na"*-, K + , or amino N is very effective in significantly shifting
population means. The realized heritability estimates (Table 5) should
provide good predictive values for other sugarbeet populations since the
parental population we used for this study was similar to the type of
commercial plant breeder would use. Perhaps of most importance, however,
was the consequence of our selections on the array of other chemical
constituents and on purity and recoverable sucrose yield. It can be
concluded from this research that selection for any one of the three
chemical constituents previously identified as important to juice quality
could significantly affect purity and extractable sugar per ton of beets.
In some cases, multiple character selection (ref. 16) may be necessary to
achieve the goal of increased extractable sugar per ton. Sugarbeets with
significantly modified levels of "impurity" components such as our low Na-*-,
K"** or amino N populations may be more amenable to processing.
Plant breeders have spent many years striving to improve purity and
sucrose in sugarbeet. Many of these breeders would agree that we have
reached a sugar yield plateau in sugarbeet. Selection for components of
purity may provide an avenue for overcoming this plateau if the correct
component or combination of components can be modified.
TABLE 5
Realized heritability (ha) estimates obtained from high and low selection
for Na, K and amino N.*
Selection group mean* Resulting population mean-*"*·

Character High Low High Low h2
Amino N 9.45 2.13 9.27 3.29 0.81

Na 29.22 6.09 11.97 6.72 0.23
K 32.50 13.06 33.53 20.77 0.66
* Values for amino N, Na and K are expressed as mg/100 ml.
■*· Means for low selection group based on 110 roots per character; means for
high selection group based on 60 roots per character.
■Resulting populations were from plants intercrossed within selection

groups; means are based on 500-800 roots for each high or low population.
19
REFERENCES
1 J. T. Alexander, Factors affecting quality, in: R. T. Johnson, J. T.

J. T. Alexander, 6. E. Rush, and 6. R. Hawkes (Eds.) Advances in
sugarbeet production: Principles and practices. The Iowa State Univ.
Press, Ames. 1972, pp. 37-381.
2 G. Rorabaugh and L. W. Norman, The effects of various impurities on the
crystallization of sucrose, J. Am. Soc. Sugar Beet Technol. 9 (1946)
238-252.
3 A. Carruthers and J.F.T. Oldfield, Methods for the assessment of beet
quality, Int. Sugar J. 63 (1961) 72-74, 103-105, 137-139.
4 G. A. Smith, S. S. Martin and K. A. Ash, Path coefficient analysis of
sugarbeet purity components, Crop. Sci. 17 (1977) 249-253.
5 G. A. Smith, R. J. Hecker, G. W. Maag and D. M. Rasmuson, Combining
ability and gene action estimates in an eight parent diallel cross of
sugarbeet, Crop. Sci. 13 (1973), 312-316.
6 Association of Official Analytical Chemists. Official methods of
Analysis, 12th ed. Washington, D.C. (1975).
7 G. W. Maag, Rapid digestion procedures for determination of metallic
ions and total nitrogen in sugarbeet samples, J. Am. Soc. Sugar Beet
Technol. 15 (1969) 361-367.
8 S . S . Martin, R. J. Hecker and G. A. Smith, Aluminum clarification of
sugarbeet brei extracts. J. Amer. Soc. of Sugar Beet Technol. 20 (1980)
597-609.
9 W. R. Akeson, D. G. Westfall, M. A. Henson and E. L. Stout, Influence
of nitrogen fertility level and topping method on yield, quality, and
storage losses in sugarbeets. Agron. J. 71 (1979) 292-297.
10 L. G. Campbell and J. J. Kern, Relationships among components of yield
and quality of sugarbeets. J. Am. Soc. Sugarbeet Technol. 22 (1983)
135-145.
11 C. W. Doxtator and H. M. Bauserman, Parent-progeny tests for sodium and
potassium content, Proc. J. Amer. Soc. Sugar Beet Technol. (1952)
319-321.
12 C. W. Doxtator, R. E. Finkner, R. H. Helmerick and P. C. Hanzas, The
performance of sugarbeet selections made for purity and chloride at three
levels of nitrogen fertility, J. Am. Soc. Sugarbeet Technol. 13 (1965)
363-369.
13 J. W. Dudley and L. Powers, Population genetic studies on sodium and
potassium in sugarbeets (Beta vulgaris L.), J. Am. Soc. Sugar Beet
Technol., 11 (1960) 97-127.
14 R. E. Finkner and H. M. Bauserman, Breeding of sugarbeets with reference
to sodium, sucrose, and raffinose content. J. Am. Soc. Sugar Beet
Technol. 9 (1956) 17-177.
15 R. R. Wood, H. L. Bush and R. K. Oldemeyer, The sucrose-sodium
relationship in selecting sugar beets, J. Am. Soc. Sugar Beet Technol.,
10 (1958) 133-137.
16 G. A. Smith, Sugarbeet, in: W. R. Fehr and H. H. Hadley (Eds.),
Hybridization of crop plants. Am. Soc. Agron., Madison, 1980.
17 S. T. Dexter, M. G. Frakes and F. W. Snyder, A rapid and practical
method of determining extractable white sugar as may be applied to the
evaluation of agronomic practices and grower deliveries in the sugar
beet industry. J. Am. Soc. Sugarbeet Technol. 14 (1967) 433-454.
20
Chapter 3
CARBONATATION PROCESS IN BEET SUGAR MANUFACTURE

M. CLEARY
INTRODUCTION
Purification of the initial aqueous extract from sugarbeet (raw juice

or diffusion juice) is accomplished through the addition of lime and carbon
dioxide; hence the term carbonatation or less accurately carbonation. This
paper describes the objectives in juice purification, a brief history of the
process, some of the more important chemistry, and finally the current
practice of carbonatation, in general terms. The product of the
purification process is called thin juice and is subsequently concentrated
in multiple effect evaporators to thick juice. This is essentially the
starting point for a three-stage crystallization process resulting in white
sugar and molasses. The thin juice must satisfy three criteria, which are
described below.
Clarity
The growth and recovery of clear clean sugar crystals requires a mother
liquor which is free of particulate or colloidal impurities. This means
that the thin juice must remain clear throughout subsequent processing
except for the presence of growing sucrose crystals. Extraneous suspended
material inevitably leads to a "false seed" for crystal growth and high
sediment levels in final product. Since the charge to the white pans, the
vessels in which white sugar is crystallized, is filtered prior to the
crystal growth, some precipitation can be tolerated during concentration of
the purified juice. Nonetheless thin juice which does not remain clear
during evaporation may lead to scaled or fouled evaporator surfaces.
Stability
Two important parameters of beet sugar technical syrups are color and
pH. Excessive amounts of color bodies in the juice will lead to
unacceptably high levels of color in the final product either as crystal
inclusions or as residue on the crystal surface. The purified juice should
be low in color and remain so through crystallization. Also the juice must
be able to maintain a sufficiently high pH to avoid hydrolysis of sucrose.
21
Purity
Sucrose comprises about 13Z of the raw juice and 85-902 of the soluble
material. Industry-wide the percent sucrose on solids value is referred to
as the purity of the juice or syrup. An objective of the carbonatation
scheme is to maximize the purity of the juice which in turn will allow more
complete recovery of the sucrose by crystallization. Thin juice purities
generally fall in the range of 88-93Z.
Early attempts to produce sucrose from sugarbeets are attributed to
Franz Achard, who 200 years ago selected the first "high" sugar varieties
from common forage beets. Available technology at the turn of the 19th
century included wine-making, sugarcane processing, and alchemy--more than
700 reagents were employed during the early 1800*s ; a sampling of the list
includes:
magnesium hydroxide bentonite
suifuric acid/chaulk alumina
lime/cows' blood soy bean meal
By the 1840's the lime/carbon dioxide combination had become the recipe
of choice and by the I860's the basic steps of carbonatation as it is now
practiced were in place. Since that time the process has been transformed
from small batch processes carried out in 200 or 300 gallon (750-1200 L)
vessels to continuous processes operating at 1000 to 3000 gallons (4000-
12,000 L) per minute and vessels of up to one hour retention time (75,000
gallons) (300,000 L ) .
Raw material
The raw juice is an aqueous mixture of the soluble material from the
counter-current extraction described in Chapter 1, and carries variable
amounts of suspended matter which combine to make the juice appear cloudy
and dark grey. Typical solids levels range from 12 to 16Z by weight; the pH
from 5.5 to 6.5; and the temperature from 5 to 75°C. It is biologically
active and chemically unstable.
A portion of the cloudiness can be attributed to soil which is not
removed prior to slicing the beets. This will vary in impact from area to
area and is a major concern when the beets have been grown in heavy clay-
bearing soils or where they have been harvested under muddy conditions. The
soil also carries micro-organisms which thrive in this medium.
22
Intercellular material from the countercurrent extraction process also

contributes turbidity. Such pectic substances, proteins, and other
mitochondrial and cell wall material, because of their gel-like behavior,
are not easily removed by filtration. Most of these substances will
coagulate at high pH or form well-behaved precipitates with calcium which
can be readily filtered in the presence of calcium carbonate.
The process is designed to deal with specific impurities while
preserving sucrose. The most important loss of sucrose comes from
hydrolysis which may be catalyzed by enzymes, acid, base, or salts, and is
referred to as inversion in reference to the effect on polarized light.
water + sucrose -» glucose + fructose
(rotates light clockwise) (rotates light counter-clockwise)
The sugarbeet contains several enzymes which facilitate this reaction
and almost all of the micro-organisms exhibit some invertase activity. The
acid-catalyzed reaction becomes important below pH 6, especially at
temperatures greater than 60°C. Raising the pH with lime effectively blocks
both pathways.
Since sucrose is not a reducing sugar, it is not readily oxidized by
the addition of lime even at temperatures of 80-100°C typical of the
purification scheme, although it can be "burned" on exposed heating surfaces
or by the uncontrolled addition of soda ash for ancillary pH control. This
reaction begins with the base-catalyzed hydrolysis of sucrose.
The hydrolysis of sucrose is one of the oldest to be found in the
literature of chemical kinetics. It represents the first association
between the velocity of a reaction and the concentration of the substrate
(ref. 1 ) .
sucrose + water -» glucose + fructose
Rate = k [sucrose]
k = ka[aH-] + kb[aOH-]°-3 + ke[S]
where ka, kb, and k s are the rate constants for the acid-, base-, and salt-
catalyzed pathways. Their values are given by:
In ka = 38.9 - E/RT
E = 24.6 kcal (activation energy)
In k b = 23.9 - E/RT
T = temperature °K
In ks = 25.5 - E/RT
23
The salt concentration is calculated (ref. 2) from the total ion

concentration, the amino nitrogen content [AmN], the alkalinity value [Ca],
and an estimate of anions which will be formed via sucrose degradation [AF].
[S] = [Total Ions] + [AmN] + [Ca] + [AF]
Values for [S] range from 0.04 equiv/1 for high purity thin juice to
0.12 for relatively low purity thin juice.
The base- and salt-catalyzed reactions are most important in the first
of typically 5 evaporator effects where the juice is subjected to
temperatures of 130°C for several minutes. The pH value for these juices
displays a marked inverse dependence on temperature, while the pOH shows
little dependence. For example, good quality thin juice with a pH of 7.8 at
80°C will have a pH of 7.2 at 130°C and the corresponding pOH values are 4.9
and 4.95 respectively. Lower purity juice shows similar behavior but the
pH's tend to be lower, even to the point that the acid-catalyzed reaction
may become very important at 130°C. Contact time of 3.5 minutes at 130°C
will hydrolyze about 0.13Z of the sucrose in high purity thin juice and
0.40Z in low purity thin juice at a pOH of 4.5. When the pOH is raised to
5.5, sucrose disappears to the extent of 0.25Z and 0.622 respectively.
The most common cation in sugarbeets is potassium which, unless the
beets have been cultivated in very saline soils, occurs in 3 to 20 fold
excess of sodium. Both are readily extracted along with the sucrose, and
together total 2500 to 3500 ppm in the raw juice and account for about 10Z
of the non-sucroses in the juice. They exhibit a marked salting-in effect
on sucrose and will reduce the final yield of crystals. This is so
important that many European and at least one American sugar producer
purchase beets on the sodium and potassium contents as well as the sucrose
content. Other cations present in juice include magnesium (1.5Z of the non-
sucroses), calcium (0.5Z) and ammonium; the latter is usually a harbinger of
protein degradation prior to processing.
Inorganic anions include chloride, nitrate, phosphate and sulfate all
of which together comprise about 10Z of the non-sucroses and may exhibit
variability with variety, soil type, growing conditions, and agronomic
practices. Nitrite, while rarely exceeding 100 ppm on juice is an important
indicator of in-process infections.
While several neutral carbohydrates have been found in diffusion juice
the most prevalent are fructose, glucose and raffinose. Fructose and
2k
glucose, the initial products of sucrose metabolism whether by infection or

the beet itself, normally occur in amounts of from 0.2 to 0.4Z on sucrose.
They tend to increase in concentration with beet storage time and always
result in thin juice with higher color and acidity.
glucose or fructose -» acids
-* color bodies
These reactions are discussed below, but it is important to note that
both are relatively fast when compared with sucrose hydrolysis (above).
Raffinose, a non-reducing trisaccharide comprised of galactose and
sucrose, exhibits solution chemistry virtually identical to that of sucrose.
In fact, when levels as high as IX or sucrose are encountered, specific
faces of the sucrose crystal exhibit retarded growth and crystal elongation
is observed. Elevated levels of raffinose are associated with beets grown
in cooler climates, although considerable varietal differences are observed.
Levans and dextrans may also be present in small but technically
important amounts as microbial by-products. They are called gums because of
their effect on filtration and are typically encountered when the beets have
been subjected to extremes of heat or frost.
Since the sugarbeet utilizes the entire complement of Krebs' cycle
acids, several are found in diffusion juice along with microbial and
oxidative products such as acetic, formic, and lactic acids. In the absence
of severe beet decomposition and infection, citric is the most abundant
organic acid (10-15Z on non-sucroses) with malic and oxalic present in
lesser amounts (3-4Z each). Acids which form insoluble calcium salts (the
latter trio for example) will be removed to varying degrees during
carbonatation; and residual amounts may precipitate as the juice is
subsequently concentrated. This is especially true of oxalic acid since the
solubility of calcium oxalate decreases as the sucrose concentration
increases.
Pectic acids (from partial hydrolysis of pectins) and their monomer,
galacturonic acid, are also found in diffusion juice. Both form extremely
insoluble calcium salts and are easily removed. Several glycosides have
been identified in sugarbeets and resultant juices. The most abundant of
these saponins has been characterized as a combination of oleanolic and
glucuronic acids. Although present in small amounts, they are only
partially removed and may precipitate on concentration. They are important
25
from a product viewpoint since a few parts per million saponins in white
sugar will cause floe on acidification in such end-use products as beverages
or jellies. Since their solubility increases as pH is raised, soda ash may
be added to the crystallization steps to prevent deposition of saponins on
white sugar.
The amino acid content of diffusion juice is about IX on juice or 15-
20Z of the total non-sucroses. Sugarbeets possess a wide range of amino
acids but the most common is glutamine which comprises half of the total
amount. Immature beets, besides having low levels of sucrose, display high
levels of glutamine which further reduces their technical value. In some
cases the process may have to be reconfigured to deal with excess glutamine.
OH
o
NH2CCH2CH2CHC02H (C^NCHô" (CH 3 ) 3 NCH 2 CH 2 OH
NH 2
Glutamine Betaine Choline
Betaine (pronounced bee-tane), or Ν,Ν,Ν-trimethylglycine, is another

nitrogen-containing compound which acts as the osmoregulating agent in
beets. It is normally grouped with the amino acids even though it has been
shown to arise from choline rather than glycine. It accounts for 5-102 of
the non-sucroses in diffusion juice and is occasionally referred to as
"interfering nitrogen" by technologists (ref. 3a).
PURIFICATION REAGENTS
The basic reagents of juice purification are milk of lime and carbon
dioxide both produced by burning limerock:
CaC03 -* CaO + C0 2
Heat for the reaction is provided by combustion (eg. coke) and thus the
"yield" of carbon dioxide is in excess of stoichometric amounts. Lime is
converted to an aqueous slurry of calcium hydroxide in the slaking process,
which is greatly enhanced by the presence of sucrose. This slurry contains
10-15 wt I lime and is referred to as milk of lime. As with any reagent its
manner of production and delivery are critical to the chemistry of the
process.
26
The use of lime and carbon dioxide serves three basic functions:
1. Source of alkalinity. High pH's protect sucrose from degradation
and virtually stop microbial action. Some alkaline degradation reactions
are crucial to production of stable juices and prevention of further
coloring. If the dissociation of calcium hydroxide is considered
Ca(OH) 2 <—* CaOH-*- + OH"
CaOH- *—> Ca-*"*- + 0H~
the source of both alkalinity and calcium is apparent. Most of the
processes in juice purification are controlled using the pH as the primary
input value, which in these juices tends to vary with temperature. The
primary reason for this behavior is that the above reactions show very
little dependence on temperature. In other words the calcium hydroxide will
serve to keep the hydroxide concentration regardless of the dissociation
constant of water.
TABLE 1
Temperature dependence of Kw.
Temperature pkw
50°C 13.26
60 13.02
70 12.77
80 12.55
90 12.34
The use of pOH as a control parameter which is relatively insensitive

to temperature is under investigation (refs. 4, 5 ) .
2. Source of calcium. Some components of the juice will form
insoluble calcium salts or coagulates and are thusly removed.
3. Source of calcium carbonate. Many of the colloids in juice,
including some of the precipitated calcium salts have a slight positive
charge and are effectively coagulated by freshly precipitated calcium
carbonate. Even colloids which are not easily filtered (gums and negatively
charged clays) can be caught using the bulk effect of calcium carbonate as a
filter aid.
C02(aq) + 2 OH- -* C 0 3 — + H 2 0
27
The first step of the reaction is the dissolution of carbon dioxide

followed by its conversion to carbonate, then there is apparently more to
the reaction than the simple reaction with calcium, although the
stoichiometry applies. When the milk of lime is produced in the presence of
sucrose the pH's are high enough to form sucrose anions (saccharates or
sucrâtes), which in turn complex with calcium.
sucrose + OH - -» saccharate- + H 2 0
When calcium precipitates as the carbonate the initial reaction is
between carbonate and the saccharate complexes of calcium to form a
saccharate-carbonate compound (ref. 6 ) .
saccharate- + 2 Ca"4"*" + C 0 3 — -* saccharate-Ca-C03-Ca-saccharate
These have also been presented as chains of (-saccharate-Ca-C03-Ca-)
monomers (ref. 7) and carbonate esters of sucrose (ref. 8 ) . In any case the
compound splits on hydrolysis and further reaction with carbonate to give a
dicalcium-carbonate species with a positive charge.
sacch-Ca-C03-Ca-sacch + 2H20 -> [Ca^COs-*"*-] + sucrose + 2 OH"
+
[Ca 2 C0 3 - ] + C0 3 -- -» 2 CaC03
This explains why the calcium carbonate exhibits a slight positive
charge and strongly adsorbs the negatively charged colloids found in juices.
Having served its purposes in the process, it is necessary to minimize
the amount of calcium remaining in the juice. This requires a consideration
of the carbonate-bicarbonate equilibria along with the Ca(OH) 2
neutralization,
CO3-- + H 2 0 <—> HC0 3 - + OH-
HC0 3 - + H 2 0 <—» H 2 C0 3 + OH-
Precipitation of calcium carbonate cannot be forced by over-
carbonating the final solution because the pH of the solution will be
lowered enough to bring the reaction into the bicarbonate range and
redissolve some of the precipitate, while redispersing anything that has
been adsorbed on its surface. There exists an "optimum" pH or alkalinity
for the juice beyond which further carbonation is detrimental; the actual pH
value depends on the non-sucrose components of the particular juice and is
best determined empirically. The equilibrium is reached five or ten minutes
after the final addition of carbon dioxide and many process schemes provide
a resting period for the juice before filtration.
28
While beyond the scope of this review, it should be mentioned that some
juice purification processes are followed by an ion-exchange process which
substitutes another cation (e.g. sodium or magnesium) for calcium (ref. 9 ) ,
as described in Chapter 5, by Shore et al.
PROCESS CHEMISTRY
Fermentation consequences
Even though fermentation reactions occur prior to the purification
steps, their products frequently have to be considered when operating the
process. Some metabolites such as lactic, acetic, and formic acids simply
add to the pool of non-sucroses (at the expense of sucrose), are not
affected by carbonatation, and increase the amount of final mother liquor
from crystallization--molasses.
TYPICAL FERMENTATION REACTIONS

sucrose -> glucose + fructose
-* levans or dextrans
glucose or fructose -» lactic acid
-» volatile acids
-» alcohols
nitrate -» nitrite
Fermentation products of interest in juice purification include
fructose and glucose (discussed below), nitrite, levans and dextrans. If
nitrite is not removed in the purification process, or if subsequent
infections are allowed to occur, nitrite may react with sulfite to form
imidosulfonate which will precipitate as the potassium salt in the first
crystallization.
N0 2 - + 3 SO3-- + 2 H 2 0 -> HN(S03)3-- + 3 OH" + SO* —
imidosulfonate
The reaction is a formal reduction of nitrogen and the reducing agent
shown here is sulfite for convenience only.
Levans and dextrans will not precipitate with calcium but they will be
absorbed on calcium carbonate. Unfortunately dextrans (and possibly levans)
reduce the size of precipitated calcium carbonate crystals and discourage
agglomeration (ref. 10). This makes the resulting sludge difficult to
filter and wash.
29
PRECIPITATION WITH CALCIUM

The most easily explained chemistry of carbonatation is the
precipitation of anions as their calcium salts. Many authors categorize
juice anions according to calcium salt solubility. The least soluble ones
include oxalate, phosphate, sulfite, and galacturonate. Proteins and pectic
substances which are present as colloids or semi-solubles are also
precipitated with calcium. Perhaps 90Z of the citrate and saponin is
precipitated and only 50Z of the malate or sulfate. Most of these salts
exhibit a decrease in solubility as the temperature is increased and the
carbonatation process usually culminates with a temperature 15 to 20° C
greater than crystallization temperatures. Also, since most of these are
more soluble in concentrated sucrose solutions, they do not crystallize
during evaporation.
Oxalate and citrate warrant special consideration after purification.
Calcium oxalate is less soluble in solutions with higher sucrose
concentration and oxalate scaling is observed in the later stages of
evaporation. Citrate, besides precipitating with calcium, will also form a
1:1 complex in alkaline solutions:
Citrate + Ca + + <—» CaCitrate-
Citrate which is not removed can "hide" calcium in solution until
crystallization at which time it can provide an unwelcome source for oxalate
precipitation (ref. 11).
ALKALINE OXIDATION AND DEGRADATION

The most vigorous condition of carbonatation is a mixture of 1.5Z lime
slurry at a pH of about 12 and a temperature of 80°C. Some processes may
enhance oxidation reactions by sparging air into the mixture. These
reactions include:
monosaccharides -* acids
sucrose -> -» acids
nitrite -> nitrate
sulfite -> sulfate
glutamine -» 2-pyrrolidone-5-carboxylic acid + ammonia
Acids identified as alkaline degradation products of sugars include
glucinic, apoglucinic, saccharummic, molassic, humic, saccharinic, and
lactic; of which only saccharummic and humic form insoluble calcium salts
30
(réf. 3b). Even though destruction of monosaccharides results in more

highly colored thin juice, this is considered preferable to allowing them to
form acids or color in the evaporation or crystallization stages.
About half of the amino acids in diffusion juice is glutamine and
unless it is deamminated in juice purification it will do so in the
evaporators and the juice pH will drop. Ammonia which is produced by this
reaction (along with ammonia from protein degradation prior to processing)
is driven off. Also glutamine which has been converted to 2-pyrrolidone-5-
carboxylic acid is unavailable for color-forming reactions.
Color formation
The final color of the processed juice is an important parameter since
it corresponds directly to the color of the crystalline sucrose which will
be produced. Color bodies have been classified into two groups—melanine
and melanoidins (ref. 12).
Melanins are represented by the colored products of the oxidation of
tyrosine to dihydroxyphenylalanine catalyzed by phenol oxidase. Other
enzymatic browning products also fall into this category, and together they
are responsible for the dark grey color of the raw juice. The formation of
melanine is stopped with the addition of lime and their removal is
accomplished via reversible adsorption onto calcium carbonate at a pOH of
2.9 to 3.0.
Melanoidins in contrast are usually produced in the process either by
the alkaline degradation of monosaccharides or their reaction with amines in
the Maillard reaction—formation of a Schiff's base followed by an Amadori
rearrangement. The reaction between glucose and glycine is typical.
HC=0
I
HC-OH
I NH2 HC=N-R HC-NH-R
HO-CH I z 1 H
HC-OH -♦ C-OH
I + CH2 ■ 1
1
1
1
HC-OH I R R
I CQ2H
HC-OH ^
I
CÔH
glucose glycine Schiffs base Amadori Product
31
The products then proceed with further color-forming reactions. Dozens

of products have been documented for the above example (ref. 13).
Unlike the melanins these products are adsorbed almost irreversibly on
calcium carbonate over a pOH range of 2.5 to 5.0. The most satisfactory
tactic in handling monosaccharide-based colors is to force them to
completion prior to the final exposure of the juice to calcium carbonate.
After the juice has been decalcified, 100-150 ppm sulfur dioxide is normally
added to deal with the residual amount of monosaccharides and with those
that will be formed during evaporation. This is accomplished by tying up
the reducing sugar as the bisulfite adduct thus making it unavailable for
decomposition or reaction with amines or ammonia.
HC=0 1
1 HC-S0 3
HC-OH + HSO3 -► 1
1 HC-OH
R 1
R
:ose bisulfite bisulfite addition
product
GENERAL PROCESS FLOW SCHEME
Many different processes have been designed to take advantage of the

chemistry described. The "classic system" described herein follows the most
commonly practiced options. A new cold liming system is described in
Chapter 4.
RAW MILK OF LIME IN

JUICE
, 1 , * * _ 1st CARB.
! ^THIN
I JUICE
-PRELIMER 1 MAINLIMER 2nd SULFI
CLARIFIER CARB. TATION
CO. CO.
RECIRCULATION
UNDERFLOW RECIRCULATION
UNDERFLOW TO VACUUM FILTERS
Fig. 1. Classic prelime system.

32
Preliming
0.25Z CaO on juice is added to bring the pH into the alkaline range.
Insoluble calcium salts are precipitated as finely dispersed colloids.
Calcium carbonate in the form of recycled first carbonation sludge is added
to begin colloid adsorption. Temperature may be cool (50°C) or hot (80°C)
depending on the temperature of the next step. The retention time is 15
minutes.
Main liming
A further 1.50Z CaO on juice is added and the juice is brought to its
maximum alkalinity. Conditions may be cool (50°C for 45 minutes followed by
80°C for 5 minutes) or hot (80°C for 10 minutes).
First carbonation
The process stream is neutralized to pOH of 3.0 with carbon dioxide.
Juice is recycled either internally or in a separate vessel to provide seed
for calcium carbonate. The retention time is 20 minutes at 80°C.
Sludge separation
First carbonation effluent is passed through a continuous clarifier
where the precipitated calcium carbonate is allowed to settle and the clear
juice flows to second carbonation. Clarifiers may be rapid (10 minutes with
the use of flocculents) or relatively slow (50 minutes using lower levels of
flocculents). The sludge which is not recycled to the pre-limer is removed
and washed by vacuum filtration, with the filtrate returned to process.
Second carbonation
Calcium is reduced to the practical minimum via the addition of carbon
dioxide at a pOH of 4.5. The temperature is brought to its maximum point in
the purification process (98°C) and the residence time is the minimum
required for neutralization (5 minutes). The sludge from this unit
operation is relatively less and much more easily filtered that first
carbonation sludge. It is removed by in-line filtration.
33
Sulfitation
Sulfur dioxide is added to the level of 150 ppm on juice to discourage
further color-forming reactions.
FURTHER CONSIDERATIONS
Ultimately the aim of carbonatation is to minimize the amount of non-
sucroses which proceed to crystallization. The relationship of sucrose and
the associated non-sucroses are embodied in the purity measurement. Purity
is the ratio of sucrose to solids (sucrose + non-sucroses) and can be used
to estimate the recovery of the crystallization process.
CALCULATION OF SUCROSE RECOVERY ON CRYSTALLIZATION

Example: Thin juice purity = 88Z
Molasses purity = 60Z (final mother liquor)
Sucrose entering : Non-sucrose entering = 8 8 : 12 = 7.33
Sucrose leaving* : Non-sucrose leaving = 6 0 : 40 = 1.50
Sucrose recovered = 7.33 - 1.50 = 5.83 or 79.6Z of 7.33
Thin juice purity = 90Z
Molasses purity = 602
Sucrose recovered = 9.00 - 1.50 = 7.50 or 83.3Z of 9.00
The calculation assumes that the ammount of non-sucroses do not change

during the process. This example indicates that the yield (production of a
factory) will be almost 5 percent higher if the thin juice purity is
90Zinstead of 88Z. Similar calculations reveal that small changes in
molasses purity do not have as much impact as those for thin juice purity.
A purity value may also be estimated for the raw juice and a non-
sucrose elimination (carb elimination) can be calculated across the
carbonatation process. It is expressed as per cent on non-sucroses entering
the process.
"leaving" unrecovered in the final mother liquor, molasses

3^
CALCULATION OF CARB ELIMINATION

Raw Juice Purity =87.0
Thin Juice Purity =90.0
Non-sucrose : sucrose in raw juice = 13 : 87 = 0.149
Non-sucrose : sucrose in thin juice = 10 : 90 - 0.111
Elim = removed/entering = (0.149 - 0.111)/0.149 = 251
To some extent this may depend more on the way the non-sucroses are
accounted for than their actual removal, but it remains one of the most
common methods of evaluating the efficiency of the process.
REFERENCES
1 L. Wilhelmy, Reaction products and mechanism in some simple model
systems. Pogg. Ann., 81 (1850) 413, 499.
2 K. Vukov, Active alkalinity of sugar factory juices. The Sugar Journal,
October 1975, 16-22 (shortened English version of the original German
translated by R. A. McGinnis).
3a P. M. Silin, Technology of Beet-Sugar Production and Refining (Engl.
Transi, by L. Markin); National Science Foundation; Washington, D.C.,
1964; p. 47; Teknologiya sveklosakharnogo i rafinadnogo proizvodstva;
Pishchepromizdat; Moscow, USSR, 1958.
b Ibid. p. 54-57.
4 J. Kysilka and T. Mcgillivray, Control of first carbonation pOH with a
differential electrode system, Presented to the 20th General Meeting
of the American Society of Sugar Beet Technologists, San Diego, CA,
1978.
5 J. Kysilka, Personal communication, February 1987.
6 J. Dubourg, Calco-carbonic purification of ternary solutions—water—
sugar--lime. Suer Belge.,70 (1950-51) 237-240.
7 J. Dubourg, Calco-carbonic purification in beet sugar manufacture.
Sucr. Franc, 93 (1952) 153.
8 G. Dorfmüller, Action of lime and carbonic acid on the ternary system
lime--sugar—water. Zucker 4 (1951) 407-409.
9 M. Shore et al., Chapter 5.
10 J.F.T. Oldfield, J. V. Dutton, H. J. Teague and E. L. Williams, Effect
of dextran on 2nd carbonatation filtration, Presented to XV Assemblee
Generale Commission Internationale Technique de Sucrerie; Vienna,
Austria, 1975.
11 M. Shore, Personal communication, March 1987.
12 K. Vukov, Adsorption of colouring matter by calcium carbonate. The
Sugar Journal, January 1977, 15-18 (transi, from German by R. A.
McGinnis originally from Zucker 29 (1976) 49-53.
13 K. Olsson, P. Pernemalm and 0. Theander, Reaction products and mechanism
in some simple systems. Acta Chem. Scand. B., 32 (1978) 249-256.
Chemistry and Processing of Sugarbeet and Sugarcane, edited by M. A. Clarke and M. A. Godshall 35
Chapter 4
AN INTEGRATED JUICE PURIFICATION SYSTEM

V. ANDERSSON AND S. BARFOED
INTRODUCTION
In thi paper, the basic processes that take place in beet juice
purification are reviewed. Some of the major improvements introduced since
the method s first put into operation are described.
The proc ess as it is at present carried out at DDS is presented in
detail; area s for potential improvement will be discussed.
Aims of iuic e purification

In general terms, purification is designed to obtain juices which
permit the production of high quality white sugar with a minimum loss of
sugar in the molasses.
The goal can be achieved by removing as much as possible of the non-
sugars from the juice. The major non-sugars in the diffusion juice include:
insoluble solids, proteins, pectins, organic acids, amides, etc., invert
sugar, inorganic compounds, and coloring compounds.
How are these compounds removed? Although the chemistry of
purification, described by Cleary in Chapter 3, is very complicated and even
today not completely understood, the basic process is fairly simple. The
major chemicals used are lime (CaO) and carbon dioxide (C0 2 ), which are made
simultaneously by burning limerock (calcium carbonate) in a kiln.
The basic process of using lime and carbon dioxide was introduced
around 1840, and was developed almost to its present form almost 120 years
ago by Perrier and Possey in 1859.
Numerous attempts have been made over the years to develop alternative
processes: it was claimed in the past that 737 different chemicals had been
tried, but lime and carbon dioxide still survive, and work on process
improvement continues.
McGinnis has a word of advice to chemists: "Almost every conceivable
variations of the liming/carbonation process itself has been considered. So
thorough has been this scrutiny that one can be almost certain that any
36
reasonable variation one may be able to devise involving beet juice, lime
and carbon dioxide will be found in the literature." (ref. 1)
In practice, however, process development has largely been empirical,
and has been accompanied by delayed development of theoretical explanations.
Improvements can still be made—sometimes it may be worthwhile to use old
findings in new contexts.
If the chemistry is old, then what is new in the DDS process? In fact,
very large improvements have been made in two major areas:
1. Continuous processing
2. Automatic process control
The traditional process was a batch process. The change to continuous
processing started about 50 years ago, when equipment for continuous
preliming and thickening of muds was introduced; since then, the complete
process has been made continuous. This development has been supported by
the development of automatic controls especially in the last 15 years.
Continuous processing controlled automatically has had a very large
effect on consumption of chemicals, labor costs, and efficiency of
purification.
Today we have a process which, compared with the old process, is both
better and cheaper.
To illustrate the "state of the art" in juice purficiation, the DDS-
system used in the five DDS factories in Denmark, and partly or entirely in
many factories around the world, is presented.
The principal steps in DDS juice purification are shown in Table 1.
TABLE 1
Principal steps in DDS juice purification.
1st Carbonation 2nd Carbonation
Cold preliming Heating and addition of soda

Cold main liming 2nd carbonation
Hot main liming Filtration
1st carbonation Additional operations:
Filtration sulphitation and concentration
Recirculation of mud of muds.
Let us look briefly at what is happening in the individual steps.

37
Cold preliming. The alkalinity is gradually increased to 0.22 CaO in

the cold prelimer; mud is recirculated, at a temperature of 40° C and total
lime content of 2.02 CaO, over a residence time of 25 min. The effects of
this step are: neutralization of acidity; precipitation of pectin and
calcium salts, and subsequent improvement of mud filterability.
Cold main liming. In the cold main liming process, lime is added to a
total of 2.22 CaO, and alkalinity of 0.502 CaO, at a temperature of 40° over
a residence time of 3/4-1 hour. The process causes precipitation of
proteins and amino acids; hydrolysis of amides, and bacteriological and
thermal stabilization. Invert is destroyed to some extent.
The liming takes place in a rather large agitated tank which acts as a
buffer tank in the process.
Hot main liming. At a temperature of 85° C and alkalinity of 0.82 CaO,
with total lime of 2.52 CaO, over a residence time of 10-15 min., further
destruction of invert is ensured.
First carbonation. C0 2 gas is pumped in to effect neutralization to
0.082 CaO. Calcium carbonate precipitates, together with floes from liming
steps. The subsequent filtration effects separation into filtered juice and
thickened mud for recirculation or concentration. After filtration the
juice goes to 2nd carbonation where soda and further C0 2 are added.
Second carbonation. C0 2 is added to 0.0252 CaO, thereby reducing the
lime salts to 0.0025-0.00402 CaO. Filtration then produces thickened mud
for concentration, and a clear juice that is ready for further processing,
including sulphitation to prevent color formation during evaporation.
Figures 1 and 2 show the details of 1st and 2nd carbonation.
This is a general description of the complete juice purification system
as it has been developed by DDS. It is of interest to look in more detail
at a part of the system which perhaps differs most from many other systems,
that is the filter station.
As is shown in Figures 1 and 2, the filter station has two functions:
1. To produce clear, filtered juice for further processing, and,
2. To concentrate the solids into a sludge, part of which is returned
to the prelimer, partly filtered on rotary vacuum filters or on automatic
filter presses.
38
To 2nd
carb
Pre-liming
J-€J
Gold main liming 1st carb filtered)

juice
Fig. 1. Schematic of first carbonation.

Filtered
juice
1st carb
filtered
juice
Unfiltered
juice
Fig. 2. Schematic of second carbonation.

39
In comparison with other widely used systems, the filters replace the
thickener, the check filter for clear juice and increase vacuum filter
capacity for mud to be recycled.
In the DDS filter station, a number of filters are run in parallel,
basically as an automatic batch operation, but with a constant and
continuous flow in and out of the station.
In fact, the operation of the individual filters is regulated by the
flow of raw juice into the 1st carbonation tank. Each time a predetermined
amount of raw juice has passed the measuring point, an impulse is sent to
the filter station starting a cleaning cycle of one of the filters.
The cleaning cycle is shown in Figure 3.
From 1st carbonation
Filtering
Filtering
To 2nd carbonation
Washing
To 2nd carbonation
Sludge pumping
Sludge
Pumping
To 2nd carbonation
Sludge
Fig. 3. Cleaning cycle.

ko
In the filtration part of the cycle, valves A and B are open, and
unfiltered juice flows from the feed tank by gravity (1 to 4 mWG). Juice
enters the filter housing, passes through the filter bags, and clear juice
exits to a filtrate tank while the solids are deposited on the outside of
the filter bags.
When the signal for washing is received, valve B is closed and valve C
is opened. Filter bags are then back-washed with clear juice from the
washing-juice tank in which the liquid level is kept constant at a higher
level than in the filtrate tank.
When the flow is reversed, the bags will expand and the mud deposited
on the outside of the bags will fall off, dropping to the bottom. At the
same time valve E is opened and thickened mud is discharged into the
hydrophore below the filter. When a predetermined amount of mud, determined
by level controls in the hydrophore, has been discharged, valve E is closed,
cleaning is continued for a short while, and juice is sent back into the
feed tank. When cleaning is complete, filtration is again started by
opening valve B and closing valve C.
During filtration, sludge is pressed out of the hydrophore by
compressed air and divided into two portions, one portion (about 60Z) being
returned to the prelimer, the rest being filtered on rotary vacuum filters
or on automatic filter presses.
How long are these filtration and washing cycles? If we look at a
6,000 tonne factory with 7,200 m 3 of juice to be filtered, 6 filters will be
needed. On average each filter will handle 50 m 3 of juice per hour.
Typically about 5 m 3 of filtered juice is produced in each filtering cycle
which means that each filter is back-washed 10 times per hour. On average
each filter is filtering for 55 minutes and backwashing for 5 minutes.
Figure 4 shows a mass balance for one filter cycle. It can be noted
3 3
that 6 m of juice is filtered and 3.2 m of filtered juice is used for back
washing. Thickened sludge (1.2 m 3 ) with a solids content of around 20Z is
taken out. The surplus of 2 m 3 is used for further cleaning of the bags
after closing of the bottom valve and is sent back to the feed tank.
An important point to be noted in the operation is that the filters are
never emptied, so filtering is begun again the moment the washing is over.
Washing j u i c e
3.2 m3
Unfiltered juice
6 m3
, j—er
Filtered [ ^ ^ 4.3 m3
ruice vJ
1.2 i 3
Sludge
Hydrophore
Mass balance for one filter cycle.
Rack with graphic juice circuits
· # -
Y Γ*+Υ
5. Control panel for juice filtration.

l±2
The control panel is shown in Figure 5. In addition to the normal

operation the filters can, if the need arises, be further cleaned, emptied
and even acid-washed automatically.
The capacity of the filters at different filtration coefficients is
shown in Figure 6.
Filtered juice
mV24h
30004
Filtering pressure 4-6 m WG
25004
20004
15004
10004
500|
2 34 6 8 10 12 14 16 Fk
Fig. 6. Capacity of filters as a function of the filtration coefficient

(Fk).
Specific capacities between 1,200 and 400 l/m2/hour correspond to 30

and 10 gal/sq. feet/hour. These very high filtration rates are achieved
because the filtration cycles are so short and washing is started before any
appreciable drop in filtration rate has taken place.
In summary, the advantages of the thickening filters are:
High capacity
Automatic operation
Juice back washing
High solids in mud
No vacuum filtration of recirculated mud
No check-filter required.
Figure 7 illustrates the effects of replacing a traditional system of

thickener and filters with DDS-filters, as experienced in our factories when
thickening filters were installed. An example of the effects of the
43
installation of a complete purification system is shown in Table 1, from the

Taber factory of BC Sugar in Canada.
TABLE 1
The effect of installation of a complete purification system on a
factory.
Cost red.
1981 1982 US$
Energy 2,900 kcal/kg S 2 ,200 kcal/kg S 196,000

CaO 2.32 o.b. 1.95 2 o.b. 60,000
Sugar loss 3.47 2 o.b. 3.28 2 o.b. 250,000
Sugar color 34 ICUMSA 26 ICUMSA
Slicing rate 3,500 t/24 h 4,200 t/24 h 152,000*
Evaporator boil
out number 54 4
2nd carb. filter
change 274 38
Various
consumptions -342 81,000
Beets app. 740,000 t app. 480,000 t
Campaign days 190 115
Total cost reduction (480,000 t beet) US$ 739,000
* Increased slicing rate resulted in a shorter campaign and

consequent labor reduction.
To summarize, we at DDS have over the last 25-30 years developed an

integrated system of juice purification, a system which is continuous and
automatically regulated to a very high degree. Large savings have been
achieved and high quality juice is produced from which white sugar is made
from the first and second boilings.
This means that we add about 852 of extra sugar from the second boiling
to the sugar produced in the first boiling without any reprocessing.
Typical sugar colors are: first product 10-18 ICUMSA; and second product
20-29 ICUMSA.
kk
metnc
Beits worked /tfoy
*>tezzzzzzzzzm
Mon minuits pr. Ion Òtti tll\
<»izzzzzzzzr
Omestont
"»""»w *·* ZStemzzËiï
004
Sugar in sludge %*tx '\
C0S9Ì Χ2ΖΣ2ΖΖΖΖΓ
Temperature loss in
the juice purification H>TZZZZZZZZ2~
Coal consumption
Lime salts in thin juice % CaO 00ia\ 1
Colour of sugar awl

v/,y,/'VÀ
,1J
Purity of thick juice I j
"'VZZZZZZZZLZZZZA
I ) Average figures from 5 OOS- factories

2 campaigns before the filters were installed
t//////\ Abroge figures from 5 00% factories

' " 2 campaigns after the fitters were installed
Fig. 7. Comparison of campaigns before and after installation of filters.
Investigations of the purification process are still going on in our

factories, and the understanding of the basic chemistry is still improving.
One example is the finding that a certain residual alkalinity is very
important in the 1st carbonation to get optimum color removal, the reason
being that a positively charged CaC03 precipitate is formed which acts as an
absorbent for colored compounds. As this absorption process will be
reversed at low alkalinities the design of the 1st carbonation vessel is
quite important; no part of the juice should experience lower alkalinity
than the final alkalinity. Therefore good mixing is very important.
45
What about the future?

Although we may not see any basic changes in the purification process,
one thing is already evident. The very rapid development of sophisticated
computerized systems of process-control will further reduce labor costs. In
the near future there will be no permanent staff in the purification
station.
Further optimization of the process to improve juice quality and reduce
consumption of chemicals will come as the fundamental understanding of the
process develops. By measuring and controlling the specific reaction
parameters rather than relying, as we do today, on more or less empirical
figures like alkalinity, total lime content, pH etc., and controlling,
through computers, a large number of variables at the same time, we will
gradually improve on the process, although today we feel that it works quite
well.
Perhaps the most important point in developing our understanding of the
fundamental processes is that we may can make a model of the purification
system, to use to determine the optimal operating parameters depending on
variations in beet quality, whether these variations are the results of
differences in growing conditions or climatic variations.
REFERENCES
1 R. A. McGinnis, Beet Sugar Technology, 3rd ed., Beet Sugar Development
Foundation, Fort Collins, Colorado, U.S.A., 1982.
Chemistry and Processing of Sugarbeet and Sugarcane, edited by M.A. Clarke and M. A. Godshall
^6
Chapter 5
ION EXCHANGE PROCESSES IN BEET SUGAR MANUFACTURE
M. SHORE, N. W. BROUGHTON, D. SARGENT, G.C. JONES AND B. W. BROWN
SUMMARY
Many processes using ion exchange resins have been proposed for use in
the beet sugar industry. After a brief introduction outlining the nature of
ion exchange resins and the trends affecting their use in the industry, this
paper concentrates on processes that have been adopted successfully and that
have stood the test of time. The processes are split into those where the
resins are used as processing aids, e.g. décalcification and decolorization,
and those where the process recovers sugar, e.g. the Quentin process and
déminéralisation. The processes are set in the context of the chemistry of
ion exchange. The ion exchange reactions that take place during the passage
of sugar juice through a resin bed, and their effect on later stages of the
process are discussed with reference to data from British Sugar's factories
and Research Laboratories. Finally the future for ion exchange processes in
the beet sugar industry is assessed.
INTRODUCTION
Some of the earliest studies of ion exchange treatment of sugarbeet

juice were by Rumpler in 1901 (ref. 1). This is comparatively recent when
compared to the application of ion exchange and absorption to water
treatment, whose history goes back to the Ancient Greek Aristotle (ref. 2),
and perhaps even the biblical prophet Moses—who could not only part the
waters but also desalinate them (ref. 3)1
However, ion exchange in sugarbeet processing is really as old as the
beet themselves, since the beet are partially composed of pectic material
which exhibits ion exchange properties (ref. 4). Beet pulp has been
advocated as an ion exchange material for treating radioactive water (ref.
5). More pertinently to the sugar industry, its ion exchange properties are
of significance in the mechanism by which chemical additives are used to
assist in dewatering the residual pulp after the sugar has been extracted
(ref. 4 ) .
Many processes using ion exchange resins have been proposed for use in
the beet sugar industry, and there are already several excellent reviews on
the subject (refs. 6-8).
47
Few of the ion exchange processes proposed on the basis of laboratory

studies or pilot-scale trials have been operated successfully on a factory
scale. To be successful, a process must be both technically feasible and
economically viable. This in turn is influenced by factors which can be
historical, geographical, economic, or political. This paper will
concentrate on processes which have been applied in sugar factories for some
years, and thus can be said to have stood the test of time.
ION EXCHANGE RESINS

Properties
Ion exchange resins consist of beads of a porous insoluble polymer to
which are attached chemical functional groups that can take part in ion
exchange reactions. Resins can be classified in a number of ways:
a) Functional groups. These fall into four basic types: sulphonic
acid, Rxx-SOs-.11"'' (strongly acidic cation exchange resins); carboxylic acid,
Rn-C02H (weakly acidic cation exchange resins); quaternary ammonium, Rn-
NR3"*".OH- (strongly basic anion exchange resins) and amino, R»-NR2 (weakly
basic anion exchange resins).
b) Resin form. This is the exchanging ion associated with the resin's
functional groups, e.g. strongly acidic cation exchange resins are commonly
used in the hydrogen, sodium or calcium form.
c) The polymer matrix to which the functional groups are attached.
This is generally either polystyrene (styrenic resins) or polyacrylate
(acrylic resins). Acrylic resins have greater resistance to organic fouling
(ref. 9) and hence are used extensively for gross decolorization of sugar
juices (refs. 10, 11).
d) The degree of crosslinkage in the matrix. This is usually expressed
as the percentage of erosslinking agent (divinylbenzene) added to the
polymer during manufacture. The amount of crosslinking determines both the
physical strength and the porosity of the resin beads—the higher the
crosslinkage the greater the resin's physical strength and the lower its
porosity, and vice versa.
e) The physical structure of the resin. Resins are either gels, with a
fairly uniform porous structure throughout the bead, or macroporous, with
additional large holes or "macropores" to give a sponge-like structure (ref.
12). Figure 1 shows electron micrographs of typical macroporous (IRA 900C)
i*8
and gel (IRA 401S) resins at 80,000 times magnification. This sponge-like
structure of the macroporous resins can be clearly seen contrasted with the
featureless surface of the gel resin. Macroporous resins can have a high
degree of erosslinking, and hence greater physical strength than gel resins,
while retaining a high porosity due to these macropores.
In the sugar industry macroporous resins are often used to treat syrups
and juices of high total solids concentration. Resin beads shrink
considerably in highly concentrated solutions and swell back to normal size
in dilute solutions, owing to the transfer of water between the resin bead
and the solution by osmosis. Hence, a great physical strain (known as
osmotic shock) is placed on the polymer matrix when, as part of the process,
the concentration of the solution passing through the resin bed changes
rapidly. This occurs, for example, when the concentrated juice is first
applied to the resin after regeneraton (sweetening-on), and also when the
juice is washed off the column at the end of the cycle (sweetening-off).
Gel resin beads tend to shatter under this kind of stress and hence are
generally only used for treating low concentration juices (for high purity
syrups, with low color and ash and long cycles. Ed.).
Swelling and shrinkage of the resin can also occur when the ionic form
of the resin is changed.
Ion exchange reactions

Ion exchange reactions are reversible and the equilibrium reactions
between a solution of a salt and a cationic (Resin-.H*) or anionic
(Re s in"*". OH - ) exchange resin can be expressed by chemical equations, of which
the following are typical for monovalent ions (refs. 6, 12):
Cation exchange resin:
Resin-.H- + M* X- - Resin-.M- + H* X" (1)
Anion exchange resin:
Resin*.OH- + M* X - = Resin*.X- + M* 0H~ (2)
For a strongly acidic cation exchange resin the equilibrium in equation
1 is displaced to the right, whereas with a weakly acidic resin the left to
right reaction is suppressed by the free acid H*X". Weakly acidic resins
have a high affinity for hydrogen ions and hydrogen form resin will only ion
exchange readily under alkaline conditions (equation 3).
49
Weakly acidic cation exchange resin:

Rxx-C02H + M- X- + OH" - RA-C02-.M- + H20 + X" (3)
Gel resin: IRA 401S ( x 60,000)
'fili *#g|P^: -lÀfe^
Macroporous resin: IRA 900C ( x 60,000)
Fig. 1. Structure of gel and macroporous resins.

50
Similarly, anionic resins are classed as strongly or weakly basic

depending on whether the equilibrium for equation (2) is displaced to the
right or left respectively. Hence weakly basic anion exchange resins in
hydroxide form will only exchange ions readily under acid conditions
(equation 4).
Weakly basic anion exchange resin;
Rrx-NR2 + M + X- + H+ lU-NIUH-.X- + M- (4)
The equilibrium of these ion exchange reactions is also affected by the
relative concentrations of the ions competing for the exchange sites on the
resin and by the affinity of the resin for each ion. If, for example, the
resin in equation (1) has similar affinities for H- and M- then the position
of the equilibrium is decided by the relative concentrations of the two
cations. The cation present in greatest concentration will be favoured to
occupy the resin's exchange sites.
The affinity of an ion exchange resin for a particular ion depends on
the size and charge of the ion. In dilute solutions, the resin has a much
greater affinity for small highly charged ions than for large monovalent
ions. The relative affinities of a number of cations with a strongly acidic
cation exchange resin are shown below (ref. 13):
Fe 3 - > Al 3 - > Ca 2 - > Mg*- > ΝΗ*+ = K- > Na- > H-
In solutions of high ionic concentration the difference between the
affinities is reduced and in extreme cases reversed, with the absorption of
monovalent ions being favoured over multivalent ions (ref. 13). The
strongly basic anion exchange resin/anion affinities have been less studied
but for some of the more common ones known they are (ref. 14):
SO**" > citrate > N0 3 ~ > PO*3- > I~ > Cl"
The situation for weakly acidic or basic resins is rather different in
that their chemical natures give them very strong affinities for hydrogen
ions (weakly acidic cation exchange resins) or hydroxide ions (weakly basic
anion exchange resins), as discussed earlier. The difference in affinities
for divalent and monovalent cations is more marked for the weakly acidic
resins than strongly acidic resins (ref. 12), so hydrogen form weakly acidic
resin will exchange with divalent cations in solution but not readily with
monovalent cations.
51
Kinetics
Another major factor in the performance of ion exchange resins is the
kinetics of ion exchange reactions. Three diffusion processes generally
control the rate of ion exchange (ref. 12). These are diffusion through the
bulk of the liquid, or to and from the ion exchange sites through a static
film of liquid around each resin bead, or through the pores of the resin.
The static film of water around each resin bead is held there by friction.
Ions diffuse through this layer to the resin at a rate that is directly
proportional to the ionic concentration in the bulk solution, and quite
independent of the resin's nature. Diffusion through the liquid film is
normally the rate determining step in water treatment (ref. 12), but this is
not necessarily the case in the sugar industry (when treating high
concentration juices or removing high molecular weight species such as
colour bodies, for example).
The working capacity of an ion exchange resin is affected by the flow
rate of liquid through the resin bed. If the flow rate is too fast to allow
the ions in solution to diffuse to and from the resin's exchange sites then
the working capacity of the resin will be reduced (ref. 12).
Significant pressure drops across resin beds are encountered at high
flow rates, and also with concentrated feed liquids. This problem is
countered by using large diameter resin beads, together with frequent
backwashing to classify the resin beds into layers of uniform particle size
distribution and to remove any resin fines. Large beads, however, have a
relatively low surface to volume ratio, and hence a decreased rate of
diffusion and ion exchange.
In the mid-1960's British Sugar started to install ion exchange
décalcification plants which operated at what were then considered quite
high flow rates (ref. 15). Therefore the ion exchange kinetics were of
particular importance in obtaining satisfactory plant performance. The
performance of resins in service was regularly monitored using tests
developed by British Sugar Research Laboratories for measuring a resin's
total ion exchange capacity and, independently, the rate of exchange (ref.
16). Subsequently these methods were adopted by the International
Commission for Uniform Methods of Sugar Analysis (ICUMSA) (ref. 17).
52
Other features
The high juice temperatures (up to 90°C) encountered in sugar factories
and refineries mean that the thermal stability of ion exchange resins is
generally of more importance for the sugar industry than in other industries
utilizing ion exchange processes. Both strongly and weakly acidic cation
exchange resins have high thermal stabilities, their maximum recommended
working temperatures being greater than 100°C (ref. 13). The quaternary
ammonium and amino groups of anion exchange resins are more susceptible to
thermal degradation than the sulphonic or carboxylic acid moieties of cation
exchange resins. The hydroxide form of a strongly basic anion exchange
resin is thermodynamically unstable even at ambient temperature, but
decomposes only very slowly at this temperature (ref. 18).
The rate of degradation increases steadily with increasing temperature,
so that the normal recommended maximum working temperatures of these
hydroxide form resins are from 40-60°C (ref. 13). Chloride form strongly
basic anion exchangers, and weakly basic anion resins in general, have
greater thermal stability, their recommended maximum working temperatures
being 75 - 80°C (ref. 13).
Another property of ion exchange resins is also of particular relevance
in the sugar industry. Their densities are low enough for them to float in
high concentration syrups.
In the U.S.A., resins used in a food industry must conform to the
appropriate Food & Drug Administration Regulations (ref. 19), and similar
regulations have been drafted for the E.E.C, (ref. 20).
Trends
Until now, the main developments in ion exchange resin manufacture have
been in response to applications in water treatment, the largest use of
these materials. However, the sugar industry has always been eager to apply
new developments in ion exchange technology: as early as 1901 (ref. 1),
Rumpler reported attempts to treat beet sugar juices with natural alumino-
silicate materials. Subsequently the development of the early organic
resins in the 1930s, based on sulphonated coal (ref. 21) or phenol
formaldehyde (ref. 22), and of polystyrene (ref. 23) resins in the 1940s,
led to many attempts to incorporate resin processes in beet sugar factories.
In turn macroporous resins (ref. 12), and more recently acrylic (ref. 24)
53
resins, were developed, largely to avoid organic fouling in water treatment,

and have been applied enthusiastically to sugar treatment processes where
their particular properties of increased physical strength or resistance to
fouling are expected to be of benefit.
Doubtless this trend will continue and any further types of resins
developed in the future, for whatever original purpose, will be examined for
possible uses in sugar processing. Recent research into the use of
inorganic ion exchangers is an example of this process (refs. 25, 26).
Often, a major factor affecting the viability of an ion exchange
process is the geographic location of the sugar factory. Climate influences
the length of the factory processing season, and problems such as evaporator
scaling that are acceptable in a factory with a short season may not be so
at another that must operate for twice as long without stopping.
In Southern Europe, beet juice purities are generally lower than
further north. For example, a British beet sugar factory might operate at a
typical thick juice purity of say 92Z sucrose on total solids, while for an
Italian factory the typical purity might be 88Z (ref. 27) or for a Sardinian
one as low as 83Z (ref. 28). As a result, in Southern Europe there is a
need for methods of increasing juice purity, and a great interest in ion
exchange techniques such as partial demineralization which achieve this.
The relative prices of white sugar and molasses will have a marked
effect on the viability of any sugar recovery process, since an increase in
recovered sugar invariably results in decreased molasses production. To a
lesser extent, the same is true of any process which alters the non-sugars
composition of the factory juice, and hence affects its overall molasses-
forming (or "melassigenic") properties. In the E.E.C, there is a three-tier
sugar price system linked to production quotas. Each country has allocated
to it 'Α' and 'Β' annual production quotas, and any further sugar produced
is exported at the lower world price. Thus a company making only 'Α' quota
sugar could obtain a higher price for any additional sugar recovered by ion
exchange than one already producing its full 'Α' and 'Β' quotas, where the
extra sugar recovered would be sold at world price.
The cost of régénérant needed for a process may also vary
geographically. For example, much of the magnesium salts in Europe are
found in the Stassfurt deposits in central Germany (ref. 29). Distance from
these deposits, and hence transport costs, has a marked effect on the
5^
profitability of the Quentin process which uses magnesium form resin (ref.
30).
Effluents and their effect on the environment are a major consideration
these days, and can affect the feasibility of operating ion exchange
processes at certain sites, particularly those far inland. This has led to
an interest in modifications to basic processes such as décalcification and
demineralization to eliminate or reduce the amount of effluent produced or
in some cases to recover it as a saleable by-product.
Resin fouling
Fouling of ion exchange resins can be defined as the presence of
substances, on or inside the resin beds, that reduces the performance of the
resin. Fouling can affect either the total ion exchange capacity or the
rate of ion exchange, or both.
Resin fouling is a major factor in determining the lifetime of an ion
exchange resin in a beet factory process, and can occur in several ways.
Appreciable amounts of suspended solids and colloids may be present in
the juice. These can block the resin bed causing a high pressure drop and
channelling of liquid through the bed. Also, they can coat the resin beads,
hindering ion diffusion into and out of the beads and reducing the rate of
exchange. These problems can be minimized by efficient juice filtration and
regular backwashing of the resin columns (ref. 15).
Similar problems occur if bacteria or algae are permitted to grow in
the column, forming slimes that can block the beds and contaminate the
product liquid (ref. 31). This problem does not occur in processes that run
at high temperatures, and where this is not possible, frequent sterilization
of the resin and the column interior walls with biocides may be required.
Fouling by precipitation of insoluble inorganic compounds inside the
resin beads can occur occasionally in juice treatment, as it does in water
treatment (ref. 32). Typical examples include removal of iron from
solutions by cation exchange resin and subsequent resin fouling by
precipitated hydrated ferric oxide; calcium sulphate fouling when
regenerating a cation exchange resin bearing calcium ions with sulphuric
acid or magnesium sulphate; magnesium hydroxide fouling when passing
alkaline solutions through magnesium form resin. In most cases inorganic
fouling is readily removed by treatment with dilute acid.
55
The most important form of fouling is irreversible absorption or

"poisoning." Some organic compounds are irreversibly absorbed onto the
exchange sites in ion exchange resins and are not removed by the normal
regeneration procedure (ref. 9). In water treatment the compounds
responsible are referred to as humic acids (ref. 33). In sugar juice
treatment, organic nitrogenous compounds such as proteins and colour bodies
are responsible for fouling of both anion (refs. 34, 35) and cation (ref.
16) exchange resins. In an investigation into poor exchange performance in
British Sugar factories in 1972, which was hypothesized to be due to fouling
of the cation exchange resins, resins with nitrogen contents up to the
equivalent of 15Z protein on resin dry matter were found (ref. 16).
Effective in situ chemical cleaning techniques were developed using sodium
hypochlorite solution (ref. 16). For example, a resin whose working
capacity had been reduced to 22Z of that of new resin could be restored to
an 82Z working capacity.
Non-removable colour bodies cause resins to go black in service within
a few years, but do not necessarily have a significant effect on the ion
exchange properties (ref. 36).
One interesting phenomenon we have occasionally observed (ref. 37), but
not found reported elsewhere, is that a strongly acidic cation exchanger may
start to behave as if it is only weakly acidic, in that contact with acid
solution markedly reduces its subsequent ability to ion exchange in neutral
solutions. It is possible that this is due to absorption onto the resin
functional groups of proteins or other organic compounds bearing free
carboxylic acid groups.
Where possible, ion exchange processes in the sugar industry are
carried out at elevated temperatures to avoid bacterial activity and reduce
syrup viscosities. However, resin functional groups, particularly for
strongly basic anion exchange resins, are chemically degraded by operation
at too high a temperature (ref. 18). This is sometimes detectable by a
fishy amine odour in the treated juice. Like fouling, thermal degradation
of resins reduces ion exchange capacity and in some cases may reduce resin
life.
56
RESINS AS PROCESSING AIDS

Décalcification
Conventional process. The removal of calcium from water, to prevent
scale formation and improve the efficacy of soaps, was the oldest commercial
application of ion exchange, and it is many years since a similar technique
was first applied to the treatment of sugar juices (ref. 38). The
prevention of scale formation in the evaporators is of great importance for
efficient factory operation. Décalcification (softening) of second
carbonatation juice by ion exchange is a common way of preventing scale
formation, which has been used in many beet sugar factories throughout the
world (refs. 39, 40), including until recently most of the factories within
British Sugar (ref. 15). Décalcification has been particularly important
for us because we have a long processing season or campaign (138 days for
one factory in 1986/87).
The first plant at a British Sugar factory was installed in 1965, and
plants were subsequently installed at a further twelve of our factories
(ref. 15).
The British Sugar plants were designed and built in-house, but they
generally resemble plants at other sugar companies. Our plants have two to
four columns (depending on factory size), each up to nine feet (2.8 m) in
diameter and holding up to 320 cubic feet (9 m Ä ) of ion exchange resin. The
plants operate continuously, with the columns being run in parallel. One
column is on regeneration, while others are treating juice. The resin used
is a strongly acidic cation exchanger in sodium form. Gel resins are
normally used--in our experience the additional cost of the alternative
macroporous resins is not justified for this process. Regeneration is with
sodium chloride solution, and there are additional stages in the operating
cycle for washing off juice régénérant. Even small amounts of precipitated
calcium salts from carbonatation escaping filtration are removed by the
resin bed which acts as a sand filter. Hence a backwashing stage is also
included in the cycle, to clean and loosen the resin bed. The régénérant
effluent has a high inorganic content but a lower B.O.D. and organic content
than some of the ion exchange effluents, e.g. in decolorization. Disposal
of this effluent can be a problem in some circumstances. The
décalcification plants removed 19kg of CaO per cubic metre of resin,
57
reducing the concentration of calcium ions to 6 - 8 ppm (ref. 15). A simple

schematic of the décalcification process is shown in Figure 2.
Exhaustion Regeneration
Second Carb. Juice 10Z NaCl
90°C 15-20 BV/hr 2-3 BV/hr

II
Strong acid cation resin||
Na"*" form
JJ
Nl/
Softened second carb. juice
to evaporators Effluent
Fig. 2. Conventional décalcification.
In 1972 British Sugar installed a commercially available continuous

moving bed ion exchange plant for décalcification at Ely factory (ref. 15).
Serious difficulties with resin transport and the balance of liquid flows
were experienced with it, due partly to the presence of insoluble solids in
the juice, and it was eventually replaced by a conventional plant.
Chemistry of decalcifcation. Décalcification (softening) of second
carbonatation juice is accomplished using a strongly acidic cation exchange
resin in the sodium form. It is often considered simply as an exchange of
calcium ions in the juice for sodium ions on the resin, according to the
following equilibrium:
2 Resin- . Na- + Ca 2 - - (Resin-)2 . Ca*- + 2Na-
In practice the resin does not remove all the calcium ions from the
juice, for two reasons:
58
(i) Some of the calcium is present as complexes which, being

negatively charged, do not participate in exchange reactions with the resin.
Calcium complexes with certain carboxylic acid anions; in British Sugar's
experience only citrate is of any quantitative significance in this respect
in second carbonatation juice (ref. 41).
(ii) Other cation exchange reactions occur, besides that of calcium for
sodium, which compete for sites on the resin or, indeed, displace calcium
from the resin (ref. 42). Potassium, sodium and ammonium are the principal
cations in the second carbonatation juice and these can be exchanged for the
sodium ions originally on the resin:
Resin- . Na+ + M + - Resin- . M + + Na +
Moreover, the high concentrations of these three cations in juice
relative to calcium can drive the main exchange reaction, by which calcium
is removed from the juice, in the reverse direction, thus:
(Resin)* . Ca 2 - + 2M"*- » 2 Resin- . M- + Ca 2 +
even though the resin has a higher affinity for calcium than for the
monovalent ions.
The various exchange reactions occurring in décalcification could have
a significant effect on the overall non-sugars composition of juice, and
therefore on the formation of molasses at the expense of white sugar
production. Several workers have investigated the relative amounts of
sodium/calcium and other cation exchange reactions taking place during
décalcification (refs. 43-45).
In a British Sugar investigation, reported in 1980 (ref. 42), it was
found that when a décalcification plant was operating under representative
conditions, the net effect during the exhaustion part of the cycle was the
removal of both calcium and potassium ions from the juice and their
replacement by sodium ions. Small quantities of ammonium ions were also
removed (see Figure 3 and Table 1). Most of the removal of potassium
occurred during the first two hours or so of each exhaustion. In the
remaining 17 hours of the exhaustion sequence a small proportion of the
potassium which had been removed was exchanged back into the juice for
calcium ions.
The observed exchanges were explicable in terms of the relative
concentrations of the various cations in juice and the relative strengths
with which they bind to the negatively charged sites in the resin beads.
59
Other décalcification processes. Several modifications and

alternatives to the conventional décalcification process described above
have been developed.
A) The Gryllus Process
The Gryllus process, developed by Mrs. Eva Gryllus in Hungary, is one
of the best known (ref. 46). It utilizes the relative affinities of cation
exchange resins for monovalent and divalent ions at different concentrations
(see above). With second carbonatation juice the décalcification resin
preferentially takes up calcium ions from juice rather than sodium or
potassium ions. This situation is reversed when softened thick juice is
passed through the resin bed, so that the calcium ions on the resin tend to
exchange back into the juice to be replaced by sodium and potassium ions.
Thus the calcium ions effectively by-pass the evaporators (Figure 4).
Output concentration (meq/100S)
35 J
30 J
0 1 2 3 4 5 15 16 17 18 19 20 mean
input
Time (hours)
Fig. 3. Typical exhaustion cycle of a décalcification column. Output
cation concentration is meq/100g apparent sucrose.
60
TABLE 1
Cation absorption during exhaustion.
Amount absorbed /keq/ m 3 resin

Cation First 10Z of cycle Complete cycle
Potassium .69 .55

Ammonium .06 .09
Calcium .03 .69
Second carb. Juice Soft thick juice
JL or Green Syrup
— ii
Macroporous strong acid ||
cation resin jj
Na-^/K-*- form jj
II
II
Softened second carb. juice Hard thick juice or green syrup
to evaporators to pans
Fig. 4. Simple schematic of Gryllus décalcification process.
The process eliminates the cost of salt régénérant, and the need to
treat spent régénérant effluent, compared to conventional décalcification.
It is also claimed to produce less molasses and more white sugar than
conventional décalcification. However, some advocates (refs. 47, 48) of
this process have tended to compare it with a theoretical décalcification
process in which only calcium/sodium exchange occurs, ignoring other
exchange reactions, and they have thereby distorted the relative merits of
the two processes. When evaluating the process, it is necessary to take
into account all the ion exchange reactions taking place, described above
61
before estimating the relative effects on juice non-sugar composition and

hence on molasses formation.
This process is in use in several countries including Hungary, Chile
and Lebanon (ref. 49).
There have been some suggestions that putting the calcium ions back
into the thick juice could cause scale formation on the vacuum pans and
hence in the white sugar produced (ref. 50). To avoid this, a later version
of the process regenerates the column with a factory syrup from a stage
after the crystallization of the white sugar (ref. 51). This variation of
the Gryllus process is used in France, Morocco and Spain.
It is necessary to use a macroporous resin for the Gryllus process
because the changes in juice concentration impose a severe physical strain
on the resin beads through osmotic shock, which gel resins are unable to
withstand.
B) N.R.S. Process
A more recent alternative to conventional décalcification is the N.R.S.
process ("New Regeneration System") (refs. 52, 53) developed by the Imacti
Ion Exchange Company, now a subsidiary of Rohm and Haas (Figure 5). This
has been installed in several factories in Western Europe in the last few
years, sometimes by conversion of conventional décalcification plants. Once
again the new process avoids the effluent loadings created by conventional
brine regeneration. For the N.R.S. process, instead of the removed calcium
being returned to juice after evaporation, it is recycled to carbonatation
for precipitation.
If a décalcification column were to be regenerated with a solution of
sodium hydroxide instead of sodium chloride, the calcium removed from the
resin would precipitate as calcium hydroxide. However, in the presence of
sucrose and under certain pH and temperature conditions this displaced
calcium remains in solution as calcium saccharate.
62
Second carb. juice Soft second carb. juice

|| + NaOH
II <50
s!/ °c
Strong acid cation

resin Na"*" form
II II
NI/
Soft second carb. juice Juice and calcium
to evaporators saccharate to carbonatation
Fig. 5. N.R.S. décalcification.
Therefore, in the N.R.S. process the resin is regenerated with

decalcified thin juice to which sodium hydroxide has been added. The
calcium displaced from the resin reacts with sucrose in the juice to give
soluble calcium saccharate. The temperature of the juice must not be
greater than 50°C, otherwise the calcium saccharate will not form. The
spent régénérant, containing sodium hydroxide and calcium saccharate, is
recycled to the carbonatation purification stage, where carbon dioxide
decomposes the calcium saccharate complex forming sucrose and calcium
carbonate precipitate. Hence, the sucrose lost in the régénérant is
recovered at carbonatation.
Sodium hydroxide is more expensive than sodium chloride, but it is
reported that less of the former is required (ref. 53). As with
conventional décalcification the sodium ions added into the juice give rise
to increased molasses formation at the expense of white sugar production.
The relative merits of these various ion exchange processes for
décalcification depend on several factors, including the cost of régénérant,
effluent treatment, any juice dilution occurring during the process, and the
effect on non-sugars composition and hence on molasses formation. What is
cost-effective in one country may not be the best option in another.
63
C) Weakly Acidic Cation Resin Process

Recently a process using hydrogen form weakly acidic cation exchange
resin to soften thin juice has been proposed (refs. 54, 55). The hot juice
is passed through the weakly acidic cation exchange resin (at high flow
rates to minimize inversion) and immediately made alkaline by addition of
active magnesium oxide. Hence the calcium ions and also some sodium and
potassium ions, are replaced by the much less melassigenic magnesium ions.
The high affinity of weakly acidic resins for calcium ions and their high
operating capacities allow the size of the plant to be appreciably smaller
than that for conventional décalcification. However, problems have been
experienced in scaling up the process, largely caused by the high flow rates
necessary for this process (ref. 55).
Decolorization
Sugar colour. Juice decolorization is not required under the range of
conditions normally occurring in beet sugar factories. In this respect the
beet sugar industry differs from the cane sugar industry, where
decolorization by ion exchange resins or carbonaceous materials has been a
standard procedure. Indeed some cane sugar refiners will refine beet raw
sugars in order to allow a maintenance period for their decolorization
plant.
Beet juice colour bodies fall into three main categories (refs. 56,
57):
i) melanoidins produced by the Maillard reaction of sugars with amino
acids;
ii) caramels formed by the degradation of sugar;
iii) melanins formed by enzymic oxidation of phenolic compounds.
The types of colour bodies in juice and their molecular weights change
through the process from raw juice to the final crystalline white sugar
(refs. 58, 59).
Studies by British Sugar have indicated that, for the normal ranges of
crystal sizes, white sugar colour consists of two components which may be
differentiated on the basis of molecular weight. These are internal colour
containing material of predominantly high molecular weight incorporated
within the crystal, and external colour, of lower average molecular weight,
6^
on the surface of the crystal resulting from a residual layer of mother

liquor (refs. 58, 59).
Removal of colour bodies from juice will improve white sugar quality
(by a reduction in white sugar colour) or have energy saving benefits from a
reduction in the amount of washing in the centrifuges.
Many of the colour bodies in beet juices are weakly anionic or
amphoteric (ref. 57), and hence can be removed by strongly basic anion
exchange resins. It has also been suggested that some colour bodies can be
attracted to the aromatic ring of the resin matrix (ref. 60). Thus the
mechanism of decolorizing is quite complex and much of the available
information is empirical.
Anion exchange resins in the chloride form are generally used for
decolorization in the sugar industry. The resin is regenerated with sodium
chloride solution, often adjusted to a slightly alkaline pH. Both styrenic
and, more recently, acrylic resins have been used. It has been reported
that acrylic resins are the more suitable for treating highly coloured
liquors, because of their greater resistance to organic fouling, and hence
their longer working life (refs. 10, 11). Styrenic resins are better for
treating lower coloured liquors (ref. 61). However, there has been a recent
report (ref. 62) that the longer working life does not always make the more
expensive acrylic resins more cost-effective overall.
Macroporous resins are often used, rather than gel resins, particularly
when decolorizing juices and syrups of high total solids, because of their
physical strength and resistance to osmotic shock.
British Sugar Studies

Laboratory studies. British Sugar has studied resin decolorization of
beet juices both in the laboratory and on a factory scale. Recent
laboratory work concentrated on the relative merits of treating different
juices and the accompanying chemical changes in juice non-sugar composition.
A styrene macroporous resin was used in most of the experiments.
The decolorization performance of the resin was strongly affected by
the total solids concentration (°Brix) of the juices (Figure 6). When 40
bed volumes of 66° Brix thick juice were treated, the bulked product juice
had a colour 65Z that of the feed juice. When the same amount of solids was
treated as a 15° Brix second carbonatation juice or diluted thick juice
65
under commensurate conditions, decolorization was considerably improved, the

product juice having a colour only 352 that of the feed juice. We
hypothesize that the much higher viscosity of thick juice probably hinders
the diffusion of the colour bodies into the resin pores, thus impairing the
performance of the resin.
As well as removing colour the resin takes part in ion exchange
reactions with juice anions. The concentrations of readily measurable juice
anions were monitored during experiments on thick juice decolorizing. The
overall changes in some of these anions for the total amount of juice
treated are shown in Table 2 and Figure 7.
Diluted Thick Juice"

(15· Bx)
— i —
100
Juice solids treated (g/cn»3 r e si n )
Figure 6. Decolorization effect of brix.

66
Anion
Concentration
.. Original
Concentration
/
/
/ / X
/
/
Bed Volumes Treated
Fig. 7. Decolorization changes in anion concentrations.
TABLE 2
Overall changes in juice anion concentrations.
Anion Change vs Untreated Juice
Chloride +81Z
Sulphate 0
Malate 0
Sulphite -16Z
Nitrate -53Z
67
Most of the ion exchange was completed during the first 10 bed volumes
of juice treated. During this period anions (including sulphate, nitrate,
malate and sulphite) were taken up by the resin on its ion exchange sites,
and chloride ions were displaced from the resin into the juice. In the
remainder of the juice run, some of the anions on the resin (e.g. sulphate,
malate) were displaced, presumably by other juice anions with a greater
affinity for the resin, so that there was no overall uptake of the sulphate
or malate ions on the resin when considering the cycle as a whole.
Nitrate is one of the smaller juice anions and it may therefore be able
to penetrate further into the resin beads. This could explain why it is
taken up by the resin for most of the cycle length, so that its overall
concentration is halved.
The sulphite concentration never recovers to its initial level, and
this can possibly be ascribed to chemical reactions such as oxidation or
combination with colour bodies during passage through the resin bed. This
loss of sulphite may require an increase in the amount of sulphur dioxide
used in the sulphitation stage of the process, in order to prevent colour
formation in the sugar end of the factory process.
The appreciable rise in chloride concentration increases the overall
melassigenicity of the juice non-sugars and may adversely affect the target
molasses purity. The increase in melassigenicity can be seen by use of the
Polish test (refs. 63, 64) in comparing samples of molasses from two British
Sugar factories before and after treatment with a decolorizing resin (see
Table 3). The impact of the replacement of less melassigenic anions by
chloride ions (thought to be one of the most powerful melassigens (ref. 65))
can be seen in the changes in the main parameters obtained from the Polish
Test. These are the molasses formation constant (a), the free water
parameter (b) and the target molasses purity (T.M.P.). The lower the T.M.P.
the greater the amount of sucrose that can be crystallized out in the
factory.
68
TABLE 3
Decolorization and Melassigenicity.
Target molasses purity
Molasses (a) (b) Factory NS/W =
Cantlev
Untreated .23 .87 54.7 37.2
Decolorized .40 .33 56.0 50.2
Bury.
Untreated .17 .93 50.3 30.5

Decolorized .26 .71 53.1 40.1
The increase in (a) after decolorization is a strong indication of

higher melassigenicity, as is the increase in T.M.P. When calculated under
the conditions prevailing at each of the factories the increase in T.M.P. is
small but significant. If the T.M.P. calculations are extrapolated to
infite non-sugar to water ratio where the T.M.P. becomes dependent on the
non-sugars (the influence of water being minimal), the data clearly show the
effect of the ion exchange reactions occurring during the decolorization
process on juice melassigenicity. The increase in melassigenicity means
that some sucrose is lost to molasses as a result of the decolorization
process.
A small increase in juice purity might be anticipated purely on the
basis of the chloride ions having a lower molecular weight than the ions
they replace in the juice. For greater precision, the change in juice
purity of the treated juice was measured indirectly by gravimetric analysis
of the spent régénérant. An increase of about 0.5 purity units was measured
over the first 5Z of the decolorization cycle for second carbonatation
juice. However, over the full cycle length the purity increase was less
than 0.1 units (within experimental error). Similar results were obtained
with thick juice. It seems likely that any increase in sugar recovery
because of this very small purity increase would be nullified by the
increase in non-sugar melassigenicity.
The laboratory results can be summarized as follows:
69
1) juice concentration has a considerable effect on the efficiency of

resin decolorization. Dilute juices (e.g. second carbonatation juice) are
decolorized with much greater efficiency than concentrated juices (e.g.
thick juice).
2) the ion exchange reactions taking place during decolorization result
in a large increase in juice chloride concentration, together with a
significant decrease in sulphite concentration.
3) the decolorization process increases the juice melassigenicity.
4) there is no significant purity increase during decolorization.
This work will be describe4 in greater detail in a paper to be
published in 1988.
Factory studies. British Sugar has operated factory-scale decolorizing
plants for treatment of thick juice, the plants being conversions of
décalcification plants (described above). The first conversion (York
factory, 1976) was temporary--the plant being used to decolorize a stored
juice during an off-season refining run. The plant reverted to
décalcification mode for the next beet decolorization from décalcification
and vice versa was found to be very practical and was repeated in subsequent
years.
Conversion of a décalcification plant to decolorization involves a
change of resin, regulation of the temperatures of syrup, régénérant and
water to below 75°C (to avoid resin damage), and the handling of sweet
waters from the sweeten-on and sweeten-off stages. The sodium chloride
regeneration system is unchanged. A simple schematic of the decolorization
process is shown in Figure 8. A second conversion (Bury factory, 1985), has
been used to obtain factory development data on decolorization of beet
juices and it has confirmed the results of our laboratory studies.
The degree of decolorization in the factory was very similar to that
obtained in the laboratory. There was no detectable change in the purity of
the treated juice, and the concentration profiles of the juice anions showed
no significant differences from the laboratory results. There was a very
good correlation between relatively small changes in the juice Brix, and the
decolorizing performance of the resin. This can be seen in Figure 9 where
the decolorization curves obtained in the factory and the laboratory are
compared. The deviations observed in the factory curve correlate exactly
with small variations in the feed juice concentration, the performance of
70
the resin deteriorating as the concentration increased and improving as it

decreased.
Thick Juice 10Z NaCl

75°C || 3 B.V./hr. 75°C || 3 B.V./hr.
II
V
i
I
II
II
|| Macroporous strong
I base anion resin
|| Cl- form
II
« i
II
II
II
Mk Effluent
Decolorized thick juice
to pans
Fig. 8. Decolorization at British Sugar, Bury Factory.
Laboratory sugar boilings on factory treated juice showed that white sugar
colour decreased in proportion to the degree of decolorization of the thick
juice. Benefits from juice decolorization are likely to be in terms of an
improvement in white sugar quality rather than any clear cut cost savings.
British Sugar Research Laboratories are currently looking further at targeted and
specialized decolorization as one potential means of improving white sugar
quality.
71
Bed Volumes Treated
Fig. 9. Comparison of factory and laboratory decolorization.
Other Plants. Very little decolorization of beet sugar thick juice has been
reported (refs. 66, 67). Almost all beet juice decolorization is performed on
dissolved raw or remelt sugars and so is similar to decolorization in cane sugar
refineries. Examples in beet sugar refining include the refineries at Arlov
(SSA, Sweden) (ref. 11), Eisdorf (Pfeifer and Langen, Germany) (ref. 68),
Tirlemont (Raffinerie Tirlemontoise, Belgium) (ref. 69) and Porkkala (Finn Sugar,
Finland) (ref. 70). Raw/remelt sugar decolorization is also carried out in Italy
(ref. 71), U.S.S.R. (ref. 72), Poland and Iran (ref. 73).
Beet juices are generally decolorized at 65 - 67° Brix and 75 - 80°C. While
this may be the most convenient way of treating raw and remelt sugars and high
concentration factory syrups, British Sugar Research Laboratories' findings
suggest that these are not the optimum conditions for decolorization. Much
better decolorization is achieved with lower concentration juices, e.g. 15° Brix
second carbonatation juice.
There is a particularly interesting process at Tirlemont (refs. 69, 74),
where a new decolorizing plant was installed in 1982 to facilitate production of
a special, high quality, white sugar (less than 4 E.E.C, points). The plant has
72
three anion exchange columns operating in parallel, with two on stream and one
regenerating at any given time. The columns are of the double deck type,
containing two 6 m 3 resin beds one above the other separated by a screen, to give
a two-stage decolorization process. The plant uses the floating bed technique
developed by Bayer with the syrup passed through in an upward flow (ref. 74).
The resin floats in this syrup, and forms packed beds at the tops of the columns.
Régénérant and water are passed through in a downward flow, and the resin sinks
to give packed beds on the baseplates. This countercurrent system gives good
efficient regeneration.
The plant uses a styrenic macroporous decolorizing resin. The resin in the
upper (polishing) bed in each column is expected to last longer in service than
resin in the lower bed, because the upper bed treats syrup that has already been
partially decolorized by the lower bed, and it is also the first bed that the
régénérant passes through. When the resin from the lower bed is eventually
discarded, it will be replaced by the resin from the upper bed, which in turn
will be replaced by new resin. In this way the second (polishing) stage in the
decolorizing treatment, in the upper bed, will always be done by resin in
relatively good condition.
Most other refineries use two columns in series for juice decolorization
(refs. 68, 70). The first column may contain acrylic resin for gross
decolorization followed by a styrenic resin for polishing. However, a recent
factory trial at the Arlov refinery in Sweden, where the performances of styrenic
and acrylic decolorizing resins were directly compared (ref. 62), indicated that
a purely styrenic two stage decolorization system may perform better than the
acrylic/styrene system. It was reported that although the styrene resin
decolorization capacity deteriorated more quickly than that of the acrylic resin,
the former removed approximately 30Z more colour from the juice over the greater
part of the normal working lives of these resins.
Powdered Resins
All the processes described so far use resin in columns, in the conventional
way. However, there is an alternative means of decolorizing. The resin can be
added into the liquor to be treated, then filtered off and discarded, in the same
way that a powdered carbon can be used. Because the resin is used only once, the
technique is not cost-effective using normal bead-form resins. However, if the
resin is in the form of a powder then the exchange process is very efficient
73
compared to a bead resin, since all the ion exchange sites are close to the
surface and therefore readily accessible. In tests on liquid sugars equivalent
decolorization was found to require about one hundred times more bead form resin
(by weight) than powdered resins (ref. 75).
Ecosorb precoat formulations of powdered resins produced by the Graver Co.
in the U.S.A. have been used in several cane and beet refineries in North
America, for the polishing stage of liquid sugar decolorizing (ref. 76).
Recently a powdered cation resin in the sodium form has also been
introduced, intended for juice décalcification (ref. 76).
British Sugar Research Laboratories have evaluated the use of powdered anion
exchange resins for decolorization of liquid sugar, in order to produce a special
high quality liquid sugar. Following development work, factory trials showed
that addition of a particular powdered resin to a 15 tonne mix of liquid sugar
(typical colour 20 ICUMSA units) reduced its colour by about 4 units per kilogram
of resin added (ref. 75).
The use of powdered resin avoids the capital cost of an ion exchange plant
that would only be used occasionally.
Miscellaneous Processes
Inversion. In acidic solutions, the disaccharide sucrose is hydrolyzed to a
virtually equimolar mixture of glucose and fructose. This process is commonly
known as "inversion" and the glucose/fructose mixture as "invert syrup."
There is a significant market for invert syrups made from sucrose (ref. 77).
Production methods include inversion by acid addition (ref. 78), by resin (ref.
77), and by invertase enzymes (ref. 79). Reported advantages of using cation
exchange resins for inversion are the ease of separation of the resin catalyst
from the sucrose (ref. 80), low colour formation (ref. 77), and no impurities
introduced into the product (ref. 77). The disadvantages include relatively high
capital costs, and the difficulty of obtaining complete inversion with resins.
The ratio of invert sugars to sucrose in the product can be varied by changing
the flow rate, or the temperature of the resin columns. Generally a solution of
granulated white sugar (or liquid sugar) is the feedstock.
Strongly acidic cation exchange resins in the hydrogen form catalyze the
inversion of sucrose by means of the hydrogen ions associated with the resin.
This catalysis has been extensively studied by several workers and is thought to
be basically similar to normal acid inversion in being a pseudo first order
7^
reaction (refs. 80-83). However, other workers have observed deviations from
first order kinetics (refs. 83-85), indicating that this is too simple a
description of the catalysis. These deviations have been at least partially
ascribed to the concomitant degradation of the products fructose and glucose by
both heterogeneous and homogeneous catalysis (ref. 83). The rate of inversion
increases rapidly with temperature, the rate constant tripling between 40°C and
70°C (ref. 80). Other factors that increase the rate of inversion are long
contact times (i.e. low flow rates), low solution pH, and highly porous resins.
Invert/sucrose mixtures are produced on an industrial scale from a variety
of grades of granulated and liquid sugars. A typical example is at the Franken
Company's Ochsenfurt beet sugar factory in West Germany (ref. 77). When a low
grade sugar feedstock is used it is decolorized and demineralized before
inversion. Typically a 50-60° Brix low-grade sucrose solution is decolorized by
passing it through one or two beds of strongly basic anion exchange resin in the
chloride form at high temperature. For inversion, the juice temperatures is
adjusted to 30-50°C (depending on the degree of inversion required) and then
passed through a number of hydrogen-form strongly acidic cation exchange resin
beds in series. The inverted juice is then neutralized by passage through a bed
of anion exchange resin in the hydroxide form before evaporation to give the
final product syrup.
pH Control. Careful control of juice pH in beet factories is necessary to
ensure efficient operation and good sugar quality. In 1970 British Sugar
Research Laboratories investigated increases in thick juice pH occurring in the
evaporators at Brigg and Spalding factories (ref. 86). These pH increases were
due to an increase in the amount of acidic compounds in the beet juice that were
eliminated during the process. The cause was thought to be the local growing
conditions during the 1969/70 campaign.
A technique by which thick juice pH could be controlled was developed on a
laboratory scale (ref. 86). This involved the controlled addition to and removal
from second carbonatation juice of a hydrogen-form strongly acidic cation
exchange resin. The cation exchange reactions on the resin released hydrogen
ions into the juice, thus preventing pH increases later on in the process. The
process is unique in that the drop in juice pH was exactly equivalent to that
which would have been produced if the acid used in regeneration had been added to
the juice. In effect the efficiency in the use of acid régénérant was 100Z
75
In partial demineralization, strongly basic anion exchange resins are

sometimes used to adjust juice pH to a desired value before evaporation (ref.
28). This stabilizes the juice in the evaporators.
Water treatment. All beet sugar factories require purified water for such
purposes as boiler feed water. Since the introduction of high pressure boilers,
purification techniques of increasing sophistication have been needed. The sugar
industry is no different from any other industry requiring purified water, ion-
exchange plants of standard design being used to provide the necessary
demineralized water (ref. 87). Apart from the normal contaminants in water
supplies (humic acids, minerals, etc) contamination with compounds such as
natural beet oils, fatty acids and sucrose can occur in sugar factories.
SUGAR RECOVERY WITH RESINS

Quentin Process
Background. For several decades alkali metal cations were suspected of
being highly melassigenic (i.e. they hold sugar in the molasses and prevent it
being recovered as crystalline white sugar). During the mid-1950s Quentin (refs.
88, 89) and Moebes (refs. 90, 91) independently confirmed this and quantified the
melassigenic effect of the alkali and alkaline earth ions. They showed that the
effect of the ions decreased in the order K > Na > Ca > Mg, potassium and sodium
ions being very much more melassigenic than magnesium ions.
In these laboratory studies, they used ion exchange to make syrups where the
majority of cations had been replaced by the specific ion under examination. It
occurred to them that the same procedure might be applied on a factory scale, and
both workers devised ion exchange processes to substitute magnesium or calcium
ions for potassium and sodium ions in factory syrups or juices with the aim of
reducing molasses purity. In the Quentin process there was direct ion exchange
of potassium and sodium ions in low green syrup for magnesium. The initial
Moebes process (refs. 90, 92) involved exchange of the potassium and sodium ions
in second carbonatation juice for ammonium ions. The juice was subsequently
limed, gassed and filtered, thus introducing calcium ions and raising the pH to
about 10.5, before conversion to thin juice and evaporation in the usual way.
After the initial development work on both processes, the greater simplicity of
the Quentin process has caused it to be considerably more favoured (further
developments of the Moebes and other ammonia processes are mentioned below).
76
The Quentin process was successfully tested on a pilot plant in Germany in

the 1955/56 campaign (ref. 89), and then put into full-scale operation. It was
patented in Germany in 1960 (ref. 93). Development and refinement was swift and
within twenty years over 50 European and American factories were using the
process.
Chemistry of Process. The Quentin process consists essentially of passage
of low green syrup through a bed of strongly acidic cation exchange resin in the
magnesium form to exchange about 502 of the potassium and sodium ions for
magnesium ions.
The treated juice has a lower potassium : sodium ratio than the feed syrup
(refs. 30, 94) indicating that a greater proportion of the potassium ions in the
juice exchange for magnesium than is the case for sodium. These results are
explicable in terms of the resin's greater affinity for potassium compared to
sodium.
Macroporous resin is used, because the high solids concentration of the
syrup puts a severe physical strain on the resin through osmotic shock.
The low green syrup treated is the feed to the final crystallization stage
in the beet sugar factory, and the additional sugar recovered is obtained at this
crystallization. The process recovers additional sugar in at least two ways:
(i) Reduction in weight of non-sugars: since magnesium ions have an
equivalent weight of 12.15 compared with 39.1 and 23 for potassium and sodium
ions respectively, the weight of non-sugars is decreased by the exchange and
hance the weight of molasses produced should be correspondingly reduced. If the
molasses purity is unchanged by the process, the expected sugar recovery for 50Z
exchange of potassium and sodium for magnesium ions would be about 0.14Z on beet.
(ii) Reduction in cation melassigenesis: magnesium ions are less
melassigenic than potassium and sodium ions, and this is usually assumed to be
due to their greater hydration compared with potassium and sodium ions (refs. 91,
94). This effectively reduces the amount of water available for solution of
sucrose in magnesium molasses and hence favours increased crystallization of
sugar. Thus molasses produced with magnesium exchange would be expected to be
more completely exhausted than without exchange, and thus should be of lower
purity.
As a further reason for reduced sugar in magnesium molasses it has been
suggested (ref. 95) that crystallization of sugar occurs more quickly than with
77
normal molasses and hence that a greater proportion of sugar than usual is
crystallized in the pans.
The total sugar recovered by these mechanisms is 0.3 to 0.5Z sugar on beet
for 40-502 exchange of sodium and potassium for magnesium (refs. 94, 96-100).
Very similar results have been obtained from British Sugar's Quentin plant at
Brigg factory (ref. 30).
British Sugar Quentin plant. The economic viability of the process depends
mainly upon the relative prices of white sugar, molasses and magnesium chloride
régénérant, as the operating and capital costs are low compared with those for
other methods of sugar recovery, such as the Steffen process. As the process in
effect recovers white sugar at the expense of loss of molasses then, in simple
terms, there has to be a sufficiently large differential between white sugar
revenue gained and molasses revenue lost to make a profit after paying capital
and régénérant costs. Owing to changes in world market conditions, this
differential greatly increased in 1974, and as a result British Sugar Research
Laboratories investigated the possibility of installing an experimental Quentin
plant in a British Sugar factory for operation in the 1975/76 compaign (ref. 30).
A Quentin plant (Figure 10) is broadly similar to a décalcification plant in
that both are for batch ion-exchange processes using a single resin, and
investigation showed that it would be feasible and relatively inexpensive to
adapt one of our décalcification plants as a Quentin plant. The resin
requirement for a Quentin plant is often stated as 6m3 per 1,000 tonnes of beet
sliced per day (refs. 94, 99, 101), and this was confirmed on the Brigg plant
(ref. 30). This is greater than is normally required for décalcification, but
our Brigg factory had a modern décalcification plant of sufficient size for
conversion. This was carried out in 1975 (ref. 30), and the plant has operated
successfully ever since. The original ion exchange columns were utilized, as
were many of the tanks and other ancillary items of plant. Additional plant was
required for handling sweet water arising from unavoidable dilution at the start
and end of each juice cycle.
78
Exhaustion Reeeneration
Low green syrup 70°Brix MgCl* or MgSO* (kieserite)

II II
90° C || 2 BV/hr. IX mg || 3 BV/hr.
II
Macroporous strong acid

cation resin Mg2"*" form
Nik
Treated syrup to flash evaporator
and crystallizer Effluent
Fig. 10. Quentin process at British Sugar, Brigg Factory.
Subsequent changes in the sugar price, and the current E.E.C, sugar
quotas, have decreased the viability of installing further Quentin plants at
present, but experience shows this process to be one of the best means of
increasing sugar recovery.
Alternative régénérants. Most of the world's magnesium chloride is
produced as a by-product of the potash industry and large amounts are
available from the Stassfurt potash deposits in Germany (ref. 29). It is
imported into the U.K. as a solution containing only 8.5Z w/w of magnesium
and thus most of the weight transported is water. Transport costs to the
U.K. are therefore a significant proportion of the cost of this régénérant.
In 1980, the régénérant cost comprised nearly 40Z of the total running costs
of our Brigg Quentin plant—a level where the profitability of the process
in the United Kingdom was principally determined by this one factor.
Furthermore, the cost of magnesium chloride had doubled from 1975 to 1978,
considerably reducing profitability.
79
Alternative sources of magnesium chloride, and of other soluble

magnesium compounds were investigated by British Sugar Research Laboratories
(ref. 30). The only alternative and sufficiently cheap bulk source of
soluble magnesium available in the U.K. was found to be kieserite, an
agricultural grade magnesium sulphate monohydrate.
Kieserite is also found mainly in the same Stassfurt deposits in
massive, compact or granular beds up to 12 ft thick with layers alternating
with potassium and sodium chlorides. As sold, the higher grades of
kieserite are crystalline solids (crude MgSO« . H 2 0) containing about 13-
162 magnesium and 6-92 insoluble matter. The quantity of sodium and
potassium salts depends upon the source. It is extensively used as a
fertilizer in the U.K. and several fertilizer manufacturers import the
material.
Kieserite is slow to wet and dissolve even in warm water, and it
easiily solidifies to a concrete-like mass when wetted or partially
extracted with water. These factors precluded the use of a standard design
saturator (as used for dissolving salt in the original décalcification
plant) for preparation of kieserite solution. However, British Sugar
designed, perfected and patented (ref. 101) a technique to dissolve all the
extractable magnesium sulphate and produce a clear and colourless solution
with a residue having little tendency to solidify. Solutions containing up
to 52 Mg w/v could be produced by these means. Laboratory studies and
factory-scale experiments showed that equivalent Quentin regeneration to
magnesium chloride was obtained with kieserite containing up to 12 K and
12 Na on kieserite. As would be expected, at higher levels of sodium and
potassium the performance deteriorates because of the effect of these ions
on the regeneration equilibrium.
Since 1980 the Brigg Quentin plant has operated successfully with
kieserite régénérant, at a substantial cost saving.
The problem of high régénérant costs was not unique to the United
Kingdom, and other workers have explored other possible alternatives (ref.
103). Regeneration with calcium salts would give substantially reduced
regeneration costs but also would probably reduce the amount of sugar
recovered. Quentin molasses (high in magnesium content) has been suggested
as a possible régénérant, in a variation of the Quentin process analogous to
Gryllus décalcification (ref. 103).
80
Demineralization
Theory of demineralization. The terms demineralization or deionization
are used to describe ion exchange processes by which ionic mineral purities
are removed from water. Similar techniques can be used in the beet sugar
industry to remove juice non-sugars (refs. 6, 77). In partial
demineralization, a proportion of the non-sugars is removed to increase
juice purity and hence extract additional sugar in the factory process. In
total demineralization, all non-sugars are removed from juice to give a
saleable liquid sugar product directly. In both cases the basic process
involves passing juice through a cation exchange resin in hydrogen form and
an anion exchange resin in hydroxide form. The hydroxide ions released by
removal of juice anions onto the anion exchange resin neutralize the
hydrogen ions released by removal of juice cations onto the cation exchange
resin, so the juice non-sugars are effectively replaced with water.
Although juice contains many non-isolated or weakly ionized non-sugars,
as well as simple cations and anions, all can be removed from the juice by a
combination of cation and anion exchange resins. For total demineralization
it is essential to include a strongly acidic cation exchange resin, because
this is the only type of resin which will remove betaine, a nitrogeneous
non-sugar present in significant amounts in beet juices (10-15Z of juice
non-sugars) (ref. 104). However, this type of resin causes inversion of
sucrose. For minimum inversion the juice is normally treated at a solids
concentration of 15-30° Brix, and requires stringent precautions against
mesophilic bacterial activity in the plant.
Complete demineralization cannot be achieved by only one column each of
cation and anion exchange resin, because the ion exchange reactions are
reversible. Further pairs of cation and anion exchange columns are
required, the number needed to produce a high purity product depending on
the purity of the feed juice. A similar effect can be obtained by
intimately mixing cation and anion exchange resins in a single mixed bed.
Hydrogen ions released by ion exchange on one bead are immediately
neutralized by hydroxide ions released from a neighboring bead. This
displaces the equilibrium reactions towards complete demineralization and
gives the effect of an almost infinite number of alternate cation and anion
exchange columns.
81
In the laboratory, thick juice can be completely demineralized to give

a pure colourless sucrose solution (plus some invert sugars) by passing it
down a single mixed bed of strongly acidic cation exchange resin and weakly
basic anion exchange resin (ref. 105). At a relatively slow flow rate of
about 0.3 bed volume/hour, 1.4 bed volumes of a water-white 99.95Z purity
(expressed as X sucrose plus invert on total solids) juice was produced from
undiluted 60° Brix thick juice, before colour and betaine started to leak
through the resin bed.
However, the advantage of mixed bed demineralization is offset by the
practical difficulties of operating such a system. The resins must be
completely separated from each other prior to regeneration. This is
normally achieved by careful backwashing (anion exchange resins are
generally less dense than cation exchange resins), and regeneration is
carried out with balanced flows of régénérant and water to prevent each
régénérant contacting the wrong resin. Juice demineralization plants use
alternate cation and anion exchange columns, with a final mixed bed column
for polishing where a high quality product is required.
It is common to follow strongly acidic cation exchange resin columns by
columns of weakly basic anion exchange resin. The high affinity of weakly
basic resins for hydroxide ions (much higher than that of strongly basic
anion exchange resins) means that less régénérant is required to restore the
resin to the hydroxide form. They can also be regenerated by ammonia
solution rather than the much more expensive sodium hydroxide solution
necessary for strongly basic resins. This appreciably reduces regeneration
costs.
Weak/strong resin combinations (either weakly acidic/strongly basic or
strongly acidic/weakly basic) require the juice to be passed through the
strongly acidic or strongly basic resin first. Weakly ionized resins need
the pH change caused by the ion exchange reactions on the strongly ionized
resins, producing an acidified or an alkaline juice, in order to function
effectively. The weakly acidic/strongly basic resin combination is more
expensive to regenerate than the strongly acidic/weakly basic combination
and the former could not be used for total demineralization as it would not
remove betaine.
Weak/weak resin combinations only work in mixed beds, where the
intimate contact of anion and cation exchange resins displaces the ion
82
exchange equilibrium towards complete demineralization, and are rarely used.

They would also not remove betaine.
An alternative approach to demineralization was developed by Moebes
(refs. 106, 107) and by Vajna (refs. 108, 109). Instead of using cation and
anion exchange resins in the hydrogen and hydroxide forms respectively,
their processes used ammonium form cation exchange resins and either
hydroxide or carbonate/bicarbonate form anion exchange resins. Thus the
process replaced juice cations and anions with ammonium hydroxide or
carbonate. These were driven off in the evaporators and recovered by
condensation, so that the treated juice was effectively demineralized. This
process has the advantage of eliminating the problem of inversion that
occurs with hydrogen form cation resins. It has been installed in a few
factories, notably at Enns in Austria (refs. 110, 111), but in general has
not been particularly successful. A recent variation of this process, which
may show more promise, has been proposed by Schoenrock (refs. 54, 55).
Factory processes. Many schemes for the demineralization of a
wide variety of beet juice syrups and dissolved sugars have been prepared
(refs. 112-114), but only comparatively few have been put into factory
operation (refs. 115-118). Demineralization is an expensive process, its
economic viability depending on a number of factors, including the relative
prices of the final product, the feedstock and the molasses no longer
produced; the possible utilization of eliminated non-sugars and spent
régénérants as
saleable by-products ; the efficiency of the plant and the number of days per
year it is operational.
The French beet sugar factory at Vauciennes has a total
demineralization plant built by the Applexion Co. (ref. 118) (see Figure
11). The feedstock is a stored thick juice so the plant can be run
throughout the year. The stored thick juice is diluted to 35° Brix with
sweet water from the ion exchange columns, and cooled to 10°C to minimize
inversion. The first stage of the process consists of three parallel
streams of two columns each, containing a strongly acidic cation exchange
resin (in hydrogen form) followed by a weakly basic anion exchange resin (in
hydroxide form). The second stage consists of two parallel streams of three
columns each. These contain a non-ionic absorbent (for colour, odour and
taste removal), a hydrogen form strongly acidic cation exchange resin and a
83
hydroxide form strongly basic anion exchange resin. The third stage is a
mixed bed of cation and anion exchange resins. The final product from this
stage, a pure sugar solution containing about 1% invert, is evaporated and
either sold as a liquid sugar or crystallized to give very high quality
white sugar.
The cation exchange resins are regenerated with sulphuric acid, the
weakly basic anion exchange resins with ammonium hydroxide and the strongly
basic anion exchange resins with sodium hydroxide. All the spent
régénérants are mixed together, concentrated and crystallized to give two
by-products. The crystalline material, consisting mainly of ammonium,
sodium and potassium sulphates, is sold as a fertilizer and the mother
liquor which is rich in nitrogen compounds and other organic non-sugars, is
sold as an animal feed supplement.
Exhaustion
Strong Weak
Acid Base Strong Strong
Absorbent Acid Base SAC/
Cation Anion
Cation Anion
Regeneration
H-SO. NaOH H2S04/foaOH
4
i _ 4-
Strong Weak Strong Strong

Acid Base Acid Base
Cation Anion Absorbent! Cation Anion
Evaporation/Crystallisation
Mother liquor Solids
Organic Nitrogen Compounde NH4,Na,K/S04
Fig. 11. Applexion demineralization at Vauciennes.

m
The French factory at Nassandres has a partial demineralization plant,

also built by Applexion. This resembles the first stage of the Vauciennes
plant, with a strongly acidic cation exchanger (sulfuric acid régénérant)
then weakly basic anion exchanger (ammonia régénérant). These are followed
by treatment on a strongly basic anion exchanger (sodium hydroxide
régénérant) to adjust the product pH in order to stabilize the juice. The
plant treats a sugar-end syrup, part of which is stored so that the ion
exchange process can be run throughout the year. Essentially, the purpose
of the demineralization plant is to remove some of the juice non-sugars and
hence recover additional sugar. The spent régénérants are reprocessed in a
similar way to those from the Vauciennes plant, to give saleable by
products.
Partial demineralization is widely used in Southern Europe, where poor
beet quality gives low thin juice purities (e.g. as low as 87-89Z in Italy
(ref. 27) and 82-842 in Sardinia (ref. 28)). The thin juice is usually
treated with a three column system similar to the Nassandres plant. A
strongly acidic cation exchange resin and weakly basic anion exchange resin
remove non-sugars, to give say a 922 purity product juice, and thereby
considerably increase sugar extraction. Finally, a strongly basic anion
resin raises the pH of the treated juice and thus stabilizes it before
evaporation.
In the U.S.A., Bichsel has worked for many years on juice
demineralization. He developed a partial demineralization system at the
Holly Sugar Corporation's Hamilton City plant (ref. 116). This involved
passing a mixture of a sugar end green syrup diluted with thin juice through
a strongly acidic cation exchange resin followed by two columns of weakly
basic anion exchange resins (see Figure 12). The output was sent back to
the feed to the white pans.
The plant was later modified on the pilot scale into a total
demineralization system designed to treat stored thick juice and produce
liquid sugar (ref. 119). This latter process was intended to operate all
year round. A two-stage demineralization process was used, with diluted and
cooled thick juice (30° Brix, 8°C) being passed through the bed of cation
exchange resin and two beds of weakly basic anion exchange resin, as before.
This was followed by two demineralization columns both containing beds of
strongly basic anion exchange resin and strongly acidic cation exchange
85
resin, in the hydroxide and hydrogen forms respectively. The beds of anion
and cation resin were separated by a screen and regenerated separately.
The weakly basic anion exchange resins were regenerated with ammonia.
The spent régénérant from these columns was passed down the cation exchange
resin column (before their regeneration with sulphuric acid), stripping off
colour bodies, betaine and other nitrogenous organic compounds. This
reconstituted non-sugars solution was then used as an animal feed supplement
by spraying it onto beet pulp. The strongly basic anion resins were
regenerated with sodium hydroxide. The spent régénérant from these columns
could be mixed with the spent sulphuric acid régénérant from the cation
exchange columns to give a by-product rich in ammonium sulphate, which could
be used as a fertilizer (refs. 120, 121).
This regeneration scheme was devised primarily to deal with the
effluent disposal problem, which threatened the viability of the process,
and also to reduce costs. Thus, a potential liability was transformed into
a positive asset. In later work, Bischel introduced a very similar partial
demineralization plant into the American Crystal Moorhead factory (ref.
122).
A novel partial demineralization process has been developed recently by
the French sugar company Generale Sucriere and the resin company Rohm and
Haas (ref. 123). One of the several unusual features is that both resins
used are types that are only weakly ionized. Weakly acidic cation exchange
resin (in hydrogen form) and weakly basic anion exchange resin (in hydroxide
form) are intimately mixed with a batch of thick juice by air agitation.
Individually, these weakly ionized resins would be unable to ion exchange to
any great extent because the chemical equilibrium favors the reverse
reaction. However, removal of the released hydrogen or hydroxide ions by
the other resin alters the equilibrium (as described above) and permits a
substantial demineralization to take place. During the next step,
sweetening off with sweet water from the previous cycle, the cation and
anion exchange resins are separated due to their relative densities compared
to that of the sweet water. The resins are regenerated by the normal
techniques for mixed beds, using sulphuric acid and ammonia, and the spent
régénérants are combined and processed into saleable by-products.
86
Exhaustion
Strong Weak Weak

Acid Base Base
Cation Anion Anion
Regeneration
Animal Feed (R.N.S.)
1 *
\ !
Strong Weak Weak
Acid Base Base
Cation Anion Anion
\' —
· ______ _ f _
fertiliser
Fig. 12. Bichsel partial demineralization; Hamilton City. (- 1st

stage regeneration; 2nd stage regeneration).
Chromatography
The idea of using ion exchange resins to separate sucrose from non-
sugars by chromatography, rather than ion exchange was suggested by Dow
Chemical Company in 1953 (ref. 124). The Chromatographie separation of
sucrose from ionic and other non-sugars on an industrial scale takes place
by two main mechanisms:
Separation by molecular size - the molecular sieve effect. If a
solution containing a mixture of large and small molecules is applied to a
column of an ion exchange resin, those molecules small enough to enter the
bead pores will be separated from the larger molecules. The large molecules
are excluded by their size from the pores and travel down the column through
the voids between the beads. Hence they have a much shorter pathway through
87
the resin bed than the molecules capable of penetrating the bead pores, and
are therefore eluted from the column first. The resin porosity is chosen to
achieve the separation required.
Ion exclusion - the Donnan membrane effect. Donnan membrane theory
states that ions in solution tend to be repelled from an ion exchange resin
by like ions occupying the resin's ion exchange sites. Hence ionic
compounds will tend to pass through the void volume of a column of resin
rather than through the resin itself, and will therefore have a shorter
pathway through the resin bed than non-ionic compounds which travel through
the pores of the resin beads. If an impure sucrose solution (dilute
molasses for example) is applied to a column of strongly acidic cation resin
in the sodium form and eluted with water, the sodium and potassium ions on
the juice will tend to be excluded from the resin pores. With a resin of
the appropriate porosity the sucrose will penetrate the resin beads and will
therefore be eluted after the ionic compounds. It should be noted that the
Donnan membrane effect does not totally exclude ions from the resin and thus
prevent ion exchange reactions taking place, but merely induces a tendency
for the ionic compounds to pass around the resin beads, rather than through
them.
These two separation mechanisms are the basis of processes for the
Chromatographie desugarization of molasses developed in several countries,
including Finland (ref. 125) (see Chapter 7), Germany (refs. 126, 127) and
Japan (ref. 128).
Ion exclusion processes require feedstocks that are very low in
multivalent ions (e.g. Ca, Mg). These ions are absorbed onto the resin's
ion exchange sites and their presence breaks down the Donnan membrane
effect, thus causing the sucrose/ionic non-sugars separation to deteriorate.
Thus the process is not suitable for molasses from factories operating the
Quentin process.
There is another separation mechanism that is employed in separating
fructose from glucose (ref. 77). An invert syrup is applied to a strongly
acidic cation exchange resin in the calcium form. The fructose forms a
complex with the calcium ions on the resin and hence is delayed with respect
to glucose, so that the glucose is eluted first.
88
THE FUTURE
It is always very risky to predict the future in any industry, and this
is particularly so in one such as ours which is subject to influence by such
unreliable factors as international politics and the weather I However it
may be possible to suggest a few trends.
First, we live in a time of great technological changes and
developments. Some of these areas of change, such as biotechnology, require
separation processes, and as a result could lead to developments in the
field of ion exchange. In turn these may also prove to have uses within, the
sugar industry—there are some strong similarities between factory juices
and the fermentation broths from biotechnology processes.
Developments in some areas of ion exchange resin technology would be
particularly welcome in the sugar industry. For example, strongly basic
anion exchange resins with greater thermal stability would be of benefit in
decolorization processes, as would a resin combining a styrenic resin's
decolorization capacity with an acrylic resin's increased resistance to
organic foulants. In demineralization a resin that will remove betaine
without inversion of sucrose would be of considerable interest. These are
just a couple of examples of where developments in ion exchange resin
technology would be of benefit to the sugar industry.
Second, new ways are still being developed for applying the traditional
types of resins in beet sugar factories of which the recent French partial
demineralization process just mentioned is an example. As our industry
changes and becomes even more efficient, as every industry must, further
opportunities and needs may arise for the use of ion exchange. The
treatment of effluents, to avoid pollution and improve our environment, is a
possible example.
So, ion exchange processes are still being developed in laboratories
for application in the beet sugar industry, and it is likely that they will
continue to be developed in the future. Which ones will prove to be
practicable on a factory-scale is often hard to say, but it is likely that
some will pass the test.
In short, ion exchange has been applied in the beet sugar industry for
many years, and we believe it will continue to be applied for many more.
89
CONCLUSIONS
1) Ion exchange processes are an important part of the beet sugar
industry and may become of greater importance in the future.
2) The major ion exchange processes fall into two groups: process
aids (e.g. décalcification and decolorization) and sugar recovery techniques
(e.g. Quentin, demineralization and Chromatographie desugarization).
3) Décalcification is a very successful process, but recently it has,
in some cases, been under pressure from the general increasing concern over
the environment (spent régénérant disposal) and competition from
antiscalants. These concerns have, at least partially, been answered by the
Gryllus and N.R.S. modifications to the décalcification process.
4) Decolorization, essential in the cane industry, is much less widely
used in sugar beet factories because it is, in the main, unnecessary and, in
normal factory processes, its economic benefits are a little tenuous. In
certain circumstances, e.g. in beet refineries or for the production of very
high quality liquid sugar, decolorization is economically viable. Further
research by both the sugar industry and by resin manufacturers is required
to develop the optimum decolorization process, one that does not introduce
undesirable ions (e.g. chloride ions) into the juice and that is not limited
by the thermal stability of the resins.
5) The practicability of the Quentin process depends on economic (the
relative prices of white sugar and molasses; régénérant costs) and
environmental (effluent disposal) factors. Some of the current research
into the Quentin process is on alternative regeneration systems which may
increase its profitability. The Quentin process remains one of the simplest
and best methods of sugar recovery, but can only recover a part of the
molasses sugar.
6) Partial demineralization is an essential part of some beet factory
processes, particularly those in areas where the quality of the beet (and
hence the beet juice) is relatively poor.
7) Total demineralization is an expensive process, its practicability,
in common with other sugar recovery processes, depending on a number of
factors. Some workers have concentrated on obtaining saleable by-products
(ranging from animal feeds and fertilizers to specific chemicals such as
betaine and amino acids) from the spent régénérants. Total demineralization
processes are appreciably more profitable when they can be run all year
90
round, with stored juices (e.g. thick juice, molasses) used as the feedstock
in the off-season. Modern demineralization processes are designed to
operate in this manner, and this is certainly the future for these
processes.
8) Chromatographie desugarization of the sugar juices is an
alternative to the Quentin and demineralization sugar recovery processes.
Much research into different Chromatographie techniques for sugar recovery
is currently being performed, and new commercially viable processes may soon
be developed.
REFERENCES
1 A. Z. Rumpier, Verin Deutscher Zucker Int. 53 (1903) 798.
2 Aristotle (ca. 330 BC) Works, Vol. 7, p993b. Clarendon Press London
1927.
3 Exodus 15, 23-25.
4 M. Shore, J. A. Adams, N. W. Broughton, N. Bumstead and G. C. Jones,
Some aspects of the chemistry of pulp pressing aids. Zuckerindustrie.
109 (1984) 215.
5 W. Th. J. P. Langenhorst, M. Tels and J. C. Vlugter, H. I. Waterman,
J., Cation exchangers on a sugar-beet pulp base. Application for
decontaminating radioactive waste water. Biochemical and
Microbiological Technol. & Eng. 3 (1961) 7.
6 A. Carruthers and J.F.T. Oldfield, Z., Methods for the assessment of
beet quality. I. Purity determination using a clarified extract from
brei. II. Determination of particular non-sugars in beet. III.
Summation of non-sugars. Zuckerin. 11 (1961) 23 and 85.
7 S. Landi and G. Mantovani, Ion exchange in the beet sugar industry.
Sugar Technol. Rev. 3 (1975) 1.
8 K. J. Parker, Ion exchange in the sugar industry. Chem. Ind.
(1972), 782.
9 R. Kunin, Acrylic-based ion exchange resins and adsorbents. Amber-Hi-
Lites. 175 (1984).
10 D. S. Martin, W. Hunter and B.W. Drean, Decolorizing resins—plant
development at Westburn refinery. Sugar Ind. Technol, 41, (1982), 69.
11 G. A. Akesson and K. A. Lilja, The sugar refining process at SSA,
Arlov division, Sweden (rebuilt 1981). Sugar Int. Technol. 43 (1984)
213.
12 T. V. Arden, Water purification by ion exchange. Butterworths, London
(1968).
13 Ion Exchange Resins, 6th Edition. BDH (1981).
14 R. Kunin and R.J. Myers, J. Amer. Chem. Soc. 69 (1947) 2874.
15 T. Lubienski and B. Mackay, Thin juice décalcification in The British
Sugar Corporation. British Sugar Corp. 22nd Tech. Conf. (1974).
16 J.F.T. Oldfield, C.W. Harvey and M. Shore, Procedures to assess and
reduce the deterioration of décalcification resins in service. Int.
Sugar J. 75 (1973) 70.
17 ICUMSA Proceedings, 16th Session (1974) 375.
18 R. Kunin, Helpful hints in ion exchange technology. Amber-Hi-Lites.
137 (1973).
91
19 U.S. F. & D.A. Regulations CFR 21 Part 173.25. Ion Exchange Resins.
20 Council of Europe Draft Resolution on Ion Exchange Resins Use in Food
Processing (1984).
21 0. Samuelson, Ion exchange separations in analytical chemistry. John
Wiley & Sons, London (1963).
22 B. A. Adams and E.L. Holmes, J. Soc. Chem. Ind. 54 (1935) IT.
23 G. D'Alelio. U.S. Patent, 2,366,007 (1942).
24 C. H. McBurney. U.S. Patent 2,591,573 (1949).
25 K. Shinbori, Y. Sugimato, T. Ando, S. Sakai, Toyo Soda Mgf. Co. Ltd.
and Japan Organo Co. Ltd. 4,264,373.
26 S. L. Cheng, Rpt. Taiwan Sugar Research Inst. 106 (1984) 31.
27 G. Assalini and G. Brandoli, Second carbonatation juice treatment
by means of ion exchangers applied to continuous equipments "Aconex"
and "Preconex." Suer. Belge 91 (1972) 107.
28 C. Barraque and R. Nicol, Deionization of ion-exchangers of second
carbonatated juice, and thorough treatment of waste water produced.
Ind. Alim. Agr. 100 (1983) 853.
29 J. Mellor, Comparative treatise on inorganic and theoretical chemistry.
Halstead Press IV.
30 J.F.T. Oldfield, M. Shore, C.W. Harvey, D. Gyte and G.C. Jones, Quentin
Process. British Sugar Corp. 24th Tech. Conf. (1978).
31 P. Pelosi and J. McCarthy. Chemical Engineering 89 (1982) 75.
32 R. Kunin, Helpful hints in ion exchange technology (cont.). Amber-Hi-
Lites. 131 (1972).
33 R. Kunin, Acrylic-based ion exchange resins and adsorbents. Part 3.
Amber-Hi-Lites. 175 (1984).
34 F. Maekawa, K. Kawasaki, Y. Horiki and T. Saito, Studies on desalting
of sugar solution with ion-exchange resin in a sugar refinery.
1. Contamination of the type 1 strongly basic anion-exchange resin by
desalting sugar solution in the reverse system. 2. Restorative
treatment on the type 1 strongly basic anion-exchange resin contami
nated by desalting sugar solution in the reverse system. Proc.
Research Soc. Japan Refineries Tech. 28 (1978) 78.
35 F. Schneider, D. Schliephake and J. Paleos, Adsorption of colouring
matter on decolorizing resins. Int. Sugar J. 70 (1968) 67 and 77.
36 G.C. Jones. Unpublished British Sugar Internal Report. (1981).
37 A. S. McGrath and G.C. Jones. Unpublished British Sugar Internal
Report (1982).
38 E. Warner, Z. Zuckerind. 1 (1952) 413.
39 P. Smit, Deliming of sugar factory juices by ion exchangers. Sucr.
Beige. 82 (1962) 82.
40 Z.D. Zhuravleva and K.P. Goncharova, Operation of the installation
for softening juice at Otrada Sugar Factory. Sakhar. Prom. 46 (1972)
31.
41 A. Carruthers, J.F.T. Oldfield and M. Shore, A.E. Wool, Pectin and
polysaccharides in beet juice and molasses. British Sugar Corp. 9th
Tech. Conf. (1956).
42 J.F.T. Oldfield, M. Shore, N.W. Broughton and G.C. Jones, Juice décal
cification and its effect on molasses production. British Sugar Corp.
25th Tech. Conf. (1980).
43 S. Zagrodski and H. Zaorska, The influence of de-liming of sugar
factory juices on losses of sugar in molasses. Suer. Belge. 81 (1961)
137.
92
44 F. Durdik and J. Burianek, Softening of thin juice by means of cation-

exchangers. I. Mechanism of ion-exchange in softening and
regeneration. Effect of softening on the quality of thin juice.
Listy Cukr. 76 (1960) 2.
45 L. Wieninger and N. Kubradinov. Jahresber. Zuckerforschungs - Institut
(Wien). (1969) 107.
46 E. Gryllus and H.F. Delavier, BMA-Zsigmond-Gryllus process: A new
process for thin juice deliming. 1. Principles of thin juice
deliming. Z. Zuckerind. 25 (1975) 493.
47 F. Durdik and H.J. Delavier, Z. Zuckerind. 28, (1977), 27..
48 E. H. Felber, New process for deliming thin juice water without
regenerating agents and waste water. J. Amer. Soc. Sugar Beet Technol.
16 (1971) 279.
49 Anon. BMA Information No. 16 (Dec.) (1977) 725.
50 J. Ponant, Gryllus Effect. Sucr. Fr. 120 (1979) 263.
51 Y. Le Henaff and D. Herve, Ind. Aliment, Agr. 94 (1977) 725.
52 P. L. Mottard, The Imacti process for juice décalcification. Int.
Sugar J. 85 (1983) 233.
53 X. Lancrenon and P. Printemps, Processes for deliming sugar juices.
Int. Sugar J. 86 (1984) 41.
54 K. W. Schoenrock, A. Gupta and D. Costesso, New approaches to the
application of ion exchange in the sugar industry. Proc. 1976 Tech.
Sess. Cane Sugar Refining Res. (1977) 118.
55 K. W. Schoenrock, Thin juice softening over the hydrogen form of a weak
catex. A.S.S.B.T. (1985).
56 D. J. McWeeny, M. E. Knowles, J. F. Hearne, Chemistry of non-enzymatic
browning in foods and its control by sulphites. J. Sci. Fd. Agric. 25,
(1974) 735.
57 H. E. Nursten, Food Chemistry. 6 (1980) 263.
58 M. Shore, N. W. Broughton, J. V. Dutton and A. Sissons, Factors
affecting white sugar colour. Sugar Technol. Rev. 12 (1984) 1.
59 N. W. Broughton, B. J. Houghton, A. Sissons, Factors affecting white
sugar colour. 2. The sources of white sugar colour. Zuckerind. Ill
(1986) 1039.
60 W. Fries, R. M. Walker, Proc. 1980 Tech. Session Cane Sugar Refining
Res. p.171.
61 W. Fries, Cane sugar decolorization by ion-exchange resins. Int. Sugar
J. 84 (1982) 325.
62 N.G.H. Hindefelt, K. A. Lilja, Acrylic resin and styrene resin for
decolorization. A full-scale comparison. Int. Sugar J. 88 (1986) 148.
63 K. Wagnerowski (translated by D. Mottard-Dabrowska). Rationalisation
du Processus d'Equisèment de la Melasse. Association pour la Promotion
Industrie Agriculture (1983).
64 J. P. Brandley. Unpublished British Sugar Internal Report (1985).
65 F. Heitz, Melassigenic power of ions. C.I.T.S. 17 (1983) 585.
66 H. Schiweck, Physical and chemical properties of an acidic colour-
substance isolated from the regeneration effluents of a decolorizing
column for remelt liquors. Zucker. 16 (1963) 555.
67 E. Steineck. Jahresber. Zuckerforschungs-Institut (Wien). (1967-8),
49.
68 H. Eichhorn. Int. Sugar J. 84 (1982) 135.
69 M. Braeckman, The new sugar house at Tirlemont refinery. 17th C.R.
Assem. C.I.T.S. (1983) 403.
93
70 L. Ramm-Schmidt and G. Hyoky, New resin decolorization station at

Finnish Sugar Porkkala refinery. Sugar Ind. Technol. 43 (1984) 241.
71 S. Landi and G. Mantovani, Operating parameters of an ion-exchange
decolorization plant in a sugar factory. Ind. Sacc. Ital. 65 (1972)
132.
72 Ya. 0. Kravets, V.N. Eremenko, G.V. Buzetskaya, L.D. Naumova,
Decolorization of refinery syrups with porous anion exchanger AV-
17-2P. Sakh. Prom. 7 (1982) 31.
73 S. Gielzynski, Use of ion-exchange columns for decolorization of
remelt syrups. Gaz. Cukrow. 89 (1981) 100.
74 G. Siegers and F. Martinola (Bayer). Paper presented at the
International Conference on Ion Exchange, Cambridge (1984).
75 G. C. Jones, A. S. McGrath. Unpublished British Sugar Internal Report
(1983).
76 A. Tavares, R. Forman and R. Kunin, Refinery and laboratory studies
on the use of Ecosorb precoat formulations for the clarification and
decolorization of cane and beet sugar syrups. Sugar Ind. Technol. 42
(1983) 192.
77 N. Marignetti and G. Mantovani, Liquid sugar. Sugar Technol. Rev. 7
(1979) 3.
78 R. Carolan, Manufacture of table syrup in Ireland. Int. Sugar J. 54
(1952) 106.
79 Anon. Susswaren 13 (1969) 186.
80 G. Siegers and R. Martinola, Sucrose inversion with cation-exchange
resins. Int. Sugar J. 87 (1985) 23.
81 I. Bodamer and R. Kunin, Heterogeneous catalytic inversion of sucrose
with cation-exchange resins. Ind. Eng. Chem. 43 (1951) 1082.
82 K. Taufel and K.S. Grunert, On the behavior of saccharides towards
ion-exchange resins. 3. Rate of hydrolysis and activation energy
in the homogeneous, heterogeneous and combined catalysis of sucrose.
Archiv, der Pharmazie 66 (1961) 439.
83 E. Berghofer, H. Kalushofer and L. Wieninger, Sucrose inversion with
cation exchangers. 1. Literature review. 2. Work carried out by
the authors themselves. Zucker 30 (1977) 196.
84 M. Takeda and T. Imura, Catalytic inversion of sucrose with ion-
exchange resin. J. Schimonoseki Univ. Fisheries 13 (1964) 103.
85 E. Berghofer, H. Kloushofer and L. Wieninger, Sucrose inversion with
cation exchangers. 1. Literature review. 2. Work carried out by
the authors themselves. Zucker 30 (1977) 187.
86 J.F.T. Oldfield, M. Shore and M. Senior. Int. Sugar J. 72 (1970) 187.
87 Betz Handbook of Industrial Water Conditioning, 6th Edition. (1968),
p. 103.
88 G. Quentin, Influence of inorganic cations on the solubility of
sucrose in molasses. Zucker 7 (1954) 407.
89 G. Quentin, Effect of cations on the solubility of sucrose in molasses
and the possibilities of technological evaluation of the differential
influence on solubility. Zucker 10 (1957) 408.
90 E. Moebes, The chemical and physical behavior of non-sugars in
molasses formation and refining. Z. Zuckerind. 7 (1957) 382.
91 E. Moebes, Influence of cations on the solubility of sucrose and the
viscosity of impure beet sugar solutions. Zucker. 10 (1957) 78.
92 E. Moebes and L. Wieninger, Effect of exchange of alkali ions for
calcium on the processing and yield of sugar juices. Zucker 8 (1955)
129.
*
93 G. Quentin. German Patent 974,408.

94 H. Neumann. Zucker 8 (1955) 129.
95 Robert Reichling and Co., K. G., 'Quentin Process.'
96 R. Pieck, J. Houssiau, R. Vandewijer, Some data on the variation in
the solubility and other physical properties of sucrose in the
presence of non-sugars in the form of K, Ca or Mg. Sucr. Belge 87
(1968) 319 and 371.
97 G. Quentin, Discussion of experience with the magnesium chloride
process. Z. Zuckerind. 14 (1964) 683.
98 J. Huberlant, Some thoughts on the Quentin Process. Suer. Belge 86
(1967) 505.
99 K. Schoenrock, Application and utility of magnesium exchange in the
beet sugar process. J.A.S.S.B.T. 16 (1971) 299.
100 Colloque Applexion. (4/6/1975) p. 95.
101 British Sugar Corporation. British patent 1,576,100.
102 V. I. Smagina, V. S. Shterman, A. R. Sapronov and G. S. Tereschenko,
Use of epsomite for increasing the sugar yield from final massecuite.
Sakh. Prom. 8 (1984) 21.
103 G. Assalini, G. Brandoli. EP 16922, A6 Mar. 1980, No. 80110028.9,
p. 15, October 1980.
104 B. P. Baker, Composition, properties and uses of molasses and related
products. United Molasses Trading Company Ltd.
105 G. C. Jones, C. W. Harvey. Unpublished British Sugar Internal Report
(1974).
106 E. Moebes, The 'Carbonate Process,' a method for the purification
of thin juice. Zucker. 13 (1960) 254.
107 E. Moebes, Pilot scale experiments with the carbonate process in the
1960/61 campaign. Zucker. 14 (1961) 497.
108 S. Vajna, Juice purification by demineralization by ion-exchangers.
Zucker 12 (1959) 186.
109 S. Vajna, Demineralization process by ion-exchange in the ammonium
cycle. Z. Zuckerind. 9 (1959) 279.
110 Sugar Chemical Co. Etablissement Fr. PI,386,961 (1964).
111 Ennser Zuckerfabriks A.-G. Fr. P.1,505,433 (1967).
112 Y. Ito, Y. Owada, S. Shinada, "Reverse two bed" system
demineratlization of sugar liquor with ion-exchange resin. Proc. Res.
Soc. Japan Sugar. Refineries Technol. 20 (1968) 39.
113 F. Perschak, Experiments on production of liquid sugar by ion-
exchange demineralization of molasses solutions. Zucker 22 (1969) 666.
114 D. Herve and X. Lancrenon, New ion-exchange processes in sugar
factories. Ind. Aliment. Agr. 98 (1981) 613.
115 S. Oikawa and A. Miyahara, Beet juice demineralization. Sugar J. 33
(1971) 19.
116 S. E. Bichsel and D. A. Hoover, Operation of the Hamilton City experi
mental ion exchange plant. J.A.S.S.B.T. 15 (1969) 324.
117 S. E. Bichsel, H. A. Davis and A. Sandre, Synopsis of the Moorhead
deionization plant operating and economic characteristics. Suer. Belge
98 (1979) 67.
118 J. Foucart and J. Paleos, The new liquid sugar process at the
Vauciennes Factory. British Sugar Corp. 28th Tech. Conf. (1986).
119 L. T. Zanto and S. E. Bichsel, Direct production of liquid sugar
from ion exchange treated thick juice. J.A.S.S.B.T. 16 (1970) 149.
95
120 J. A. Levad, T. D. Carpenter and S. E. Bichsel, Utilization of

concentrated ion-exchange non-sugar waste on dried beet pulp.
J.A.S.S.B.T. 16 (1971) 348.
121 R. F. Olson, A. M. Sandre and S. E. Bichsel, Utilization of a
deionization plant waste régénérant stream to produce a valuable
granular fertilizer by-product. Suer. Belge 98 (1979) 11.
122 S. E. Bichsel, H. A. Davis and A. Sandre, Synopsis of the Moorhead
deionization plant operating and economic characteristics. Suer. Belge
98 (1979) 67.
123 G. Rousseau. In: Ion Exchange Technology, eg. D. Naden, M. Streat,
Published by Horwood, Chichester (1984) 381.
124 R. M. Wheaton and W. C. Bauman. Ind. Eng. Chem. 45 (1953) 228.
125 H. Hongisto and M. Loisa, Finnsugar separation process for
desugarization of molasses. Z. Zuckerind. 27 (1977) 279.
126 Anon. Zuckerind. 108 (1983) 636.
127 H. G. Schneider and J. Mikule, Recovery of sugar from beet molasses
by the P[fiefer] and L[angen] exclusion process. Int. Sugar J. 77
(1975) 259.
128 K. Sayama, Y. Senba, T. Kawamoto, S. G. Kenkyukaishl, Chromatographie
separation of molasses constituents. 1. Recovery of sucrose from
molasses. Proc. Res. Soc. Japan Sugar Refineries' Technol. 29 (1980)
1.
Chemistry and Processing ofSugarbeet and Sugarcane, edited by M.A. Clarke and M.A. Godshall
96
Chapter 6
TREATMENT AND DISPOSAL OF EFFLUENTS OF THE SUGARBEET FACTORY
JOHN A. RICHMOND AND ROBERT W. STRICKLAND
INTRODUCTION
"All substances are poisons, there Is none which Is not a poison. The
right dose differentiates a poison and a remedy." (Paracelsus 1493-1541).
The United States Environmental Protection Agency (EPA) was created in
December of 1970 to consolidate and strengthen federal efforts to protect
the health and welfare of the American people, by rehabilitation and
protection of the environment. The most important function of EPA is to
assist state and local governments in their efforts to establish regulations
deemed necessary to protect their local environments. Localization of
environmental regulation is not conducive to across-the-board solutions to
waste water handling in the beet sugar industry which is spread over 5
federal regions and 16 states. Large climatic, geographic, socio-economic
and political variances result in a wide range of regulations, limits, and
control methods.
The American beet sugar industry is essentially rural. Most of its
beet sugar factories operating today were built between 1916 -1965 in
comparatively open space near streams or rivers which could furnish a water
supply and also accept waste water discharge. Now, however, factories find
themselves surrounded by residences. This coupled to their close proximity
to rivers and streams places additional environmental pressures on effluent
management. Management of factory effluent streams is further complicated by
a number of factors, including variations in processes, types of equipment,
water usage, length of processing time and climatic conditions.
Resolution of any problem should start at the cause, not the effect.
Most beet sugar factories operating today have streamlined their process for
maximum utilization of water entering the process in an effort to reduce
overall discharge. Discharge streams have been separated into treatment
systems designed for specific waste streams. Thus a factory may have more
than one type of treatment process for different effluent discharges.
97
SOURCES OF DISCHARGE
Flume Water
Flume water, the water which is used for transporting and washing
beets, is normally the waste stream accounting for the greatest volume, and
in many cases has the greatest Biochemical Oxygen Demand (BOD) loading.
Miscellaneous waste streams generated in the factory are, at many locations,
combined with the flume water system. As such, flume water is generally
considered the main factory waste stream (Figure 1).
[ PILED ^V
RAW OVERFLOW
f SUGAR I
WATER
BEET FLUME
V
DEWATERINQ SCREEN
J ÜJ
I
BEET WASHER
MUD MUD
EFFLUENT SETTLING
MISC. FACTORY WASTE
POND POND
WASTE WATER TREATMENT ΙΓ
Fig. 1. Flume water circuit.
Composition of flume water is greatly affected by harvest conditions,

which determines the amount of trash and mud delivered with the beets, and
fluctuates considerably from season to season. The physical soundness of
beets contributes to the overall BOD loading of flume water. Constituents
of the beets, mainly sucrose, are leached from cut and damaged surfaces into
the water. Beets which are frozen or rotten contribute large amounts of
soluble organic matter to the water system, which ultimately increases the
98
BOD loading with which treatment systems have to deal. BOD values in excess
of 10,000 mg/1 are not unknown (Table 1).
Condenser Water
As shown in Table 1, condenser water is the second largest component of
factory waste water discharge. Large volumes of water are required in jet
or barometric condensers to produce reduced pressures for operation of
evaporators and vacuum pans. Normally condenser water ranges in temperature
between 50° and 65° with a B0Ds of 50 mg/1 or less.
For factories with adequate water supply, condenser water has been used
on a once through basis with disposal following cooling, directly back to
the source. Because of its only slightly degraded condition and elevated
temperature, condenser water is often used in various other locations
throughout the beet sugar process, often as a source for flume water make up
and factory wash water. In recent years cooling towers, spray ponds and
other mechanical cooling systems have enabled recycling of the condenser
water. Complete recycle of condenser water is difficult to achieve where the
quality of raw water is poor or marginal. Scaling properties increase as
the concentration of solids increases with evaporation and with increased pH
due to ammonia absorption. Certain states permit, under NPDES (National
Pollutant Discharge Elimination System) regulations, the direct discharge of
condenser water into rivers, provided certain conditions are met. These
NPDES permits are being more closely scrutinized as the concern for
environmental quality increases and one could almost predict that these
permits will soon be disallowed.
Due to tremendous volumes of water used in condensers, careful
consideration has to be given when planning treatment programs. Raw water
treatment has become an integral part of the overall water treatment.
However, cost effectiveness and types of treatment must be carefully
evaluated so that the treatment itself does not interfere with later
disposal programs (Figure 2).
Lime Cake
Lime cake, a waste product generated in the carbonatation purification

process, is almost always handled separately from other factory wastes.
Normally this material, mostly calcium carbonate, sugar, organic matter and
99
water, is pumped to a pond by slurrying with water. In the pond, generally

called a lime pond, solids settle out leaving a liquid phase on the surface.
In most beet sugar factories this liquid phase is drained off to separate
lagoons or is combined with the flume water disposal system. Lime cake
supernatant liquid appears to furnish some beneficial nutrients, such as
nitrogen and phosphates to the flume water.
CONDENSER WATER TO :
BEET FLUMES, DIFFUSER,
Dn
OTHER HOT WATER USES, OR
STEAM DISCHARGE
' >C BOILERS I I—l
EVAPORATORS
gia RAW
f \ : ^ * 4 VACUUM PANS t—j I \£jjÊmmA \ WATER
JCONDENSATEK- V / I I ^-^ I I
V TANK J Γ^ II
HEATERS
4 J v^/~i
Fig. 2. Condenser water circuit.
Solid material from lime ponds can be recycled back to the factory and
by re-burning, used again. However the energy requirements for this process
are seldom feasible. There are other uses for this material such as soil
conditioning, especially in acidic soil conditions. Lately this product has
been given serious consideration for S0 2 scrubbing in fossil fuel boilers
(Figure 3).
Miscellaneous Wastes
In addition to waste streams described above, there are lesser

additional ones, which if not controlled can cause severe upsets to the
overall treatment program. Water from boiler blowdown, gas scrubbers,
100
LIME MUD
SETTLING EFFLUENT LIME MUD SLURRY
POND POND JUICE
DRIED LIME MUD TO
■I"' SALES PROCESS
Fig. 3. Lime mud water circuit.
pumps, other equipment, cooling requirements, factory spills and washing,

sanitation and various other sources ultimately end up in the treatment
facility. While not of the volume of flume or condenser water these sources
have to be monitored closely because materials not conducive to treatment
may enter the overall water discharge system,
Composition of Wastes
Because of variations in each factory's volume and composition of its
waters, a "typical" waste water composition as presented in Table 1 should
be used for illustative purposes only.
TREATMENT
Evolution of treatment designs has been mandated by an ever increasing

public awareness of possible pollution sources and the increasing complexity
of groundwater relationships to surface impoundments. A balance between
social, economic and regulatory requirements has forced the American beet
sugar industry to examine, from a grass roots basis, the composition of
materials being discharged.
101
TABLE 1
Wastes from streams 1-4 (réf. 1)
Suspended
Waste water BOD/ton Suspended solids/ton
Stream flow/ton beets BOD beets solids beets
# sliced (gal) (ppm) slices (lb) (ppm) sliced (lb)
1 2600 210 4.5 800-4300 17-93

2 2000 40 0.67
3 90 8600 6.5 120,000 90
4 290 1300 2.8 720 2
Pressure generated by environmental concerns and by the economic burden

of discharging effluent wastes to municipal treatment facilities has created
a worldwide demand for proven, reliable and cost effective methods of
treating organic strength industrial waters. For the past few decades,
wastewater treatment in the American beet sugar industry has been dominated
by aerobic processes using bacteria which require oxygen to metabolize
dissolved organic compounds and convert them into carbon dioxide and
settleable solids. These aerobic treatment ponds were, by and large
successful. Careful attention to BOD loading is required as these systems
operate on a very fine microbiological balance. Adjustments to trace
nutrients available to the system are often necessary.
As the environmental picture started to change, major changes were made
in waste water planning. Generally, volumes were reduced by recycling and
process steps initiated in an attempt to decrease the amount of organic
loading sent to the waste water system. Energy costs also increased which
provided further incentives to reduce overall volumes. However, as these
volumes were being reduced, actual loading remained fairly constant as the
water being discharged was more concentrated. Upsets to the delicate
microbiological balance were common and energy costs for aerobic treatment
continued to escalate. This made a new look at discharge and treatment
procedures mandatory.
ANAEROBIC TREATMENT PROCESSES

Anaerobic treatment over the last few years has become one of the most
promising solutions to the waste water treatment dilemma. Hydraulic
102
retention times and sensitivity to fluctuating loads were thought to be a

major handicap for anaerobic type systems. Several systems are now
available, which have overcome the initial problems associated with
fluctuating load and retention times, which may incorporate both an
anaerobic and aerobic process producing an effluent of high quality. By
utilizing anaerobic followed by aerobic treatment a cost effective facility
can be designed.
Some apparent advantages of an Anaerobic/Aerobic system are:
1. Waste streams high in carboydrates and proteins can be purified to
more than 99Z.
2. Energy requirements are small. In some installations energy is
collected and utilized as a major energy source for the treatment system.
3. Sludge produced is stable and volume is generally small.
4. The process is stable, simple to operate and can be designed for
varying loads.
5. Nutrient requirements for the anaerobic process are only about 102
that of aerobic processes. A BOD: N : P ratio of 100 : 0.5 :0.1 is normally
required for anaerobic systems compared to a ratio of 100: 5 : 1 for an
aerobic system (ref. 1)
The basic biochemistry involved may be summarized as follows (ref. 2):
Hydrolysis
Non-soluble organic compounds such as cellulose, starch, proteins and
lipids are hydrolyzed into sugars and amino and fatty acids by enzymes
excreted by acidifying bacteria. This process is rather slow and is
often regarded as the rate controlling step for the rest of the anaerobic
treatment.
Organic acid formation

Hydrolyzed compounds from step 1 are converted into organic acids such
as acetic, propionic, butyric and lactic by acid forming bacteria. Nitrogen
and sulfur groups of amino acids are decomposed into ammonia and hydrogen
sulfide.
Acetic Acid Formation
Organic acids from the previous steps are converted into acetic
acid, hydrogen and carbon.
Methane Formation
Acetic and remaining organic acids are converted by methane
bacteria into methane gas and carbon dioxide.
Steps 2 and 3 are often referred to as the acidogenesis stage while
step 4 is referred to as the methanogenesis stage. While the above may
present an oversimplification of the actual process, it does give the basic
principles involved in an anaerobic treatment system. Hydraulic retention
times required for each step are very dependent upon contaminant loading.
Systems must be engineered for each application and be based on reliable
analysis of effluent flows and loads. Effluents from anaerobic systems may
be further treated in a simple aerobic post-treatment pond to produce a
final effluent with more than 99Z of organic compound removed.
Several anaerobic processes have been developed and utilized in the
beet sugar industry (ref. 3) for waste water treatment and may incorporate
both acidogenesis and methanogenesis or the methanogenesis stage alone.
The ANAMET System, developed in Sweden by Sorigona AB and the Swedish
Sugar Company (ref. 4) is currently being used at several American beet
sugar factories. In these plants, acidogenesis and methanogenesis take
place in one digestion vessel. The digester is totally enclosed and
utilizes a side mounted agitator to insure continuous mixing with incoming
influent material. Retention time varies between two and fifteen days
depending upon the loading. The digester may be used for the methanogenesis
stage only, thus giving a factory the option of first utilizing anaerobic
stabilization ponds and allowing for more flexibility in water utilization.
Control of pH and temperature is of major importance to assure the
bioactivity necessary and must be taken into account when ponds are used as
part of the overall anaerobic treatment system. After discharge from the
digester, the mixture of water and biomass is separated in an external
sludge settling device and the concentrated sludge returned to the digester.
Biogas is collected in the roof dome and either vented or used as a fuel
source. Effluent may be further treated to a very high standard under
aerobic conditions.
iCfc
There are several other manufacturers of anaerobic systems which offer

alternate methods of treating waste water, either by separating the two
stages, or utilizing existing equipment for one stage and adding the
equipment necessary for the second stage. Table 2 shows typical Anamet
process data from treatment of some waste waters in the beet sugar industry.
(ref. 4) (Figure 4).
SOLID WASTE
WASTE
I EXTRACELLULAR ENZYMES
WATER
] [
ï
SOLUBLE WASTE BIOSLUDGE
ACH) FORMING BACTERIAS
ORGANIC ACIDS ♦ NH3 ♦ H 2 S
METHANE FORMING BACTERIAS l r ^

] t
METHANE ♦ CARBON DIOXIDE
X
I AMMONIA REDUCTION
1
I
REMAINING ORGANIC WASTE ♦ O j * N ♦ P
]
± ] [
Ì
CO 7 H20
] t CLEANED WASTE WATER BIOSLUDGE
Fig. 4. ANAMET process.
TABLE 2
Results of Anamet treatment to beet sugar waste water.
Sugar factory
waste
Raw Material Processed 6,000

Waste water flow, gal/d (1/d) 720,000 (2,979,000)
BOD7, mg O2/I 4,000
BOD7, tons 02/day 12
Methane Production, ft3/d (m3/d) 134,196 (3,800)
Stable sludge prod gal/d, 101 d.m. (1/d) 2,270 (8,600)
Electrical power comsumption, kWh/d 1,500
BOD7 after Anamet Process, mg 0a/l 40
BOD7 reduction in Anamet Process X 99
BIOAUGMENTATION
Since the beet sugar industry has in the past few years primarily used
anaerobic processes to accomplish the major reduction of organic loadings to
its waste water system, one major problem has arisen: odor (refs. 5, 6).
To effectively illustrate this point, following is an example of complaints
written to a local newspaper in a community where a beet sugar factory is
located: "I'm a visitor in world capitals - like Clinton, Iowa being the
Walleye Capital, and a town in Louisiana being the Crawfish World Capital.
(...) could also be a World Capital: the World's Capital of Stench. I've
been all over the world, but never in a place that stinks worse." This
community also included a good size stock feed lot so not all the blame can
be placed on the beet sugar factory; however, for those who have been
associated with the waste water facilities at beet sugar factories, these
articles are a reminder of what can be.
The naturally occurring micro-organisms in an anaerobic waste water
system produce hydrogen sulfide (H2S) as a by-product of their metabolic
process. In high levels, hydrogen sulfide poses serious health hazards, is
highly corrosive and has a negative impact on the health of the entire
biomass. HÄS is considered the most common source of community odor
complaints, although studies have shown that due to the complex mixure of
water being treated, odors other than hydrogen sulfide are generated. Odors
from lime settling ponds seem more severe than those from flume mud settling
ponds. This may be because lime settling ponds have much slower settling
and drying rates. The study further points to a correlation between odor
and pH and odor and total dissolved solids. Classification of odors emitted
from sugar beet waste waters indicate that certain mercaptans and methane
also contribute to the overall odor emitted.
Numerous techniques have been employed to combat hydrogen sulfide,
especially in recycled water being used in flumes and areas inside the
factory. Some of these techniques include:
1. Addition of sufficient oxygen to the water to eliminate anaerobic
conditions which produce the hydrogen sulfide.
2. Prechlorination of flume water which helps in two ways:
a) Chlorine reacts with hydrogen sulfide to form nonodorous compounds.
b) Chlorine is toxic to all bacteria; it kills those which produce the
hydrogen sulfide as well as many others.
106
3. Use of strong oxidizing agents such as hydrogen peroxide and potassium

permanganate.
4. pH adjustment. At high pH levels the dissolved concentration of
hydrogen sulfide is low; pH adjustment is usually accomplished by
adding lime to the water.
5. Use of biocides. Eliminates hydrogen sulfide by killing all microbes.
Some of the above have one thing in common: they treat a symptom by
killing microbes which produce hydrogen sulfide. These same microbes enable
organic material to be converted in the acidogenesis and methanogenesis
processes described earlier, not a desirable effect for the waste treatment.
As some control of flume water systems is almost always necessary, the
selection of that treatment should be conducive to the overall treatment
program.
Within the last five or so years new biological products have been
tried at several beet sugar factories (bioaugmentation), both in the United
States and abroad. These products consist of specialized bacterial strains
which have the ability to utilize dissolved hydrogen sulfide in their
metabolic process and which do not produce hydrogen sulfide as a by-product.
Complete water systems, both in-house and in waste lagoons may now be
utilized for anaerobic digesters without the odor from hydrogen sulfide
posing a health hazard and source of community complaints.
Mazer Chemical markets one such product, called MAZON BIO-ACT (ref. 7).
This product is a blend of naturally occurring facultative anaerobic
bacteria which thrive under anaerobic conditions. This facultative property
allows the product to thrive under anaerobic conditions and also under
aerobic conditions up to 2 ppm dissolved oxygen. BIO-ACT can help reduce
odors caused by hydrogen sulfide because it consumes hydrogen sulfide in its
metabolic processes and incorporates the sulfur-containing compound into its
cell walls.
The claimed benefits are:
1. Elimination of complaints due to hydrogen sulfide.
2. Reduction of soluble hydrogen sulfide in wastewater.
3. Elimination or reduction of the need for special chemicals, such as
bleach or potassium permanganate to oxidize the hydrogen sulfide.
4. Power savings by reducing aeration required to oxidize hydrogen sulfide.
107
5. Help in meeting the BOD and Total Suspended Solids standards for water
discharge.
6. Lowering effluent BOD and Total Suspended Solids for surcharge savings.
7. Reduction of fats, oils and greases.
8. Improved recovery from organic or toxic shock loads that upset a waste
water treatment plant.
9. Improvement in overall plant efficiency, particularly in the cold months
when bacterial metabolic rates are slowed down significantly.
FIELD EXPERIENCE
The effectiveness of utilizing specialized bacterial strains for
organic sulfur compounds is documented in several instances (ref. 7). Due
to extreme climatic variations encountered by the American beet sugar
industry, use or design of any treatment sytem must be tailored to meet
local specific needs. In parts of the Mountain and Plain states, factories
start their operation in the fall when temperatures may range from the mid-
thirties to upper seventies. Operations continue through much of the winter
where temperatures drop to forty below zero. Waste water treatment in
environments as severe as these increase the complexity of designing systems
which will achieve desired goals. For anaerobic systems to be effectively
utilized, water temperature is critical. Maintaining this temperature often
creates prohibitive energy costs, or prevents the facility from treating all
the waste water. Ponding a significant amount of this water for later
treatment is quite common. Major problems are encountered when these ponds
start to thaw in the spring. Bacterial action has been essentially dormant
and the water is almost completely oxygen depleted.
A test of bioaugmentation was conducted at a facility where conditions
were similar to those described above. The hydrogen sulfide level was
recorded when the pond first thawed and a shock treatment of BIO-ACT (ref.
7) added by distributing around the perimeter of the pond. This was
followed by the recommended daily maintenance dosage. Daily dosage of
BIO-ACT was maintained until the hydrogen sulfide level remained at 1 ppm
for a two week period. Monitoring continued and additional feed was added
if levels of hydrogen sulfide indicated the necessity. After the spring
thaw, water in the pond was kept in an anaerobic state for approximately two
months, then utilized for irrigation. The final test for the treatment was
108
to confirm with local city officials that the number of odor complaints they
received from the community had tapered off. This was confirmed. A review
of the local newspaper showed that during the spring and early summer no
letters to the editor were received protesting the "stench." It should be
noted that the trial was conducted at the same location where the two
excerpts from the paper were earlier reported.
A case history of sugar factory waste water treatment using BIO-ACT as
reported in technical literature published by MAZER indicates similar
results (Table 3).
TABLE 3
Case history of sugar factory waste water using Bio-Act (ref. 7).
Inoculation level One Time Shock

Maintenance dosage 5 gal. per day (GPD) (19 1/d)
Storage capacity of flume system 40 mill. gal. (151 mill.l)
Flow rate of flume system 1 mill. GPD (3.8 mill, 1/d)
BOD range 1000-4000
H*S prior to treatment 25 ppm
Fecal bacteria count prior to treatment 100,000
H2S after treatment 0-1 ppm
Fecal bacteria count after treatment 3,300
Derived Benefits:
1. Odor complaints were eliminated.
2. Soluble H2S was reduced to 0-1 ppm.
3. Fecal bacteria were reduced to 3,300
4. BOD was lowered so that 32 mill. gal. (121 mill. 1) of water could be
discharged to river.
ENVIRONMENTAL CONCERNS
Environmental requirements have become, or are quickly becoming, the

deciding factor in decisions focused on waste water treatment. Groundwater
contamination regulations and the public's concern for exposure to
contaminants which may affect their well being have far reaching level and
financial implications to every manufacturer in the United States today.
Each facility must, if they plan to survive, anticipate and plan for the
future with regard to their respective waste water facilities. Treatment
systems with effluents 99Z pure may not be adequate when certain chemicals
are present in that effluent and the material is going to unlined storage
ponds.
An example of the shift in regulations is exemplified by the new
California Safe Drinking Water and Enforcement Act, "The Act." After
reviewing this act it is easily understood why treatments must be designed
to meet state and county requirements.
THE ACT
Proposition 65, the Safe Drinking Water and Toxic Enforcement Act of
1986, was overwhelmingly adopted by California's voters on November 4, 1986.
An initiative statute, Proposition 65 represents one of the toughest toxic
statutes ever enacted by any state or the federal government. These new
legal rules dramatically shift the burden to businesses using chemicals to
demonstrate that their products, process and wastes are safe for human
exposure. This initiative may be the most difficult initiative in
California's history to implement, simply because it depends so heavily upon
science for its basic operation. The initiative demonstrates a policy shift
from a balanced effort of trade offs in using chemicals in today's society,
i.e. as reflected in the cost/benefit risk management approach used in
current regulations to a more aggressive policy of human health as a higher
priority. There is no provision in the initiative to balance the economic
costs and health benefits. To accomplish this "higher priority" the act
enunciates four new rights:
1. "No discharge unless safe." Protect the people and the water they drink
against chemicals that cause cancer, birth defects or other
reproductive harm.
2. "Right to know." To be informed about exposures to chemicals that cause
cancer, birth defects or other reproductive harm.
3. "Right of private citizen to sue alleged violators." To secure strict
enforcement of the laws controlling hazardous chemicals and deter
actions that threaten public health and safety.
4. "Bounty hunter." To shift the cost of hazardous waste cleanups more
onto offenders and less onto law-abiding tax payers.
The "no discharge unless safe" prohibition establishes a mandate that

no state-listed chemical may be discharged which may enter into a source of
110
drinking water, unless it can be shown by the discharger that the amount of
chemical discharged is safe. A bill pending before the California
legislature now would declare that all ground water is to be considered as
potential drinking water. The initiative defines a "significant amount" as
any "detectable amount," does not allow for standards established by FDA to
be used, and places the burden to prove that a chemical is safe on included
chemicals which are common to naturally occurring substances, such as
limerock.
Although the above is only a brief recap of the initiative, it does
show how complex waste water treatment may become in the very near future.
In the past, regulations passed in California have later been adopted by
other states, and in some cases, by EPA. This may be an indication of what
is to come, and makes treatment of the beet sugar industry's waste water
even more complex.
REFERENCES
1 R.A. McGinnis, Beet Sugar Technology, 1982, 677-688.
2 BIOPAQ Waste Water Treatment Systems. Technical Literature.
3 M. Shore, N. W. Broughton and N. Bumstead, Anaerobic Treatment of Waste
Waters in the Beet Sugar Industry, Presented to Institute of Water
Pollution Control, March, 1983.
4 AB Sorigona, The ANAMET Process. Technical Literature.
5 F.A. Brunner and Sverdrup & Parcel and Associates, Inc., Investigation
of Sugar Beet Waste Odors. Presented to 20th General Meeting of the
American Society of Sugar Beet Technologists, San Diego, CA, 1978.
6 J.W. Davey, J.F.T. Oldfield, D. Van Dover and L. Falkman, New Paths to
Odor Control, Presented to the 24th General Meeting of the American
Society of Sugar Beet Technologists, Phoenix, AZ, 1987.
7 Mazer Chemical, Technical Bulletin, Mazer MAZON BIO-ACT.
Chemistry and Processing of Sugarbeet and Sugarcane, edited by M. A. Clarke and M. A. Godshall 111
Chapter 7
ION EXCHANGE FOR DESUGARING OF MOLASSES AND BYPRODUCT ISOLATION
L. H. RAMM-SCHMIDT
INTRODUCTION
There are several traditional ways to reduce the sugar content in beet
molasses. These processes are mainly based on ion exchange, where the
sodium and potassium ions are substituted with less melassigenic ones, e.g.
magnesium and calcium. A more modern and efficient process was developed
some 15 years ago based on ion exclusion (the Finnsugar-Pfeifer & Langen
process). Commercial processes are in operation in Finland, Germany and
Belgium. When sugar prices are low, the economics of this process is
dependent on local conditions, e.g. toll regulations and the price level of
domestic sugar production. The profit of the Chromatographie process may,
however, be increased by isolating other compounds from the molasses besides
the sugar. Such compounds may be different amino acids. "Tailor making" of
molasses used for yeast production has also been done resulting in higher
yeast yield and lower pollution. Some important nutrients for the yeast
growth may also be isolated from previous separations and added to the
fermentation molasses. All these things have been practiced in the Finnsugar
Naantali desugaring plant during its 10 years of troublefree operation.
BACKGROUND OF THE DESUGARING PROCESSES

The fairly high sugar content left in the molasses after
crystallization has encouraged the sugar producer to look for methods to
extract more sugar out of the molasses in order further to increase the
yield and profits of the total process. There are two major ways to do
this: either by ion exchange, where the aim is to increase the
crystallization yield of the main processes, or by direct desugaring of the
molasses.
Ion exchange
The first attempts were made in the Netherlands during World War II
with cation exchange resins. In this case the sodium and potassium ions in
the B-sugar juice were exchanged for calcium ions, using a resin bed in
hydrogen form followed by an immediate neutralizing of the sugar juice with
milk of lime. The less melassigenic calcium ions increased the extraction
112
of sugar in the C-boiling. An improved treatment was achieved using

magnesium ions. This led to development of the Quentin process. The
regeneration of the resin bed is accomplished with magnesium chloride.
Another application of ion exchange is to demineralize the thin juice in
mixed bed filters. As part of the melassigenic alkali ions are removed,
yield increases in all crystallization steps. A great variety of other
methods have also been developed, some of them described in Chapter 3.
Desugaring of molasses
One process is the Steffen sucrate treatment, where the sucrose is
precipitated as a calcium sucrate, which is used instead of lime in the
juice treatment in order to reduce chemical costs.
The newest alternative for molasses desugaring is the Finnsugar-Pfeifer
& Langen Chromatographie separation process. Molasses is separated into two
fractions: one contains the main portion of the sugar in the molasses and
the other the nonsugars. Though the adsorbent used is a cation exchange
resin, it is not used for ion exchange, but as an adsorbent for molecular
fractionation only. No regeneration chemicals are needed for normal
operation. The phenomenon which is based on ion exclusion and the so called
partition chromatography or molecular sieve phenomenon has been described
earlier by Schneider (ref. 1) and the separation process as a whole by
Hongisto & al. (refs. 2,3).
Comparison of the processes
Comparison of the different processes reveals the following:
Demineralization in general:
- Low yield
- Corrosion problems
- High use of regeneration chemicals
- Heavy waste water load
Quentin process:
- Low yield
- High quality requirements on the regeneration chemicals (MgCl2) (see
Chap. 3).
- Heavy waste water load
- High Mg content in the molasses may adversely affect the utilization and
selling price of the molasses.
Steffen process:
- Molasses has to be diluted

- Raffinose is recirculated together with the sucrate affecting
the crystallization yield. Thus a portion of the molasses has to be
withdrawn untreated. Enzymatic hydrolysis of the raffinose may be applied
to avoid this problem.
- The process does not work with cane molasses.
Finnsugar-Pfeifer & Langen process:
- Molasses has to be diluted

- The process can be in operation independent of the campaign
- Cane and refinery molasses may also be treated
- Special separations with isolation of byproducts and nutrients
can improve profitability of the process
- No regeneration chemicals are needed
- Low waste water production
A closer comparison of the processes has been made by Hongisto e_t al.
(réf. 4 ) .
MOLASSES COMPOSITION
Molasses is a multicomponent liquid. Composition varies with type and
origin: beet, cane or refinery molasses, and also from soil type, crop
maturity and crop fertilization. In Finland where the beet growing season
is probably the shortest in the world, we have experienced a much higher
amino acid content in the molasses compared to more Southern European
molasses. The average composition of beet molasses as well as residual
molasses from the separation is shown in Table 1.
LARGE SCALE CHROMATOGRAPHY
Because the basic principle of molasses desugaring in the

Finnsugar-Pfeifer & Langen process is Chromatographie, all components in the
molasses are removed from the column in different ways. Components that
have lower affinities for the resin move faster, along with the eluant phase
and are removed first at the bottom of the column. Components with strong
affinities are retarded and move slowly. The molecular weight of the
component also affects the behavior in the column. In this case high
11Λ
XABLE 1
Composition of beet molasses and desugarized beet molasses (expressed as
percentage of dry substance).
Residual Molasses
Higher sugar Low sugar

Beet Molasses content content
SUGARS 66.5 21. 10.

Sucrose 63.5 18. 6.
Raffinose 1.5 1. 1.
Other Sugars 1.5 2. 2,
OTHER ORGANIC COMPOUNDS 23.0 54.5 62.0
Glutamic acid + pyrrolidone 4.0 9.4 10.8

carboxylic acid
Other amino acids 3.0 7.0 8.1
Betaine 5.5 15.0 17.8
Organic acids 5.5 13.0 14.9
Pectins and others 5.0 9.6 10.4
INORGANIC COMPOUNDS 10.5 24.0 28.0
K20 6.0 14.0 16.1

Na 2 0 1.0 2.3 2.6
CaO 0.2 0.6 0.7
MgO 0.2 0.4 0.5
AI2O3 Fe 2 O a 0.1 0.25 0.3
S1O2 0.1 0.25 0.3
Cl 1.7 3.4 3.9
S02 + S0 3 0.5 1.2 1.4
P*05 0.1 0.2 0.2
N205 0.4 0.9 1.0
Others 0.2 0.5 1.0
molecular weight components passing through the resin bed are excluded from
the resin pores and travel through the column faster than the low molecular
weight components. Fractions are thus eluted in the following order: First
the salts and the large molecular weight substances including some coloring
agents, soluble gums, waxes, proteins and amino acids, followed by the
oligo-, di- and finally the monosaccharides.
Feed H.O
Molasses
Resin
Product Fraction
Recycle
Nonsugars
Nonsugars Recycle Product

Fraction
Fig. 1. Separation scheme for desugaring of molasses in a Chromatographie

column. (I, II and III are consecutive batches.)
The best separation result is achieved when the flow pattern in the
column is as close to plug flow as possible. This can be achieved with a
relatively small resin particle size and an appropriate flow rate. Resin
with a narrow coefficient of variation is essential. The feed device and
the bottom construction must be such that back mixing is minimized. In
Figure 1, the separation scheme in a column is shown and in Figure 2 the
Chromatographie separation curve for beet molasses is shown. By controlling
the outlet valves in Figure 1, it is possible to change the composition
purity of the collected fractions according to their intended use.
116
Tlme/mln
Fig. 2. Chromatographie separation curve for beet molasses separation.

(1 = nonsugar; 2 = raffinose; 3 = sucrose; 4 = monosaccharides; 5 = betaine;
6 = total concentration.)
Intermediate fractions are normally collected for dilution of the raw

molasses or for recycling to the top of the column. The total separation
scheme for consecutive feed batches is shown in Figure 3.
USE OF DIFFERENT FRACTIONS

Sucrose rich fractions
The sucrose rich fractions (also called product fractioni from the
separation may be treated in two different ways, producing either liquid
sugar or granulated sugar. When producing liquid sugar * further
purification is carried out by demineralization and decolorization in ion
exchange columns followed by polishing with activated carbon. When
granulated sugar is to be produced there are several possibilities. During
the beet campaign the product fraction can be mixed with raw juice before
juice purification, or it can be evaporated and directed to A or B boiling,
depending on the product fraction purity chosen in the separation. Outside
117
3000
I I 24.0 « -I 10« »
DA B C DA
FroclionoMon Sample number, templing interval 5 mm. AB Rei· molostes
B-C Recycle 1
C D : Product
D-A Recycle 2
Fig. 3. Fractionation of molasses by Chromatographie separation,

consecutive batches.
the campaign the product fraction may be stored in the same way as thick
juice. Long storage time requires ample storage capacity. If it is not
available, the storage can be compensated for by short "intercampaigns"
outside the normal campaign.
Residual molasses fraction

The major byproduct from the desugaring is certainly the residual
molasses fraction. Though this fraction is low in sugar content, its energy
value is close to that of untreated molasses because of other organic
material. The enrichment of proteins and other nonsugar matter will
increase its nutritive value to even greater than that of normal molasses.
There are many possible uses for residual molasses;
118
- Direct use as animal feed in different molasses mixtures.

- Mixed with beet pulp, brewers grain and husks.
- Use as a fertilizer.
The best use in different cases is certainly the one that gives the highest
selling price for the residual molasses.
Direct use as animal feed in different liquid molasses mixtures. The
simplest way to use residual molasses would be to use it as is for animal
feed purposes. The low sugar and high potassium contents often restrict
this use though animals in several cases have got used to plain residual
molasses. Local fodder legislation also puts limits on sugar content and
ash. Therefore, mixing it together with other molasses or molasses type
material is a better solution. In Finland, we make a certain molasses mix
using imported cane molasses and vinasses, refinery molasses, different wood
molasses from the paper and pulp industry and wheat molasses from the wheat
starch industry. Some of these special molasses have a high content of
reducing sugars elevating the total sugar content of the mixture to the
required level. In this case, the selling price for the residual molasses is
the same as for untreated beet molasses. If the salt content is still too
high, it is possible to crystallize out part of the potassium salts as
potassium chloride or potassium sulfate from the residual molasses.
Mixed with beet pulp, brewers grain and husks. In Finland, the major
part of residual molasses is used for the production of low sugar molassed
beet pulp. When the sugar content in normal molassed beet pulp is about
20Z, this product has a sugar content of 10Z. The name for this product is
"molassed beet pulp 10" (MBP 10). An analytical comparison is shown in
Table 2.
"Molassed Beet Pulp 10" is sold in bulk or pelletized. A new type of
molassed beet pulp with sugar content of only 7.5% has also been released.
The Finnish Agricultural Research Center has performed extensive tests with
the nutritional value of residual molasses and its suitability as fodder for
domestic animals. The results achieved have been very good. The advantages
improved digestibility, palatability and higher crude protein content. The
texture of the pelletized type is also softer compared to normal pelletized
molassed beet pulp. Thus, the product has been sold for the same price as
the traditional type. The molassed pelletized beet pulp is also superior to
119
TABLE 2
Analytical comparison of normal molassed beet pulp and "molassed
beet pulp 10."
Normal molassed beet pulp MBP 10
Sugar 20.72 1 0 . OX
Protein 12.42 15. 42
Ash 8. 32 1 2 . OX
Water 7.0% 10.02
Other org. 51. 62 52.62
substances
Total 100.02 100.02
plain pelletized beet pulp because the pellets are stronger (thus the
product contains less fines), volume weight is higher and protein content is
also higher. Tests have been carried out in Finland also on brewers grain
and husks. These mixtures have given excellent animal feeds.
Use as fertilizer. If there is no market available locally for the use

of residual molasses as animal feed, it can still be used as fertilizer.
Residual molasses contains all the minerals and humic matter the plant has
taken from the soil in the same proportions, thus these can be returned to
the fields by irrigation. In this case, the separation should be controlled
so that the sugar content in the residual molasses is minimized (below 1 0 2 ) .
Experiments have been made in Finland with fertilization of beet fields
showing that the beet yield per hectare was higher with residual molasses
fertilization compared to general fertilization. The selling price for
residual molasses as fertilizer is naturally lower than that for animal feed
use.
Special fractions
Special fractions may be isolated from the Chromatographie separation
either for "tailor making" of molasses used in the fermenting industry (e.g.
yeast, alcohol and citric acid) or for the industrial production of pure
individual components present in the molasses. Because the separation
process is controlled by a microcomputer, it is easy to program it so that
certain fractions are directed to separate collection tanks.
120
Fermenting fractions. Some of the components in the normal molasses do

not ferment and are thus left in the bottoms after the fermentation. These
compounds often cause waste water problems and lower the total yield of the
fermentation. Another problem with normal molasses in the yeast industry
can be high color and salt, e.g. nitrites. Reduction of color bodies and
salts would improve the yeast quality and color. Normal molasses contains a
certain amount of nutrients necessary for yeast growth. Such compounds are
inositol, pantothenic acid and biotin. These substances may be collected
from the separation and added to the fermentation molasses or sold as
special "nutrient packages" to the industry. "Tailor making" of a molasses
lot for yeast production could be done in the following way:
- Part of the nonsugar fraction where the salt content and color
is highest is removed.
- Yeast growth-promoting nutrients collected from earlier
separations are added to the molasses.
- Part of the sugar peak could be removed to adjust the sugar
content to an optimum. The sugar content is otherwise elevated
due to removal of the nonsugar fractions.
According to Finnish experience the yeast yield of this "tailor made"
molasses calculated on the original molasses dry solids is the same or
higher despite the fact that a certain amount of the material is removed.
Thus, the selling price for such molasses is correspondingly higher than for
standard molasses. If the yeast producer buys a lot of standard molasses he
could have it specially treated at a desugaring plant for the cost of the
dry solids removed. These solids would serve as a payment for the
treatment. The very pure sugar fraction is used for sugar production and
the nonsugar components are added to the bulk of residual molasses.
Individual components. The Chromatographie separation process can also
be used to isolate single components from the molasses. To obtain a pure
fraction several consecutive separation steps have to be made; pure
raffinose and some amino acids have been produced in this way. The
profitability and market of such products have to be checked thoroughly in
each case.
121
ECONOMIC ASPECTS
The economics of the desugarization process is highly dependent on
local conditions. In sugar exporting countries the profitability is not
good as long as the sugar world market price is low and the price margin
between molasses and sugar is very narrow. But in most sugar importing
countries, with domestic production covering part of the consumption, toll
regulations and legislation may change the price situation totally. Local
farming politics normally creates a need for different "umbrella prices."
The situation is also highly dependent on the use and price level of
molasses. The profitability of a separation plant may, however, be greatly
increased by doing special separations. A thorough cost evaluation should
be performed in each case. In Finland, the running of the Naantali
desugaring plant has so far been very profitable. As long as the sugar
world market price is low, only domestic molasses have been desugarized.
During periods of high sugar world market price (e.g. 1980-1981) imported
molasses were also processed for re-export of the sugar.
EXPERIENCE WITH NAANTALI DESUGARING PLANT

General
The plant has been in service since 1978. The standard capacity is
125 tonnes of molasses per day, giving an annual capacity of about 40,000
tonnes. Due to special separations and periodical shortness of domestic
molasses, utilization has been 60-90Z annually. The amounts of molasses
processed and product fraction produced annually are shown in Figure 4. The
performance of the separation process has been very stable during many
years. A slight reduction in column capacity was realized during the first
years due to resin aging, but since then values have been almost unchanged.
Typical data representing present performance are shown in Figure 5. Total
recovery of sugars (granulated, bag basis) extracted from the molasses has
averaged 70Z. An average of 15Z goes into the residual molasses and the
final 15Z into the secondary molasses.
122
PRODUCT
FRACTION
MOLRSSES
Fig. 4. Molasses processed and product fraction produced by Naantali plant

on a dry solids basis.
Special separations
The following special separations have been made:
- tailor making of molasses for yeast production
- raffinose separation and purification
- isolation and purification of different amino acids and nutrients
(pantothenic acid, biotin and inositol)
- removal of potassium chloride and potassium sulphate
from residual molasses in a forced circulation evaporator.
Technical aspects
A list of the main equipment is presented in Table 3.
Pretreatment. Softening of the molasses to a calcium content of 0.1Z
has been done with sodium carbonate and filtration using a small amount of
filter aid (diatomite or perlite). Average consumption expressed as
123
percentage of molasses dry solids has been 1.5% sodium carbonate and 0.65%
filter aid.
TABLE 3
List of the main equipment in Naantali desugaring plant.
Pretreatment filtration:
- 1 Seitz Sirius pressure filter
Separation:
- 8 Columns, resin volume 35 m3 each

- 1 Column, resin volume 60 m3
- Total resin volume 340 m3
Thin liquor evaporation:
- 1 Five-stage Rosenlew LTV-rising film evaporator

- Residual molasses evaporated in stages 1-2-3
- Product fraction evaporated in stages 4 - 5
Residual molasses final concentration:
- 1 one-stage FC-evaporator
Process tanks:
- 14 stainless steel tanks 30 m3 each

- 8 smaller tanks 2-12 m3
Process pumps:
- Standard centrifugal pumps

- 2 Mono-pumps
Control system:
- Separation: 9 microcomputers controlled by one host computer

- Process control: Damatic distributed digital process automation
system.
1*
PREHEATED BEET MOLASSES
Sucrost 61 %
Nonsugirs 39 %
Dry solids 100 %
PRODUCT FRACTION NONSUGAR FRACTION

Sucrose 52 % Sucrose 9 %
Nonsugars S% Nonsugars 34 %
Dry solids 57 % Dry sol Ids 43 %
SUGAR IN THE BAG SECONDARY MOLASSES

Sucrose 43 % Sucrose 9 X
Nonsugars 5 %
Ory solids 14 %
Fig. 5. Typical performance data for the Naantali desugaring plant.
Separation. Specific separation capacity has been 300 kg molasses

3
d.s./m resin/day. No regeneration of the resin since start-up has been
necessary. During service, the resin reaches an equilibrium with sodium,
potassium, calcium and magnesium ions, but provided the pretreatment is done
properly, the amount of harmful divalent ions is kept low enough.
Backwashing of the resin bed has been done at 4-8 weeks intervals. The
resin life was originally expected to be about 10 years, but the practical
life seems to be much longer. About 30Z of the resin has been replaced
mainly because of gradual resin loss. Capacities have been unchanged for
the last 8 years.
Evaporation. Thin liquid concentrations have been: product fraction
18-20Z dissolved solids basis; residual molasses 7-8Z dissolved solids
basis.
The evaporators have worked almost troublefree without any serious
scaling problems. Chemical and water jet cleaning is done once a year.
Crystallization. The product fraction has been processed during the
beet campaign or crystallized outside the campaign to beet raw sugar for a
second crystallization at Porkkala refinery. Double crystallization at
Naantali factory has also been done producing a high quality granulated
white sugar.
Waste water production. A loss of about 2.5 tonnes of dry substances
per week to the sewer has been measured. This corresponds to 0.3Z of the
feed molasses dry solids. No other waste water load is produced by the
separation process.
Personnel. The total desugaring process is run by 15 persons in the
following way:
- 2 shift operators on a five shift basis
- 1 day man
- 1 laboratory assistant on a two shift basis
- 1 instrument technician
- 1 supervisor
CONCLUSIONS
It has been shown that the Chromatographie molasses separation process
is superior to traditional desugaring processes, mainly due to a high sugar
yield, but also due to the fact that special fractions simultaneously may be
isolated from the molasses either for further purification, or for "tailor
making" of molasses used in fermentation.
The residual low sugar fraction may not be regarded as waste, because
it can be utilized in several ways for the same selling price as normal
molasses.
The Finnsugar Naantali desugaring plant has been working for ten years
without any significant reduction in capacity though the replacement of
resin has not been necessary to the extent originally estimated. Thus, the
profitability of the plant has been satisfactory.
126
REFERENCES
1 H. G. Schneider, Ion exclusion in cane sugar refining, Proc. Sugar
Industry Technol., 37 (1978) 110-131.
2 H. J. Hongisto, Chromatographie separation of sugar solutions; the
Finnsugar molasses desugarisation process, International Sugar
Journal, 79 (1977) 100-104, 131-134.
3 H. J. Hongisto and P. Laakso, Application of the Finnsugar-Pfeifer
& Langen molasses desugarisation process in a beet sugar factory,
Paper presented at the 20th Meeting of the A.S.S.B.T., San Diego,
1978.
4 H. J. Hongisto, and M. Loisa, Finnsugar-Trennverfahren zur Melasse-
Entzuckerung, Zeitschrift für die Zuckerindustrie, 27 (1977) 279-
283.
Chapter 8
ULTRAFILTRATION USED IN A NOVEL FLEXIBLE PROCESS TO PRODUCE HIGH FRUCTOSE

SYRUP FROM DIFFERENT RAW MATERIALS
T. R. HANSSENS AND K. KOERTS
ABSTRACT
The economic production of high fructose syrup from plant tubers

containing carboydrates, such as chicory, Jerusalem artichoke or sugarbeet,
in a flexible process, is the aim of an elaborate research program at SUIKER
UNIE of the Netherlands.
The process will make use of recent developments in membrane filtration
for juice purification, such as inorganic ceramic membranes.
A brief description of the new process is given in this paper.
Experience with ultrafiltration systems and their economics are outlined.
INTRODUCTION
SUIKER UNIE is a cooperative society of sugarbeet growers that

processes the bulk of Holland's harvest. The amount totals approximately
4,500,0,00 tonnes per year, representing 700,000 tonnes of white sugar.
At present the sweeteners used in the European Economic Community (EEC)
are dominated by sugar from sugarbeets. Relatively small quantities of
starch based sweeteners (excluding dextrose) and synthetic sweeteners are
marketed within the EEC, used mainly for dietetic purposes. The reason for
the difference from the USA is the market regulations of the EEC. It is
expected that this might change in the next decade or by the turn of the
century at the latest. However, there is now a need to look into additional
crops, since a limited number of crops such as sugarbeets, potatoes and
wheat are not sufficient for the farmers, because of the heavy load on the
soil. Jerusalem artichoke (Helianthus tuberosis L.) is considered a
promising crop by agronomists and has been under study for several years
(ref. 1).
PROCESS
SUIKER UNIE has developed a new process (refs. 2,3) which produces
sweetener from different raw materials. Let us consider the Jerusalem
artichoke (JA), a rich source of inulin, a polyfructan, as an example. In
October the first tubers are harvested and sent to the factory (see Figure
1). The tare, or attached soil and debris, of the raw JA has to be removed
128
in a washing process. The cleaned JA is grated in a mill and the pulp is

then sent to the juice separation section.
raw Irieoept l on
s
N Ml 1 1 ine mm^mr^m t, 1 o n
I V V
material *
^7
MUltlpl·- Ultj-ei-
•ff#Ct- A. EX» 1 o n i z u t l o n Λ
f i l t r a t l o n
m v a p o r · t o r * · VI Vj
f r u c t o s e syrup
Fig. 1. Scheme of the process.
The juice/pulp separation section consists of a number of solid-bowl

decanter centrifugals in series, to expel the more or less dry solids at one
end and the juice at the other. To obtain an acceptable loss of inulin, the
main stream of solids is counter-currently washed with water or diluted
juice.
The purification of the juice is done in two steps, the first being
ultrafiltration and the second deionization. The deionization process is
rather straightforward, but should also include the recovery of the spent
régénérant chemicals and the "non-sugars", which will have been absorbed at
the surface of the ion-exchange resins. The recovered dry spent régénérant
chemicals are valuable as fertilizer and the recovered non-sugars as fodder
components. This process has been in operation in Vauciennes (France) for
several years.
129
Recovery of the spent chemicals is necessary for environmental reasons.

By producing fertilizer, the cost of the chemicals is almost canceled out by
income from the fertilizer.
Two options to hydrolyze the inulin exist: hydrolysis using free or
immobilized enzymes, or hydrolysis using acid. The point in the process
where hydrolysis will take place will vary with the technique selected.
The raw juice is converted by the process to a clear waterlike sweet
syrup with 70Z solids, containing 80Z fructose and 20Z glucose. It is also
possible to crystallize the fructose. The product has the advantage that
there is none of the after-taste that is common in starch based fructose
syrups.
Other processes for JA have been very well summarized by researchers in
Canada (ref. 4 ) .
Membrane Filtration
After the separation into carbohydrate-free pulp and raw cell-juice,
which is slightly diluted with water, the juice has to be totally purified.
We will emphasize our experience with ultrafiltration. After
systematic analysis of various systems, we had opted for tubular systems and
organic membranes. Just a year or two ago, ceramic membranes became
commercially available and it was then that we started testing them.
The selection of tubular systems, instead of plate-and-frame, spiral-
wound, or hollow fiber membranes, has been based on the assumption that the
raw juice might still contain particulates e.g. pulp or sand which will
cause plugging. The cleaning of the membrane surface could become
prohibitive in non-tubular systems (ref. 6). Another advantage of tubular
systems is that when a normal cleaning procedure is ineffective because of
severe fouling, mechanical cleaning by sponge balls is possible.
To summarize, the criteria for the selection of an ultrafiltration-
system are:
- permeate must be free of precipitable component
- flux level maximal
- flux decline in operating time minimal
- cleaning time minimal
- membrane replacement minimal
- personnel requirement minimal
130
- energy consumption minimal

- maintenance minimal
- capital cost minimal
This should lead to a system with minimal operating costs. Many
different systems have been examined and some results of our investigations
will be shown.
The ultrafiltration installations

A flow scheme of the ultrafiltration test-rig is shown in Figure 2.
The installation is just one stage and contains the most important elements
for studying the operation of this unit under different operating
conditions. From the results, some of which will be mentioned here, it
should be possible to design the full-scale unit.
-b-i
Ultrmfil-
tretion-
modules
Permeate
Feed
Retentate
Feed
pump Sooeter-
pump
i ;
Cooler
Pressure-
relief-
vilve
Fig. 2. Diagram of the ultrafiltration scheme.

131
One of the elements used to control performance is the combination of

the booster pump and the valve in the retentate outlet. With the booster,
the flow rate of the recycle stream can be adjusted to determine the stream
velocity of the concentrate along the membrane surface. It has been
observed that this influences the flux level and the flux decline with time.
For fluids having a relatively high capability of fouling we have observed
that this velocity should be higher than 2 m/sec; perhaps 4 m/sec is better.
With the pressure-relief valve, the system pressure can be set to a
chosen value, and this influences the flux level to a certain degree. The
valve is also a means of setting the output rate of retentate and, allowing
the possibility of controlling the recovery (also called the concentration
factor) of a stage. Some results are summarized in the graphs in Figure 3.
The graph (3.1) of the average product-flux versus the concentration
factor shows the results of many trials: they form a rather wide band. The
next curves (3.2) of product-flux versus time show the fouling rate at
different concentration factors. The course of the flux is influenced by
the load of the solids-content and by the linear velocity along the
membrane. The lower the load and the higher the velocity, the better the
flux-level will be. A higher velocity will of course give a higher pressure
loss. The clean water flux time-lapse (3.3) is a measure of the permanent
fouling increase. Like many others, we have not observed a relation between
clear water flux and product flux. Graph 3.4 shows the relation of the flux
versus the system pressure. This relation is important for the proper
selection of system pressure; long modules lead to substantial loss of flux,
due to pressure loss, when operating in the bend of the curve.
When designing the full-scale installation, one should know what
recovery is required. The level of recovery is directly responsible for the
loss of carbohydrates. Figure 4 shows such a system. It consists of about
seven stages, in a merry-go-round system, with each stage set at a certain
concentration level. The stage where the feed is introduced operates at the
lowest concentration; the last stage at the highest. Of course the fouling
rate of the last stage is also the highest, so after a rather short time
this stage is shut down to undergo cleaning. A cleaned stage will enter the
introduction-phase and all the other stages will be promoted one stage
higher. Many test runs with the one-stage installation are required to see
the flux-decline with time at different concentration levels. A test
132
program is thus very extensive because there are so many variables, some of
which are uncontrollable, e.g. the variation in the composition of the raw
material, which is shown in tables 1, 2 and 3. Other important variables
are the temperature and pH levels.
Cleaning procedures also formed part of the trials; important
parameters are temperature, concentration and type of detergents and
antiseptics.
Currently most suppliers of membrane systems have computer programs to
calculate an optimal system. In addition to test results, the input into
such a program should include such limiting or boundary conditions as sugar
loss.
Composition of feed, product (permeate) and retentate
Some typical compositions of slightly diluted raw juices from different

raw materials are shown in Table 1. Because of variations in the materials,
even within one raw material, it is necessary to show the figures for each
test.
The dry matter is expressed as the refractometric dry substance (brix),
calibrated with pure sugar solution. The sugar content in sugarbeet juices
is determined by a standard method using polarization. The sugars, or
rather carbohydrates from inulin, are determined first by hydrolysis of the
inulin and then by analysis of the reducing sugars by the Luff-Schoorl
method.
Nitrogen is determined by the Kjeldahl method; sodium and potassium by
the flame-photometric method; calcium and magnesium by complexometric
titration (EDTA).
The compositions of the resulting products or permeate have been
summarized in Table 2; that of the retentate or concentrate in Table 3. The
conditions during the trials are given in comments below those tables. Code
P and C give an indication of the membranes used; polymeric and ceramic
respectively. For all membranes tested, the value of the cut-off rate given
by their suppliers was 20,000 daltons. This has been confirmed in the
laboratory by using HPLC technique on permeates. We have concluded that
there are no differences in quality of permeates. The tested polymeric
membranes were based on polysulfone or polyacrylonitrile.
133
1
Avg.Productflux Product-flu« v * . increase!
vs. conc-fictor time at d i f f e r e n t Cone.- j
\ conc-factors factor
flux in 1/ffiî.h F
\\ ^ 1
1 ^\ \ flux in 1 / m * · h
u
Graph 3 . 1
3U X^^-J
Graph 3 . 2 \ . 2fi
4 β *"^ |
1 1 1 Ì 1 _J !
Ί T ime ( h )
100 ->
Cone T f a c t f c r ( l o g )
1 Product-flux vs.
1 . Graph 3 . 4 1
S 1 system-pressure
40 J flux in l/m*.h
%
ò^f77< 1
y o >:>:>>>
X < < <; <' <; <: < 1
Clean water / » :> :> > :> > :> 1
Graph 3 . 3 / < <. <. <: <. <. <; I
flux vs. time / » ;>.>.>>>.>
/ (. <: <. c <: <. <. 1
/ ·>> > > >>
flux in l/m*.h.bar *J /
/
<; t; <; <; <; <. <.
»>*>>>>>
I
1 (* <* <' < < < <.
/ o..>,>x>>.
1 20
Ί
40
1 Γ
/
/
/
»">">">'>>>
■ (<
>>>>>>>
< < < < <.
\Time (days) Pressure

->
Fig. 3. System c h a r a c t e r i s t i c s .
13^
τ PRODUCT
fri
RETENTATE
FEED
—(D—M- τΜ >
OIA-
-M
/ /i~I
J HxH
ZZ®-*
i
wat
—<D—w i
I /
C.I.P.
I Γ
I i
i r
ii X.
Fig. 4. Flow diagram of a 7-stage ultrafiltration system.
As can be seen from the tables, the purity coefficients of sugarbeet

and JA increase more than 4Z absolute; that is not so for chicory. The
explanation is the difference in the distribution of the molecular weight of
inulin found in JA and chicory (ref. 7). The retention of carbohydrates in
chicory was found to be slightly higher than that of JA.
The permeates produced did not form any precipitate during de-
ionization, fulfilling one of the aims of applying ultrafiltration.
The decrease of nitrogen is probably due to the protein component.
Also remarkable is the drop in calcium content in permeate, which can be
ascribed to pectins removal.
135
TABLE 1
Composition of feed of ultrafiltration.
Sugarbeet Jer. Artichoke Chicory
Dry matter w/w X 14.5 14.9 15.57 14.06 15.7 17.4

Sugars w/w X 12.28 12.9 12.9 12.0 14.5 14.9
Nitrogen w/w X 0.10 0.13 0.14 0.15 0.15 0.14
Purity w/w X 84.7 86.6 85.3 85.3 91.7 85.6
Potassium mgeq/kg 34.4 35.3 101.4 99.4 68.8 60.7
Sodium mgeq/kg 7.3 8.2 4.2 3.2 10.9 10.4
Calcium mg/kg 85 140 140 65 180 135
Magnesium mg/kg 153 185 105 140 60 55
U. F.-system/comment P/1 C/2 P/3 C/4 P/5 C/6
TABLE 2
Composition of permeate.
Dry matter w/w X 13.61 13.75 13.84 13.94 13.7 18.2

Sugars w/w X 12.58 12.65 12.5 12.52 12.7 14.9
Nitrogen w/w X 0.07 0.08 0.10 0.12 0.14 0.1
Purity w/w X 92.1 92.0 90.3 89.8 92.4 85.6
Potassium mgeq/kg 34.3 33.5 95.6 99.0 68.7 58.9
Sodium mgeq/kg 6.8 7.9 2.6 3.4 9.0 10.1
Calcium mg/kg 8.0 11.0 68 95 134 122
Magnesium mg/kg 168 140 118 120 115 93
U.F.-system/comment P/1 C/2 P/3 C/4 P/5 C/6
TABLE 3
Composition of retentate.
Dry matter w/w X 17.6 17.4 20.3 19.8 19.0 20.9

Sugars w/w X 13.2 12.5 14.4 13.9 17.5 16.5
Nitrogen w/w X 0.23 0.31 0.46 0.39 0.19 0.32
Purity w/w X 74.9 71.8 70.9 70.2 92.1 78.9
Potassium w/w X 37.3 39.2 107.0 99.3 70.8 64.4
Sodium mgeq/kg 7.6 8.2 3.5 3.3 11.4 10.1
Calcium mgeq/kg 445 371 200 98 265 491
Magnesium mg/kg 190 393 135 135 75 233
U. F. - system/ comment P/1 C/2 P/3 C/4 P/5 C/6
136
Comments (see Tables 1. 2 and 3):
1. Temp. 50°C, pH - 4.5, u - 4.5 m/s, CF - 4.4 - 5.5, continuous.

2. Temp. 60°C, pH » 5.5, u » 4 m/s, CF - 4, batchwise.
3. Temp. 40°C, pH - 6.4, u » 2.4 m/s, CF » 5, continuous.
4. Temp. 60°C, pH = 5.0, u » 4 m/s, CF - 1 - 9, batchwise.
5. Temp. 50°C, pH » 6.0, u - 2.7 m/s, CF - 2, continuous.
6. Temp. 62°C, pH - 6.0, u « 4 m/s, CF = 1 - 9, batchwise.
P - Polymeric membrane
C = Ceramic membrane
Economic considerations
We have tested two systems: one during three campaigns and the other
during just one, since ceramic membranes became available on the Dutch
market only two years ago.
All the tests have been done in close cooperation with the suppliers of
the systems. This had led to two systems and offers, which are summarized
in Table 4.
TABLE 4
Economics.
Ceramic Polymeric
Capacity in tonnes/h 72 72
Concentration factor 20 20
Dia-water in m 3 /h 7 7
Investment in DF1 6.8 milj. 8.4 milj
Membrane-area in m* 932 4600
Membrane replacement in DFl./y' 317,000 800,000
Electricity in DFl./y 467,000 774,000
Maintenance in DFl./y 20,000 122,000
Deprec. & interest in DFl./y 931,000 914,000
Running costs DFl./y 1,731,000 2,610,000

As seen from this table some interesting points can be noted:
i The membrane filtration area of the ceramic system is much lower than
that of the polymeric system. Two reasons will be mentioned here; first
the system pressure is at a lower level and second the product flux is
higher for the ceramic. In the process design of the ceramic system the
flux can be kept constant by decreasing the pressure difference over the
membrane. This method is not possible for the polymeric membranes,
because variations in pressure difference can deteriorate the membrane.
No compaction has been observed at the ceramic system.
ii Membrane replacement is less frequent with the ceramic system because

the membranes are practically insensitive to chemicals and microbial
attack. The expected lifetime is much longer.
iii Energy consumption is at a lower level, because the ceramic system is

more compact and operates at a lower pressure level.
The investment in an integrated production facility for invert sugar

syrup has been estimated by several engineering contractors. It was on the
same order of magnitude as for a modern, but conventional sugarbeet factory.
As a rule of thumb, one can say that for every 1000 tonnes per day of
sugarbeet the investment is about DF1. - 30 m to 35 m (Feb. 1988: 3 DF1 -
U.S.$1).
A striking difference between this novel process and the conventional
one is the reduction in loss of sugar to molasses; this process produces no
sugar in molasses at all. That is equivalent to 2.4Z of sugar based on
beets. Thus, for the same quantity of end product 12 X less sugarbeets are
needed. The total running costs are 16Z lower compared to the conventional
process.
CONCLUSIONS
The process described makes it technically and economically feasible to
produce high fructose syrup, liquid invert, or sugar from different raw
materials.
The process is easily adaptable for different kinds of materials, such
as Jerusalem artichoke, sugarbeet or chicory.
Ultrafiltration is a feasible technique to purify juices which are
heavily loaded with fouling substances, especially juices of variable
compositions.
138
Ceramic membranes have processing and cost advantages over the

polymeric type of membranes.
ACKNOWLEDGEMENTS
This elaborate research was possible only with the aid of many
colleagues, who are not mentioned by name in this report. I would like to
take this opportunity to thank management and these colleagues.
REFERENCES
1 J. van Kasteren, NRC-Handelsblad (16 October 1985), 7: In France about
9000 ha JA is planted at present. High yield varieties of 80 tonnes/ha.
Avg. yield in France 40-45 t/ha. Cost of cultivation, fertilizer and
harvesting included DF1. 1500/ha; yield DF1. 3500/ha; total cost of JA
DF1. 5000/ha.
2 T. R. Hanssens and K. Koerts, Process for the recovery of disaccharides
from disaccharide-containing tuberous plants, European patent 0 126
512.
3 Ibid., Process for the recovery of monosaccharides from poly-, oligo-
and/or disaccharide-containing plants, Eur-patent 0 126 513.
4 N. Kosaric, A. Wieczorek, G.P. Cosentino, Z. Duvnjak. Adv. in Biochem.
Eng./Biotechnol., 32 (1985) 1-24.
5 S. E. Fleming, J.W.D. Groot Wassink, C.R.C. Crit. Rev. Fd. Sci. & Nutr.
1979, 1-25.
6 W. Kofod Nielsen, S. Kristensen, R.F. Madsen, Sugar Techn. Rev.,
9(1982) 59-117.
7 R.H.F. Beck, W. Praznik, Starch/Starke 38 (1986) 391-394.
Chemistry and Processing of Sugarbeet and Sugarcane, edited by M.A. Clarke and M.A. Godshall139
Chapter 9
UTILIZATION OF FIBER AND OTHER NON-SUGAR PRODUCTS FROM SUGARBEET
JAN TJEBBES
The economics of sugar production from beet is traditionally based

almost completely on the revenue from the sugar as summarized in Table 1.
The process does produce a few other commodities but their part of the
balance sheet is small. It is also very important to minimize the amount of
molasses produced.
TABLE 1
Beet sugar manufacturing.
Income
Yield/
Product ton beet Sw. Kroner X
Sugar 149 kg 715 92

Molasses 37 kg 41 5
Pulp 50 kg 19 3
During the lifespan of the industry, many attempts have been made to
find other sources of income from the sugarbeet. The best known alternative
is probably the making of alcohol from either the beet juice or the
molasses.
This presentation will concentrate on attempts to utilize components of
the beet other than sugar. I don't think that I will have the pleasure of
telling you anything really new, since our industry is so mature that many
ideas have been discussed already. However, technologies and conditions do
develop and continually give us the impetus to work with the challenge of
finding alternative uses for the sugarbeet.
Also, at least in Europe, the political pressure today creates a threat
to the income from sugar as such. We thus need new sources of income from
the sugarbeet.
Components of suearbeet
In Tables 2 and 3 are listed some of the major and minor sugarbeet
components.
1Λ0
TABLE 2
Dry matter in the beet, Z of dry matter.
Component Z
Sucrose 74
Ash 3
Soluble organic non-sugars 18
Insolubles (marc) 5
TABLE 3
Organic non-sugars in sugarbeets.
Carbohydrates 0.3
Organic acids 0.2
Amino acids 0.2
Amides 0.1
Protein 0.1
Betaine 0.3
Saponins 0.1
Lipids 0.02
Hemicellulose 2.0
Cellulose 1.0
Pectin 2.0
Lignin 0.5
The inorganics, i.e. the ash, can be left out of the discussion. To my
knowledge no attempts have been made to make any economic use of the
inorganics of sugarbeet. The lime sludge from the process is nowadays sold
as a soil improver, mainly because of its content of lime. The value of
this product is increased by the presence of elements such as boron that the
beet brought out of the soil where it grew, but it is a bit strained to call
this an economic use of the beet.
The organic non-sucrose components (Table 3) are more interesting.
However, the carbohydrates, i.e. the hexoses and the raffinose, have so far
been considered as negative for the process. The same applies to
ιΛΐ
polysaccharides such as dextran that normally are present only when the beet
is damaged by frost.
There are several organic acids in the beet, the most abundant being
citric and oxalic. Both have an important influence on the technological
value of the beet, but they are not isolated for any separate purpose,
although citric acid is produced by fermentation, from beet molasses.
The amino acids and the amides in the beet are also of importance for
their technological value. Glutamic acid and glutamine can both be isolated
either separately or as glutamic acid after hydrolysis. Normally separation
by ion exchange resins is utilized. This had been used on an industrial
scale in France and the United States, but to my knowledge there is no
activity in this field today.
In addition to the small amount of real protein in the sugarbeet, much
of the nitrogenous substances are considered as protein when calculating the
fodder value, and thus add to the economical value of the beet. There is
therefore no commercial interest in isolating them. Beet leaves are rich in
"protein"; many projects exist to make industrial use of the leaves,
although none has resulted in production. The leaves, however, are not the
subject of this paper.
Betaine or trimethylglycine is of interest: this zwitterion can be
isolated by resin chromatography, as is done in Finland. Betaine finds uses
in the metal plating industry, in cosmetics and as a vitamin in the breeding
of chickens.
There are several saponins in the beet; best known is the glucuronic
glucoside of oleanoic acid. The saponins are negative to the process and to
the quality of white sugar, but there are no methods to isolate them from
process streams. Saponins could possibly be used as detergents for washing
or in cosmetic products.
The lipids or fats content of the beet is very low and has a slight
negative influence on the quality of the fiber products that is discussed
later in this paper.
The gums in the beet exhibit interesting properties for several food
applications. The pectin can be used for gelling jams and marmalades, but
its gelling properties are less intense than those of citrus or apple pectin
because of differences in methylation. Pectin is isolated from an acid
extract of beet pulp. If this extraction is performed rather vigorously and
1^2
followed by removal of unpleasant taste and odor substances an interesting

mixture of beet gums is obtained. This mixture could be called Beet Gum,
and is a mixture of pectin (a-l,4-polygalacturonic acid), araban and
galactan. It should be comparable to gum arable in its rheological
properties. A solution of Beet Gum has an astonishingly low viscosity but
gels well with sugar at high concentration. This can be used in a variety
of foods and sweets like throat tablets, toffees and caramels. Beet Gum
also finds use as an emulsifier in margarine.
Beet pulp and fiber
The substances in the last third of the table in Table 3 not only show
the highest concentrations but are also the most interesting ones when
considering their future commercial value. In this group we find both
completely insoluble compounds and some with a low but not negligible
solubility in water. The sugar process is run so as to minimize the
concentration in the beet raw juice of all these compounds. They are for
this reason often referred to as insolubles.
This means that substances belonging to the four last groups in Table 3
remain in the exhausted beet cossettes which usually are called beet pulp.
They actually are the main components of the pulp so if we want to make
money out of the pulp we must make use of those four groups of non-sugars.
Some of the possible uses of beet pulp include: fiber for animals (pressed
or dried, with or without molasses), dietary fiber, and beet gums.
The conventional way is to use the pulp as a cattle feed. The pulp is
first pressed and the presswater is normally returned to the sugar process.
The pressed pulp with or without molasses can then be dried into a storable
product. This drying is done by flue gases from a burner fired with oil or
gas. With the rapidly increasing energy costs in the seventies, a search
for energy saving technologies started. Two lines were followed:
1. Not drying at all.
2. New methods of drying.
During this development, a réévaluation of the pulp and its components
also took place. The hemicellulose and the pectin give the pulp technically
and physiologically interesting properties, which were partially lost by the
older drying methods. Important factors include the structure of the
cellulose and the water binding and gelling capacity of the hemicellulose
and pectin.
An example is the difference in the milk producing capacity of cows
kept on different feeds: a cow fed on pressed pulp will produce 10Z more
milk than a cow fed on conventionally dried pulp.
Beet pulp, that has not been damaged by flue gas drying, can also be
used as a raw material for food. However, raw beet pulp has a taste that is
not very pleasant; also, the structure of the cellulose is different from
that with the most desirable properties. Taste and odor problems can be
caused by several components, including Maillard products, aldehydes,
hydroxy-acids, pyrazines, saponins and geosmin.
Our company has developed a steam drying process that has a positive
influence on the cellulose structure and that removes substances that cause
flavor problems. The product is marketed under the trade name FIBREX. The
composition of this product is listed in Figure 1. The pectin and other
gums in the beet provide a high total dietary fiber content as they are not
digested by the human stomach. The water holding capacity gives the beet
fiber a high functional value in bread and other foods.
The low-fat fiber is gluten free, and therefore suitable for diets of
gluten-allergic people. The fiber is also free of phytic acid, thereby
reducing risk of minerals adsorption by phytate that affects body minerals
absorption.
COMPOSITION
(% of total)
_ ^ ^ ^ ^ /HEMICELLULOSE 29
CELLULOSE 18 - ^ ^ ^ ^ ^
WATER 10
FAT 0.3-
SUGARS 3.5-
MINERALS 3 . 5 / - < - - > ^ | r U^^~ P E C T i N 22
PROTEIN 10 LIGNIN 4
Total Dietary Fibre 7 3 %
(22 - 25% soluble thereof)
Fig. 1. Composition of FIBREX beet fiber.

1M
Beet fiber, FIBREX, is compared to dietary fibers from other sources in
Table 4.
TABLE 4
Comparison of composition and properties of selected dietary
fibers (Z dry solids (D.S.)).
Beet Wheat Soy Citrus Pea

fiber bran hulls fiber fiber
Dietary fibers
Cellulose 20 12 50 20 ~60
Hemicellulose 32 22 14 ~20 ~15
Pectin 25* (2)** - ~30 ~-2
Lignin 4 6 5 ~2 6
(5)*** (13)*** (7)*
Total dietary fiber 81 42 69 72 85

(30-60)
Soluble thereof 22-25 3 4 ~29 ~7
Other substances
Protein 11 17 12 7.5 3.4
Starch - 20 4 - ~8
Sugars 3.4 1 1 6.7 < 1
Fat 0.3 5 2.5 0.3 1.8
Minerals 3.4 7 5 2-3 2-3
Gluten - ~10 - - -
Phytic acid - 3-4 1-2 - -
Energy
content/100 g DS 265 kJ 835 kJ 385 kJ
(dietary fiber (63 (200 (92
not counted) kcal) kcal) kcal)
Water holding 5 1.5 1..5 1-2

capacity
(weight of water/weight of D,,S.)
* Only polygalacturonic acid - carbohydrates, bound to the

galacturonic acid have been counted with the hemicellulose
** Uronic acids (but not in the form of pectin)
*** Lignin content in X of dietary fibers.
Odor and taste problems have been mentioned several times in this
paper. Anyone who has ever tasted a sugarbeet is probably not surprised.
Although sweet, the beet tastes rather earthy. Several compounds are
responsible for this; one is the above mentioned saponin. Another compound,
geosmin, is not as well known. Geosmin, shown in Figure 2, is present in
the beet and remains in the pulp after sugar extraction. Its presence has
been established by using GC-MS techniques. It is removed by the FIBREX-
process, but it must be remembered that the level of geosmin needed to
produce an off-flavor is about 1 mg/tonne.
CH3
\x i \y
OH |
CH3
Fig. 2. Structure of geosmin (Geo - earth, osme - smell).
1A6
Chapter 10
ANALYTICAL METHODS OF SUGAR FACTORIES—NEW DEVELOPMENTS
H. SCHIWECK AND G. STEINLE
INTRODUCTION
Süddeutsche Zucker AG, the largest German sugar company, with

7 factories and an annual beet input of 6 million tonnes, produces
1 million tonnes of sugar per year, and also liquid sugar, fructose syrup
and crystalline fructose. Isomalt, brand name Palatinit, which we developed
recently, is in the process of obtaining FDA-approval for use in the USA.
At Südzucker, routine chemical in-process control is done by the
factory laboratories. Coordination, trouble shooting and development are
done by the Central Laboratory, which is located at the Offstein factory
near Mannheim. Among other things, this research laboratory develops new
analytical methods designed for in-process control and routine analyses.
These new methods are designed to increase precision and accuracy and/or
reduce the time and personnel usually needed for measurements.
Many of the traditional methods used in the beet sugar industry, such
as polarimetry and refractometry, have been, and indeed still are, used
mainly for determining the mass flow of dry substance and sugar entering the
factory and accounting for it as it leaves the factory as sugar, molasses
and dried pulp, and as losses during the process. Other techniques, such as
titrimetric and conductimetric methods and pH measurement, have been used
mainly for setting and monitoring various empirically derived process
parameters.
Technical processes have to meet ever-inereasing requirements as to
their efficiency and effectiveness and therefore need sophisticated
analytical controls. An example is the determination and control of sucrose
losses, due to chemical and bacterial action, that produce coloring matter
and substances that negatively affect sugar quality and recovery. We also
have to control the optimum use of process aids.
Over the last years new processes have been introduced, for example
molasses sugar recovery by liquid distribution-chromatography, which can be
optimized only under constant analytical control.
As a result, analytical processes are increasingly automated, and in-
process control is introduced in order to augment the number of samples and
reduce the time needed for each analysis without affecting quality.
The introduction of computer control of the processing plant requires
reliable analytical data without the delay associated with laboratory
methods. Such a demand can only be met by means of continuous
instrumentation: sampling, monitoring the essential parameters and
reporting directly, in digital format, to the controlling computer. If the
total analytical process is divided into different steps: sample taking,
sample transportation, sample analysis and data representation and
exploitation, there are two alternatives for automation of process control,
which can be used wherever and whenever appropriate: automation of the
factory laboratory and on-line measurement in the factory.
Automation of the factory laboratory. In this case, analyses are
carried out automatically by a central facility located as a rule in the
factory laboratory. Samples are transported to the analytical unit either
manually, over a ring line, or pneumatically. The transmission and
processing of measured values can be done automatically via display of the
data in the factory or with the help of computers and control units.
On-Line measurement in the factory. In this case the analyses are
carried out on-line within the factory. There is no transportation of
samples. The display of data effects results directly in the factory by
serving as a basis for manual or automatic control of the process.
The first applications of automation using a centralized facility are
for control of the raw material, i.e. for beet quality determination, the
control of the final products and pollution control.
On-line measurements and automatic data registration within the factory
are the most efficient ways of supervising and controlling the fabrication
steps.
We expressed this opinion for the first time in 1968 (ref. 1). At that
time, we started controlling extraction by way of continuous automatic
polarization of presswater, of extraction liquid at any point in the
diffusion system, of raw juice and other juices in the factory.
Furthermore, we developed a direct determination of optical rotation and
refractive index for determining the purity of intermediate products in the
boiling station. We also explained the measuring techniques which allow.
1A8
direct determination of technological values during processing, i.e.

determination of dry matter of pressed pulp by means of reflection in the
infrared range; determination of the concentration of potassium, sodium and
calcium by selective glass electrodes, and determination of alkalinity and
total content of calcium oxide in juices by means of automatic titration.
In the following we will describe the development of analytical control
methods in sugar factories introduced since our initial trials in 1968. We
will furthermore point out in which areas the suggestions made in 1968 have
been complied with, and in which areas further developments have been
achieved.
Before starting out with the supervision and control of a process we
should answer two basic questions.
First we have to find out why control is necessary; whether automation
is appropriate, and at which points of the process the control is to be
operated. Secondly, we have to decide which parameters to measure and which
analytical methods to use efficiently in order to gather the information
needed at reasonable costs and with sufficient accuracy and speed.
To answer these questions we have to analyze the production process in
order to determine the parameters which are to be controlled. The following
steps in the process will be treated in detail:
a) Beet quality determination.
b) On-line measurement of the dry substance of pressed and dried pulp.
c) Control of the purity of juices and syrups.
d) Control of the crystallization process.
The first analytical measurement required for process control is the

sugar content of the beet at the beet reception point. In many cases it is
combined with the determination of other quality parameters, required to
calculate expected yield in white sugar.
Figure 1 shows the beet reception and tare house operations as they are
practiced in many cases. After the beets have been sampled, weighed and
washed, they are sawed up to prepare the brei. After adding basic lead
acetate or aluminum sulfate solution, the brei is extracted and filtered.
The quality parameters are then determined one at a time using the clarified
filtrate. A different technique is used in the analytical facilities of the
Enns factory in Austria and in the tare houses of Südzucker (Figure 2),
sampling polarisation
i
weighing
i K 4-Na
washing
♦
brei
oc -amino N
1 preparation
♦
addition BLA
1 orAI2(S04)3
♦ NO3"
extraction
♦ ,
filtration
Fig. 1. Flow diagram of beet reception and tare house operations showing
sequential determination of quality parameters.
where we measure polarization, potassium, sodium and a-amino-nitrogen

simultaneously. In this case the filtrate is taken out of the polarimeter
funnel with the aid of peristaltic pumps.
Polarimetrie analysis is traditionally employed to determine sucrose
content of beets and other products in the sugar factory. We are aware of
the fact that this method measures the rotation of all optically active
substances in the solution: not only the rotation of sucrose but also the
rotation of raffinose, glucose, fructose, arabinose, amino acids and
fragments of pectins. For this reason, different methods have been
suggested in order to determine specifically the sucrose content of beets.
150
sampling
♦
weighing
♦
washing
♦
brei
preparation
♦
addition BLA
or AI 2 (S04)3
♦
extraction
♦
filtration
-♦ .
'1 Y
K ♦ Na polarisation cC -amino N |
Fig. 2. Beet reception and tare house operation showing simultaneous

determination of quality parameters.
These include: isotope dilution, gas chromatography, high performance

liquid chromatography and enzymatic methods.
However, the sugar industry has been reluctant to adopt these methods
because of cost considerations, and/or the time needed for the analysis of
each sample, and/or the complexity of these techniques. As far as the other
quality parameters are concerned, it is usual to use flame photometry to
determine sodium and potassium, ion-selective electrodes for nitrate
determination and colorimetrie methods for the determination of a-amino-
nitrogen.
In the case of a-amino-nitrogen, there has been a tendency to change
from the "blue number", the colorimetrie measurement of the blue copper-
complexes, to fluorometric methods, which can also be used in yellow colored
samples clarified with aluminum salts.
Fluorometric measurements are carried out on compounds resulting from
the reaction of the amino group either with fluorescamine (Figure 3) or with
151
o-phthalaldehyde (Figure 4). These methods register all existing amino

groups and not just the a-amino-nitrogen. Therefore the values obtained
with fluorometric methods are higher than those measured with the
colorimetrie "blue number" method. To apply the copper complex method to
beet extractions with aluminum salts, it is necessary to eliminate sample
absorbance by incorporating a sample blank in the analytical method. This
method was applied during the 1986 campaign with good results in one of our
beet laboratories.
Fig. 3. Reaction of fluorescamine with primary amino groups.
Mercaptoethanol Fluorophore
HS-CH2-CH2-OH
=0
o-Phtha!aldehyde
C=0
H
H2N-R
Isoindole
Amine
DÏV"
Fig. 4. Reaction of o-phthalaldehyde with primary amino groups.
152
The beets enter the factory production process at the diffusion plant.
Diffusion is the first step in the process for which on-line analytical
measurements and computer control are needed. Parameters to be controlled
are: draft of raw juice, sugar loss, and lactic acid formation.
For total automatic control of the extraction process, i.e. to minimize
sugar loss and optimize draft, it is necessary to measure the polarization
of pulp press-water and the quantity of raw juice.
The computer program we have developed and which is used in our
Plattling factory, adjusts the draft and, if necessary, the slice rate.
To measure the polarization, the pulp-press water is filtered
continuously through a membrane filter which eliminates particles over 1-2μ.
The sugar content is measured with an automatic polarimeter.
An important and unavoidable source of sucrose loss in diffusers is
infection with bacteria. L-lactic acid is the most important product of
bacterial metabolism in a diffuser. The determi nation of L-lactic acid and
pH-measûrement offer a tool to monitor the bacteriological condition of
diffuser and to minimize sugar losses. We use two methods for off-line
measurement of lactic acid: determination by ion-chromatography and
enzymatic analysis.
Since its introduction in 1975, ion-chromatography has become a
versatile method capable of analyzing a tremendous number of ionic
compounds. The separation of the compounds is done by way of ion exchange,
ion pair separation and ion-exclusion in the case of organic acids (Figure
5).
Electrolytes are used as eluents for the ions. The ions in the sample
move at different rates and separate into discrete bands according to their
relative chemical affinity for the resin material used in the separating
column. The eluant then carries the discrete ion band through a post-column
supressor device that reduces the interference of the eluant and converts
the analyte ions to a highly conducting form while simultaneously converting
the eluant ions to a less conducting form. Detection is accomplished via
conductivity.
153
Huent Reservoir
Delivery
Mode
Separator Columa
Data Mode
Fig. 5. Scheme of an ion Chromatograph.
Figure 6 shows an ion-chromatogram of a mixture of organic acids. This

chromatogram illustrates the advantage of the IC-method, which determines in
addition to lactic acid other acids of the Krebs-cycle, the lower carbonic
acids of fermentation, and pyrrolidone-carboxylic acid, produced by
saponification of glutamine.
15^
min
20
citric
tarlane
malic
25 pyruvic
lactic
formic
30 acetic
propionic
PCA
35
butyric
40-
Fig. 6. IC-Chromatogram of carboxylic acids. Apparatus: Dionex-IC;

column: HPICE AS-1; Eluant: 6mM octanesulfonic acid - 5Z acetonitrile;
flow rate: 0.5 ml/min; detector: conductivity. PCA = pyrrolidone-
carboxylic acid.
We cannot, however, distinguish between the D- and L-forms of lactic

acid with this method. This is a very important point because sucrose loss
due to bacterial action increases L-lactate, whereas the chemical
decomposition of invert sugar in raw juice leads to a racemic mixture of D-
and L-lactate. If we want to distinguish the amounts of lactic acid
produced by bacterial and chemical processes, we have to resort to the
enzymatic determination of D- and L-lactate. In this case L-lactate is
oxidized to pyruvic acid by NAD which is reduced to NADH. The reaction
proceeds under the control of the enzyme L-lactate dehydrogenase (L-LDH).
An analogous reaction with D-lactate and NAD proceeds under the control of
D-lactate dehydrogenase (D-LDH):
L-LDH
L-Lactate + NAD- = = = = = = pyruvate + NADH + H"·-
D-LDH
D-Lactate + NAD- = = = = = = = = = = = = pyruvate + NADH + H-
This method is very useful for routine control in the factory

laboratory.
The diffusion process also produces wet beet pulp. This pulp is
pressed and dried to increase its dry substance content.
Analytical control of pulp presses and pulp dryers focuses on dry
substance measurement, which is the main determinant of the fuel cost of
pulp drying. Dry substance is also a quality parameter for dried pulp and
pulp pellets.
In practice this measurement is done with the aid of infrared light
(Figure 7). The principle of the measurement is that infrared light at 1.9
and 1.4|im is absorbed by water. From infrared light with for instance 1.4
and 1.6μπι, falling on a sample, the Ι.βμπι is not absorbed. Measuring the
remitted infrared light with the wavelength which is not absorbed as a
reference by turning a filter wheel, we receive an indication of the water
content of the sample (Figure 8).
The last stage of the sugar production process is the vacuum pan
station where sugar juices and syrups are concentrated to supersaturation
and then crystallized by evaporation. In the vacuum pan station, automatic
control is ensured by way of three different measurements. We measure the
purity of the different products, the superaturation of the syrups before
crystallization and the grain size distribution in crystal magma.
Further refinement of our first on-line purity measurement in 1968,
mentioned above, over recent years has led to a system of continuous purity
determination (Figure 9). This system is composed of: a balance to realize
a 1:1 dilution of the syrups; a mixing chamber with stirrer; a polarimeter;
a refractometer.
The entire facility is controlled by seven valves. This control unit
allows determination of the purity of a particular syrup.
Figure 10 depicts a system which allows the determination of four
syrups, one at a time. In the analysis unit each product has its specific
chamber for weighing, diluting, homogenizing and temperature adjustment.
After this the products are analyzed by only one refractometer and one
polarimeter. Simultaneous sample preparation allows a measurement every 3
to 4 minutes. The products are transported to the analysis unit via a ring
line.
156
1 IR-light source
2 concave mirror
3 lens
4 deflection mirror
6 4 5 filter wheel
6 detector
\#^-ο 1- principal ray

?
sample
Fig. 7. IR-moisture determination ray path.
filter wheel
M test filter
V?
\° CO R reference filter
L perforatori «*»ΓΛΑΓ\
2? *
c
Φ
c
■_n M0 Ri L _J~L· Ro
!
0° 60° 120 ° 180° 240° ang
Mi R0
Measuring Principle
Fig. 8. IR-moisture determination.

f H l l * l l l l «^
purity measurement
Ô V6
r- pobrimeter
V
refractometer |
ZD
Fig. 9. System of continuous purity measurement.
Analytical on-line control in the vacuum pan station focuses on

determining supersaturation and crystal content, the essential guiding
parameters for the evaporative crystallization of sucrose.
Since supersaturation correlates with the viscosity of syrups, this
long been used for measuring supersaturation. There is also a good
correlation between supersaturation and density, and recently the
measurement of density has been used for this purpose.
158
Purity measurement f or four product s
1 I 1 j
n
U
■■■
Δ
balance
ΛΛ h n , ,
m ® ®| ®.
IH* Il II II II II II
l i zï Ì A A A
L 1
!( 3 rtfractometer . ' .
computer
1
( ) polarimeter 1 ■ 1
1
L |
y
Fig. IO. Purity measurement for four products.
However, if density is used as a means of determining supersaturation,

a high degree of sensitivity has to be ensured. This requirement can be met
with radiometrie density measurement. Economical and technical reasons
prevented the application of this method earlier since the detection of the
radiation by a scintillation counter required quite a few electronic
devices. The density measurement arrangement has become competitive only
recently thanks to cheaper electronic components.
Figure 11 shows the elements of the density measurement arrangement,
consisting of the radiation source with dip pipe, the shielding and the
scintillation counter. The radiation of the a-37Cs-gamma-source is
attentuated as it passes through the magma, the attenuation being an
exponential function of density.
Fig. 11. Measurement of radiometrie density.
The radioactive material in the vacuum pan is fixed in a glass bead

eliminating the risk of emission in case of a breakage.
The measurement of crystal size distribution in a magma is an important
criterion for evaluation of crystallization work, particularly in regard to
the degree of secondary crystal formation.
Sieve analysis, which is generally used, provides inaccurate results in
this respect, and in the last few years two other methods of measurement
160
were developed. The most widely established method uses the Coulter
principle, diagrammed in Figure 12.
Fig. 12. Scheme of a Coulter counter.
In the Coulter counter the particles, suspended in a conducting liquid,

are passed through a capillary with a superimposed electrical field. As
soon as there is a particle in the orifice, a voltage impulse, proportional
to the volume of the particle, is generated. Thus the capillary to be used
has to have a definite relationship to the particle size to be measured.
The size limits for the technique have been quoted as being from 0.6 to
400 firn. The magma sample to be measured is diluted with a glycerin solution
containing an electrolyte.
At Südzucker we carry out grain size measurement with a Zeiss-
Mikrovideomat using either a microscope or a macroscope depending upon the
grain size range to be measured. The particles seen by the objective are
taken up by a television camera projected onto a monitor screen and analyzed
there according to their sizes.
Since a direct measurement of magma samples is not possible, these have
to be diluted with special dilution fluid, containing aqueous sucrose and
i6i
fructose solutions. This suspension fluid renders individual crystals

easily discernible.
CONCLUSION
In conclusion, I would like to reiterate what was said at the beginning
of this chapter. The methods of analysis applied in the sugar industry
support and improve the processing steps involved in sugar production. The
aim of the manufacturing process, which starts with extraction and ends with
crystallization, is the isolation of a pure final product without
significant loss of sugar.
We are of the opinion that these two requirements can be fulfilled only
by controlling the individual processing steps. The analysis of end
products cannot change the quality. Errors made during processing also
cannot be corrected by analysis afterwards.
Therefore I would like to end with a quotation from an article on
production improvement in Fortune magazine, from Warner-Lambert: "Quality
comes from the tight control of production processes, not after-the-fact
inspection."
REFERENCE
1 H. Schiweck, Zeitschr. Das automatische kontinuierliche Messen einiger
technologischer Größen während des Produktions-prozesses in der
Zuckerfabrik. Zucker, 21 (1968) S. 494-510.
I62
Chapter 11
SUGARCANE PROCESSING: RAW AND REFINED SUGAR MANUFACTURE
M. A. CLARKE
INTRODUCTION
In 1987-88, over 65 million tonnes of sugar were produced from

sugarcane, on a worldwide basis. The tonnage of sugarcane grown is more
difficult to estimate: production of ethanol, as in Brazil, and of non-
centrifugal sugars, as on the sub-continent, all increase the total
production of sugarcane above the 700 million tonnes required for
centrifugal sugar production. Countries where sugarcane is grown
commercially are shown in Figure 1. Harvest seasons run from year-round in
Hawaii, Cuba, Colombia and other tropical areas, to the brief 3 month crop
seasons of Louisiana and the sub-tropics.
Fig. 1. Countries where sugarcane is grown are shown in solid color.

163
Photosynthesis converts water and carbon dioxide to sucrose in all

green plants, and most efficiently so in the giant grasses of Saccharum
species. The sugarcane plant is the most efficient collector of solar
energy in the plant kingdom, fixing 21 of available solar energy into
chemical bonds of organic compounds, chief among them sucrose.
Isolation and purification of as much as possible of the sucrose from
the plant are the goals of sugarcane processing. Sugar production is a
series of separations of non-sugars from sucrose. Process chemistry is
designed for maximum removal of non-sugars with minimum destruction of
sucrose.
Processing has traditionally taken place in two stages:
1. Extraction of juice from sugarcane and conversion to raw sugar
(94-98Z sucrose), at factories (mills) in areas where sugarcane is grown.
2. Refinement of raw sugar, shipped from areas of production to areas
of consumption, to white and brown refined products (white sugar of 99.9Z
sucrose).
The sugarcane plant cannot be stored after harvest: sucrose begins to
decompose shortly after the stalk is cut. The two-stage production process
evolved because of the need to convert cane rapidly into a product that
could be stored.
A third major traditional process has been the production of
"plantation white" sugar, a white sugar of over 99Z sucrose produced
directly from sugarcane.
The last two decades have seen changes and new trends in these
traditional systems, as consumption of sugar in the tropical producer
countries increases (and decreases in the United States because of
replacement by cheap corn syrups), and as these producer countries, and
their growing food processing industries, demand higher quality sugars.
Most notable are the increase in refining done in the tropics, and the new
processes for production of high quality white sugar directly from cane
juice.
The traditional systems will be presented briefly in this paper from
the point of view of the chemistry of sugarcane juice and the production and
refining processes. The relationship of the papers in this symposium to the
process will be noted; the paper is an overview of the sugar manufacturing
process as a background for the subsequent papers.
l6k
RAW SUGAR PRODUCTION

In the tropics and semitropical areas where sugarcane is grown, cane is
cut by man or machine, hauled to the mill by vehicles ranging from bullock
carts to railway trains and trailer trucks, and processed in factories of
capacity ranging from a few hundred tonnes per day to the huge 20,000 tonnes
per day factories in Florida, and giant complexes in Brazil that can grind
40,000 tons of cane per day, though using over half the juice for ethanol
production. Cane, either whole stalk or cut into 15B - 20" (35-50 cm)
billets, is stripped of leaves before it reaches the mill and, in some
areas, is washed by water sprays as it enters the mill. The influence of
cane variety of its composition is discussed by Legendre in Chapter 12.
Juice is extracted either by conventional milling, or, less usually, by
diffusion. A milling train, or tandem, encompasses up to seven sets of
three roll mills, arranged in a great variety of designs (refs. 1-4); water
is applied to one or more of the later mills and circulated to improve
extraction.
Diffusion is continuous and countercurrent; many designs of diffuser
are in use (refs. 1-4). Diffusion, in general, gives greater extraction of
sugar from the cane than does milling, but also greater extraction of non-
sugars, increasing loss of sugar to molasses. The overall recovery of sugar
from cane is higher with a diffusion process, but water usage is greater.
The raw sugar process is outlined in Figure 2.
i 1
WASH MULTI-STAGE
MILL Ί EVAPORATION VACUUMI
CRYSTALLIZATION
CAME | OR JUICE CLARIFICATION , , $ γ ρ υ ρ
DIFFUSER 10-1β%
70%
~Γ~ / \
BAGASSE RAW SUGAR MOLASSES
Fig. 2. Raw cane sugar manufacturing process in sugar mill or factory.
The composition of whole, millable sugarcane, without leaves and trash

is shown in Table 1.
165
TABLE 1
Composition of sugarcane.
Millable Cane Cane (Z)
Water 73-76
Solids 24-27
Soluble solids 10-16
Fiber (dry) 11-16
The chemical composition of juice solids and their mineral constituents

are tabulated in Chapter 12.
Among other compounds found in juice (ref. 1) are series of substituted
hydroxybenzolc acids and phenolic acids, in total concentration range of 100
- 1000 ppm. These compounds are often colored yellow to brown, or become
colored during subsequent processing, and represent the part of the sugar
colorant complex that originates in the cane plant (ref. 1 ) . Some of them
along with dimethylsulfide (from the cane plant) and acetic acid make up the
characteristic brown sugar and molasses flavors, as discussed in detail by
Godshall in Chapter 15. Of the amino acids, aspartic and glutamic, on the
order of 0.1Z each, are usually those in greatest concentration in sugarcane
juice. These, and other compounds to be removed in processing, are outlined
in Table 2.
TABLE 2
Juice components removed in raw sugar manufacturing process.
Average Concentration, X on solids
Component Juice Raw Sugar
Invert (glucose +
fructose) 3-5 0.205
Organic acids 2-4 0.204
Polysaccharides 1-2 0.05-0.15
Inorganic ash 2-4 0.3-0.5
Protein 0.5-0.6 0.03-0.08
166
Juice as it comes from cane has pH 5.0 to 5.5 and contains invertase
enzyme, which both contribute to the breakdown, by hydrolysis, of sucrose.
First processing steps are heating, to inactivate enzymes, and raising the
pH, with lime (liming), the latter usually added to juice to form a solution
of calcium saccharate for easier mixing.
Clarification
Juice is heated to 95°-100°C, and lime, or calcium saccharate is added
to raise juice pH to 6.7 - 7.3, depending on maturity and freshness of cane.
Higher pH is generally preferable for purification, but the extra calcium
adds to evaporator scale later in process. The goal is a pH of 7 in
clarified juice. There is always a pH drop over clarification because of
removal of basic salts. Addition of too much lime can cause decomposition
of monosaccharides to organic acids and increase the pH drop. In some
areas, juice is heated after liming.
Lime, colloidal and suspended solids, mud and bagasse particles settle
out, aided by addition of polyacrylamide flocculant at a level of 2-5 ppm on
juice. The choice of polyacrylamide of suitable molecular weight and degree
of hydrolysis for best flocculation depends on juice chemistry, as discussed
thoroughly by Whayman (refs. 5, 6). The presence of phosphate ion,
naturally-occurring, or added to a level of 300 ppm, greatly increases
clarification through calcium phosphate bridge-linking of flocculating
particles (ref. 7). The chemistry of the nature of calcium phosphates
formed and conditions determining their formulation are another interesting
area of study (ref. 1).
Clarification transforms dark, muddy-looking whole cane juice to a
light-colored translucent solution, with chemical removal of non-sugars as
outlined in Table 3.
TABLE 3
Juice components removed in clarification.
Inorganic ash Polysaccharides (starch, dextran,

(sulfates, phosphates, silicates) cell-wall polysaccharides)
Iron and aluminum salts Gums (hemicellulose)
Organic acids and colorants Waxes and lipids
Amino acids and protein Suspended mud particles
Bagacillo
Sucrose Z solids concentration increases slightly, but some sucrose
remains in the settled muds which are separated and filtered, and must be
washed for sucrose recovery.
Clarification systems, new and old, are described in detail in Chapter
17 by Trott.
The pH is a major point of control: pH too far below 7 causes
inversion and loss of sucrose; too high a pH causes decomposition of glucose
and especially fructose to organic acids and colorant molecules.
Chapter 16, by Richards, describes recent observations on decomposition
reactions of sucrose, which explain some of the so-called "unknown loss" of
sucrose both in raw and refined sugar manufacture. These reactions can
occur in juice, syrup or refinery clarification, evaporation and
crystallization.
Clarified juice, at 13-18 Brix (X solids), similar to raw juice Brix,
is then evaporated in a series of multiple effect evaporators of any one of
a great variety of designs (ref. 1-4) to a syrup of 60-70 Bx. Inorganic
salts in the juice, particularly calcium sulfates, silicates, oxalates, are
removed in this process through formation of evaporator scale, which lowers
heat transfer efficiency in the evaporator body and must be periodically
removed (a controlling factor in operating schedules). The pH of clarified
juice, ideally at 7, will drop further to 6.5 - 6.8 in syrup because of
further removal of basic salts. This pH range is appropriate for efficient
crystallization—the next purification step in raw sugar production.
Crystallization
Syrup from the evaporators, at 65 - 70 Brix, is concentrated under
vacuum, to enable operation at lower temperatures and minimize sucrose
decomposition; seeded with finely ground sugar, and crystallized. The
mixture of crystals and mother liquor is separated in a basket centrifuge,
and mother liquor is re-concentrated to give a second crop of crystals.
These serial crops of crystals are taken in vacuum pans; the last run-off
syrup (or molasses) is heated and cooled at atmospheric pressure in
crystallizers; crystals from this strike are remelted. The final molasses
from this strike, or blackstrap, is sold as a byproduct.
168
Crystallization is the major purification step, where most non-sugars

concentrations are reduced from syrup solids to crystal sugar by an order of
magnitude, and the cheapest because the major requirement is heat, and raw
sugar factories generate steam from burning their own bagasse. Vacuum
crystallization equipment and practices with a great variety of systems
patterns for recycling syrups and molasses are discussed in detail in many
references (ref. 1-4). In achieving the maximum yield of high quality sugar
crystals, efforts are made to minimize sucrose loss to molasses and maximize
non-sugar concentration. Invert, or reducing sugars, can replace sucrose in
molasses on a partition basis; therefore, in beet sugar processing, where
there is almost no invert, the loss of sucrose to molasses is greater than
in cane processing. Every non-sugar has a "melassigenic coefficient," that
is, the number of parts of sucrose per part of non-sucrose that will be
taken into molasses at saturation (ref. 9). The major inorganic salt in
cane juice, syrup and molasses is potassium chloride and the order of
melassigenic coefficients is: K > Na > Ca > Mg (ref. 1). Approximate
composition of cane blackstrap molasses is shown in Table 4 (ref. 1).
Johnson, in Chapter 21, presents a new membrane technology system for
removing potassium ions from molasses which would improve molasses
exhaustion, as well as molasses fermentability.
TABLE 4
Approximate composition of cane blackstrap molasses.
Normal percentage range

Component by weight
Water 17-25
Sucrose 30-40
Glucose 4-9
Fructose 5-12
Total reducing substances (as invert) 10-25
Polysaccharides, starch and gums 2-5
Ash, as sulfates 10-21
"Crude protein" (as N x 6.25) 2.5-4.5
True protein 0.5-1.5
Amino acids, principally aspartic and
glutamic 0.3-0.5
Aconitic acid (1-52), citric, malic, oxalic,
glycolic acids 1.5-6.0
Mesaconic, succinic, fumarie, tartaric acids 0.5-1.5
169
Raw sugar is washed (sprayed with water while being separated in the
centrifuge), dried on belts and by slingers, and stored in bulk. It is
shipped by ocean-going vessel, rail car or truck to the refinery.
Major classes of non-sugars that are removed from juice in raw sugar
manufacture, and also removed in refining, are color and colorants,
discussed by Riffer in Chapter 13, and polysaccharides, including those that
are part of the sugarcane plant, such as starch, and those that are products
of microorganisms on sucrose, such as dextrans, discussed by Kitchen in
Chapter 14.
There has been a trend in recent years to produce raw sugars of 98°
pol, rather than the traditional 96 pol, and in some areas, a very high pol
raw of over 99 pol (réf. 10). The additional purification involved in their
production is discussed in several of the following chapters, notably in
that of Trott, who discusses syrup clarification, the newest unit process to
be used in raw sugar manufacture. Syrup clarification removes an additional
fraction of colorant, polysaccharide and turbidity from raw sugars.
The raw sugar factory produces as byproducts bagasse fiber and
molasses. These and other minor byproducts, and their myriad uses and
applications are well described by Paturau (ref. 11); many are discussed in
detail in this volume. Another byproduct, found with increasing frequency,
is electricity: a cane factory burns bagasse from its own cane as fuel and
is energy self-sufficient. There is approximately 20Z bagasse in excess
over standard factory requirements: this is burned and the excess
electricity generated sold to the local grid, at first in Hawaii, Mauritius
and Florida, and now in many areas.
The best known byproduct, or alternative product because often it is
the major and not a secondary product, is alcohol. The manufacture of rum,
the beverage alcohol, is described in detail by West and Harris, in Chapter
20. The enormous cane-based fuel alcohol production in Brazil is outlined
and discussed by Vaz-Rossell in Chapter 22.
The use of sucrose as a chemical feedstock is well established
(refs. 11, 12), although at present only in restricted practice. Khan, in
Chapter 23, discusses current and proposed systems for the use of sugar as
raw material for chemical process.
170
CANE SUGAR REFINING

Raw sugar, traded on a 96 pol basis, is converted by sugar refineries,
large processing plants of capacity from 300 to 3000 tonnes per day,
operating year-round, to a highly purified product, suitable for a diversity
of food, beverage and pharmaceutical uses, and stable under long-term
storage conditions. Refineries were traditionally located in cities or
other areas of high market density in North America and Northern Europe; the
few in cane-growing areas have in recent years increased in number as the
demand for refined sugar of consistently high quality by food processing
companies grows in those areas.
The refinery process is, again, a series of separations of non-sugars
from sucrose. Colorants and polysaccharides are major classes of non-
sugars: their nature and chemical behavior in process are discussed in
chapters 13 and 14), respectively. Ash, or inorganic components, is the
other major class of non-sugars. Raw sugar also contains more invert than
refined, and throughout all refining processes, decomposition of sucrose to
invert is a hazard. Although refining processes are generally run at pH
above 7, that is, at higher pH than the raw sugar factory, inversion can
occur at pH's up to 8.5.
The refinery process is outlined in Figure 3, and a comparison of the
composition of raw and refined sugars is presented in Table 5 (ref. 13). It
will be noted that a major non-sugar in raw sugar, and in refined, is water.
Of the organic non-sugars in raw sugar, polysaccharides and colorant are
each present on the order of 1000 ppm.
Raw Raw
Sugar Clarification Dacoiorization Vacuum I
Minaler . • ü ü l J Malfar Liquor Crystallization
ana* I Filtration'
Affination
Rhosphatation
Carbonation
Granular Carbon
Bona Charcoal
T~X
Refined White
and
Molasses
and
Ion Exchange Rasin
Brown Sugars Syrups
Fig. 3. Refined sugar manufacturing process in sugar refinery.

171
TABLE 5
Average analysis of typical raw and refined sugar.
Component Raw sugar (2) Refined sugar (Z)
Sucrose 97.73 99.95

Invert 0.56 0.006
Ash 0.45 0.007
Organic non-sugars 0.62 0.014
Water 0.64 0.023
Color 2000-5000 ICUMSA units 10-50
Particulate matter (bagacillo, silicates, waxes), which is not listed

in this analysis of soluble raw sugar components, is removed to levels well
below 100 ppm in refined sugar.
In addition to white sugars, in a range of crystal types, and in cubes,
tablets and powdered sugar forms, refineries produce various brown sugars
(known as "soft sugars" because of their flow characteristics) and syrups.
These contain some of the colorant and ash that are in the raw sugar and add
to the organoleptic properties of soft sugars. Flavor is an important
selling point for these products: flavor in brown sugars, among other sugar
products is explained and discussed by Godshall in Chapter 15.
Refinery processes
Raw sugar crystals are coated with molasses (mother liquor) from which
they were crystallized. The first refinery step is affination: the
mingling of raw sugar with a heavy syrup to a mixture that is spun in a
centrifuge with sprays of hot water washing off the syrup coating. Table 6
shows the effect of affination upon removal of non-sugars.
The next step is to melt (or dissolve) the washed raw sugar, to a melt
liquor of concentration of 68-73 Brix, which concentration is maintained, for
the rest of the process.
Clarification, where inorganic components, some polysaccharides and
colorant, and turbidity are removed, is accomplished either by phosphatation
or carbonatation process (known as carbonation outside North America).
These processes are discussed so thoroughly by Trott, in Chapter 17, that
they will not be discussed further here, other than to show an average
removal of non-sugars by the respective processes in Table 7 (ref. 13).
172
TABLE 6
Effect of affination process on raw sugar components (ref. 13).
Component Raw sugar (Z) Affined sugar (Z) Purification (Z)
Sucrose 98.45 99.52

Invert 0.50 0.17 66
Ash 0.45 0.15 67
Organic 0.60 0.16 73
Color
(ICUMSA units) 2000-6000 700-1500 66
TABLE 7
Removal of non-sugars by carbonatation and phosphatation (ref. 13).
Melter liquor Carbonatated liquor Phosphatated liquor

Component analysis (Z) analysis removal <*> analysis removal (Z)
Sucrose 99.52 99.60 __ 99.57

Invert 0.17 0.14 18 0.15 12
Ash 0.15 0.12 20 0.14 7
Organic 0.16 0.14 12 0.14 12
Color
(ICUMSA units) 12,000 700 42 800 33
The next process step is pressure filtration, through diatomaceous

earth filter aid after phosphatation, and through calcium carbonate mixed
with filter aid after carbonatation. Pressure filtration removes remaining
turbidity and suspended material and assists in color and polysaccharide
removal. The liquor emerging from the filter is clear and sparkling, but
still yellow in color.
Decolorization
The pale yellow liquor must be decolorized to a level low enough to
give refined sugar quality color upon crystallization (10-50 ICUMSA units).
To reach this level, some 80-90Z of color in the clarified-filtered liquor
must be removed. Decolorization processes, using both traditional bone char
and granular activated carbon adsorbents, ion-exchange resin (bead and
powdered forms) and the new SURE technology, are described by Trott in
Chapter 17. Bone char and desalting ion-exchange resins will remove
inorganic salts as well as color, and all adsorbents will remove some
turbidity, by simple physical adsorption during filtration through a bed of
medium particles.
Crystallization is the final purification process in the refinery. A
series of four crystallizations under vacuum are usually performed, with
crystals from each stage containing a level of about 10Z of the level of
impurities in the mother liquor, and yield at each step of about 50Z by
weight. A standard four-boiling system, as used at California and Hawaiian
sugar refinery in Crockett, California, is shown in Figure 4. There are
many excellent reference books on sugar boiling and vacuum pan technology
(refs. 1-4, 8, 9). Good pan circulation is most important to prevent
decomposition of sucrose (see Chapter 16) and minimize loss of sucrose with
concomitant color formation.
»0.1 LIQU01 » 0 . 2 + 3 LIQUOR
7É-77· Iris 62· Sri*

4,000 CPH
1
Γ
1
1A STRIKE
1
IB STRIKE 1C STRIKE j ID STRIKE
3 Fane Syrup t 2 Fana Strup r 1 Fan .syrup. J Π«

14 Strlkaa/Day 14 ftrlkea/Day 14 Strikes/Bay | 12 Strike/Day
i 1 J SYRUP TO
REMELT j
STATION '
CRAWUTED SUGAI 1LEHD 1
25-35 Color Volt·
VET SUGAR
(VI
I VET SUGAR TO WtTIMC MELT SYSTEM
Fig. 4. Granulated sugar crystallization system; C and H Sugar, Crockett

Refinery.
Mechanical stirrers are strongly effective in increasing circulation,

and fortunately are being utilized now almost routinely, both in refineries
and raw sugar factories.
17^
Syrups from the fourth strike and other low purity liquors are
recirculated through a remelt or recovery system; a low-purity dark colored
sugar is crystallized, remelted, and sent through process again.
The first four strikes of white sugar, dried and cooled, make up the
refined white sugar sold in 1, 2, 5, and 10 lb bags to the home market in
the U.S., and in bulk or large bags to the industrial market (see Table 5
for composition of sugar). Refined brown sugars (Chapter 15) are produced
from liquors that are purified but not decolorized.
The importance of sugar as a food processing ingredient is explained by
Tsang, in Chapter 18. Bollenback, in Chapter 19, discusses some questions
on sugar, diet and health that have sometimes been raised in the United
States, although seldom in the tropical countries where overall shortage of
food creates an increasing demand for sugar each year. As standards of
living increase, the demand for sugar, and for refined white sugar,
increases. Much of the increased demand comes from food processing
companies, which supply the soft drinks and processed foods that are
consumed in increasing amounts when living standards and disposable income
increase. Food processors demand high quality and set tight specifications.
These demands are being met by increased export of refined products from
North America and the European Economic Community, and by increased refinery
capacity in the tropics.
In some areas, new refineries have been built to create this capacity.
These are usually attached to a sugar factory, to share the available energy
from bagasse burning, and to simplify the refinery process by recycling
lower purity materials back to the raw sugar factory (ref. 14).
There has traditionally been white sugar, called plantation white or
mill white, produced in cane-growing areas, usually by sulfitation of cane
juice, often in combination with carbonatation (refs. 1, 2, 4, 8, 10). In
the last 10 years new juice and syrup clarification systems have become
available to improve the quality of direct production white sugars. In
Chapter 17, Trott describes these new milling and refinery processes,
particularly the Talodura syrup clarification process, and Blanco Directo
white sugar process, of Tate and Lyle Ltd. (ref. 14).
The following chapters are intended to acquaint the reader with the
basic process chemistry and innovations in technology in the manufacture of
sugars from sugarcane, with additional information on added-value
175
byproducts, such as fuel alcohol, chemicals and rum, on new products, and on
the food technology uses and health and nutrition aspects related to
sucrose.
REFERENCES
1 G. P. Meade and J. C. P. Chen, Cane Sugar Handbook, 11th ed., Wiley-
Inter science, New York, 1984.
2 V. E. Baikow, Manufacture and Refining of Raw Cane Sugar, 2nd ed.,
Elsevier, Amsterdam, 1982.
3 J. H. Payne, Unit Operations in Cane Sugar Production, Elsevier,
Amsterdam, 1982.
4 E. Hugot, Handbook of Cane Sugar Engineering, 2nd ed., Elsevier,
Amsterdam, 1972.
5 E. Whayman and 0. L. Crées, Mechanistic studies of cane mud floccu-
lation, Int. Soc. Sugar Cane Technol., 1974, 1175-1182.
6 G. S. Shephard, The influence of raw cane juice constituents on juice
clarification, Int. Sugar J., 83, (1981) 330-334.
7 M. C. Bennett, Flocculation technology in sugar manufacture, Sugar
Ind. Technol., 34 (1975) 22-32.
8 P. Honig, Principles of Sugar Technology, Vol. 1, Elsevier, N.Y.
1953; Vol. 2, Elsevier, N.Y., 1959; Vol. 3, Elsevier, N.Y., 1962.
9 P. M. Silin, Technology and beet-sugar production and refining, Transi.
from Russian, Israel Program for Scientific Translation, Office of
Technical Services, U.S. Dept. of Commerce, Washington, 1964.
10 M. A. Clarke, The future of raw sugar quality, Sugar y Azucar,
(1985) 32-49.
11 J. M. Paturau, By-Products of the Cane Sugar Industry, 2nd Ed.,
Elsevier, Amsterdam, 1982.
12 V. Kollonitsch, Sucrose Chemicals, Int. Sugar Res. Foundation,
W.S. Cowell, Ltd., Ipswich, 1970.
13 J. C. Abram and J. T. Ramage, Sugar Refining: Present Technology and
Future Developments, in: G. G. Birch and K. J. Parker (eds.),
Sugar: Science and Technology, Applied Science and Technology, London,
1979, pp. 49-96.
14 M. A. Clarke, The sugar refining industry today, Sugar y Azucar Year
book, Ruspam Press, New York, 1983, 71-94.
176
Chapter 12
VARIETAL DIFFERENCES IN THE CHEMICAL COMPOSITION OF SUGARCANE.
BENJAMIN L. LEGENDRE
A g r i c u l t u r a l Research S e r v i c e , USDA, Sugarcane Research U n i t , Houma,

Louisiana 70361
SUMMARY
The sugarcane stalk consists of approximately 75% water with the remainder
divided between fiber and soluble solids. Fiber is a general term to
describe the solid residue left after the juice (water and soluble solids) is
extracted by milling, and consists primarily of cellulose and lignin. The
proportion of both fiber and soluble solids increases as the cane matures,
and the relative rate of increase and the relative level of these components
at maturity are heritable. Sucrose, the major component of soluble solids at
maturity and the economically most important constituent of sugarcane, can
reach a concentration in the juice of over 20%. Besides water and sucrose,
other constituents of the juice are glucose, fructose, minerals, protein, gum
and polysaccharides, organic acids and miscellaneous minor constituents.
These constituents may vary in concentration with age of stalk, varieties,
weather patterns and plant growth, nutrient uptake and availability, crop
damage due to insects, diseases, wind damage and early freezes, and
application of chemical ripeners. This paper will discuss the differences in
chemical composition of the juice of the cane stalk as it relates to
varieties, with emphasis on sucrose, glucose, fructose, deterioration
products, mineral constituents and colorants.
INTRODUCTION
The cane stalk, without top, leaves, stubble or roots, is composed of
approximately 75% water, and the remainder is divided between fiber and
soluble solids. According to Irvine (ref. 1), the amount of each of these
three components is genetically determined and varietal differences are well
known. The noble varieties (Saccharum officinarum) contain a greater
proportion of water and are relatively low in fiber and sucrose. The
interspecific hybrids are higher in both fiber and sugars. The chemical
composition of the cane plant varies greatly and is influenced by several
internal and external factors, including variety, age of cane, conditions
under which the cane is grown, soil type, fertilizers, water and pests and
diseases (ref. 2). The soluble solids of the juice of the cane stalks, in
order of abundance, are sucrose, glucose and fructose; minerals; waxes, fats,
and phosphatides; and miscellaneous minor constituents generally listed as
percentage of juice solids (Table 1).
177
TABLE 1
Composition of sugarcane and juice solids*
Millable cane Cane (%)
Water 73 - 76
Solids 24 - 27
Soluble solids 10 - 16
Fiber (dry) 11 - 16
Juice Soluble
constituents solids (%)
Sugars 75 - 92
Sucrose 70 - 88
Glucose 2-4
Fructose 2-4
Salts 3.0 - 4.5
Inorganic acids 1.5 - 4.5
Organic acids 1 - 3
Organic acids 1.5 - 5.5
Carboxylic acids 1.1 - 3.0
Amino acids 0.5 - 2.5
Other organic nonsugars
Protein 0.5 - 0.6
Starch 0.001 - 0.050
Gums 0.3 - 0.6
Waxes, fats, phosphatides 0.05 - 0.15
Other 3-5
* Irvine, Cane Sugar Handbook, 10th ed., 1977, p. 16.
When cane is cut and cleaned by hand, and delivered fresh, the processor
receives the best possible starting material for sugar production; however,
in many areas of the world, including Louisiana, sugarcane is cut and loaded
by machine and the cane is cleaned by burning. As a result, sugarcane
delivered to the mill may contain tops, leaves, stubble and roots, as well as
soil, water and other extraneous material. Irvine (ref. 1) has shown that
juice from tops - including the stem tip as well as leaf blades, sheaths, and
rolls - contains less than 1% sucrose and is relatively rich in starch,
soluble polysaccharides and reducing sugars (Table 2).
178
TABLE 2
Carbohydrate constitutents of sugarcane parts, CP 65-357*
Portion Juice analyses (mg/ml)

of Soluble
Part whole (%) polysacch. Sucrose Fructose Glucose
Leaf blade 11.1 5.40 7.72 3.75 6.76

Leaf sheath 4.3 4.03 14.20 3.33 5.92
Leaf roll 2.0 5.58 6.85 7.56 13.60
Stem tip 1.6 5.90 14.82 12.94 17.52
Millable cane 66.1 1.51 163.07 4.74 5.69
Stubble 9.3 2.01 152.50 3.01 5.94
Roots 1.3 1.28 8.76 1.25 2.50
Dead leaves 4.3 5.42 0 0 0
* Irvine, Cane Sugar Handbook, 10th ed., 1977, p. 17.
The differences in chemical composition of the juice of the cane stalk are
considered in the remainder of this paper as it relates to varieties grown in
Louisiana with specific emphasis on sucrose, glucose, fructose, deterioration
products, mineral constituents and colorants. These differences may be
regarded as representative for varieties grown all over the sugarcane
producing world.
Sugars and other carbohydrates
Sucrose in the juice and cellulose in the fiber are the two main
constituents of sugarcane, and both are formed through the bonding of simple
sugars. The simple sugars glucose (dextrose) and fructose (lévulose) occur
free in sugarcane, usually in lesser amounts than sucrose. The commercial
production of sugar from sugarcane juice is based on the crystallization of
sucrose while glucose and fructose remain dissolved.
Sucrose
Sugarcane varieties in Louisiana are grouped into three intergrading
maturity classifications based on their relative sucrose content - early,
mid-season and late (ref. 3 ) . Sucrose levels generally rise rapidly during
the first four to six weeks of the harvest season followed by a moderate
increase or even a slight decrease during the remaining weeks of the harvest
season. The relative changes in juice quality during the harvest season of
179
seven varieties, CP 63-357, CP 70-321, CP 72-356, CP 72-370, CP 74-383,

CP 76-331 and CP 79-318, is shown in Table 3. As an example, the variety
CP 65-357 continues to accumulate sucrose throughout the sampling period,
while CP 72-370 does not after midseason. The sucrose content of many
varieties can be modified by the use of chemical ripeners.
TABLE 3
Average normal juice sucrose (NJS) of seven commercial
varieties in the plant cane crop grown in Louisiana
during 1986
NJS (%)
Dates of harvest
Variety Oct. 15 Nov. 15 Dec. 15
CP 65-357 11.79 C c 14.06 B ab 14.97 A a

CP 70-321 11.76 B c 13.96 A b 14.14 A abc
CP 72-356 11.56 B c 13.76 A be 13.37 A c
CP 72-370 12.93 B a 14.34 A ab 14.13 A abc
CP 74-383 11.31 C c 13.09 B c 13.56 A be
CP 76-331 12.61 B ab 14.73 A a 14.84 A a
CP 79-318 11.97 C be 14.33 B ab 15.03 A a
Means followed by the same letter are non-significant at
P - 0.05. Capital lettering, read across; lower case,
read down.
Table 4 shows the effects of the chemical ripener glyphosate on three

varieties, CP 65-357, CP 70-321 and CP 72-370. When applied four to six
weeks prior to harvest, the sucrose content of most varieties can be
increased by as much as 30% (ref. 4 ) . The delay of harvesting beyond six
weeks after application will usually mean a reduction in the advantage from
its use, due to the deleterious effect of the ripener on plant growth.
180
TABLE 4
Varietal differences in response to the chemical ripener
glyphosate as measured by average increases in recoverable
sugar per stalk over time after treatment*
Recoverable sugar/stalk (g)

Treat Treatment-harvest interval ([weeks )
Variety ment 0 2 4 6 8
CP 65-357 Control 44.4 65.3 85.7 90.2 101.5

Glyphosate — 70.5 103.0t 118.4t 110.4
CP 70-321 Control 49.9 66.9 79.5 93.8 112.3

Glyphosate — 77.9 101.Of 111.6t 138.0t
CP 72-370 Control 67.7 78.3 80.1 106.0 117.1

Glyphosate — 88.2 108.6t 120.4t 115.0
* Glyphosate applied 09/04/86 at 0.34 kg/ha
t Indicates significant difference from control P β 0.05
Glucose and fructose

As sugarcane matures the concentration of sucrose usually increases and
both glucose and fructose decrease. In a 1983 study (unpublished data), the
concentrations of glucose and fructose as determined by high performance
liquid chromatography (HPLC) were significantly different among three
varieties, CP 65-357, CP 72-356 and CP 72-370, at the beginning of the
sampling period and decreased differentially during the sampling period
(Table 5). The early maturing, high sucrose variety, CP 72-370, had about
two thirds the concentration of glucose and fructose on the early sampling
date when compared to the mid- to late-maturing varieties, CP 65-357 and
CP 72-356. By the last sampling date, when all three varieties had matured,
the concentration of both sugars in all three varieties was low.
181
TABLE 5
Average glucose and fructose content in juice of
three commercial varieties in the second ratoon
crop grown in Louisiana during 1983*
Dates of harvest
Oct. 1 Nov. 1 Dec. 1
GLU FRÜ GLU FRU GLU FRU
%
CP 65-357 1.10 1.01 0.36 0.45 Tt 0.06
CP 72-356 1.12 0.93 0.53 0.59 0.06 0.28
CP 72-370 0.68 0.66 0.16 0.22 T 0.12
* GLU s Glucose; FRU s Fructose
t T * Trace
Deterioration products
The quality of sugarcane varies with age, increasing toward an optimum,
then gradually declining. Once the cane is cut, deterioration begins almost
immediately. Deterioration may also begin before harvest in pest-ridden cane
or in fields affected by fire, freezes or wind storms.
After cutting, sugarcane begins to lose water, giving an apparent increase
in sucrose content. At the same time, the activity of the enzyme invertase
becomes more evident as some sucrose is inverted to reducing sugars. The
rate of inversion is a genetic trait and can be minimized by variety
selection. Sucrose inversion varies with temperature and moisture and is
most rapid in hot, dry periods.
In addition to inversion, serious loss of sugar from other causes can
occur in harvested cane. Burning, freezing, and mechanical damage affect
deterioration in a similar manner. According to Irvine (ref. 5), these
conditions damage both the protective rind of the stalk and the underlying
sugar storage cells. The stalk is invaded and the sugars from the dead and
dying cells is utilized by Leuconostoc spp. bacteria that can change sucrose
into dextran. A small amount of dextran causes increases in juice viscosity,
slows crystallization, reduces sugar yield, and produces needle-grain
crystals. Dextran reduces quality throughout processing, and carries over
into the refined sugar.
182
Legendre et al. (réf. 6) measured changes in juice composition of eight

varieties as affected by post-freeze deterioration in Louisiana. Christmas
Day, 1983, brought a record daily low temperature for December of ~10.6°C
(13°F) at Houma, LA, as the harvest was being completed. Results of this
study showed that cane sampled just prior to the freeze produced juice of
excellent quality with the expected significant varietal differences in
TABLE 6
Varietal differences in the chemical composition of
sugarcane juice following a severe freeze*
Constituents
Days after freeze (-10.6°C)
Variety =2 10 12 15
Dextran (ppm, dry weight x 1000, Roberts' test)

CP 70-321 0.206 0.941 6.837 18.958
CP 72-370 0.771 8.077 15.210 24.226
L 65-69 0.246 14.243 10.547 10.751
LSD .05 NS 6.148 6.887 10.045
Titratable acidity (ml 0.1N NaOH/10 ml juice)

CP 70-321 1.96 2.14 3.20 6.68
CP 72-370 2.14 2.82 3.83 7.84
L 65-69 2.38 3.21 5.56 7.79
LSD .05 0.24 0.32 0.68 NS
Sucrose (%, HPLC)

CP 70-321 18.26 16.87 15.15 12.57
CP 72-370 17.89 16.18 14.74 11.94
L 65-69 16.31 14.21 9.66 7.02
LSD .05 1.01 0.87 2.51 4.70
Fructo se (%, HPLC)

CP 70-321 0.19 0.59 1.33 1.74
CP 72-370 0.17 0.74 1.68 2.18
L 65-69 0.31 1.42 2.90 3.43
LSD .05 0.02 0.63 0.80 1.27
Glucose (%, HPLC)

CP 70-321 0.18 0.46 0.51 0.52
CP 72-370 0.03 0.35 0.42 0.46
L 65-69 0.27 0.33 0.25 0.28
LSD .05 0.06 NS 0.14 0.07
*Legendre, et al., Proc. of the 1984 Sugar Processing
Research Conf., 1986, p. 95, 97, 99.
sucrose and titratable acidity and with low dextran content. Samples taken
at 10 and 12 days after the freeze showed extensive changes in juice
composition and, by the 15th day, all varieties were unacceptable for
processing. There were significant varietal differences in sucrose,
titratable acidity and dextran content. A closer look at the composition of
sugarcane juice of three of the varieties, CP 70-321, CP 72-370 and L 65-69,
as affected by post-freeze deterioration showed an increase in fructose on
each date of analysis (Table 6). On the other hand, the glucose content
increased relatively much less among the pre-freeze and post-freeze sampling
dates. Total invert sugar (glucose and fructose) was lowest on the
pre-freeze harvest date while sucrose was highest; after the freeze, the
concentration of invert sugar (predominantly fructose) and dextran increased;
the concentration of sucrose decreased. The loss of sucrose is much greater
than the increase in other sugars.
Mineral constituents
Irvine (ref. 1) stated that the mineral nutrients of greatest concern in
processing are those dissolved as salts in the juice (Table 7) rather than
those that are constituents of organic molecules. The mineral content (as %
solids) tends to increase with the plant1s age, or at least, to remain
TABLE 7
Mineral concentrations in sugarcane juice*
Concentration
Constituent (% solids)
Potassium (K20) 0.4 - 1.4

Sodium (Na20) 0.03 - 0.10
Sulfate (SO3) 0.11 - 0.52
Chloride (Cl) 0.10 - 0.29
Calcium (CaO) 0.17 - 0.32
Magnesium (MgO) 0.20 - 0.33
Silica (S1O2) 0.06 - 0.71
Phosphate (P 2 0 5 ) 0.01 - 0.40
Iron (Fe203) 0.06 - 0.14
Sulfated ash 3.6 - 4.4
Conductivity ash 3.4 - 4.4
* Clarke, Cane Sugar Handbook, 11th ed.,
1985, p. 29
181*
constant. Potassium is the most abundant mineral element in the juice;

unlike others, it is most abundant in the younger parts and decreases in that
from the older parts of the stalk (ref. 1). The high ash content of tops and
the effect of potassium on sucrose crystallization give processors incentive
to avoid milling cane tops and associated green leaves.
In some early experiments in 1937, Fort and Holmes (ref. 7) found
differences between two varieties in soluble ash, potassium, nitrogen and
phosphorus in both the top and the stem of sugarcane. In a more detailed
study two years later, Fort and McKaig (ref. 8) found differences in ash
content among eight varieties.
Colorants
Phenolics are important from a processing standpoint because of their
reactivity with metals and their ability to produce highly colored reaction
products. As much as two thirds of the color in cane juice may be due to
enzymatic browning of phenolic acids (ref. 9). In a recent study, Godshall
and Legendre (ref. 10) showed that there was a significant difference in the
phenolic content among four varieties, CP 65-357, CP 70-321, CP 74-383 and
NCo 310, with CP 70-321 having the lowest total phenolic content and
CP 65-357 the highest (Table 8). The study also showed that phenolics
increased as the cane matured in all four varieties.
TABLE 8
Concentration of phenolics in sugarcane varieties
in Louisiana during maturation period*
Mean caffeie acid equivalent (ug/ml)

Dates of harvest
Variety Sept. Oct. Nov. Dec.
CP 65-357 479 a 816 a 653 a 1027 a

CP 70-321 296 c 262 c 155 c 628 c
CP 74-383 368 be 459 b 539 b 807 b
NCo 310 430 ab 360 c 456 b 820 b
Means within a column followed by the same letter
are non-significant at Pe0.05
* Godshall and Legendre, Int. Sugar J., 90(1),
1988, 16-19.
185
REFERENCES
1 James E. Irvine, Composition of cane and juice, in: George P. Meade and
James C.P. Chen (Ed.), Meade-Chen Cane Sugar Handbook, John Wiley and
Sons, New York, 1977, pp. 15-29.
2 C. Van Dillewijn, Botany of Sugarcane, Chronica Botanica, Waltham,
Massachusetts, 1952.
3 B.L. Legendre, Changes in juice quality of nine commercial sugarcane
varieties grown in Louisiana, J. Am. Soc. Sugar Cane Technol. 4 (1985)
54-57.
4 B.L. Legendre and C.K. Finger, Response of sugarcane varieties to the
chemical ripener glyphosate, Proc. 14th Annual Plant Growth Regulator
Soc. of Am. Meet., Honolulu, HA, Aug. 2-6, 1987, pp. 479-484.
5 James E. Irvine, Sugar Cane, in: George P. Meade and James C.P. Chen
(Ed.), Meade-Chen Cane Sugar Handbook, John Wiley and Sons, New York,
1977, pp. 1-14.
6 B.L. Legendre, W.S. Charles Tsang and M.A. Clarke, Changes in juice
composition of sugarcane as affected by post-freeze deterioration in
Louisiana, Proc. 1984 Sugar Processing Res. Conf., New Orelans, LA, Oct.
16-18, 1984, USDA-ARS, ARS-49, 1986, pp. 92-107.
7 C.A. Fort and R.L. Holmes, Preliminary studies on the composition of the
whole cane of varieties Co 281 and Co 290, Sugar Bull., 16(2) (1937)
2-4.
8 C.A. Fort and M. McKaig, Comparative chemical composition of juices of
different varieties of Louisiana sugarcane, U.S. Dept. Agric. Tech. Bull.
688, (1939) 1-68.
9 B.C. Goodacre and J. Coombs, Formation of colour in cane juice by
enzyme-catalyzed reaction. Part II. Distribution of enzymes and colour
precursors, Int. Sugar J., 18 (1978) 323-326.
10 M.A. Godshall and B.L. Legendre, Phenolic content of maturing sugarcane.
Int. Sugar J., 90(1) (1988) 16-19.
186
Chapter 13
THE NATURE OF COLORANTS IN SUGARCANE AND CANE SUGAR MANUFACTURE

RICHARD RIFFER
SUMMARY
The colored substances in sugarcane are flavonoids subject to enzymatic

oxidation to more highly colored species during early stages of processing.
Other classes of colorants are those resulting from Maillard reactions and
chemical degradation. The latter is especially significant at high pH, and
one subclass, caramelization, is generally considered separately, because it
is primarily a thermal process.
The colorant fraction encountered in the raw sugar refinery is highly
heterogeneous. The simple phenolic substances most accessible to study
represent only a small fraction of the total. Colorants in the higher
molecular weight range (up to about 50,000 daltons) are without definite or
constant composition and hence difficult to study. These substances have
nonetheless been characterized, using such criteria as polarity, molecular
size, functional groups, and pH sensitivity. Such data are valuable to
refiners, who have the task of removing color from raw sugar efficiently and
at low cost.
INTRODUCTION
The colorants in cane sugar products include not only naturally

occurring plant pigments but also a large number of substances produced
during processing. The colorant fraction is highly heterogeneous. Because
of differences encountered in raw juice composition, from genetic and
environmental factors, and almost limitless possibilities for variations in
color-forming reactions during sugar manufacture, it is likely that no two
colorant fractions are exactly alike. Furthermore, because no definite or
constant composition can be expected in most of the colorants, usual
criteria of purity cannot be applied to these substances.
Most of the work done toward elucidation of the structures of specific
colorants has focused upon low molecular weight compounds, because these are
most easily isolated and studied. However, such materials account for only
a small portion of the colorant fraction. These relatively simple molecules
are nonetheless of great interest because they are likely to be incorporated
into more complex species by further reaction. Moreover, some of these
materials exhibit remarkable stability, surviving at trace levels even in
granulated sugar.
167
COLOR MEASUREMENT
The current practice of measuring sugar color at 420 and 560 nm can be
traced back at least 60 years to Peters and Phelps (ref. 1 ) . By way of
comparison, commercial instruments for UV/visible spectrometry go back only
about 40 years. The 560 nm absorbance is typically much smaller than that
at 420 nm, because non-sugars that absorb in the visible zone do so
primarily at the violet to blue end of the spectrum, that is, at the edge of
the UV region. Thus the visual color is yellow to yellow-orange. For this
reason, a single measurement at 420 nm will often suffice for
characterization of a sugar color, even though most of the visible spectrum
is excluded.
Readings at 560 nm are sometimes made for dark solutions, for which the
420 nm reading would be off scale except under high dilution that could
introduce large error. However, since the visual 560 nm color, violet, is
not ordinarily encountered in sugar products, clearly what is being measured
is the tail end of a peak with a maximum rather distant from 560 nm.
Measurements at 420 and 560 nm are obviously not comparable, so a wavelength
must be specified for any reported reading.
It is sometimes recommended that sugar colors be measured in the UV
region of 270-280 nm, where most colorant fractions exhibit maxima.
However, from a practical point of view, it seems preferable to characterize
sugars in the visible region, since the ultimate basis for specification of
sugar color is, after all, the visual appearance to the human eye. It
should be noted that the spectral similarities of various colorant fractions
exist because they contain carbonyl groups in common (^m.ac 270-285); UV is
insufficiently informative to permit differentiation between the fractions,
which differ morphologically in important ways.
The absorbancy of a non-scattering solution obeys the familiar Lambert-
Beer law. However, Deitz (ref. 2 ) , noting the turbidity of commercial sugar
solutions, defined an attenuation term to express the loss of light from a
beam source by both absorption and scattering. Thus sugar solution colors
are sometimes reported using an "attenuation index" rather than a specific
absorptive index.
Higher wavelength measurement, such as at 720 nm, is sometimes used to
eliminate absorption for determination of turbidity by attenuation of the
source beam. Although it is true that some sugar color is colloidal (some
168
caramel components have radii approaching 50Â), a portion of this is due to

complexation of colorants with macromolecules such as polysaccharides and
proteins. Turbidity and color are mutually interfering, but they result by
and large from distinctly different chemical species. The turbidity
components are often eliminated (together with a portion of the color) by
filtration through a 0.45μ membrane prior to color measurement. Although
turbidity measurement is beyond the scope of this article, it should be
noted that nephelometry (measurement of 90° scatter) is far more sensitive
to low turbidities than is attenuation.
The colorant fraction can be sufficiently heterogeneous that no single
measurement can fully characterize it. For example, Riffer (réf. 3) found
that both 420 and 560 nm readings are insensitive to color resulting from
catechols complexed with iron. Such substances are believed responsible for
greenish hues in brown sugars with a high iron content (ref. 4).
Sugar products displaying low color levels generally appear yellow. In
dark samples containing high molecular weight polymerization products, these
highly absorbing substances contribute a black pigmentation, resulting in
the familiar brown hues.
Measurement of sugar color is generally made at pH 7.0, the target pH
of most sugar processing. This pH is not necessarily optimal, because (a)
sucrose is actually most stable closer to pH 8; and (b) color measurement
has a high pH dependence and a curve of color vs. pH displays a large
derivative near pH 7 (ref. 5), with an inflection point near pH 8.5 at
420 nm. On the other hand, colors measured in the pH range 10.5 to 12 are
insensitive to small pH changes but relatively unstable with respect to
time.
Brown sugars are sometimes characterized by reflectance color, because
brightness and luster are considered desirable attributes of such products.
In this case, the "color" depends not only on colorants present but also on
grain size, moisture, and the presence of suspended solids and colloidal
substances in inclusions and in the syrup coating the crystals. Red source
lamps are commonly used for such measurements because red reflectance seems
to correlate best with a product that is attractive to the eye; that is, one
that is glossy golden brown or reddish brown. Such reflectance color will,
of course, be highest when green impurities, usually associated with iron
contamination, are absent.
Isolation of color components
Decolorization in the refinery can be carried out most efficiently when
knowledge of the properties of colorants is applied to process design and
operation. Toward this end, numerous studies have been carried out to
isolate sugar colorants and fractionate them into individual components.
Attempts using dialysis (refs. 6, 7), electrophoresis (refs. 8-11),
molecular sieves (refs. 12, 13), ion exchange chromatography (ref. 14), and
two dimensional TLC (ref. 15) have met with varied success. Molasses has
often been used as a concentrated source, but this practice is open to the
criticism that molasses colorants are not representative of those in raw
sugar. In addition, published articles on this subject don't always
distinguish between cane and beet colorant sources, from which phenolic
components are very different--catecholamines in beet, flavonoids in cane,
phenolic acids in both.
Studies on colorant fractions indicated that they could be categorized
into distinct groups, viz.. naturally occurring phenolics and their reaction
products; melanoidins from reactions between carbohydrates and amino acids;
and sugar degradation products. This last division includes distinctly
different colorants depending upon pH of formation, and contains a subclass,
caramelization products, generally considered separately because they result
from thermal processes, whereas the other reactions proceed slowly even at
room temperature.
Some colorants do not fit into these categories. For example,
phenolics can also react with amino compounds, and polyphenols can form
adducts with proteins, for which they have a high affinity. Likewise,
polysaccharide preparations frequently contain colored material; such
complexation is probably one mechanism by which some color is removed during
clarification and filtration. Roberts and Godshall (ref. 16) found that
some polyphenolic acid colorants are attached to an indigenous cane
arabinogalactan by ester linkages.
Smith (ref. 13) estimated that for a 98.5° pol raw sugar, colorants
accounted for about 16-202 of the weight of the non-sugars. For granulated
sugar, he estimated the colorant level to be about 30 ppm (ref. 17).
190
pH sensitivity
The raw sugar colorants that the refiner must deal with vary widely in
ease of removal. Everyone on the technical side of the industry has heard
of raws containing seemingly normal color levels but which stubbornly
resisted decolorization. Smith (ref. 18) found that colorants exhibiting a
high pH sensitivity were most easily removed by bone char. The "indicator
value" or I.V., so named because of the color change observed with pH, is a
measure of this sensitivity; it is the ratio of 420 nm liquor color at
pH 9.0 to that at pH 4.O. Thus "good" raws — those easily decolorized on
bone char--are those with high I.V. values.
The very fact of indicator-like behavior suggests certain chemical
features of the colorants. The second derivative of 420 nm color vs. pH
displays a single maximum near pH 8.5 ± 1. This figure suggests, but does
not prove, that phenolics are responsible for the pH sensitivity; in fact,
there are also non-phenolic contributors. The color change is primarily one
of intensity rather than hue, unlike the sharp, brilliant color changes
exhibited by members of the sulfonephthalein series.
The colorants displaying the highest degree of pH sensitivity are in
fact polyphenols, free flavonoids, and their glycosides. Caramelization and
alkaline degradation products are intermediate in I.V., with melanoidins at
the low end of the scale. Sensitive colorant can be converted under
suitable conditions to darker insensitive colorant. This suggests a
condensation involving phenolic groups. In the case of caramels, those
formed under acid conditions often contain highly pH-sensitive components,
but these can be destroyed under alkaline conditions, leaving less sensitive
residual products.
Smith (ref. 18) found that liquor pH sensitivities increase as a result
of affination, defecation, and crystallization. In char column tests, he
found that pH-sensitive color is adsorbed at a greater initial rate than
insensitive color. The pH-sensitive color was also found to have a greater
tendency to be included in the crystal during crystallization of bone char
decolorized liquors.
In other work, Smith (ref. 13) observed that high molecular weight,
less pH-sensitive colorants show a higher absorption over most of the
visible region than low molecular weight, more pH-sensitive colorants.
Thus, of two liquors with the same 420 nm color, the sample containing less
191
-sensitive colorant will be visually darker. In other words, comparison

420 nm colors is valid only for samples containing similar distributions
colorant I.V. This stipulation is almost never observed in practice and
Lustrates one of the pitfalls of using 420 nm color alone as an indicator
expected ease of refining a raw sugar.
sctral and molecular weight data

Molecular sieves and dialysis are among the techniques that have been
3d to provide molecular weight data on colorants. Smith (ref. 13), using
5 former, found that colorants from various refinery streams ranged in
Lecular weight from below 350 to above 50,000 daltons. Using dialysis,
Derts and Godshall (ref. 19) observed that 46-932 of the color in refinery
reams had a molecular weight above 12,000, 83-93Z greater than 8,000, and
-99Z greater than 3,500. The implication was that flavonoids are
Latively minor contributors but could nevertheless be of great
jnificance if they persist in later stages of refining. On the other
id, Smith (ref. 20) estimated that about half the color of Hawaiian raw
jar was due to naturally occurring pH-sensitive colorants. Tu (ref. 21),
/ever, reported that although the principal colorants in cane juice and
:up were of relatively low molecular weight (<5,000), the major colorants
raw sugar were of molecular weight above 5,000. Tsuchida and Komoto
if. 22) found that most molasses color was in the molecular weight range
10,000-50,000.
Despite certain inconsistencies, these data generally support the view
it flavonoids are the principal colorants in fresh cane juice, but that
jher molecular weight reaction products begin forming at early stages of
)cessing, until such substances predominate in dark streams late in the
)cess. The yellow colors observed at early stages are of phenolic origin,
»reas later high molecular weight species appear to contain linear
ljugated systems containing at least six double bonds, the minimum
îonance pathlength requirement for absorption in the visible region. High
.ecular weight colorant can be characterized by absorption spectra, and
:hough these tend to be only moderately informative, ethylenic and
rbonyl unsaturation are clearly indicated.
The colorants on the crystal surface film are of significantly lower
.ecular weight than those occluded within the crystal (ref. 16, for
192
example). Cane sugar colorants are known to have a much higher molecular
weight range than those from beet (ref. 23). Since such colorants are
selectively occluded in crystal formation, white sugars can be obtained from
darker beet liquors than cane liquors. Color adsorption in the cane crystal
also occurs on different faces from those in the beet crystal (ref. 24).
Acidity and charee density

The pK values of caramels, melanoidins, and alkaline degradation
products have been estimated to be 4.7, 6.9, and 5.1, respectively (ref.
25). The phenolic pigment fraction would be expected to have a value near
that of catechol, 9.4.
Using these figures and the Henderson-Hasselbalch equation, one can
calculate to what degree these classes of compounds are ionized at refinery
pH, say 7.2:
X ionized
Caramels 99.7
Melanoidins 66.7
Alkaline degradation products 99.2
Flavonoids 0.6
These values suggest that for decolorization mechanisms dependent

wholly upon ionic charge, caramels and alkaline degradation products would
be removed most effectively, then melanoidins, and last flavonoids. For
phenolic substances to be ionized to a degree of 66Z, a pH of about 9.7
would be required. Removal of unionized substances would require van der
Waals or dipole-dipole interaction, or non-polar hydrophobic bonding. These
non-ionic mechanisms might appear to predominate in bone char
decolorization, because color removal is improved by lowering the pH of the
liquor before treatment (ref. 13). However, other effects also contribute to
this improvement, because colorant solubility and molecular size are also
influenced by pH.
Considerations of acidity alone cannot fully characterize the ionic
charge on colorants. Most colorants contain more than five anionic charges
per molecule (ref. 26). Williams (ref. 27) found that colorants display an
almost linear relationship between molecular weight and net charge in the
range of 300-1000 daltons, but of course this region encompasses only a
small fraction of the total color.
193
Under conditions of high pH, or in the absence of salts, high

molecular weight colorant molecules are expanded by the mutual repulsion of
their charged groups to extended rod-like configurations of large rotational
cross-section, similar to the polyethylene glycols. With high ionic
strength, however, or at low pH where the colorants approach their
isoelectric points, the molecules contract to a random-coil or globular
configuration, so that their molecular size is close to that of dextrans of
the same molecular weight (refs. 28, 29). Such transformations obviously
influence both kinetic and thermodynamic components of the adsorption
process.
More polar colorant solutes should be best adsorbed at polar sites,
such as on a bone char hydroxyapatite surface or at resin ion exchange
sites. Less polar species are best removed at non-polar sites, such as on
carbonaceous adsorbents or on a styrene resin backbone. Both surface types
are available in bone char and anionic resin, hence their high degree of
effectiveness in decolorizing sugar liquors.
These considerations are thermodynamic ones: the interaction between
solute and adsorbent must be stronger than dispersive forces for
decolorization to take place. In addition, high molecular weight colorants,
particularly near their isoelectric points, have an unfavorable entropy of
solution, due to ordering of water molecules around the solute, which serves
as a driving force for removal. However, kinetic factors are important:
diffusion to the adsorption site is rate limiting for high molecular weight
colorants.
Naturally occurring pigments in sugarcane

A number of cane colorants were identified in 1970 by Färber and
Carpenter (ref. 30), summarized in Table 1. In addition, they identified
several related uncolored phenolic acids: jD-hydroxybenzoic acid, 4-hydroxy-
3,5-dimethoxybenzoic (syringic) acid and 4-hydroxy-3-methoxybenzoic
(vanillic) acid.
19^
TABLE 1
Naturally occurring pigments identified in sugarcane.
Cane pigment Color characteristics
Chlorogenic acid Orange in alkaline solution

Caffeic acid Yellow, fluorescent
p-Hydroxy cinnamic acid Yellow, blue fluorescence
(p-coumaric acid)
4-Hydroxy-3-methoxy- Yellow
cinnamic acid (ferulic
acid)
4-Hydroxy-3,5- Yellow in alkaline solution,
dimethoxycinnamic Blue fluorescence
acid (sinapic acid)
7-Hydroxycoumarin Blue fluorescence
(umbelliferone)
Kaempferol Yellow
The colorants were isolated from sugarcane leaf extracts, but many of
them were found to survive into the refined granulated sugar. Cinnamic
acids and their derivatives appear to be especially resistant to removal by
known refining techniques. Chlorogenic acid was found in every sample of
granulated sugar examined, which is why even the whitest of sugars can turn
yellow in alkaline solution.
Flavonoid pigments are the major components of pH-sensitive color,
whereas colorants formed during processing are relatively pH-insensitive.
It has been estimated that such phenolics contribute as much as 2/3 of raw
sugar color at pH 7 (ref. 31).
Cane anthocyanins, red under acid conditions, are decomposed above pH 8
by opening of the pyran ring and oxidation, so that they do not survive
clarification. In contrast, flavones survive factory clarification because
their high pH anionic forms are very stable. They are also stable under
mild acid conditions, and their glycosides require boiling with acid to
bring about even partial hydrolysis.
Carpenter's group later identified additional colorants: coniferin (a
cinnamic acid glucoside); coumarin; esculin (a coumarin glucoside); and
quercetin and rutin (flavonols) (ref. 32). These types of compounds are
generally yellow, and more highly colored at higher pH, with the flavones
usually being more deeply colored than cinnamic acid derivatives.
By 1985, four flavonols and 25 flavones had been identified in
Saccharum and in sugar products. All of the flavones were derivatives of
tricin, luteolin, and apigenin (ref. 31), shown in Figure 1. Note that only
the luteolins contain the catechol function, and even in this group the 3'-
hydroxyl is in some cases methylated.
TRICIN
API&ENIN
Fig. 1. Basic flavone structures found in Saccharum.
Flavonoid color, unlike the other categories, does not originate in

sucrose. Although flavonoids are responsible for a large fraction of raw
sugar color, they also play subtle roles in plant physiology by regulating
enzyme activity, attracting pollinators, and providing resistance to disease
and predators. Chlorogenic acid, for example, is a growth stimulator.
196
Breeders select varieties primarily for sugar yield, although such factors
as drought tolerance and specific disease resistance are also important
considerations. It is likely that flavonoids play an important role in all
of these attributes of a desirable cane variety. Development of resistant
varieties that at the same time exhibit good growth characteristics may have
resulted in cane with increased phenolic content. However, it should be
borne in mind that the flavonoid content of a cane juice sample depends not
only on variety but also on the freshness of the sample, the maturity of the
stalk, the position of the stalk, the presence of disease, and environmental
factors during cultivation.
Fluorescence
Commercial cane sugars always exhibit at least some slight
fluorescence, resulting from trace levels of colorants as well as uncolored
constituents. Many sugars display a fluorescence maximum near 440 nm, which
is a summation peak from a number of compounds (ref. 33).
Fluorescence, the absorption of light at one wavelength and the
emission of it at a longer one, increases much faster than color in
caramelization and invert degradation reactions, but both increase at the
same rate in melanoidin-forming reactions. Thus the fluorescent components
of granulated sugars are not necessarily phenolics, although these appear to
be more difficult to remove and hence more likely to survive the refining
process.
Enzymatic browning
One way to limit the phenolic contribution to raw sugar color without
genetic manipulation of the flavonoid fraction is by rapidly inactivating
the enzymes in pressed juice. The benefit is obvious: color that is
prevented from forming need not be removed.
Enzyme-catalyzed color formation results from the action of o-diphenol:
0Z oxidoreductase on phenolics, particularly on chlorogenic acid. The
enzyme may be inactivated by heat or inhibitors such as thioglycolate.
Goodacre .et. al. (ref. 34) showed that over half the original color in
pressed juice may be derived from the interaction of amino acids with
enzymatically generated quinones.
The ο,-diquinone resulting from enzyme-catalyzed oxidation of
chlorogenic acid is chemically reduced by a secondary ο,-diphenol, and the
secondary quinone thus formed polymerizes to form color. Alternately the
chlorogenic quinone can react with amino acids or other amino compounds that
polymerize to intensely colored substances (ref. 35). These high molecular
weight products have reduced pH sensitivity and increased tendency to boil
into the crystal preferentially in the manufacture of raw sugar. Smith
(ref. 17) found about two dozen colorants in the enzyme-inactivated first
expressed juice, of which he estimated that about five were the main enzyme
substrates, since these substances were either absent or significantly
reduced in concentration in normal juice.
Caramelization
Maillard reactions can occur even at room temperature, but
caramelization normally takes place at elevated temperatures. Because
process temperatures rarely exceed 100°, thermal carbon-carbon σ-cleavage
does not occur, and therefore the degradation is not a true pyrolysis.
Traces of impurities strongly catalyze the thermal degradation of sucrose,
which is one reason why its reported melting point covers the range of
160-188°.
Sucrose caramel is a complex mixture of mono-, oligo-, and
polysaccharides, together with colored substances. The composition is
dependent upon reaction time, pH, temperature, and the presence of
impurities}' certain of the components are colloidal. The reactions involved
appear to be a mixture of first-order dehydrations and second-order
condensations. Although the reactions are highly complex, it is generally
agreed that the first step is the splitting of the glycosidic linkage of
sucrose.
Even if the reactions that occur represent only a small sucrose loss,
color increases can be significant. The caramels typically are present in
small quantities but tend to attach themselves to the crystal surface and
can contribute substantially to solution color (ref. 36).
Many carbohydrates exhibit very similar patterns of caramelization.
For example, thermal degradation of sucrose, starch, cellulose, lactose,
glucose, and ß-D-glucosides all yield 1,6-anhydro-ß-D-glucopyranose
(levoglucosan), which can be further dehydrated to l,6-anhydro-3,4-dideoxy-
198
A3-ß-D-pyranosen-2-one (levoglucosenone) (ref. 37). Nucleophilic addition

to the double bond in the latter could result in further reaction.
Levoglucosan itself is known to pyrolyze to branched-chain species with
molecular weights up to 5 x 10 A (ref. 38), perhaps via the levoglucosenone
intermediate. A possible scheme is shown in Figure 2. Note that water
elimination, a predominant reaction, does not require solid-state sugar but
only fairly high liquor densities.
Oxygen does not influence the tinctorial power of the caramel formed
(ref. 39), but can affect its solubility in water or acid.
SUCROSE
INVERT
various ^*
routes
3,4-DIDEOXY- ^.EVOGLUCOSENONE
GLUCOSULOSE-
3-ENE (OGU)
HOCH,,
POLYMERS, COLOR
5-(HYDR0XYMETHYL)-
2-FURALDEHYDE (HMF)
Fig. 2. Some possible pathways of sucrose thermolysis (refs. 58 and 59).
Maillard browning
Maillard browning is more than a simple Schiff base condensation of

carbonyl and amino compound; it is in fact a complex scheme of many
reactions. There is considerable evidence for free radical participation in
199
an accompanying oxidative process. Amino acids do not react in their

cationic forms, only slightly in their zwitterion forms, and fully only in
their anionic forms. If the amino compound is in fact an amino acid,
Strecker degradation usually takes place, whereby C0 a is liberated and an
aldehyde or ketone is formed with one carbon atom fewer than the original
amino acid. The browning can be inhibited by sulfite addition, but such
usage in the United States may be expected to decline as a result of FDA
concerns about residuals.
The nitrogen content of sugarcane juice amounts to only a few
hundredths of one percent, of which about half is present in forms
potentially active in color forming reactions. The initial step of the
Maillard reaction is believed to be formation of N-(D-glucosyl) amino acids,
which are converted to D-fructose amino acids by the Amadori rearrangement
(ref. 40):
HOCH2
1
OH HO
Numerous studies have been carried out with model compounds, in

attempts to elucidate the browning reaction mechanisms and rates. Much
information has been obtained from these investigations, but the actual
process is far too complex for any comprehensive laboratory simulation.
Binkley (ref. 41), reporting on melanoidins in cane final molasses, found an
empirical formula of Ci7-ieH26-270ioN, wtuch suggests two six-carbon units
bound to a 4-5 carbon amino acid residue. Other researchers have reported
similar empirical formulas. Binkley also found an average molecular weight
of about 5,000, with a range from 10,000 to 50,000. The high C/H ratio
suggests a high level of unsaturation, probably resulting from a combination
of carbon-carbon double bonds, carbonyls, rings formed by cross-linking, and
the inclusion of aromatics.
The familiar condensation of amino acids with carbohydrate carbonyls
contributes not only to complex color-forming reactions but also yields
colorless volatile products that are important contributors to molasses
200
aroma (réf. 42). Similar compounds contribute to the aroma of many roasted
or browned foods, such as chocolate, baked bread, roasted meat and coffee.
Other chemical degradation

When sucrose or fructose is heated under mildly acid conditions, the
eis and trans isomers of 3,4-dideoxyglucosulose-3-ene (DGU) are formed via
various routes. DGU, which displays a bright green fluorescence, is the
immediate precursor of 5-(hydroxymethyl)-2-furaldehyde (HMF). Although the
principal decomposition products of HMF are the colorless levulinic and
formic acids, it is also an important color precursor, polymerizing to brown
products (ref. 43).
Under alkaline conditions, a completely different set of reactions
takes place. In the high pH region, sugars easily isomerize and condense
with amino compounds. For sucrose, the first step is a dehydration of the
fructose moiety. The degradation proceeds via a complex scheme of
aldolization, Lobry de Bruyn-van Ekenstein isomerization, ß-elimination,
α-dicarbonyl cleavage, and benzilic acid rearrangement. However, most of
the reaction products are uncolored species. It should be clear from this
discussion that there is considerable overlap between the categories of
caramelization, Maillard reactions, and chemical degradation.
One reason why there is no continuum between high and low pH
degradation products is that the acid-catalyzed sequence proceeds via
reductones, which are unstable at pH 6 and above (ref. 44):
I
c=o
C-OH
II
C-OH
I
Calcium ion inhibits high pH color formation, whereas sodium and
potassium enhance it. At low pH, sodium and phosphate inhibit color
formation, and calcium enhances it. Chloride has no impact. Citrate has a
striking effect on color formation: complete inhibition at low pH and
strong enhancement at high pH (ref. 43). It should be noted that citrate is
the immediate precursor, by dehydration, of aconitate, the principal organic
anion in raw sugar. In some cases these additives probably do not alter the
201
rate of summed sugar degradation reactions so much as re-direct them toward

more or fewer colorless products.
The role of iron

Iron is present in sugar liquors primarily in the form of complexes
with non-sugars, with intensely colored catechol adducts displaying a high
degree of stability. As a result, iron is readily removed in the
manufacture of granulated sugar, but removed only with difficulty from brown
sugar liquors, which contain relatively high levels of chelating groups.
Iron catalyzes the rate of color formation from hexose degradation,
particularly at moderate pHs (6-8). At lower pHs, there is accelerated
oxidation of invert, but the products are largely colorless. At pH 9 and
above, iron is less effective at catalytic darkening, perhaps because of low
solubility of an active form. Iron also catalyzes color formation in brown
sugars and liquors, acting as a carrier for auto-oxidation. There is
evidence for phenolic involvement in these reactions (refs. 3, 4). However,
in an earlier study, Parker and Williams (ref. 45) reported no evidence for
iron catalysis in color formation.
Iron participates not only as a catalyst but also as a constituent of
resulting chromophores. Even when phenolics are absent, iron can increase
the absorbance of browning products, sometimes strikingly. Hashiba .et al.
(ref. 46) studied synthetic Amadori compounds and found that the colors of
these substances could be reduced by as much as 90Z by removal of iron. The
site of complexation in such adducts is probably enolic.
Phosphate plays an important role in modulation of iron activity.
Colored iron adducts can be decolorized at high phosphate levels, where the
iron-phosphate complex predominates. Phosphate complexes can also disrupt
the propagation of autoxidative free radical chains (refs. 47, 48). On the
other hand, removal of iron complexes on bone char is inhibited at high
phosphate concentrations.
DECOLORIZATION OPERATIONS
In the refinery, dark colored liquors can result from exceptionally
dark raws, or from light raws that decolorize poorly. Smith (ref. 18) found
that in both cases, pH sensitivity was responsible; this is the basis for a
rapid laboratory test to characterize raws for ease of decolorization (ref.
202
49). Kennedy and P. Smith (ref. 50) found that monomeric, pH-sensitive
color was removed primarily by affination, bone char, and resin. However,
N. Smith (ref. 20) reported that affination removed such material only with
difficulty. On the other hand, both groups found that affination and
clarification effectively removed larger molecular weight, pH-insensitive
color. N. Smith also observed that colorants most poorly eliminated by
affination were also poorly removed by clarification. Of course, there is
no a. priori reason why these two unit operations should produce similar
results, because clarification is most effective for colloidal color,
whereas affination can remove only surface film color.
The effectiveness of bone char and anionic resin stems from their
combinations of ionic and non-polar surfaces, as described earlier.
However, the sites for high I.V. color on bone char may be quickly
saturated, after which the char selectively adsorbs pH-insensitive color
(ref. 18). Discrepancies in the reports of different observers regarding the
behavior of various adsorbents toward specific colorant types appear to be
based largely on capacity factors, rather than on thermodynamic or kinetic
considerations.
In decolorization mechanisms dependent upon adsorption at anionic
centers in colorants, divalent anions such as sulfate may be serious
competitors. Sulfate is much smaller than anionic color and can therefore
diffuse more readily to the adsorption site. Uncolored low molecular weight
organic anions are also able to reach active sites much faster than
colorants and can inhibit decolorization.
Kennedy and Smith (ref. 50) found that styrene resin is better than
bone char for removing tricin derivatives, but poorer for those of luteolin.
Riffer (ref. 3) observed that styrene resin was also better than granular
carbon for removal of iron adducts of colorants. Paton and Smith (ref. 51)
reported that although granular carbon had a high affinity for flavonoids,
it was not particularly good for amino-nitrogen derivatives. Rader and
Anderson (ref. 52) found that styrene resin is very efficient in removal of
alkaline degradation products, but less so for melanoidins and caramels,
which behave similarly. They observed that a weakly basic phenol
formaldehyde resin with a flexible structure to accommodate very large
molecules performed better than styrene for melanoidins and caramels.
203
However, such resins have not achieved notable commercial use in the sugar
industry.
Vorona £t. a_l. (ref. 53) found that alkaline degradation products
contain a considerable proportion of colorless compounds which are not
easily adsorbed by carbon. Caramel also contains a large amount of such
material, while melanoidins are mostly colored and, Vorona et al. found,
unlike Paton and Smith cited above, easily adsorbed. The adsorption of
caramelization products on vegetable carbon through hydrophobic bonding
would be expected from the fact that carbohydrates pyrolyze to a
graphite-like residue, which is indeed the route to commercial carbons.
Colorant partition during crystallization

Most investigators agree that high molecular weight, pH-insensitive
color is most likely to boil into the crystal, but some researchers have
reported the opposite phenomenon, preferential inclusion of pH-sensitive
color. Smith (ref. 54) has shown that these seemingly irreconcilable
observations can be explained.
He noted that in the raw factory, high molecular weight pH-insensitive
enzymatic browning products tend to boil into the crystal because they
predominate over low molecular weight pH-sensitive, naturally occurring
phenolics. A high proportion of the latter go into the molasses syrup,
whereas most of the former go on into the refining process. By the time the
sugar has been through the refinery, however, colorants predominate that are
of lower pH sensitivity than the original enzymatic browning products, as a
result of further reaction and preferential removal of pH-sensitive material
by bone char. These colorants formed in the refinery have a lower tendency
to boil into the crystal than either residual phenolics or enzymatic
browning material. Consequently, in the refinery it is the relatively
pH-sensitive colorant that tends to boil into the crystal; hence the
persistence of traces of flavonoids even in highest quality granulated.
The above difference has been obscured by the fact that the pH
sensitivity of washed raw is higher than that of whole raw sugar, leading to
the erroneous conclusion that in both raw and refined sugar, pH-sensitive
color is preferentially occluded. The relatively low pH-sensitivity of
whole raw sugar results from chemial changes in the film color after
2C&
boiling, with the loss of pH-sensitive material and the formation of darker
pH-insensitive colorants.
Laboratory studies indicate that alkaline degradation products remain
primarily in the mother liquor and are mainly adsorbed on the surface of the
sugar crystals, with only a slight tendency to be incorporated within the
crystal. Melanoidins are occluded to a degree dependent on the amino
component but also have a noticeable tendency to be deposited on the crystal
surface. Washing removes color due to alkaline decomposition products and
caramels, but not if melanoidins are present (ref. 36).
Chemical additives
Numerous additives have been studied as adjuncts to adsorbent
decolorizers. Because most colorants are anionic at process pHs, they can
be precipitated by long-chain alkylquaternary ammonium ions. This is the
basis of the Talofloc process, which is in common use in conjunction with
refinery clarification, to reduce the color burden on resin, carbon, or bone
char.
Oxidants such as hydrogen peroxide and sodium hypochlorite (ref. 55)
effect decolorization by cleavage of conjugated unsaturated sites,
shortening the resonance pathlength and forming carboxylic acids. Phenolics
are oxidized to quinones and acylic products. However, the decolorization
is accompanied by sugar loss through oxidation, particularly from invert.
Bisulfite adds to unsaturated carbonyl groups (ref. 56) and certain
flavonoids (ref. 57), again resulting in a shortened resonance pathlength in
affected chromophores and shifting the absorption out of the visible region
into the UV. Bisulfite also inhibits new color formation via Maillard
pathways.
Hydrosulfite decolorizes by reduction of diketones, quinones, and iron
complexes. The reduction of quinones arrests their oxidative polymerization
to brown-black pigments. EDTA decolorizes by removing iron from colored
complexes.
With the exception of quaternary précipitants, none of these additives
have found practical use in the cane sugar industry in the United States.
Drawbacks include the costs of the additives, sugar loss, and concern about
residuals. Bisulfite in its various forms has wider use in the processing
of sugarbeets.
205
REFERENCES
1 H. H. Peters and F. P. Phelps, Color in the sugar industry, Technological

Papers of the Bureau of Standards, No. 338, 1927.
2 V. R. Deitz, Investigations of Bone Char Research Project, Inc., 21st
quarterly report, Jan.-March. National Bureau of Standards, Washington,
D.C. 1951.
3 R. Riffer, Recent laboratory studies at C & H Sugar, Proc. 1986 Sugar
Processing Res. (in press).
4 R. Riffer, The chemistry of iron in the sugar refinery, Proc. 1984 Sugar
Processing Res. Conf., 1985, 231-251.
5 F. G. Carpenter and V. R. Deitz, Investigations of Bone Char Research
Project, Inc., Technical Report No. 65, National Bureau of Standards,
Washington, D . C , 1962.
6 G. A. McLaren, Jr., 8th Congr., Intern. Ind. Agr., Brussels, 1950.
7 W. W. Binkley, Hydrogenolysis of the dialysed browning products of
final cane molasses, Int. Sugar J., 59 (1957) 64.
8 D. Gross, Proc. C.I.T.S. 1957, 121.
9 C. C. Tu and J. H. Payne, Colored compounds in evaporator syrup,
Proc. Intern. Soc. Sugar Cane Technol., 12 (1965) 1671.
10 E. J. McDonald and J. P. Madacsi, A yellow component in sugar colorant.
Proc. 1966 Tech. Sess. Cane Sugar Refin. Res., 1967, 136-141.
11 L. Färber, E. J. McDonald and F. G. Carpenter, Separation of colorants
from cane sugar, Proc. 1968 Tech. Sess. Cane Sugar Refin. Res., 1969,
85-104.
12 B. Cortis-Jones, Methods for fractionations of impurities in cane juice
and mill syrups, Intern. Sugar J., 64 (1962) 133, 165.
13 N. H. Smith, Fractionation of sugar colorants with molecular sieves.
14 C. C. Tu and K. M. Onna, The non-sucrose constituents of Hawaiian raw
cane sugar crystals. Intern. Soc. Sugar Cane Technol., 1959, 291.
15 D. Linecar, N. H. Paton and P. Smith, Techniques for the isolation of
cane sugar colorants, Proc. Tech. Sess. Cane Sugar Refin. Res.,
1979, 81-91.
16 E. J. Roberts and M. A. Godshall, Some observations on the high molecular
weight colorants in sugar, Proc. 1978 Tech. Sess. Cane Sugar Refin. Res.,
1979, 68-80.
17 N. H. Smith, Fractionation of sugar colorants by high pressure liquid
chromatography, Proc. 1976 Tech. Sess. Cane Sugar Refin. Res., 1978,
19-34.
18 N. H. Smith, The pH sensitivity of sugar colors and ease of decoloriza-
tion, Proc. 1964 Tech. Sess. Cane Sugar Refin. Res., 1966, 14.
19 E. J. Roberts and M. A. Godshall, Color in refinery products, Proc. 1980
Tech. Sess. Cane Sugar Refin. Res., 1981, 50-59.
20 N. H. Smith, Gel filtration for determining efficiency of color removal
in processed raws, Proc. 1972 Tech. Sess. Cane Sugar Refin. Res., 1975,
1-7.
21 C. C. Tu, Sources of coloring matter in commercial sugar., Intern.
Sugar J. 76 (1974) 3-6.
22 H. Tsuchida and M. Komoto, Fractionation and some characterizations of
the non-dialysable colorant in cane sugar refinery final molasses by
DEAE-cellulose column chromatography, Proc. Res. Soc. Japan Sugar
Refineries' Technol., 22 (1970) 65-76.
206
23 M. Shore, N. W. Broughton, J. V. Dutton and A. Sissons, Factors affecting

white sugar colour, Sugar Tech. Rev. 12 (1) (1984) 1-99.
24 G. Mantovani, G. Vaccari, G. Sgualdino, D. Aquilano and M. Rubbo,
Coloring matter inclusions in sucrose crystals, Proc. Intern. Soc. Sugar
Cane Technol., 1986, 663-669.
25 G. A. Chikin, V. I. Sigova and T. M. Makeeva, Determination of the
ionization constant of sugar manufacture colorants, Teoriya i Praktika
Sorbtsionnykh Protsessov 11 (1976) 101-2.
26 Tate and Lyle, Ltd., Group Research and Development, Annual Report 5,
1971.
27 J. C. Williams, Properties of colorants produced by the degradation
of reducing sugars, Proc. C.I.T.S. 15 (1975) 319-327.
28 Tate and Lyle, Ltd., Group Research and Development, Annual Report,
44, 1973.
29 D. V. Freeland, R. Riffer and J. G. Penniman, Electrokinetics applied
to sugar refining, Intern. Sugar J. 81 (1979) 196.
30 L. Färber and F. G. Carpenter, Identification of sugar colorants,
Proc. 1970 Tech. Sess. Cane Sugar Refin. Res. Res., 1971, 145-156.
31 P. Smith and N. H. Paton, Sugarcane flavonoids, Sugar Technology Rev.,
12 (2-3) (1985) 117-142.
32 L. Färber and F. G. Carpenter, Plant pigments as colorants in cane
sugar, Proc. 1972 Tech. Sess. Cane Sugar Refin. Res., 1975, 23-31.
33 J. H. Wall and F. G. Carpenter, Fluorescence in commercial sugars.
34 B. C. Goodacre, J. Hutson and J. Coombs, Enzyme catalysed formation of
colour in cane juice, Intern. Sugar J., 83 (1980) 11.
35 D. Gross and J. Coombs, Enzymic colour formation in beet and cane
juices, Intern. Sugar J., 78 (1976) 69.
36 V. Prey and H. Wesmer, Research on the behavior of different groups of
coloring matter during crystallization, Z. Zuckerind., 25 (1975) 341-
346.
37 Y. Halpern, R. Riffer and A. Broido, Levoglucosenone (1,6-anhydro-3,4-
dideoxyl-A3-ß-D-pyranosen-2-one). A major product of the acid catalyzed
pyrolysis of cellulose and related carbohydrates, J. Org. Chem. 38
(1973) 204.
38 R. D. Guthrie, Glycosans and anhydro sugars, în: the Carbohydrates.
Chemistry and Biochemistry, W. Pigman and D. Horton (Eds.) Academic
Press, New York and London, Vol. IA, 1972, 431.
39 J. Palasinski, P. Tomasik and S. Wiejack, Thermolysis of carbohydrates
in oxygen-free atmosphere. Part 1. Caramelization of mono- and
disaccharides. Starch/Starke, 37 (9) (1985) 308.
40 H. Paulsen and K.-W. Pflughaupt, Glycosylamines. in: the Carbohydrates.
Chemistry and Biochemistry, W. Pigman and D. Horton (Eds.). Academic
Press, New York and London, Vol. IB, 1980, 899.
41 W. W. Binkley, Browning polymer, a major colorant of cane final
molasses, Z. Zuckerind., 20 (6)(1970) 291-5.
42 J. Pokorny, L. Dvorakova, A. Marcin, H. Bulantova and J. Davidek,
Sensory profiles of reaction products between amino acids and D-fructose
or D-arabino-hexosulose, Die Nahrung 23 (9-10) (1979) 921-7.
43 F. G. Carpenter and E. J. Roberts, Colorant formation under refining
conditions, Proc. 1974 Tech. Sess. Cane Sugar Refin. Res., 1975,
106-115.
44 H.-D. Belitz and W. Grosch, Food Chemistry (Lehrbuch für Lebensmittel
chemie, 2nd edition), Springer Verlag, Berlin, 1987, pp. 213-219.
207
45 K. J. Parker and J. C. Williams, The isolation and properties of sugar

colorants, Proc. 1968 Tech. Sess. Cane Sugar Refin. Res., 1969, 117-127.
46 H. Hashiba, A. Okuhara and N. Iguchi, Oxygen-dependent browning of
soy sauce and some brewed products, Prog. Food Nutr. Sci. 5 (1964)
93-113.
47 A. Grinberg, An introduction to the chemistry of complex compounds.
D. H. Busch and R. F. Trimble, Jr. (Eds.), Pergamon Press, 1980,
249-264.
48 P. Neta and M. G. Simic, Chemical changes in food during processing,
AVI Publishing Co., 1985, 63-72.
49 M. A. Clarke, R. Blanco and M. A. Godshall, Color tests and other
indicators of raw sugar refining characteristics, Proc. 1984 Sugar
Processing Res. Conf., 1986, 284-302.
50 A. M. Kennedy and P. Smith, Color in refineries, Proc. Sugar Ind.
Technol, 1976, 156-160.
51 N. H. Paton and P. Smith, An HPLC study of the changes in colorant
composition following factory decolorization of raw liquors with
bone char, resin and granular carbon, Proc. 1982 Sugar Processing Res.
Conf., 1983, 1-23.
52 J. E. Rader and R. E. Anderson, The rate of reaction of synthetic
colorants and resinous adsorbents, Proc. 1970 Tech. Sess. Cane Sugar
Refin. Res., 1971, 114-124.
53 L. G. Vorona, A. K. Kartashov and G. P. Pustokhod, Sakh. Prom. 40 (8)
(1966) 24.
54 N. H. Smith, Unpublished communication, 1974.
55 R. Riffer, A study of chemical additives for use in sugar refining,
56 E. E. Royals, Advanced Organic Chemistry, Prentice Hall, 1954, 639-642.
57 L. Jurd, Reactions involved in sulfite bleaching of anthocyanins.
J. Food Sci., 29 (1) (1964) 16.
58 A. Ohnishi, E. Takagi and K. Kato, Identifications of 1,6-anhydro-3-
deoxy-ß-D-glucopyranosen as a thermal decomposition product of cellulose,
Chem. Let. (Japan), 1974, 1361-2.
59 F. Shafizadeh, R. H. Furneaux and T. T. Stevenson, Some reactions of
levoglucosenone, Carb. Res., 71 (1979) 169.
208
Chapter 14
POLYSACCHARIDES OF SUGARCANE AND THEIR EFFECTS ON SUGAR MANUFACTURE
R. A. KITCHEN
SUMMARY
The major proportion of polysaccharides entering the mill or the

refinery are generated by the sugarcane plant. The remaining
polysaccharides are formed by microbiological contamination, which can occur
either in the field or during processing. Recycling of these
polysaccharides occurs during processing, and in refineries can result in a
15 fold increase in content from the raw sugar to the final molasses. All
of these polysaccharides have a detrimental effect on sugar manufacture, and
this effect is not easily diminished as very few of the processing steps
remove polysaccharides.
INTRODUCTION
A comprehensive review in 1972 by Imrie and Tilbury (ref. 1) discussed

the polysaccharides in, and associated with, sugarcane. This current
review, therefore, will be largely concerned with the literature after that
date, although some of the more interesting earlier material will be
discussed.
Many different kinds of polysaccharides are found in the sugarcane
plant. These polysaccharides are all long chain molecules made up of simple
carbohydrate units; the linkages between the units and the structure of the
molecules, straight or branched, can vary. In this article polysaccharides
are considered to contain more than ten carbohydrate units.
Cellulose and hemicellulose, which are components of the cell wall,
give structural strength to the standing cane plant. They are not
considered in this review, however, as they are not soluble in water and
consequently are not likely to affect sugar manufacture. Starch, which is
involved in the metabolic activity of the growing plant, largely occurs in
sugarcane as insoluble granules. Because these granules can be solubilized
during processing, starch has an effect on the manufacturing process and
will therefore be discussed. Other soluble polysaccharides, such as
sarkaran, indigenous sugarcane polysaccharide, Roberts' glucan ,
galactomannan CP, molasses polysaccharide and dextran will also be
discussed. A recent article reviewed these same sugarcane polysaccharides
(with the exception of starch and dextran), but did not examine in detail
their chemistry in, and effects on, sugar manufacturing (ref. 2).
POLYSACCHARIDES OF SUGARCANE AND SUGARCANE PRODUCTS

Starch
Starch in sugarcane is largely in the form of insoluble, nearly
spherical granules varying widely in size, but averaging about 5 μπι. The
distribution of the starch granules in mature sugarcane stalks occurs
largely in the cane top, and in the nodal regions, where it disappears
rapidly once growth processes reach their peak. The quantity of starch
between varieties and within any one variety of cane varies widely. Some
poor varieties contain as much as 2,000 ppm starch, while a normal sugarcane
juice usually contains 50 ppm starch. Other than climate, the starch level
in sugarcane has been found to be affected by soil type, soil pH and the
potash levels in the soil (refs. 1,3).
During milling the starch granules are normally separated from the
plant tissue and carried into the cane juice. The effects of heat and lime
gelatinize the granules, and the solubilized starch passes into the
clarified juice. During subsequent boiling, starch and other high molecular
weight polysaccharides have a tendency to be incorporated into the raw sugar
crystals, which are then transferred to the refinery. The refinery thus
becomes the recipient of any problems associated with starch (ref. 3).
A recent investigation (ref. 4) on non-USA raw sugars found insoluble
starch that occurred in two size ranges. The larger granules were 5 μπι in
average size, whereas the smaller granules were 1 Jim average size, and were
attached to plant material. These latter granules did not gelatinize in
water or sucrose solution up to the boiling point of the solution, whereas
the larger granules did. Both granules were believed to have come from the
sugarcane plant.
Once the starch granules have been solubilized, two different
polysaccharides, amylose and amylopectin, are released. Amylose is
essentially an a-(1,4) linked, linear glucan having a helical structure, one
complete turn containing six glucose units. Amylopectin, as well, contains
a-(1,4) linkages, but it is a highly branched glucan with the branches
attached by a-(l,6) linkages (Figure 1). The considerable difference in
210
structures between the two glucans results in different chemical and

physical properties, as are shown in Table 1.
The ratio of amylose to amylopectin varies with the sugarcane variety,
and can also change as the starch passes through the manufacturing process
(refs. 1, 3). Part of the changes occurring during manufacture can be
attributed to the rétrogradation of amylose (ref. 5 ) ,
- OH CLÜC > - 0 - {> GLUC V 0 - ( s GLUC " " V o -s m S T ^ - O - f V CLÜC V O -

4 1 4 Ί 4 Ί 4 Ί 4 \
-0-j GLUC ^ - { " I Î L Û C " " " ^
0
I·
T- 0~C GLUC ">-0-( s GLUC V 0 - ( GLUC ) - 0 - { GLUC }
4 1 4 Ί C \ 4 "--—11
0
- 0 - ^ GLUC j-Q-^C GLUC > 0 H GLUC > < K GLUC

I·
^Q-Û^V-Q-
AMYLOPECTIN
CH.0H
G L U C : Of-D-GLUCOSE A( ^-
HO
3 Y0M
OM
Fig. 1. Representation of amylose and amylopectin.

211
TABLE 1
Chemical and physical properties of amylose and amylopectin.
Property Amylose Amylopectin
Iodine reaction Intense blue Red -violet

x max of iodine complex About 650 nm About 540 nm
Molecular weight (daltons) 10" - lo6 107-10*
Chain length* More than 2,000 19-28
Solubility in water Variable Soluble
Stability of aqueous solution Retrogrades Stable
* Average number of glucose units per non-reducing end group.

Reproduced from Imrie and Tilbury (ref. 1). with thanks.
Sarkaran
Sugarcane which is not processed immediately after harvesting undergoes
deterioration and is classified as stale cane. A polysaccharide was
detected in such cane in South Africa, and its structure was determined.
The preliminary investigation involved the detection of various
compounds found in deteriorated cane versus fresh cane as a means of
monitoring the condition of harvested cane. Volatile acids, alcohols,
non-volatile organic acids, amino acids, starch and polysaccharide contents
were determined over a period of days. The starch decreased and the
polysaccharides increased with time, while the other analyses were
ineffective as indicators of deterioration. The analysis for soluble
polysaccharides was therefore judged to be the best indicator (ref. 6).
The polysaccharide was assumed to be a dextran, which along with lactic
acid, is produced from sucrose by bacteria. However, no lactic acid was
found (ref. 6). Large quantities of stale cane were milled; the
polysaccharide was isolated by alcohol precipitation and dialysis, and was
then compared to a starch sample, and to a dextran produced by a pure
culture of Leuconostoc mesenteroides. The composition and structure were
determined by periodate oxidation, acid hydrolysis followed by paper
chromatography, exhaustive methylation followed by hydrolysis and GLC
analysis, and by infrared spectroscopy. The polysaccharide was shown to be
a straight chain D-glucan having 25% a-(1,6)and 75% a-(1,4)linkages
(Figure 2 ) , a specific rotation of t160" and a molecular weight of
212
approximately 20,000 daltons. The results showed clearly that the

polysaccharide was not a starch or a dextran (ref. 7).
j» ,
C GLUC > < ^ GLUC ><K CLUC > Q < GLUC Λ
0
0
C GLUC ffi GLUC *><K GLOC Λ
<K
0
Fig. 2. Representation of sarkaran.
The stale cane polysaccharide showed some structural similarities to

pullulan, a glucan produced by the mold Aureobasidium pullulans. Samples of
the polysaccharide and pullulan were therefore hydrolyzed by pullulanase, an
enzyme specific for a-(1,6) glucan linkages. Paper Chromatographie
investigations indicated that, after the hydrolysis, pullulan yielded only
maltotriose, whereas the cane polysaccharide yielded 49Z maltotriose, 38Z
maltotetraose and 13Z other a-(l,4) linked glucose oligosaccharides.
Pullulan was clearly different from the cane polysaccharide, which,
therefore, was considered a new a-glucan and was named sarkaran (ref. 8).
In Queensland, Australia, sugarcane is normally grown and harvested
annually. Occasionally cane is left in the field, i.e. stand-over cane, and
is processed in the next milling season. Stand-over cane is quite often
difficult to process, a problem which has been attributed to the presence of
a sugarcane polysaccharide. Later, a polysaccharide was isolated from such
cane, and its structure was determined.
Stand-over cane samples were shredded and crushed, and the
polysaccharide was isolated from the juice by alcohol precipitation. The
composition and structure of the polysaccharide were determined by the
methods previously used in the South African investigation, but by newer
instrumental techniques as well. These included the use of reflective
densitometry to detect the oligosaccharides on paper chromatograms after
enzymatic hydrolysis of the polysaccharide, combined gas chromatography-mass
spectrometry to identify the alditol acetates after permethylation and
213
13
hydrolysis of the polysaccharide, and C and ^H nuclear magnetic resonance
spectra of the stand-over cane polysaccharide, pullulan and South African
sarkaran. The results from this investigation indicated that the stand-over
cane polysaccharide was sarkaran, with a specific rotation of +167° and a
27Z a-(1,6) linkage content. Sarkaran, therefore, was much more widely
spread in cane than had been indicated by the stale cane work (ref. 9).
Molasses samples obtained from mills which had experienced difficulties
in processing stand-over cane were found to contain an impure glucan.
Purification of this glucan proved difficult because of the high
concentration of colorants. Structural determinations on the material
indicated that the polysaccharide was sarkaran (ref. 10).
A procedure for monitoring the concentration of sarkaran in cane
billets was developed using the enzyme pullulanase. Standard graphs, made
by plotting the concentration of pure sarkaran versus the reducing sugars
produced by the enzymatic hydrolysis, showed linearity and precision. The
application of the method to stand-over cane juice was time-consuming,
however, as the sarkaran had first to be isolated by alcohol precipitation.
The use of high performance liquid chromatography to monitor the formation
of oligosaccharides was ineffective, as sarkaran samples showed considerable
variation in their constituent maltodextrins with different conditions of
storage (ref. 11).
A survey of stand-over cane showed one area with high concentrations of
sarkaran, up to 0.13Z on cane. The cane in this area was in poor condition
with excessive stalk splitting and a color different from healthy cane. The
randomness with which sarkaran occurred in this area of similar climatic
conditions, and the relatively long development period before the glucan's
appearance, indicated that sarkaran was not a plant product. It was
suggested that the glucan could be formed by a microorganism, possibly a
yeast (ref, 12).
Gel permeation chromatography was used, with standard dextrans as
references, to determine the molecular weight of sarkaran. Values of
185,000 and 200,000 daltons were obtained; the difference in values from
those of South Africa were attributed to different isolation and
purification procedures, or to different conditions of biosynthesis (ref.
12).
21Λ
Indigenous sugarcane polysaccharide (I.S.P.)

In 1964, a polysaccharide was isolated from fresh cane juice in
Louisiana under conditions which prevented inclusion of starch or the
formation of dextran. After hydrolysis, the material was found to contain
arabinose, galactose, glucose, mannose, xylose and small amounts of rhamnose
(ref. 13). In later years, the composition of this material was further
investigated, galactose being found to be the major component, followed by
arabinose, with glucose, mannose and xylose in trace quantities. This
heterogeneous polysaccharide, which was termed I.S.P., was found to vary
with cane variety, cane age and the method of purification. It was not
completely removed during processing, and could be detected in all sugarcane
products (ref. 14). It had a negative specific rotation of -46° to -50°,
and a molecular weight ranging from 100,000 to 300,000 daltons. Glucuronic
acid was also found as a component of I.S.P. (ref. 15); its concentration
was in the 11 to 8Z range when freshly extracted (ref. 2). It was the
presence of this acid which was probably responsible for the early
literature reporting that pectin, composed of galacturonic acid, had been
found in sugarcane. Currently, no galacturonic acid has been found in any
polysaccharide isolated from a sugarcane product, which suggests that the
glucuronic acid had been misidentified and pectin does not exist in
sugarcane (ref. 15).
Indigenous sugarcane polysaccharide was also isolated from fresh cane

juice in Queensland, Australia. Purification of the I.S.P. was performed by
a combination of alcohol precipitations, gel permeation and ion exchange
chromatography. Gas chromatography was used to identify the constituent
carbohydrates, with arabinose and galactose being in highest concentrations.
Gel permeation chromatography employing standard dextrans as references
indicated a molecular weight of approximately 70,000 daltons; this value was
not substantiated by high performance liquid chromatography which indicated
two components of molecular weight 1-2 million and 20-25 thousand daltons.
Permethylation of the gel purified material, followed by hydrolysis and gas
chromatography of the methylated products indicated that the arabinogalactan
had a most unusual structure. The backbone consisted of a chain of
ß-(l,3)-D-galactose units to which every two out of three had attached (via
the primary hydroxyl) a D-galactosyl or L-arabinosyl side chain (Figure 3 ) .
A specific rotation of -56° was attributed to a ß-configuration in the
D-galactose and an a-configuration in the L-arabinose (ref. 16).
The Queensland group also investigated one of two fractions isolated
from a sample of the Louisiana I.S.P. They found essentially the same
structure as before, a backbone of D-galactose units with attached galactose
and arabinose side chains. The other low concentration monosaccharides,
rhamnose, xylose, mannose, glucose and glucuronic acid, were thought to be
attached to the arabinose and galactose side chains through the primary
hydroxyl groups of the latter. It was noted that variations in the
arabinose : galactose ratio had occurred, and that the glucuronic acid
concentration was higher in the Louisiana sample. Gel permeation
chromatography indicated that the I.S.P. sample had a molecular weight of
110,000 daltons (ref. 17).
OM
OH
GAL: 0 - 0 - G A L A C T O S E ARAB f : Of-L-ARABINOSE
#COOM
£7
ARAB p : a-L-ARABINOSC
G A:0-O-GLUCURONXC A C I D
Fig. 3. Generalized structure of indigenous sugarcane polysaccharide.

216
Roberts* glucan
During purification of I.S.P. by gel permeation chromatography, several

investigators have reported the appearance of two peaks (refs. 16,17).
These peaks indicated fractions of different molecular weight, but the
composition of both fractions had not been identified.
In Louisiana, therefore, juice from freshly cut sugarcane was treated
to isolate the I.S.P., and the purified product was then fractionated by a
hollow fiber dialyzer with a 50,000 molecular weight cut-off. The
polysaccharide isolated by this technique represented 401 of the original
I.S.P. weight.
The composition of this polysaccharide was established by
permethylation followed by gas chromatography, periodate oxidation,
treatment by the debranching enzymes pullulanase and isoamylase [specific
for a-(1,6) linkages], thin layer chromatography of the hydrolyzate, and
a-amylase treatment of the material which remained at the origin of the thin
layer plate.
The polysaccharide had a specific rotation of +120°, a molecular weight
of 15,000 - 50,000 daltons and, by acid hydrolysis, contained 98Z glucose,
with traces of arabinose, galactose and mannose. The proposed structure for
the D-glucan consisted of a straight chain backbone of a-(1,4) linked units,
with a-(1,4) linked side chains ranging in length from glucose to
maltooctaose. The side chains were attached to the backbone by a-(1,6)
linkages (Figure 4). The structure was highly branched, similar to
amylopectin and glycogen, but was of lower molecular weight (ref. 18).
Pending determination of the structure, the compound has been named
"Roberts' glucan" after the scientist who first isolated it.
C CLOC ^ CLÜC ^
0
I·
-O^CLÜC J^GUJC J(,
-Onf CLUC )>0Η[ GLOC

I· ÔjT^LüC ^
GUK CL0C
"°^ ì*$ ^ΐ C ™χ \
A k Λ k J: k A k ± k I; k
-O^f CLUC >0< GLUC y*< CLUC > Q - T c U « " > » < CU« > ^ CLOC ^
Fig. 4. Proposed structure of Roberts' glucan.

217
Polysaccharide C.P.
Raw sugar shipped from Mackay, Australia, was refined in Japan, and
some of this sugar was used for the preparation of carbonated beverages.
Considerable amounts of floe were found in these beverages. Therefore, a
study was initiated in Japan to determine the materials responsible. The
flocculent material was precipitated out of raw or refined sugar by storage
in carbonated water. Purified material was analyzed for polysaccharides,
protein and inorganics. Raw sugar gave floe samples containing nearly equal
quantities of polysaccharides and protein, whereas refined sugar floe was
largely polysaccharide in nature. Acid hydrolysis, followed by paper and
gas-liquid chromatography, indicated the flocculent contained rhamnose,
arabinose, xylose, mannose, glucose and galactose (réf. 19).
Samples of Australian, Philippine, Cuban and South African raw sugars
each produced a floe in the Japanese carbonated beverage test. The floes
were isolated, purified and acid hydrolyzed; the component carbohydrates
were found to be mannose, glucose, galactose, arabinose, xylose and
rhamnose, the first three being in highest concentration. One of the
Australian floes was also fractionated on DEAE cellulose, and each of the
fractions acid hydrolyzed, and analyzed. The ratio of mannose to glucose to
galactose varied with each fraction, and the floe was classified as a
galactoglucomannan, or a mixture of glucomannan and galactomannan (ref. 20).
Subsequent work on a polysaccharide designated CP involved removal of
trace amounts of starch and dextran by enzymes, alcohol precipitation of the
polysaccharide, followed by dialysis. Structural analysis of the purified
polysaccharide involved gel filtration, permethylation followed by
gas-liquid chromatography-mass spectrometry, periodate oxidation, enzymatic
analysis, and acid hydrolysis followed by thin layer chromatography.
Gel filtration produced two fractions, fraction one being a
galactomannan with a D-mannose to D-galactose ratio of 2.3 : 1.0, a specific
rotation value of +97.3° and a molecular weight of 3,500,000 daltons. This
fraction was structurally analyzed; fraction two was mainly
glucose-containing with traces of galactose, arabinose and xylose.
218
Ç MAN Ì1 I
KAN : a-D-MANNOSE
GAI. : a-0-GALACTOSE
Fig. 5. Proposed structure A of galactomannan CP.
Two possible structures A and B were proposed for the galactomannan:

(A) The main chain consisted of a-(1,6) and a-(1,2) linked mannose
units. Single galactose-mannose residues in the ratio of 3 : 1 were attached
as side chains to 861 of the a-(1,6) linked mannose units by way of a-(1,2)
linkages (Figure 5 ) .
(B) The main chain consisted of only a -(1,6) linked mannose units, 86Z
of which were branched at 0-2 (Figure 6). Mannose residu.es were a-(l,2)
linked in the side chains, which were terminated by non-reducing mannose or
galactose groups (ref. 21).
219
i Œ*
Fig. 6. Proposed structure B of galactomannan CP.
Molasses polysaccharide
A polysaccharide was isolated from a high viscosity molasses sample by

separation with cetyl trimethylammoniurn bromide, and purification by alcohol
precipitations and gel filtration. Analysis by " C nuclear magnetic
resonance spectroscopy and by gas-liquid chromatography-mass spectrometry
indicated the polysaccharide was composed of glucose-galactose-arabinose in
the ratio of 3 : 1 : 1, with trace amounts of mannose and xylose. The
backbone of the polysaccharide consisted of a-(1,3) linked glucose units,
with galactose branch chains attached by a-(1,6) linkages to the glucose,
220
and terminal arabinose units linked a-(1,6) to the galactose (Figure 7 ) . No

specific rotation, molecular weight or proposed source of the polysaccharide
was reported (ref. 22).
CH20H
QSSli GAL : Cr-D-GALACTOSE
OH
QÇLÛÇ),
ARAB : L-ARABINOSE
0 OH
C•L
GAL ]1 HOCHJ"
C biucli
5 l
Fig. 7. Proposed structure of molasses polysaccharide.
Dextrans
Dextrans are polysaccharides which, although associated with sugarcane,

are not considered to be present in healthy, sound cane. They are, however,
readily produced in damaged sugarcane, or during sugar manufacture under
conditions of poor housekeeping. An excellent review by Imrie and Tilbury
deals specifically with the dextrans arising from deteriorated cane or cane
juice, the effects of these dextrans on processing, and how to minimize the
effects (ref. 1 ) . A very detailed review by Sidebotham discusses the origin
and the structures of dextrans bacterially produced on a sucrose substrate,
and includes those produced in the sugar industry (ref. 23). A further
review outlines the bacterial production and structural determination of
dextrans; the review also considers the commercial uses of dextrans, which
represents areas of usage normally avoided by the sugar manufacturer (ref.
24).
Dextrans are glucose-containing polysaccharides in which the main chain
residues are a-(1,6) linked, with a-(1,4) and a-(1,3) linkages at the branch
points (ref. 24), A dextran (Figure 8) is defined as a glucan which has at
least 50-60Z of a-(1,6) linkages (refs. 25, 26). Most dextrans have high
221
molecular weights, the range being 100,000 to 10,000,000 daltons. They are
normally soluble in water, insoluble in alcohol, and are highly
dextrorotatory with specific rotations of +200° and greater (ref. 1).
Dextrans are produced by bacteria which grow almost exclusively on
sucrose-containing media; these bacteria are confined to the genera
Lactobacillus, Leuconostoc and Streptococcus (ref. 23). Two of the
most common species of bacteria involved in the formation of dextran are
Leuconostoc mesenteroides and Leuconostoc dextranicum. Both of these
microorganisms have been identified in deteriorated mill juices, but so far
only the former has been found in contaminated refinery products (ref. 27).
Leuconostoc mesenteroides is a species of bacteria which contains many
different strains. Jeanes et al. (ref. 28), who characterized and
classified the dextrans produced from ninety-six different strains of
bacteria, found that dextrans with unlike properties were produced from
approximately eighty different strains of Leuconostoc mesenteroides. The
strain of bacteria that produced any particular dextran was found to affect
the percentage of a-(1,6) linkages, the type and percentage of branch
points, the molecular weight, and the specific rotation value. Even the
solubility varied with the strain type, greater contents of a-(1,6) linkages
increasing the water solubility, greater contents of a-(1,3) linkages
decreasing the water solubility (refs. 23, 28, 29).
Leuconostoc mesenteroides is found in most soils, is airborne, and thus
is also on the sugarcane plant (ref. 25). Contamination of the plant is
therefore possible, but generally requires prior damage to the rind of the
cane stalk by storm, burning (refs. 30, 31), insect problems or freezing
(refs. 32, 33). Any delays which occur between burning and harvesting, or
between harvesting and processing (refs. 34-37), can also create conditions
which are favorable for the growth of Leuconostoc (ref. 38). Contamination
by this microorganism represents a decrease in sucrose yield and, therefore,
an economic loss (ref. 26), but the accompanying formation of dextrans has
even more far-reaching effects on the manufacture of raw sugar and on the
refining processes (ref. 39).
222
( Uuc jj
1 >, o
Λ Uwe ii J GÜ^tf
0 (UOl
3 0
0
Γ c^ i
It 3 Ò
tGLUC
GLUC It
0
I
Fig. 8. Representation of dextran (Leuconostoc mesenteroides strain

B512).
Summary
Although the sugarcane plant is known as a highly efficient producer of
sucrose, the plant is also host to a wide variety of polysaccharides. These
materials differ considerably in their structure, their constituent
monosaccharides, their molecular weights and other properties, and in their
ability to affect the isolation and purificaton of sucrose. It is this last
category which is the subject of the second half of this review.
EFFECTS OF SUGARCANE POLYSACCHARIDES ON SUGAR MANUFACTURE
In the early days of sugar manufacture, many processing problems were

attributed to the presence of starch. The early investigations were
hindered by the lack of an accurate starch analysis method, and by
223
disagreement as to what starch content caused a problem. There was also

disagreement whether the problems were caused solely by starch, or by a
combination of starch and other non-sugar impurities. It is now believed
that the presence of starch in cane juice and in raw sugar has its greatest
effect on the filterability of those materials.
As well as starch, other cane polysaccharides have a deleterious effect
on processing, usually by affecting viscosities, polarization values and
evaporation processes. The polysaccharides which are produced by
microbiological contamination, whether in the field, factory or refinery,
also cause difficulties in sugar production. The formation of dextrans,
primarily by the bacteria Leuconostoc mesenteroides. not only affects every
stage in sugar manufacture, but also represents a loss of sucrose.
The effects of all of these polysaccharides on sugar processing have
been thoroughly reviewed by Imrie and Tilbury (ref. 1 ) . Reviews dealing
specifically with dextrans include a paper discussing their formation and
their effects on processing in Australia over a thirty year period (ref.
39), a detailed account by Sidebotham of their effects in mills and
refineries (ref. 23), a discussion of the problems in sugar production (ref.
25), and a paper dealing solely with the effects on refinery processes (ref.
27). Other articles deal with the laboratory evaluation of dextran in
various refining processes (ref. 40), or on the passage of raw sugar dextran
through the refinery (refs. 41, 42).
Polarization values
Excluding the dextrans, the specific rotation values of most of the
polysaccharides associated with sugarcane are not overly dextrorotatory or
overly levorotatory. They should not, therefore, be expected to affect the
polarization of sugar samples unless present in very high concentrations.
The dextrans, however, are highly dextrorotatory, with specific
rotation values of +200°, and higher (refs. 1, 28). This value is at least
three times that of sucrose (+66.54°), which suggests that dextran-
containing raw sugar samples should show an enhanced polarization value.
Such an effect was observed when refined and raw sugar samples, which were
known to contain dextran, were dialyzed to remove the sucrose, and the
specific rotation/weight dextran determined (ref. 40). The enhancement in
polarization was about 0.3°/l,000 ppm dextran (refs. 40, 42).
22^
Polarization of dextran-containing cane juice or raw sugar samples,

after clarification by basic lead nitrate, did not show enhanced values,
while polarization after clarification with basic lead acetate did (ref.
43). In further studies, clarification by basic lead acetate prior to
polarization determination was found to co-precipitate some of the dextran
(ref. 44); in other studies, basic lead acetate clarification was found to
preferentially remove the higher molecular weight dextrans (> 40,000
daltons) leaving lower molecular weight dextrans (10,000 daltons) to affect
the polarization value (ref. 45). Thus, polarization values are elevated by
the presence of dextrans, but prior clarification by basic lead acetate
drastically decreases the effect.
Viscosity increases
With the exception of Roberts' glucan which showed little effect on
viscosity, the solubilization of sugarcane polysaccharides in juices and
liquors results in viscosity increases. Sarkaran (ref. 10) and molasses
polysaccharide (ref. 22) were both isolated from highly viscous products,
the high viscosity values being attributed to their presence. Starch
solubilized during milling also results in increases in viscosity (ref. 46).
The polysaccharides most thoroughly examined, however, are the
dextrans; their presence in juices and liquors increases the viscosity to
such an extent that many manufacturing processes are affected (ref. 47).
Furthermore, after the initial viscosity increase, the effect can be
worsened by the recycling and further dextran build-up via low purity
streams (ref. 41). Thus, viscosity increases can occur in the magma (ref.
42), massecuites (refs. 25, 48) and molasses (refs. 25, 48).
In laboratory studies, mixtures of cane dextran (molecular weight
5,000,000 daltons) and sucrose showed that the viscosity increased at a
slightly faster rate than the concentration of dextran increased (ref. 49).
This effect was found to be temperature independent (ref. 50). An increase
in viscosity also occurred with an increasing molecular weight of the
dextran added (ref. 51); furthermore, the viscosity was affected by the
structure of the dextran; the higher the degree of branching, the weaker was
the connection between molecular weight and viscosity (ref. 50). In 65Z
sucrose solutions at 80°C, the presence of 1Z dextran increased the
viscosity by 100-130Z, while 3Z dextran gave a 250-3502 increase (ref. 46).
225
Dextrans in mill juices often resulted in poor clarifications of those

juices (refs. 35, 52). In refineries, where the dextran levels were much
lower, it was expected that the polysaccharides would be removed by the
clarification procedures. It was reported, however, that both carbonatation
and phosphatation (with and without cationic surfactants) were ineffective.
The increased viscosities due to dextran particularly influenced
phosphatation clarification, as the rates of flocculation, coagulum
flotation and scum compression were all adversely affected (ref. 40).
Decolorizations of feed liquor by bone char, granular or powdered
carbon, and ion exchange resins also did not affect the dextran
concentrations. Their presence in the feed liquors, and the accompanying
viscosity increases, adversely affected the decolorization processes, as
viscosity influences the rate of adsorption and the diffusion process (ref.
40). A recent publication described the complete blocking of bone char
adsorbents by a white gel-like material, presumably a dextran of unusual
structure (ref. 53).
The high viscosities produced in cane juices and liquors by the
presence of dextrans result in a decrease in sucrose crystallization rate
(refs. 37, 47, 48), the decrease being most pronounced in low-grade boilings
(ref. 54). The effects of these high viscosities are decreased heat
transfer in the evaporators, crystallizers and pans, increased in-boiling, a
decrease in both sucrose yield and quality, and increased energy costs due
to longer processing times (refs. 25, 40).
Filterability
Press filtration is used in sugar manufacture as a direct filtration

process or as a liquor polishing step. As these filtrations are carried out
throughout the manufacturing process, any decrease in filterability will
decrease throughput and increase energy and handling costs.
Starch is known to adversely affect filterability. In the milling
process, starch is released into the extracted juice as insoluble granules.
Clarification gelatinizes most of these granules; the soluble starch passes
into the clarified juice, and during crystallization is occluded in the raw
sugar. Dissolving of the raw sugar in the refinery releases the starch,
which in clarification by carbonatation causes severe filtration problems.
The amylopectin portion of starch, being negatively charged, interacts with
226
anionic sites on the growing crystal, and is immobilized in the crystal

matrix. The amylose, being neutral, accumulates on the surface of the
growing calcium carbonate crystal. This prevents the formation of
agglomerates, which normally provide a loose packing arrangement and faster
filtrations. Amylose, therefore, was considered one of the main filtration
impeding impurities in raw sugar (refs. 3, 55-57).
Other workers confirmed that the filtration rate after carbonatation
clarification was severely retarded by even small amounts of amylose.
Laboratory prepared samples of cane and potato amylopectins also retarded
filtrations, and this effect was attributed to blockage of the filter medium
due to their molecular size. No difference in size was noted for
agglomerated calcium carbonate crystals formed with, and without, the
addition of amylopectin (ref. 58). Amylopectin was also found to retard
liquor polishing rates in refinery-type phosphatation clarification, by
upsetting the coagulation of the scums (ref. 59).
The effect of starch on carbonatation filterability was found to be a
function of molecular weight and the degree of starch hydrolysis. The
molecular weights were expected to cover a wide range in raw sugars due to
the action of natural and added enzymes on the cane starch (ref. 60).
Dextrans with molecular weights no larger than 2 x 10 6 daltons had no
significant effect on carbonatated filterability, whereas a dextran of
molecular weight 5-40 x 10 6 daltons had very serious effects. The impedance
was markedly reduced by treatment of the polysaccharide by dextranase (ref.
60). Samples of different molecular weight dextrans (i.e. from 10,000 to
40,000,000 daltons) added to refined sugar showed that an increase in
dextran concentration gave a decrease in filterability, the effect being
most pronounced with the higher molecular weight dextrans. The
filterability was almost unaffected by low molecular weight dextran (ref.
49).
Dextran from deteriorated cane was also found to decrease the
filterability of raw sugar liquors. Plotting of the dextran content versus
the filterability for a number of samples gave graphs which showed an
inverse relationship: as the dextran concentration increased, the
filterability decreased, and vice versa (ref. 61).
Filterability variations with raw sugar were attributed to minor
constituents in the sugar, dextran and starch being the most undesirable.
227
The principal effects on filterability were filter plugging and increased

viscosity. Starch, as ungelatinized granules, caused plugging, but in
solution, after gelatinization, increased the viscosity. The presence of
dextran in raw sugar often doubled or tripled the normal viscosity of the
syrup (ref. 46). In cases where the dextran concentrations did not affect
the viscosity, filter impedance was attributed to a surface tension effect
due to the dissolved dextran. Ultracentrifugation to remove insolubles gave
sugar solutions which still filtered more slowly than a pure sugar solution,
and it was concluded that filtration could be impeded by soluble as well as
insoluble components (ref. 62).
Filtration, therefore, appeared to be affected by viscosity and by
total solids (ie. dissolved and suspended). The viscosity due to dextrans
was subsequently found to have very little effect on the filtration
impedance of washed raw sugars at 25°C, the major contribution being as
dissolved solids. The total soluble sugarcane polysaccharides had less
effect on filtration impedance than dextran alone, again as dissolved solids
rather than through viscosity. The molecular weight of the dextrans was
also important, as increased molecular weight resulted in increased filter
impedance. Although the impedance was lower with the lower weights, there
was still considerable filter impedance at the lowest molecular weight
(70,000 daltons), and therefore a potential problem for the refiner. The
suspended solids, isolated by centrifugation, were found to cause severe
filtration impedance. They were either polysaccharides (75-90Z acid
hydrolyzable) or soil particles (non-hydrolyzable), with the major fraction
of particles under 10 μια. in size (ref. 63).
The filter impedance of washed raw sugars was also examined at 70°C and
80°C. As with the room temperature experiments, the suspended solids showed
the greatest filter impedance. In the dissolved solids fraction, the effect
of total soluble sugarcane polysaccharides increased as the temperature
increased, but dextran remained the most significant polysaccharide to
affect filtration. The low molecular weight dextrans still showed
considerable filtration impedance, with very little difference being found
between the 10,000 and 40,000 dalton molecular weights. With starch, the
filterability decreased with gelatinization of the starch granules, the
soluble starch effect being already evident at 70°C. The solutions
228
containing ungelatinized starch showed extremely high filtration impedance,

the starch granules still being intact (ref. 64).
A similar study discussed the various filter impeding impurities in raw
sugar, and concluded that although a particular impurity may decrease the
filterability in one raw sugar, another raw sugar with the same impurity may
be unaffected. It was concluded that filterability studies involved complex
mechanisms and interpretations, and the results may not even be applicable
to different raw sugar samples from the same country (ref. 65).
Sucrose crystal shape

During crystal growth, the shape of sucrose crystals can be influenced
by the presence of impurities. Polysaccharides are one of several
impurities which are preferentially adsorbed on a crystal face, causing that
face to grow more slowly. The dextrans are the only polysaccharides
associated with sugarcane which are believed to undergo such an adsorption,
the result being crystals elongated in a specific direction.
When dextran was added to pure sucrose solutions, and boiled, the
sucrose crystals showed c-axis elongation. This elongation was increased
when the dextran was added to a good quality syrup, and further increased
when a refractory syrup was used. Removal of the high molecular weight
material from the refractory syrup, followed by boiling, gave sucrose
crystals in which the c-axis elongation had been eliminated. The addition
of this high molecular weight material to standard sugar, and subsequent
boiling, gave elongated crystals identical to those produced by the addition
of dextran (ref. 66).
The c-axis elongations were enhanced by increasing the concentration
and the molecular weight of the added dextran, and by increasing the
temperature. The effect was not particularly strong with pure sugar
solutions, which required high levels of dextran to produce significant
elongations. Dextran produced much greater elongations in factory syrups,
which suggested that materials in the syrup interacted with dextran and
increased its effect as a habit modifier (ref. 67). Other workers found
similar effects, limited elongation in solutions of sucrose and dextran, and
extensive elongations in solutions with dextran and liquor impurities. They
suggested that dextran and the liquor impurities exerted a synergistic
effect, leading to the extensively elongated crystals found in needle grain
229
(refs. 68, 69). This joint "dextran-syrup component" elongation was thought
to occur by a preferential adsorption onto the faces which hindered b-axis
elongation (ref. 66), thereby ensuring that growth occurred more rapidly
along the c-axis (ref. 70).
The structure of the dextran was also important, c-axis elongations
occurring only with dextrans having 83Z and higher a-(1,6) linkages (refs.
67, 71). Polysaccharides with low percentages of a-(l,6) linkages [ie. high
a-(1,4) linkages] did not cause c-axis elongation (refs. 67, 72).
There is some literature which questions the degree of influence that
dextran, oligosaccharides and polysaccharides have on sucrose crystal shape,
as will be briefly described. Syrups prepared from deteriorated cane were
fractionated into oligo- and polysaccharides. Each of three
polysaccharides, two being glucans, was found to cause c-axis elongation in
sucrose-polysaccharide crystallizations. The oligosaccharides did not, and
therefore, were not considered to be a primary cause of crystal elongation
(ref. 73). The mono- and oligosaccharides produced by partial or extensive
acid hydrolysis of dextrans were found to decrease the elongation effects as
the molecular weight decreased, suggesting that the oligosaccharides were
less significant than the polysaccharides (ref. 74). Other workers found
that a variety of oligosaccharides did not affect the axial ratios of
sucrose, while dextran did (ref. 75).
An oligosaccharide and a polysaccharide isolated from refinery molasses
were tested in sugar crystallization experiments, and the oligosaccharide
was found to cause elongation (ref. 76). Further experiments concluded that
the oligosaccharides had more influence on the elongation of sucrose
crystals than the polysaccharides, and that microorganisms present in juices
produce oligosaccharides which can cause c-axis elongations (ref. 77).
Recent publications reported that oligosaccharides caused c-axis elongation
in raw sugar factories and refineries, while polysaccharides at higher
concentrations also contributed to some of the elongation in the factories.
The treatment of the oligosaccharides by invertase removed their elongating
properties; tentative identification of the oligosaccharides indicated at
least two fructosyl-sucroses and at least two glucosyl-sucroses (refs. 78,
79).
230
A study with raffinose- and dextran-modified sucrose crystals found the

b-axis elongated by the former, and the c-axis elongated by the latter. The
b-axis elongated crystals were needle-shaped, while the c-axis elongated
crystals were square shaped (ref. 80), such as are normally produced in
massecuites. These square-shaped crystals were already elongated in the c-
direction, as pure sucrose crystals grown in water are approximately twice
as long in the b-direction as in the c-direction. The observations that
only limited elongations occurred with dextrans in pure sucrose solutions
can probably be attributed to workers accepting the square-shaped crystals
in massecuites as being the proper sucrose crystal shape (ref. 78).
The extensive c-axis elongation which occurs in needle grain makes
sucrose recovery by centrifuging very difficult. Firstly, the crystal shape
affects both the packing of the crystals in the fugai basket and their
subsequent washing, elongated crystals being less densely packed and less
easily washed (ref. 81). Elongated crystals are also more fragile and may
break during centrifuging (ref. 70). Massecuites with elongated crystals
are usually poorly purged, with the crystals either passing through the
fugai screens or plugging the screen holes (ref. 82). Low grade massecuites
are even more difficult to centrifuge. Soft massecuites with these crystals
are also difficult to purge, and the resulting sticky (gummy) soft sugars
are troublesome and time-consuming to package (refs. 40-42).
Melassigenic effect
Dextran has a melassigenic effect which decreases the yield of sucrose
(ref. 25). Furthermore, high concentrations of dextran in the final white
syrups and in the affination syrup ultimately end up in the soft sugars and
blackstrap molasses (ref. 42). The entry of any large quantities of dextran
into a refinery, therefore, can lead to elongated crystals, poor centrifugal
separations and, coupled with the melassigenic effect, will produce molasses
of higher than normal purity (refs. 40, 41). The other polysaccharides of
sugarcane are not known to cause a melassigenic effect.
Acknowledgements
The author is grateful to the senior management of B.C. Sugar for the
opportunity to prepare this chapter, to Mrs. Anne Kitchen for her excellent
diagrams and enthusiastic proof-reading, to Sandra Daly for her diligent
231
typing and retyping, and to Mr. Dan O'Connell for his usual perfection in
slide preparation.
Literature Cited
1 F.K.E. Imrie and R. H. Tilbury, Polysaccharides in sugar cane and its
products. Sugar Technol. Rev. 1 (1972) 291-361.
2 M. A. Clarke, E. J. Roberts, M. A. Godshall and F. W. Parrish,
Non-starch, soluble polysaccharides of sugarcane. Proc. Sugar Process.
Res. Conf., 1986, in press.
3 J. B. Alexander and M. Matic, Starch: its occurrence, importance, and
removal in sugar manufacture. Proc. Tech. Sess. Cane Sugar Refining
Res. (1974) 13-28.
4 F. W. Parrish, W. R. Goynes, E. J. Roberts and M. A. Clarke, Recent
observation on starch and sugarcane products. Proc. Sugar Process. Res.
Conf. (1984) 53-59.
5 E. C. Vignes, Notes on cane starch and its determination. Proc. Intern.
Soc. Sugar Cane Technologists 15 (1974) 1288-1295.
6 J. Bruijn, Deterioration of sugar cane after harvesting. 1. Changes
in juice composition. Intern. Sugar J. 68 (1966) 331-334.
7 J. Bruijn, Deterioration of sugar cane after harvesting. 2. Investi
gation of the polysaccharide formed. Intern. Sugar J. 68 (1966)
356-358.
8 J. Bruijn, Deterioration of sugar cane after harvesting. 3. Enzymatic
hydrolysis of the polysaccharide formed. Intern. Sugar J. 72 (1970)
195-198.
9 J. D. Blake and J. Littlemore, A water-soluble polysaccharide from
stand-over cane. 1. Isolation and structural characterization from
the field. Intern. Sugar J. 86 (1984) 222-226.
10 J. D. Blake and J. Littlemore, A water-soluble polysaccharide from
stand-over cane. 2. Isolation and structural characterization from
molasses. Intern. Sugar J. 86 (1984) 235-240.
11 J. D. Blake and M. L. Clarke, A water-soluble polysaccharide from
stand-over cane. 3. Studies on the measurement of sarkaran. Intern.
Sugar J. 86 (1984) 255-259.
12 J. D. Blake and M. L. Clarke, A water-soluble polysaccharide from
stand-over cane. 4. Studies on the physical properties and origin
of sarkaran. Intern. Sugar J. 86 (1984) 276-279.
13 E. J. Roberts, J. T. Jackson and J. H. Vance, Progress in research on
the soluble polysaccharides of sugarcane. Proc. Tech. Sess. Cane
Sugar Refining Res., 1964, 76-84.
14 E. J. Roberts, M. A. Godshall, F. G. Carpenter and M. A. Clarke,
Composition of soluble indigenous polysaccharides from sugar cane.
Intern. Sugar J., 78 (1976) 163-165.
15 E. J. Roberts and M. A. Godshall, Identification and estimation of
glucuronic acid in indigenous sugar cane polysaccharide. Intern. Sugar
J. 80 (1978) 10-12.
16 J. D. Blake, M. L. Clarke and P. E. Jansson, An arabinogalactan from
sugar cane. Carbohydr. Res. 115 (1983), 265-272.
17 J. D. Blake and M. L. Clarke, Observations on the structure of I.S.P.
Intern. Sugar J., 86 (1984) 295-299.
232
18 E. J. Roberts, M. A. Clarke, M. A. Godshall and F. W. Parrish, A glucan

from sugarcane. Proc. Sugar Process. Res. Conf.,1984, 60-71; idem,
Intern. Sugar J. 87 (1985) 227-231.
19 T. Miki, S. Saito and M. Kamoda, Composition of polysaccharides in
carbonated beverage floe. Intern. Sugar J. 77 (1975), 67-69.
20 T. Miki, S. Saito, H. Ito and M. Kamoda, Composition of poly
saccharides in carbonated beverage floe from raw cane sugar. Proc.
Intern. Soc. Sugar Cane Technologists, 17 (1980) 2751-2762.
21 T. Miki, A galactomannan in carbonated-beverage floe from raw cane
sugar. Carbohydr. Res. 129 (1984) 159-165.
22 J. A. Cremata and L. R. Orozco, A new polysaccharide isolated from
high viscosity final molasses structural study. Proc. Intern. Soc.
Sugar Cane Technol. 17 (1980) 2546-2555.
23 R. L. Sidebotham, Dextrans. Advances in carbohydrate chemistry and
biochemistry. Academic Press Inc.: New York, 1974; Vol. 30, 371-444.
24 P. T. Murphy and R. L. Whistler, Industrial gums: polysaccharides and
their derivatives, Academic Press Inc.: New York, 1973, 513-542.
25 E. E. Coll, M. A. Clarke and E. J. Roberts. Dextran problems in sugar
production. Proc. Tech. Sess. Cane Sugar Refining Res.,1978, 92-106.
26 M. A. Clarke, E. J. Roberts, M. A. Godshall, M. A. Brannan, F. G.
Carpenter and E. E. Coli, Sucrose loss in the manufacture of cane
sugar. Sugar y Azucar 75 (1980) 64-68.
27 R. A. Kitchen, Dextrans: their effects on refinery processes. Proc.
Intern. Dextran Workshop, Sugar Processing Research, Inc., New Orleans,
1984, 53-61.
28 A. Jeanes, W.C. Haynes, C. A. Wilham, J. C. Rankin, E. H. Melvin,
M. J. Austin, J. E. Cluskey, B. E. Fisher, H. M. Tsuchiya and C. E.
Rist, Characterization and classification of dextrans from ninety-six
strains of bacteria. J. Am. Chem. Soc. 76 (1954) 5041-5052.
29 E. J. Bourne, R. L. Sidebotham and H. Weigel, Dextrans and dextranases.
X. Types and percentages of secondary linkages in the dextrans
elaborated by Leuconostoc mesenteroides NRRL B-1299. Carbohydr. Res.
22 (1972) 13-22.
30 D. H. Foster, P. A. Inkerman and K. E. McNeil, Studies on cane
deterioration in Australia. Proc. Intern. Soc. Sugar Cane Technologists
17 (1980) 2204-2220.
31 D. H. Foster, Trends in green cane processing. Proc. Australian Soc.
Sugar Cane Technologists, 1979, 11-17.
32 G. Singh and S. Singh, Effect of freeze temperature on quality
indicators of juice in sugar cane. Intern. Sugar J. 77 (1975) 131-132.
33 B. L. Legendre, W.S.C. Tsang and M. A. Clarke, Changes in juice
composition of sugarcane as affected by post-freeze deterioration in
Louisiana. Proc. Sugar Process. Res. Conf., 1984, 92-107.
34 W. D. Wells and G. P. James, Rapid "dextrans" formation in stale cane
and its processing consequences. Proc. Queensland Soc. Sugar Cane
Technologists, 1976, 287-293.
35 R. H. Tilbury, Dextran and dextranase. Proc. Intern. Soc. Sugar Cane
Technologists, 14 (1971) 1444-1458.
36 R. P. Fulcher and P. A. Inkerman, Further studies on the deterioration
of cane and cane juice. Proc. Queensland Soc. Sugar Cane Technologists,
1974, 161-169.
37 J. S. Keniry, J. B. Lee and C. W. Davis, Deterioration of mechanically
harvested chopped-up cane. 1. Dextran: a promising quantitative
indicator of the processing quality of chopped-up cane. 2. The Rate of
dextran formation. Intern. Sugar J. 69 (1967) 330-333, 357-360.
38 E. B. Lillehoj, M. A. Clarke and W.S.C. Tsang, Leuconostoc spp. in
sugarcane processing samples. Proc. Sugar Process. Res. Conf., 1984,
141-151.
39 P. C. Atkins and R. J. McCowage, Dextran--an overview of the Australian
experience. Proc. Sugar Process. Res. Conf., 1984, 108-140.
40 C. C. Chou and M. Wnukowski, Dextran problems in sugar refining: a
critical laboratory evaluation. Proc. Tech. Sess. Cane Sugar Refining
Res., 1980, 1-25.
41 M. J. Fowler, Raw sugar dextran flow through a cane sugar refinery.
Intern. Sugar J. 83 (1981) 74-77.
42 K. R. Hanson, The effect of high-dextran-content raw sugars on refinery
performance. Proc. Sugar Ind. Tech. 39 (1980) 152-159.
43 B. Guzman, Polarimetrie analysis in the sugar industry: influence of
clarifying agents and of polysaccharides on the polarization of cane
juices and molasses. Intern. Soc. Sugar Cane Technologists. 16 (1977)
2897-2907.
44 6. A. Bradbury, R. M. Urquhart, J. H. Curtin and R. J. McCowage.,
The effect of dextran on raw sugar polarization. Sugar J., 48 (1986)
11-13.
45 M. A. Clarke, E. J. Roberts, T.B.T. To and W.S.C. Tsang, Update on
dextrans amd dextran analysis. Proc. Sugar Ind. Tech. 45 (1986)
253-266.
46 G. M. Elgal, Minor constituents of sugar and filtration. Proc. Sugar
Process. Res. Conf., 1982, 104-117.
47 W. Mauch, E. Farhoudi, Quality factors in commercial white granulated
sugar. Sugar Technol. Rev. 7 (1980) 87-171.
48 J. A. Cremata, F. Cumuna and L. Corderò, Polysaccharides as sucrose
crystalline habit modifiers. 1. Influence in crystal 'c' axis
elongation. Proc. Intern. Soc. Sugar Cane Technologists. 18 (1983)
433-454.
49 R. P. Fulcher and P. A. Inkerman, Dextranase. 4. The effect of cane
dextran on the filterability of sugar. Proc. Queensland Soc. Sugar Cane
50 P. F. Greenfield and G. L. Geronimos, Effect of dextrans on the
viscosity of sugar solutions and molasses. Intern. Sugar J. 80 (1978)
67-72.
51 G. L. Geronimos and P. F. Greenfield, Viscosity increases in
concentrated sugar solutions and molasses due to dextrans. Proc.
Queensland Soc. Sugar Cane Technologists., 1978, 119-126.
52 P. N. Stewart and C. A. Rehbein, A survey of methods of analysis for
sour juices. Proc. Queensland Soc. Sugar Cane Technologists., 1964,
255-261.
53 R. A. Kitchen, M. A. Clarke and A. J. DeLucca III. Gel-like material
produced in the processing of colored liquors. Proc. Sugar Process.
Res. Conf., 1986, in press.
54 J. A. Watson, The recovery of sucrose from low-grade refinery syrups.
Sugar Technol. Rev. 8 (1981) 81-147.
55 J. P. Murray, Filtering quality of raw sugar: influence of starch and
insoluble suspended matter. Proc. Ann. Congr. S. African Sugar Tech.
Assoc. 46 (1972) 116-132.
56 J. Bruijn, Research at Sugar Milling Research Institute, South
Africa. Sugar J. 37 (1975) 7-9.
57 J. P. Murray, F. M. Runggass and M. Vanis, Filtering quality of raw
sugar. Sugar J. 39 (1976) 14-20.
58 E. Whayman and A. L. Willersdorf, Cane starch. 3. Effects of starch
on carbonatation. Intern. Sugar J. 78 (1976) 99-100.
59 E. Whayman and T. W. Meredith, Effect of raw sugar components on
refinery filtration. Proc. Intern. Soc. Sugar Cane Technologists. 16
(1977) 2581-2585.
60 P. Hidi and R. J. McCowage, Quantification of the effects of different
raw sugar impurities on filtration rates in carbonatation refineries.
Proc. Sugar Process. Res. Conf., 1984, 186-208.
61 G. P. James and J. M. Cameron, Influence of deteriorated cane on raw
sugar filterability. Proc. Queensland Soc. Sugar Cane Technologists,
1971, 247-249.
62 G. N. Richards and P. N. Stewart, Contribution of soluble and insoluble
components to raw sugar filterability. Proc. Queensland Soc. Sugar Cane
63 J. A. Devereux and M. A. Clarke, Non-sucrose components of cane sugar
and efficiency of press filtration. Proc. Sugar Ind. Tech. 43 (1984)
36-59.
64 J. A. Devereux and M. A. Clarke, Observations on filtration impedance
in raw sugar. Proc. Sugar Process. Res. Conf., 1984, 209-230.
65 C. C. Chou, Filtration-impeding impurities in raw sugar. Proc. Sugar
Ind. Tech. 42 (1983) 56-94.
66 D. N. Sutherland, Dextran and crystal elongation. Intern. Sugar J. 70
(1968) 355-358.
67 D. N. Sutherland and N. Paton, Dextran and crystal elongation: further
experiments. Intern. Sugar J. 71 (1969) 131-135.
68 M. Saska and J. A. Polack, Effects of dextran on sucrose crystal shape.
Proc. Sugar Process. Res. Conf., 1982, 134-139.
69 G. Mantovani, C. A. Accorsi and G. Vaccari, Non-sugars affecting
sucrose habit modifications. Proc. C.I.T.S., 15 (1975) 399-405.
70 A. VanHook, Growth of sucrose crystals. A review. Sugar Technol. Rev.
8 (1981) 41-79.
71 M. T. Covacevich, G. N. Richards and G. Stokie, Studies on the effect
of dextran structure on cane sugar crystal elongation and methods of
analysis. Proc. Intern. Soc. Sugar Cane Technologists. 16 (1977)
2493-2507.
72 M. Matic, Dextran problems. Proc. Tech. Sess. Cane Sugar Refining Res.,
1980, 121-122.
73 G. J. Leonard and G. N. Richards, Polysaccharides as casual agents in
production of elongated sucrose crystals from cane juice. Intern. Sugar
J. 71 (1969) 263-267.
74 M. Saska and J. A. Polack, Effects of dextran hydrolysate on sucrose
crystal elongation. Zuckerind. 107 (1982) 941-943.
75 F. W. Parrish and M. A. Clarke, Effects of oligosaccharides and poly
saccharides on sucrose crystallization. Proc. C.I.T.S. 17(1983)
467-482.
76 M. F. Kamoda, H. Onda, T. Ito, T. Shirasaki, M. Takeshi and T. Ando,
On the formation of needle-shaped sugar crystals. Proc. Intern. Soc.
Sugar Cane Technologists. 13 (1968) 362-373.
235
77 J. H. Montenegro, B.E.M. Recort, E.L.R. Suarez, A. L. Guerra and M. D.

Ramos, Nonsugars and their influence upon the sucrose crystal habit.
Proc. Intern. Soc. Sugar Cane Technologists. 18 (1983) 477-505.
78 P. G. Morel du Boil, Sucrose crystal habit in a refinery. Proc. Ann.
Congr. S. African Sugar Tech. Assoc. 59 (1985) 33-38.
79 J. Bruijn and P. G. Morel du Boil, Sucrose crystal deformation caused by
impurities in refinery and raw house products. Proc. Sugar Process
Res. Conf. 1986, in press.
80 T. H. Shah and H. J. Delavier, Study of the influence of some non-
sucrose substances on habit modification of crystallized sucrose.
Z. Zuckerind. 24 (1974) 27-31.
81 F.H.C. Kelly, The morphology of sucrose crystals. Sugar Technol. Rev.
9 (1982) 271-323.
82 S. Stachenko, Symposium on remelt crystallization, systems and tech
niques. Proc. Sugar Ind. Tech. 29 (1970) 176-187.
236
Chapter 15
FLAVOR AND ODOR IN SUGARCANE PRODUCTS
M. A. GODSHALL
SUMMARY
Edible sugarcane products, such as table syrups, molasses and brown

sugar, possess, in addition to the sweet taste of sucrose, characteristic
flavors and odors which are derived from the sugarcane plant and from
sucrose reactions during processing. These include both volatile and
nonvolatile constituents. In this review, the identification of important
flavor compounds is described and their sources indicated. The contribution
of significant factors such as phenolics, volatiles, ash, off-flavors,
polysaccharides, and color to the flavor profile are discussed.
INTRODUCTION
A small, but significant and steadily growing, proportion of the

world's sugarcane production is utilized, not as white refined sugar, but as
high value-added, specialty products, the highly colored and flavored end
products of sugar manufacturing and refining.
As industrial consumers, food processors, and home consumers become
more sophisticated and demanding, traditional quality parameters cannot
always assure that these products will possess the desired flavor qualities
and be consistent over a period of time. It is, therefore, of interest to
understand the factors that contribute to desirable and undesirable flavors
in these products, and the influence of various processing steps on the
final flavor.
Christianson and Anhaiser (ref.1) emphasized the importance of flavor
in brown sugar and suggested that flavor quality could be controlled by
monitoring key (unspecified) processing steps. Much of the time, the flavor
of brown sugar products is a happy circumstance of traditional processing.
With regard to flavor, the Cane Sugar Handbook states only that brown
sugar flavor should be free of harsh, bitter or salty tastes, and discusses
odor only in the context of undesirable odors in refined sugars (ref.2).
Studies in the development of flavor in sugar products and the
identification of key flavor compounds has taken a secondary place to the
study of color development and identification of colorants. This is not
surprising when one considers that the vast bulk of sugar production is
white sugar, required to be free of color, odor and extraneous flavor.
237
Color formation is, however, integral to the formation of many important

types of flavors, and so the study of flavor development under processing
conditions is complementary to the study of color in sugar products. Color
formation is discussed in detail by Riffer in Chapter 13.
Types of brown sugar products

Brown sugar products exist in a variety of forms, each type reflecting
a different method of production.
Brown sugar products can be divided into two main categories—those
that are produced directly from cane juice at the place of origin, and those
that are produced during the refining of raw sugar. The first category
includes demerara, muscovado, and turbinado sugars and a variety of
cane factory molasses and syrups. They are especially valuable in the baking
industry as flavoring, coloring, sweetening and glazing agents (ref. 3 ) .
In the second category are the boiled or coated brown or "soft" sugars of
various color gradations, manufactured demerara, and a variety of refinery
molasses, golden syrups, and other syrups.
Demerara sugar, named after the county in Guyana where it was first
produced, is prepared from the first crystallization of cane syrup and has
large golden crystals and a slightly sticky texture. Turbinado sugars are
produced in a manner similar to raw sugar; much of the sticky molasses film
on the crystals is washed off in centrifuges, hence the name, producing a
mild flavored, light golden, large crystal, free-flowing product.
Muscovado, also sometimes known as Barbados sugar, is the product of
the third crystallization. It is dark brown, with a small grain and sticky
texture. There are several grades.
Refinery brown sugars, the ones most familiar to the household
consumer, are also called "soft" sugars because their very small, free-
flowing grain gives the impression of a soft texture. These are produced
either by boiling crystals from a dark liquor or by coating refined sugar
with a thin film of molasses; these latter sugars are also sometimes
referred to as "painted" sugars. In some cases, caramel coloring and invert
sugar are added in small amounts to enhance the color and texture.
Manufactured demerara is a coated sugar with a large crystal.
238
ANALYTICAL METHODS
Methods for isolating flavor and odor compounds from sugar products are
the same as those used in other fields of flavor research. The primary
requirement is to isolate the very small quantities of material that are the
flavor compounds, which may exist in the sub-parts per million range, from
the bulk of the product, which is sucrose. Methods exist which exploit the
chemical properties of the flavor compounds: acid-base relationships,
specific functional group affinities, and differential solvent solubilities.
It is also necessary to preserve compounds unaltered during extraction and
to determine relative quantities.
Once isolated, the most important flavor compounds should be identifed
and their contribution to flavor and odor determined.
The most common isolation technique is liquid-liquid extraction, and a
variety of solvents have been used, including ether, chloroform, and ethyl
acetate. As early as 1936, Takei and Imaki (ref. 4) used ether to extract
characteristic raw sugar odor from cane molasses. Japanese workers have
reported the extraction of hundreds of kilograms (up to 1 ton) of cane
molasses with acetone followed by ether to obtain an "oleoresin" (ref. 5).
Solventless extraction methods have included vacuum distillation (ref. 6)
and steam distillation (ref. 7) of molasses.
Fractionation of extracts into acid, basic and neutral fractions has
been reported (ref. 5), and adsorption onto clay or silica gel with
subsequent elution into various polarity fractions using solvents (refs.
8,9). Godshall and Roberts found that XAD-2 resin has particular affinity
for phenolic and furan type compounds in brown sugars (ref. 10).
The invention in 1979 of an external sampling device for direct elution
of volatiles into a gas Chromatograph (GC) (ref. 11) allowed rapid and easy
isolation of the most volatile compounds in foods. This device was used, in
conjunction with mass spectrometry (MS), to isolate and identify important
volatiles in cane molasses (ref. 12), cane leaves and juice (ref. 13) and
brown sugars (ref. 14).
SOURCES OF FLAVOR
The sugarcane plant
Freshly squeezed cane juice has a characteristic mild, sweet,
green-cane flavor and aroma. It quickly darkens in color from light green
to dark brown due to the activity of polyphenol oxidase and begins to
undergo the many changes caused by processing that ultimately lead to the
products under discussion.
Sugarcane leaves contain volatile components that contribute to the
typical cane taste and aroma. The most significant of these, isolated
from crushed and slightly heated cane leaves, are listed in Table 1 (ref.
13). Perhaps the most important is dimethylsulfide (DMS), which is
responsible for characteristic "cane molasses" odor at the level of about
3 ppm (ref. 12). Apparently, DMS is enzymatically released from a precursor
when leaf tissue is damaged. DMS content of cane leaves increased up to a
thousand-fold with prolonged heating of crushed cane leaves (ref. 13). All
of the compounds in Table 3 can continue through the process and have been
found in factory molasses and raw, turbinado, and brown sugars.
TABLE 1
Important volatile constituents identified in cane leaves
Odor contribution
Constituent to fresh cane juice Type of odor
Acetaldehyde Moderate Fresh, fruity, green

Ethanol Slight Alcohol, bitter
Dimethylsulfide Strong Green cane
3-Hexen-l-ol Strong Green leaves
l-Hexen-3-ol Strong Green leaves
Cane juice contains 16-19Z sucrose and 1-2Z invert. The leaves and
rind contribute to the non-sugar load of the juice, which includes
waxes, fatty acids, phytosterols, chlorophyll, flavonoids, anthocyanins,
phenolic acids and free amino acids. Protein, polysaccharides, and organic
acids enter the juice upon disruption of the tissues during milling. While
most of the waxes, chlorophyll, anthocyanins, and protein are removed or
destroyed during liming and clarification, the remaining organic and amino
acids, invert, polyphenolics and polysaccharides can contribute, alone or as
reaction intermediates, to color, taste, and viscosity, and so become
components of the over-all flavor profile. Inorganic salts enter the
process with the cane juice; these will also influence flavor.
240
Cane juice contains numerous polyphenolic compounds. Färber and

Carpenter (ref. 15) identified 21 benzoic and cinnamic acid derivatives.
Australian workers have extended the list to include a number of flavonoids
(ref. 16). While the contribution of phenolics to the flavor of cane juice
has not been studied, many of the benzoic and cinnamic acid derivatives
identified in cane juice have bitter and astringent properties (ref. 17).
Maga reported that these types of compound exhibit a strong synergistic
effect on lowering the flavor threshold when used in combination with each
other (ref. 18). It is known that they survive through the process.
Tokitomo et al. (ref.19) examined one fraction from a silicic acid
separation of a fresh cane juice ether extract which possessed a sweet,
sugary odor. They found several compounds that may be important to the aroma
and taste: hexanoic acid, benzyl alcohol, 2-phenylethanol, 3-phenyl-
2-propenol, A-vinylphenol, and 4-hydroxy-2-methoxybenzaldehyde.
Processing
Sucrose and the minor constituents listed above have the potential to
react with each other under process conditions of varied temperature and pH
to produce both acceptable and unacceptable flavors. Figure 1 outlines the
interactions.
I
|| KEY FLAVOR PRECURSORS > PROCESSING > PRODUCT
I
|| Sucrose Crushing Color
il Amino Acids Clarification Flavor
Invert Sugars Evaporation Viscosity
Phenolic Acids Filtration Over 150 compounds
Polysaccharides Crystallization (Maillard compds.)
Iron (High temp.) (Caramel compds.)
Ash (pH changes) (Acetic acid)
(Enzymatic changes) (Strecker aldehydes)
(Microbial activity) (Organic acids)
(Volatiles) ||
= = ■ ' = = M =^=^==ass i. ii = B = ± = U
Fig. 1. Flavor precursors interact under processing conditions to

produce a highly flavored product.
241
During processing, sugar undergoes a variety of degradations

which produce numerous volatile and non-volatile flavors and odors.
Extraneous influences can play a part as well as microbial activity.
Sugar degradation. Sucrose, by itself, has an enormous potential for
degradation. During processing, it can undergo thermal degradation with
caramelization as well as acid and alkaline degradation. These degradations
result in the formation of numerous volatile components, including
aldehydes, furans, and alcohols; acids and polymeric brown pigments also
result. Hodge summarized the known volatile compounds developed during
caramelization and listed their odor and taste properties (ref. 20). Most
of these have been identified at on time or another in cane sugar products.
Acidic degradation of sucrose during processing can be minimized by
maintaining the solutions at or slightly above neutrality. During acidic
decomposition, sucrose, in a self-catalyzing mechanism, produces, among many
other compounds, lactic acid, glycolic acid, pyruvaldehyde, acetic acid,
diacetyl, and, finally, 5-(hydroxymethyl)-2-furaldehyde (HMF), which
polymerizes into brown pigments. The result can be harsh and bitter flavors
in the final product.
During alkaline degradation of sucrose, fragmentation into 3-carbon
compounds leads to many of the same types of compounds seen with acid
degradation. Alkaline degradation of sucrose can proceed very rapidly,
going through an array of pleasant aromas before culminating in strong,
bitter flavors. In general, it can be said that acid degradation of sucrose
produces undesirable flavors while alkaline degradation, if stopped in time,
produces very desirable flavors.
Maillard reactions. The reaction of sugars with amines is one of the
most important reactions in flavor chemistry and also the most complex.
Numerous possible flavors can result, depending on the starting reactants,
their concentration, pH, temperature, and length of reaction. Almost any
type of cooked flavor can be reproduced with the sugar-amine reaction:
bakery, chocolate, caramel, meaty, roasted nut, and so on (refs. 21,22).
Among the compounds formed, of interest to the sugar industry, are
pyrazines (nitrogen heterocycles with nutty aromas); the Strecker
aldehydes (deamination and decarboxylation of a-amino acid by oxidants such
as carbonyl compounds which produce aldehydes and ketones with one less
2^2
carbon than the parent amino acid); and oxygen heterocycles such as furans
and pyranones.
Microbiological activity. Microbiological activity is kept to a
minimum in sugar production and refining by the high temperatures and high
concentrations of the process solutions. However, there are areas where
such activity can occur, as for example, in cane milling and during
recycling of low purity sweetwaters. Owen (ref. 23) has outlined the
microbiology of sugar processing. The microbiologically produced compounds
of most interest to the flavor profile are ethanol, acetic acid, butyric
acid, lactic acid and diacetyl. Lactic acid and diacetyl are always found
in brown sugars and molasses; they also originate from sucrose
decomposition.
Godshall and DeLucca (ref. 24) found that acetic acid is the major
volatile constituent of brown sugars, and were able to trace its origin to
bacterial action in recycled sweet waters. In this situation, acetic acid
was a desirable constituent of the flavor profile of brown sugars.
Butyric and formic acids have been identified as their benzyl esters in
molasses in the amount of 100 ppm and 1000-4000 ppm, respectively (ref. 25).
These same acids were identified as their phenacyl esters in one brown sugar
with poor flavor characteristics (11 ppm and 57 ppm, respectively) (ref.
26).
Extraneous effects. The final category of process-derived flavors
encompasses those that arise from other than action upon the sugar.
These would include such things as:
. Volatile amines that result from resin breakdown (old resin or
thermal or osmotic shock to resins), giving fishy odors;
. Iron entering from the equipment and producing metallic or acid
tastes, depending on its ionization state;
. Salty or bitter tastes from high potassium and other ash
components resulting from excessive fertilization or drought;
. Odors absorbed from packaging materials;
. Stale or flat taste that results sometimes from grinding of sugar
to make confectioner's sugar.
FLAVOR IN MOLASSES
More work has been done to identify flavor and odor constituents in
cane molasses than in any other sugar product. Molasses flavor is complex
and variable and so there is interest in understanding its properties.
In 1978, Godshall et al., listed 75 compounds previously identified in
molasses that contribute to aroma and flavor (ref. 13). Many of these were
the types of things found as a result of the various sugar degradation
mechanisms discussed above.
There are basically two types of molasses—factory molasses, generally
called blackstap, and refinery molasses, with different flavor
characteristics, reflecting their respective origins. Factory molasses is
characterized by "cane" smells, which arise from DMS and 3-hexene-l-ol
("leaf alcohol", see Table 1). Refinery molasses, by contrast, lacks DMS and
leaf alcohol or else has small quantities which are masked by other flavors.
The characteristic aroma of factory molasses is influenced, then, by two
major factors: the strong "green" or "cane" component attributed to DMS; and
the sweet, caramel component attributed to diacetyl and furans (ref. 13).
Both cane and especially refinery molasses can develop what is known as
"licorice" flavor. Some licorice flavor is desirable in molasses; the
tolerated level is very much a function of cultural and geographic factors.
This flavor arises from heating process liquors for a long time at fairly
low temperatures (ref. 27).
Molasses may become harsh and bitter, possibly from condensation of
phenolics and furans into browning polymers. High ash content also
contributes to salty and bitter tastes characteristic of both types of
molasses.
FLAVOR IN REFINERY BROWN SUGARS AND FACTORY TURBINADOS

Compounds listed in Table 2 contribute to flavor and aroma in refinery
brown sugars (refs. 10,29).
Factors that correlate with the flavor of refinery brown sugars and
factory turbinados were examined by Godshall et al. (ref. 14). Chemical,
sensory, and volatile profiles were developed for 27 sugars, including light
and dark brown coated and boiled refinery softs and factory turbinados.
Sensory profiles were obtained using the descriptive analysis method.
Trained panelists rated 13 flavor attributes, and four were found to be
2*J4
TABLE 2
Compounds identified in brown sugars that contribute to flavor and aroma..
Acids Furans Phenolics

Acetic acid Furfural Benzole acid
Butyric acid Furfury1 alcohol 3,4-Dihydroxybenzoic acid
Propionic acid 5-(Hydroxymethyl) 4-Hydroxybenzoic acid
-2-furaldehyde 4-Hydroxycinnamic acid
Alcohols 2-Methyl-3-furanone 3-Hydroxy-phenylacetic acid
Ethanol Pentyl furan 2-Methylbenzoic acid
3-Hexen-l-ol Phenylacetic acid
Isoamyl alcohol Ketones Syringic acid
Diacetyl Vanillic acid
2,3-Pentanedione
Aldehydes Miscellaneous
Acetaldehyde Acetyl formoin
Hexanal Maltol
3-Methylbutanal
2-Methylpropanal
Pentanal
statistically significant; that is, sugars could be distinguished on the

basis of these flavors. The significant flavors were molasses, char,
licorice, and green. These agreed well with flavor descriptors developed by
Christianson and Anhaiser (ref. 1 ) . Char, licorice and green were negatively
correlated to flavor score, and molasses flavor was positively correlated.
Regression analysis of the chemical data and the flavor scores showed
that a-amino nitrogen, chloride, ash, color, and phenol correlated most
highly with flavor score. These correlations were all negative, indicating
that flavor quality deteriorated as levels of these variables increased.
Refinery brown sugars differed from turbinado sugars. Only the
turbinado sugars had significant "green" flavor. It was absent or extremely
low in the brown sugars. The GC profiles of the volatile constituents
showed that the turbinado sugars had small quantities of leaf alcohol,
absent in the refinery sugars, confirming the sensory results. Table 3
compares the volatile profiles of brown sugars and turbinado sugars.
The seven volatile compounds listed in Table 3 appear in all the
volatile profiles of brown sugar products and are quite common in foods.
The refinery sugars were higher in diacetyl, pentanedione, and furfural,
products of sugar caramelization, which contribute to brown, sweet, and
caramel flavors. The turbinado sugars were higher in 2-methylpropanal and
2k5
3-methybutana1, which are Strecker degradation products of amino acids.

These aldehydes have fruity and green-fruit aromas.
Table 4 compares the chemical data of light brown boiled sugars and
turbinado sugars. The sugars differed significantly in all areas. Acetic
acid is particularly important for perception of normal brown sugar flavor
(ref. 14). Its much lower levels in turbinados helps to explain some of the
differences in the flavor profiles.
TABLE 3
Comparison of the volatile profiles obtained by GC of sugars.
Integrator counts x 10~ s
Compound Brown Sugars Turbinado Sugars
Acetaldehyde 44.0 15.9

Ethanol 10.0 0
Diacetyl 28.4 14.8
2-Methylpropanal 26.9 49.1
Pentanedione 17.7 2.9
3-Methylbutanal 24.8 46.3
Furfural 9.8 1.0
Total volatiles 186 167
Total acids 1853 646
TABLE 4
Comparison of the chemical profiles of sugars.
Chemical variable Light brown sugars Turbinado sugars
Color (ICUMSA) 5732 1787

Phenolics (ppm) 1071 344
Iron (ppm) 6.39 2.71
Ash (Z) 1.29 0.21
Acidity (mEq) 6.98 1.03
Acetic acid (ppm) 477 47
Amino nitrogen (ppm) 279 950
Invert (Z) 2.18 0.076
Sucrose (Z) 93.7 99.5
2^6
Contribution of specific fractions to the flavor of brown sugars. An

experiment was carried out to determine the contribution of specific
fractions to the flavor of light brown sugars (ref. 28). These fractions
were: the ethyl acetate extract of acidified solutions of brown sugar; the
XAD-2 extract (ref. 10); and the nondialyzable material with molecular
weight above 12,000 daltons. The extracts were dried by evaporation to
remove all traces of solvent and added back to a weight of white sugar equal
to the amount from which the extract was obtained. These solutions were
presented to several panel members along with solutions of plain white sugar
and the brown sugar from which the extract had been obtained. They were
asked to compare the flavor of the extracts to the original flavor. Results
are summarized in Table 5.
The ethyl acetate extract contributed surprisingly strong acidic and
fruity flavors. In spite of the strong flavor, its taste was not like the
sugar from which it was extracted. The flavor of these compounds (acids and
esters) is mostly masked by the other components; contribution to flavor is
minor except possibly in turbinado sugars.
The XAD-2 extract was of particular interest because previous work had
shown that this resin has an affinity for phenolic and furanoid type
compounds which contribute to brown sugar flavor (ref. 10). Tasted alone,
XAD-2 extracts of brown sugar are almost intolerably bitter (ref. 10).
However, when added to sugar, they were perceived as very sweet, aromatic,
and caramel-like, with some burnt, bitter and cardboard notes. The
panelists felt that this was close to the total brown sugar flavor.
The dialyzate had a bland, hay-like flavor with some green notes. It
n
was also described as having a "creamy texture. These descriptors pick up
the polysaccharide component of the dialyzate, which may have some flavor
compounds and plant phenolics bound to it during process. The main
contribution to the flavor profile is its effect on viscosity, which adds
desirable mouthfeel. It is also a major contributor to the color because it
contains the high molecular weight brown pigments.
FLAVOR INTERACTIONS OF SUCROSE

Sucrose interacts in many interesting ways with common food
constituents and these should be kept in mind when considering the question
of sucrose in flavoring systems. Interactions include masking,
247
modification, enhancement, inhibition and synergism, among others. Some of

the interactions of sucrose are briefly outlined below.
Aroma absorption. Aroma substances are sorbed onto amorphous sugar at
a greater rate, by several orders of magnitude, than on crystalline sugar
(ref. 30). Sucrose also retains volatiles on drying to a degree higher than
other mono- and disaccharides, and the presence of an electrolyte increases
retention (ref. 31).
TABLE 5
Flavor characteristics of selected fractions extracted from light brown
sugars sold in North America.
Amount of extract obtained (2 of sugar)
Sugar XAD-2 extract Dialyzate Ethyl acetate extract
0.14 0.44 0.09

0.18 0.43 0.17
0.12 0.31 0.35
Extract Color Perceived flavors Contribution to flavor
XAD-2 Orange Brown sugar, caramel, Strong contribution;

molasses, butterscotch, Caramel flavors;
chemical, aromatic, Some color
woody, bitter, burnt,
cardboard, warm, maple
Dialyzate Brown Bland, sweet, creamy, Little flavor contr.

grassy, straw-like Color
Viscosity
Ethyl Pale Floral, fruity, citrus, Minor flavor contr.

acetate yellow gumball, acid, faint Flavors are masked
brown sugar
Sweetness suppression. Perceived sweetness decreases as solution

viscosity increases because diffusion of sugar to the taste buds is hindered
by the hydrocolloid matrix. Sweetness intensity can decrease as much as 25Z
in guar gum solutions (ref 32).
Sweetne s s enhancement. Addition of maltol to a sweetened product
enhances perception of sweet flavor so that up to 152 sucrose can be reduced
2^8
(ref. 33). Other caramel compounds such as furaneol and diacetyl also
appear to have sweetness-enhancing potential. This may explain the sweeter
taste of brown sugar compared to white sugar.
Suppression of metallic taste. The threshold for perception of ferrous
salts in sucrose solution at pH below 5.5 is only 1.5 ppm but increases to
26.2 ppm at pH > 7.5. Similar effects occur with ferric salts, with
thresholds being 3.9 and 248.7 respectively (ref. 34). The increase in
threshold is caused by complexation of sucrose with iron, masking the
metallic taste.
Effect on aromas in solution. Volatility of compounds can change
markedly in carbohydrate solutions, with some increasing (ethanol,
3-pentanone, butanol), some decreasing (ethyl formate, 4-heptanone), and
some unchanged (propanal) (ref. 35). (These data were for concentrated
invert and fructose solutions.)
CORRECTION OF FLAVOR PROBLEMS

Off-odors and flavors do occasionally cause problems with sugar
products. These are often caused by extraneous factors such as those
discussed above: resin breakdown, migration of volatiles from packaging
material, and excessive microbiological activity. These difficulties can
usually be overcome by correcting the source of the problem.
Sometimes, however, problems are caused by an imbalance in the flavors
that arise during process. The flavor of brown sugar products is a complex
blend of several significant factors, such as salty, bitter, licorice,
green-cane, and caramel flavors. When these elements are out of balance,
the result can be too much flavor, too little flavor or not the "right"
flavor. It is somewhat more difficult to correct these types of situations.
Some possible solutions include blending with other products to
increase or decrease flavor, as is done to adjust color. Changing heating
or boiling times and temperatures temporarily may be helpful. Sometimes pH
adjustments are required.
Another possible process control is to use specific absorbents for
specific problems. Some laboratory scale work along these lines showed
common absorbents could modify flavor in molasses (ref. 36). The removal of
phenolics and amino nitrogen and their effect on the flavor and odor of a
factory molasses with harsh and bitter flavor are summarized in Table 6.
249
Carbon and acid (cation) ion exchange resin improved the flavor
considerably. Anion exchange resin made the flavor worse.
Polyvinypyrrolidone (used in clarifying beer and wine) and Bentonite-NG
(used in refining oils) also had positive effects. XAD-2 improved the
flavor somewhat, but is not approved for food use. However, XAD-4 resin,
with similar properties (ref. 14), does have such approval.
TABLE 6
Result of absorbent treatment on the flavor of cane molasses.
Decrease in component (Z)
Absorbent Phenolics Amino N Flavor/odor characteristics
None — Very bitter, harsh cane molasses.

XAD-2 17.5 14.9 Less bitter, less acid, mellowed.
PVP 39.4 25.2 Most bitter and harsh flavors
removed; cane flavors remain;
fruity odors are stronger.
Alumina 15.8 14.7 Little change
Florisil 11.3 6.7 Little change
Bentonite-NG 80.8 20.7 Little flavor and odor remained
Bentonite-160 17.8 9.4 Little change
Anion resin Extremely harsh and bitter; acidic
958 (NaCl) 14.3 and fruity flavors removed.
Cation resin More acidic flavor; strong green-
118H (HC1) 0 54.2 cane odor; less flavor.
Cation resin More acidic flavor; much of bitter
118H (NaCl) 38.7 37.0 flavor removed.
Carbon 12.8 30.1 Mild, improved molasses flavor;
less bitter; caramel and some
green-cane flavor and odor.
FUTURE TRENDS
What does the future hold for the flavor of brown sugar products?
Several intriguing possibilities come to mind:
Specialty flavors might be tailor-made by exploiting the Maillard
reaction; that is, by adding a certain amino acid to develop a particular
flavor. Another possibility is to use specific cane varieties, such as
those high in amino acids, or harvested at such time as to exploit the
increased presence of natural flavor precursors. No research has been done
in the area of effect of cane variety on flavor, but there is a perception
250
in the industry that mature canes develop more pronounced and desirable
molasses flavor than immature canes.
More research needs to be done in the area of altering the process to
strengthen or weaken the flavor, as in producing a high or low-licorice
flavor molasses, and in the use of resins and absorbents to correct flavor
problems.
Bacteriological action and fermentation could be used to produce
unusual flavors. Bio-technology research is underway to produce flavors
by tissue culture and controlled fermentation.
Addition of flavor enhancers such as maltol, ethyl maltol or diacetyl
may be considered, although both the economic and labelling requirements may
be unfavorable. Other added flavors such as coffee or chocolate may be
popular if properly positioned in the market, with adequately explained
applications.
Some specialty products do already exist with flavor as their selling
point. Several companies in North America sell dry-flow molasses, some with
soy bran and wheat flour to act as drying and flavoring agents. A European
sugar company makes caramel and coffee-flavored brown sugars. In the
Orient, cane juice is pasteurized and packaged or canned for sale as a
beverage.
REFERENCES
1 G. Christianson and L. Anhaiser, The importance of flavor in brown sugar
for consumer goods application, in: Proc. Sugar Industry Technol., 39
(1980) 177-185.
2 Cane Sugar Handbook, 10th Edition, (ed) G.P. Meade and J.C.P. Chen,
John Wiley & Sons,Inc., New York, 1977, pp. 500, 699-700.
3 J. Hassett, Tantalise taste buds with real brown sugar, Bakers Review,
(March, 1979) 21 and 23.
4 S. Takei and T. Imaki, Bull. Inst. Phys. Chem. Research (Tokyo), 15
(1936) 124.
5 E. Abe, Y. Naktani, T. Yamanishi and S. Muraki, Studies on the "sugary
flavor" of raw cane sugar, Proc. Japan Acad., 54, Ser. B (1978) 542-547.
6 M. Yokota and I.S. Fagerson, The major volatile components of cane
molasses, J. Food Sci., 36 (1971) 1091-1094.
7 T. Shirasaki, H. Ito and M. Kamoda, Studies on the odor of molasses,
Part II, Volatile carbonyl compounds in refinery molasses, Proc. Res.
Soc. Japan Sugar Refineries' Technologists, 16 (1965) 44-49.
8 W.W. Binkley and M.L. Wolfrom, Chromatography of Cuban blackstrap
molasses on clay; some constituents of an odor and pigment fraction,
J. Amer. Chem. S o c , 70 (1948) 290-292.
251
9 C A . Shafer, The use of instrumental and sensory analyses to optimize

the determination of important aroma fractions in cane molasses from
different countries, Dissertation Abstracts International, B 45 (2) 504,
(1984) 144 pp. Food Sci. Technol. Abstr., 17(9) (1985) L-15.
10 M.A. Godshall and E.J. Roberts, Phenolics in sugar products: Their role
in flavor and color production, Proc. 1982 Sugar Processing Res. Conf.,
1983, pp. 47-69.
11 M.G. Legendre, G.S. Fisher, W.H. Schuller, H.P. Dupuy and E.T. Rayner,
Novel technique for the analysis of volatiles in aqueous and nonaquous
systems, J. Amer. Oil Chem. S o c , 56 (1979) 552-555.
12 M.A. Godshall, E.J. Roberts and M.G. Legendre, Identification of
volatile constituents responsible for characteristic molasses aroma by
unconventional gas chromatography, J. Agric. Food Chem., 28 (1980)
856-858.
13 M.A. Godshall, M.G. Legendre and E.J. Roberts, The identification of
volatile constituents in sugarcane and cane sugar products, Proc.
1978 Technical Session on Cane Sugar Refining Res., 1979, pp. 46-67.
14 M.A. Godshall, C.H. Vinnett and V. Chew, Sensory analysis of brown
sugars and its correlation with chemical measurements, Proce 1984 Sugar
Processing Res. Conf., 1986, pp. 22-52.
15 L. Färber and F.G. Carpenter, Plant pigments as colorants in cane sugar,
Proc. 1972 Technical Session Cane Sugar Refining Res., 1975, pp. 23-30
16 P. Smith and N.H. Paton, Sugarcane flavonoids, Sugar Technol. Rev.
12 (1985) 117-142.
17 J.A. Maga, Simple phenols and phenolic compounds in food flavor,
CRC Critical Rev. Rood Sci. Nutrition, 10(4) (1978) 323-372.
18 J.A. Maga and K. Lorenz, Taste threshold values for phenolic acids which
can influence flavor properties of certain flours, grains, and oilseeds,
Cereal Sci. Today, 18 (1973) 326-328, 350.
19 Y. Tokitomo, A. Kobayashi and T. Yamanishi, Aroma components of fresh
sugar cane juice, Agric. Biol, Chem, 48 (1984) 2869-2870.
20 J.E. Hodge, in: Symposium on Foods: The Chemistry and Physiology of
Flavors, H.W.Schultz, E.D. Day and L.M. Libbey (Eds.), AVI, Westport,
Conn, 1965, pp. 465-491.
21 M.J. Lane and H.E. Nursten, The variety of odors produced in Maillard
model systems and how they are influenced by reaction conditions, in:
The Maillard Reaction in Foods and Nutrition, ACS Symp. Ser. 215, G.R.
Waller and M.S. Feather (Eds.), ACS, Washington, D.C., 1983, pp.
141-158.
22 A.F. Mabrouk, Flavor of browning reaction products, in: Food Taste
Chemistry, ACS Symp. Ser. 115, J.C. Boudreau (Ed.), ACS, Washington,
D.C., 1979, pp. 205-245.
23 W.L. Owen, The Microbiology of Sugars, Syrups and Molasses, Barr-Owen
Research Enterprises, Minneapolis, MN, 1949, 275 pp.
24 M.A. Godshall and A.J. DeLucca,II, Acetic acid, a major volatile consti
tuent of brown sugar: Its origin and measurement, J. Agric. Food Chem.,
32 (1984) 390-393.
25 K.W. Healey and J. Carnevale, Determination of volatile fatty acids in
molasses by gas-liquid chromatography of their benzyl esters, J. Agric.
Food Chem., 32 (1984) 1363-1366.
26 M.A. Godshall and E.J. Roberts, Isolation and identification of consti
tuents contributing to odor and flavor in syrups and brown sugars, Proc.
1980 Technical Session Cane Sugar Refining Res., 1981, pp. 26-48.
27 L. Anhaiser, personal communication (1980).
252
28 M.A. Godshall, unpublished results (1985).

29 M.A. Godshall, Flavor components in sugar products, Sugar Indus.
Technol., 40 (1981) 178-187.
30 E.A. Niediek and L. Babernics, Aroma sorption properties of amorphous
sucrose and lactose, Gordian, 79 (1979) 35-36, 42-44.
31 J. Flink and M. Karel, Retention of oganic volatiles in freeze-dried
solutions of carbohydrates, J. Agric. Food Chem., 18 (1970) 295-297.
32 B. Launay and E. Pasquet, Sucrose solutions with and without guar gum:
rheological properties and relative sweetness intensity, Progr. Food
Nutr. Sci, 6 (1982) 247-258.
33 J.C. Johnson, Specialized sugars for the food industry, Noyes Data
Corp., Park Ridge, New Jersey, 1976, pp. 297-298.
34 H. Cross, T. Pepper, M.W. Kearsley and G.G. Birch, Mineral complexing
properties of food carbohydrates, Starke, 37 (1985) S. 132-136.
35 A.G. Wientjes, The influence of sugar concentrations on the vapor
pressue of food odor volatiles in aqueous solutions, J. Food Sci.,
33 (1968) 1-2.
36 M.A. Godshall, Flavors from beet and cane products, 1986 Conf. Sugar
Processing Res., Savannah, Georgia, October 19-21, 1986 (in press).
253
Chapter 16
EFFECTS OF IMPURITIES ON DEGRADATION OF SUCROSE UNDER PROCESSING CONDITIONS
G. N . RICHARDS
INTRODUCTION
During the past seven years, in a series of papers from the James Cook
University of North Queensland, we have published studies leading to a detailed
understanding of the chemical mechanisms of thermal and of alkaline degradation
of sucrose and have exploited the understanding of the thermal degradations to
develop new synthetic methods of general application, especially to
fructofuranosides (ref. 1 ) . Host of these studies however, have involved
kinetics and syntheses in aprotic solvents such as methyl sulfoxide and it may
not immediately be obvious how they apply to the real world of production and use
of sucrose. Thermal degradation of sucrose may conceivably occur in harvesting,
milling and refining and also in many subsequent food process operations. The
degradations may be divided into two types, firstly amorphous sucrose with little
or no water present (crystalline sucrose is very much more stable), and secondly,
concentrated aqueous solutions of sucrose.
DEGRADATIONS OF AMORPHOUS SUCROSE

To consider first the case of amorphous sucrose, with little or no water
content, this may occur in the burning of cane at harvest, where dried down juice
may produce amorphous sucrose, which as a result of fire temperatures is likely
to yield kestoses (see below). Molasses is often subjected to long periods at
relatively high temperatures and this situation will result in similar chemistry
to that described below. Finally, in many food process operations, sucrose is
initially dissolved in water, dried down during the process (e.g. baking), often
to produce amorphous sucrose, which may be exposed to the level of thermal
treatments and chemistry described below. Some of the results on such systems
have been described previously (ref. 2 ) . In the case of concentrated aqueous
solutions of sucrose, the obvious dominant reactions are hydrolysis and
subsequent degradation of the resultant fructose and glucose, especially the
former. This may occur in sugar boiling, during milling and refining, and also
in many food process operations. In such circumstances, impurities such as
reducing sugars and inorganics are normally present and the results described
25^
below show that both types of impurity have dramatic effects on sucrose
degradation.
CH2OH
HO
fc> 1
OH
H
T>-J > :
K^>•
[ ^ CHjOH
HO
I /°\ I
''
+ ^
<3/
[ ^
I
CH2OH I
CH2OH ~HS^
λ-\
HO
HOCH,
>^ 3
COH \ + U^ no sL » non-specific
lNL/1. J/ f degradation
HO
>I
OH 0 HO X^H"
2 ^v HOÇHj 0
+
H i \{ H O V ^ O H , C Η,ΟΗ
sucrose-OH
\ HO
CH2OH
Ilo HOÇHj c
\
s.
f-O-sucrose
t<L> H>
mr
y "C-CH.OH
2
HO ^ j
OH
OH
T
HO
I
Fig. 1. Mechanism of thermal degradation of sucrose.
The predominant mechanisms of thermal degradation of sucrose (1) are shown

in Figure 1. It has been unequivocally shown that in methyl sulphoxide solution
a-D-glucopyranose is an initial product which subsequently anomerizes (ref. 3 ) .
The same initial reaction occurs very rapidly in melts of sucrose at 190°C (ref.
4). The first step in the degradation is very sensitive to catalysis by traces
of acid and there is no doubt that this catalysis occurs by protonation of the
glycosidic oxygen in (1) (ref. 5). The fructose carbocation (2) which is
liberated in the first scission reaction, is lost rapidly by several reaction
channels. It may cyclize to form the anhydride (3), undergo nonspecific
degradation to a wide range of products such as hydroxymethylfurfural, add
255
hydroxyl ion to produce fructose, or it may add to one of the hydroxyl oxygens of
another sucrose molecule to produce a trisaccharide (the kestoses).
In preliminary experiments it was observed that powdered pure sucrose
crystals would survive for many hours at (say) 150°C without any detectable
degradation beyond slight darkening and evidently this stability is associated
with relative absence of molecular mobility in the crystalline lattice. In
practice, however, sucrose will not always be crystalline when subjected to heat.
An experimental procedure was therefore designed (ref. 2) to produce amorphous
sucrose (a melt) at temperatures well below its crystal melting point, and under
these conditions the sucrose degrades very much more rapidly than the crystalline
material. The loss of sucrose in the melts is shown in Figure 2. The rates are
too rapid for meaningful analysis at the higher temperatures, but at temperatures
of 150°C and lower the most dramatic conclusion from these curves is that there
is an unequivocal lag phase in the degradation.
Figure 2. Degradation of sucrose melts at various temperatures.

256
The products of the degradation of pure sucrose at 135°C are shown in Figure
3. Glucose is formed in relatively high yield, while fructose is always produced
in lesser amount and is itself more subject to further thermal degradation than
glucose, thus producing the maximum in the fructose curve. Trisaccharides
(kestoses) are formed in smaller yield and traces of anhydrofructose (ref. 3) and
of disaccharides other than sucrose are also observed but are not plotted on
Figure 3.
Figure 3. Degradation of sucrose melt and formation of major products at 135°C:

S = sucrose, G = glucose, F β fructose and K = kestoses.
The effects of impurities on the rate of loss of sucrose were studied at

120°C and are shown in Figure 4. Of the impurities used, only sodium carbonate
had the effect of reducing the rate of thermal decomposition of sucrose. The
presence of 10Z glucose in the sucrose melt shortened the lag phase and
257
accelerated the degradation. With 5Z glucose plus 5Z fructose, both effects were
increased and 10Z fructose (not shown in Figure 4) was even more potent in
accelerating the degradation. Addition of a neutral salt (sodium chloride) to
the sucrose melt was also remarkably effective, at less than IZ concentration, in
reducing the lag phase and accelerating the decomposition.
100
" * ^N/V> \ \ c
A""""-
80 \ ^D
f.
Is
\E
s
1 40
20
» » » '
Figure 4. Effect of impurities on the rate of degradation of sucrose melts at

120°C: A - pure sucrose, B - sodium carbonate (0.04 mole/mole sucrose), C «
glucose (10Z), D - glucose (5Z) + fructose (5Z) and E - sodium chloride (0.05
mole/mole sucrose).
The probable explanation of the lag phase is that the initial degradation of
pure sucrose is extremely slow, in fact too slow to be detected by the methods
used in this study. However, traces of initial degradation products formed
during this phase may themselves be subject to more rapid degradation reactions
and some products of such secondary reations may be acidic. Such products (e.g.
258
formic and levulinic acids) could be extremely small in amount, yet could result
in protonation of sucrose and hence in an increasing rate of degradation. As the
degradation proceeds the carbocation (2) would be increasingly produced and would
partly undergo non-specific degradation reactions to produce more acid catalysts.
The same non-specific degradations would also include reactions which generate
water which would provide the hydrogen and hydroxyl ions required for further
reaction as shown in Figure 1.
"When reducing sugars are present in the sucrose melt they undergo thermal
degradation much more rapidly than sucrose. Because fructose degrades more
rapidly than glucose it is much more potent as an impurity in reducing the lag
phase in the sucrose degradation. In confirmation of this, it was found that a
fructose melt, after heating at 150°C for 30 min., when dissolved in water showed
pH 4.5.
The above hypotheses all require the initial slow formation of acidic
degradation products which increasingly catalyze the sucrose degradation and
result in accelerating decomposition. The stabilization of sucrose by the
presence of a small amount of sodium carbonate (Figure 4) is therefore
interpreted as due to neutralization of the traces of the secondary acidic
products. The dramatic influence of very low levels of sodium chloride in
degradation of sucrose is more difficult to explain. However, all of the
mechanisms involved in Figure 1 are heterolytic and it is conceivable that the
effects of a small amount of sodium chloride may operate through increase in the
dielectric constant of the sucrose melt.
DEGRADATION IN AQUEOUS SOLUTIONS

The hydrolysis of sucrose by water (i.e. inversion) has been extensively
studied, especially with added acid catalysts in dilute solution. However, in
sucrose processing, sucrose is most frequently exposed to heat in very
concentrated aqueous solution at_ about neutral pH. The evaporation of solutions
to crystallization during milling and refining is an obvious example where even a
low level of hydrolysis can have important economic impact. In food processes
also, we frequently find examples of exposure of concentrated aqueous solutions
of sucrose to high temperatures for significant periods. In all of these
circumstances the sucrose solution generally contains impurities such as reducing
sugars and inorganic salts and it is this type of system which we have now
studied.
259
w
■ vvlwl vwv I
\j ' ο ι ι ι ν ι ι υ ο α υ υ ι ι α » IVI*?
in water, 100°C
100iL——
I %
o\° \
CD I · \ \
c
Έ I \ \
\
\
\
'(5 \
E glucose\ \
2 I ·
^ pure\
Φ h80 \
(0 \ sucroseV
o fructose,
v.
\ \
O [-70 \ N \
3 X
\1 2 \
1
_1J I A 1
Time, hours
Fig. 5. Degradation of sucrose in water at 100°C; effect of added glucose and

fructose.
Figure 5 shows the rate of loss of sucrose at 100°C in a solution containing

20 g of sucrose and 7.5 g of rigorously deionized water. The rate is measured by
HPLC against an internal standard (ethanol) added to a sample of the solution
after heating for the required time in a sealed glass tube. As with the thermal
degradation of amorphous sucrose, a lag phase is observed in sucrose loss and we
postulate the same type of explanation. That is, that the initial rate of
hydrolysis of sucrose by water is extremely slow, but finite. The products are
fructose and glucose, and the former especially degrades relatively rapidly under
these conditions to produce a mixture of products which includes some acids, such
as levulinic and formic (of course hydroxymethylfurfural is the major degradation
product). These minor acids form the initial hydrolysis products, then induce
the autocatalytic form of the curve shown in Figure 5. To verify the above
hypothesis, the addition of 10Z glucose (based on sucrose) reduces the lag phase,
while the same amount of fructose almost removes the lag (Figure 5) and in the
same experiment, fructose was observed to be lost much more rapidly than glucose.
260
SUCROSE :WATER:SALT/100°C
20g 7.5g.
(moles) 1 7 · 0.05
100
Time, hours
Fig. 6. Degradation of sucrose in water at 100°C; effect of sodium acetate and

sodium chloride.
The effect of a small amount of a weak base on sucrose hydrolysis at 100°C

is shown in Figure 6, where 0.05 mole of sodium carbonate per mole of sucrose is
seen to confer complete stability (within the accuracy of the experiment) on the
sucrose for more than four hours. This is interpreted as due to neutralization
of secondary acidic degradation products which would form from any traces of
primary hydrolysis products.
The effect of sodium chloride on the sucrose hydrolysis is also shown in
Figure 6. This is observed as a dramatic shortening of the lag phase in sucrose
261
loss (i.e. an acceleration of the hydrolysis). This effect is produced by a

ratio of only one mole of sodium chloride to 20 moles of sucrose. There are two
possible types of explanation for this effect. Either the sodium chloride
accelerates the initial slow hydrolysis of sucrose and thus increases the initial
rate of formation of primary products and hence the rate of formation of acidic
secondary degradation products; or alternatively, the sodium chloride accelerates
the rate of degradation of primary products (glucose and fructose) to acids. It
is possible that both effects operate, and on the present evidence we are not
able to reach a definite conclusion, but in general the first interpretation is
favored at this stage.
Sucrose:Water:Salt(chlorides),100°C
1 s 7 '· 0 . 0 5 moles
100k - ^
Time, hours
Fig. 7. Degradation of sucrose in water at 100°C; effect of salts.
The influence of other salts is shown in Figure 7. Cations have been chosen
which are common and often abundant in sucrose processing; the anion is chloride
throughout and the mole ratio of salt to sucrose has been maintained at 1:20. It
is evident that calcium ions are more effective than sodium and that magnesium
ions are much more effective in accelerating the sucrose hydrolysis. The other
alkali metal chlorides were also studied, but within the accuracy of our
262
experiments, they produced the same effect as sodium. Preliminary experiments

also indicated that use of other halides produced only small effects.
G ^F
G + F+
Fig. 8. Hydrolysis of sucrose.
The explanation of the influence of magnesium ions on sucrose hydrolysis

must be speculative at this stage. The first step in hydrolysis of sucrose is
most probably the scission of the oxonium ion shown in Figure 8, where G is
glucose and F is fructose (other oxonium ions will form at alcohol hydroxyl
groups). Any effect which increases the concentration of oxonium ion will
increase the rate of hydrolysis and of course this is the basis of acid catalysis
of sucrose inversion.
One of the most familiar differences or trends between sodium, calcium and
magnesium salts is the increasing tendency of the cation to form stable hydrates
and this provides a possible explanation for the effects shown in Figure 7. In
this system, it should be noted that the number of potential hydrogen-bonding
sites of sucrose molecules exceeds the number of water molecules and that there
will be competition for water molecules between sucrose and the cations.
The magnesium ions are likely to be especially effective in this competition
and the water molecules which are hydrated to magnesium will have oxygen-hydrogen
bonds which are more polarized (i.e. more acidic) than free water molecules.
Thus, as shown in Figure 9, there will be an increased tendency for transfer of
263
hydrogen ion from a hydrated water molecule to any other electron donor such as
the glycosidic oxygen of sucrose. This will have the effect of increasing the
concentration of the oxonium ion shown in Figure 8 and hence increasing the rate
of hydrolysis.
The effect is most pronounced in concentrated aqueous solutions where there
is competition for water molecules between sucrose and magnesium ions. Thus,
Figure 10 shows that magnesium chloride is much less potent in increasing sucrose
hydrolysis when the solution contains 50 moles of water per sucrose molecule than
with 7 moles of water per sucrose, while the sucrose:magnesium ratio is kept
constant. The same figure shows little or no effect of variation of
sucrose:water ratio in pure water.
Mg ++ (OH 2 ) x
Fig. 9. Protonation of sucrose by hydrated magnesium ion.
The above experiments are relevant to any sucrose process in which sucrose
is heated with water, especially in the presence of impurities such as reducing
sugars and salts. They indicate a need for particular concern when magnesium
ions are present in significant amount, as may occur especially in sugarbeet
processing. It should be noted however, that in "real life" the anions will not
necessarily be halides. In juices especially, carboxylic acid anions such as
acetate, lactate, citrate, etc. are present and these (and anions of any other
weak acid) may exert alkaline buffering effects which will effectively stabilize
the sucrose towards hydrolysis in the same way as the sodium carbonate shown in
Figure 6.
264
I Sucrose ·· water : MgCL / 1 0 0

from 1=7=0 to 1=50 = C)molar
100
Έ
σ>
e
'<ô
VX ^ \
E
\J=50=0.05 \
Φ
α>
Léo \
co
o
Vl=7=0.05 \ \
k.
o
3
(/)
ko l \ 2 \
l i i i \—
Time, hours
Fig. 10. Influence of water content on catalysis of sucrose hydrolysis by

magnesium chloride.
ACKNOWLEDGEMENTS
Part of the work described above was carried out at James Cook University of
North Queensland with the experimental assistance of G.B. Hawes and J.G. Kelly
and financial support from the Sugar Research Institute, Mackay. The remainder
of the experimental results, especially those involving aqueous solution, owe
much to the experimental assistance of T.L. Lowary.
REFERENCES
1 W. Moody and G. N. Richards, Formation and equilibration of D-fructosides
and 2-thio-D-fructosides in acidified dimethyl sulfoxide: synthetic and
mechanistic aspects. Carbohydr. Res., 124 (1983) 112-113 and earlier
references therein.
2 G. N. Richards, Initial steps in thermal degradation of sucrose. Int.
Sugar J., 88 (1986) 145-148.
3 L. Poncini and G. N. Richards, Thermolysis of sucrose in dimethyl
sulfoxide solution. Carbohydr. Res., 87, (1980) 209-217.
4 G. N. Richards and F. Shafizadeh, Mechanism of thermal degradation of
sucrose. A preliminary study. Aust. J. Chem. 31 (1978) 1825-1832.
5 W. Moody and G. N. Richards, Effect of traces of acids on reaction of
sucrose in dimethyl sulfoxide. Carbohydr. Res., 108 (1982) 13-22.
265
Chapter 17
CLARIFICATION AND DECOLORIZATION PROCESSES
R. R. TROTT
INTRODUCTION
The quality of refined or white sugar of any kind tends to be judged by

most consumers, whether industrial or domestic, first and foremost on its
whiteness, or absence of color. Insoluble impurities in the sugar are very
undesirable; even a small proportion tends to give the crystals a dull
appearance.
In the refinery, the removal of color and turbidity takes place in a
number of successive steps in the process. The overall reduction of color
between the raw cane sugar and the refined crystalline product is about 99Z;
typically from about 3,000 to 30 color units. The removal of turbidity is
even more efficient.
The first purification process in a refinery is the removal of the
surface layer of molasses from the relatively pure sucrose crystal. This
surface layer contains a high proportion of the total impurities in the raw
sugar, and the washing, or affination, process typically removes 65-70Z of
the color, ash and non-sucrose sugars present in the original raw sugar.
The washed crystals are dissolved in water to yield a syrup, often referred
to as melt liquor, which goes forward for further purification.
The final purification process in the refinery is the separation of the
sucrose in a very pure form by crystallization. Although the
crystallization process is very efficient, the melt liquor must be both
clarified and decolorized to produce a syrup from which good quality refined
sugar can be crystallized. This paper reviews the traditional processes
which have been, and still are, successfully used for melt liquor
clarification and decolorization. New processes which have been, or are
likely to be, useful in making the clarification or decolorization process
more efficient or less expensive are also discussed.
The sole aim of sugar factories in many cane sugar producing countries
in the past was to produce a brown raw sugar for export to off-shore
refineries in developed countries. In such cases, simple juice
266
clarification using lime was the only treatment needed before

crystallization. However, in recent years, it has been necessary for sugar
factories to produce higher quality raw sugar for export and good quality
refined or semi-refined white sugar for direct consumption (ref. 1 ) .
Consequently, clarification and decolorization processes used in sugar
factories have had to be improved. This paper reviews the traditional
methods of purification employed and discusses some of the changes which
have taken place in these processes to enable sugar factories to produce
good quality white sugar directly from cane juice.
TRADITIONAL REFINING PROCESSES

Clarification
Simple clarification of the melt liquor to remove finely divided
particles of insoluble and semi-colloidal matter and to prepare the liquor
for the subsequent decolorization process can be effected by mechanical
filtration with filter aids. However, it is now almost universal for
refineries to use either a carbonatation or a phosphatation process to
clarify the melt liquor. These chemical processes, as well as removing the
insoluble particles, also eliminate significant amounts of color and other
non-sucrose impurities which would otherwise have to be removed at greater
expense in the subsequent decolorization or crystallization processes.
The traditional carbonatation and phosphatation processes and recent
improvements to these processes using new chemicals and techniques are
described and discussed below.
Phosphatation. Clarification of melter liquor by phosphatation
originated in refineries in the United States. Basically, the process
consists of precipitating insoluble calcium phosphates of variable
composition in hot melt liquor by first adding phosphoric acid, or an acid
phosphate, followed by calcium hydroxide, either as a slurry with water or
dissolved in sugar syrup as lime sucrate, to give a final pH of 7.2 - 7.4.
Originally, a batch phosphatation process was used in which the
phosphates were precipitated in the liquor as outlined above, the treated
liquor then heated to a temperature of 85 - 90°C and allowed to stand until
the insoluble phosphates formed a floe. Diatomaceous earth filter aid was
mixed in as a body aid and all the liquor filtered.
267
This batch process was superseded by a simple continuous flotation

process, which achieved a similar degree of decolorization and purification,
but avoided the difficult removal of the floe by filtering all the liquor
and therefore required less filtration equipment. In the continuous process
the melter liquor is treated with phosphoric acid or acid phosphates at a
higher level than in the batch process, namely, 200-500 parts per million on
brix solids as P 2 O s , depending on the color and general quality of the
melter liquor being processed. Lime slurry or lime sucrate is added to give
a pH of 7.1 - 7.4. The treated liquor is aerated and then flows to a
continuous flotation clarifier where, during a retention period of 20-50
minutes, the liquor is slowly reheated, and the insoluble phosphates are
carried to the top of the liquor by the expanding air. The scum so formed
contains inorganic impurities, suspended solids and color adsorbed and
trapped in the calcium phosphate floe; it is gently raked from the surface
and the clarified liquor is removed from the bottom of the clarifier.
Many different designs of flotation clarifier have been developed since
the early 1920*s, but all are essentially rectangular tanks based on the
original principle of the Williamson clarifier. The methods of aeration,
heating and scum removal vary with the different designs. Various clarifier
designs and the main process parameters for this type of continuous
flotation clarification are discussed by Baikow (ref. 2) and Saranin (ref.
3).
The clarified liquor from such flotation clarifiers contain traces of
insoluble suspended matter, and this is normally removed by polish
filtration to protect the following decolorizing medium such as bone char,
granular carbon or ion-exchange resin.
The volume of scum is generally between 5Z and 102 of the volume of
liquor treated, and the sugar liquor in the voids between the floe particles
is quite considerable, particularly if the scum layer is not well
compressed. The scum is desweetened by diluting with water to about 202
solids content, adding filter aid and filtering using a precoated filter.
Carbonatation. Clarification of melter liquor by carbonatation has
been used in Britain since cane sugar refining was established there over
100 years ago and is also the normal clarification process used in many
other refineries in Europe, Australia and South Africa. It has more
recently been adopted by some refiners in North America.
268
Basically, the process consists of precipitating calcium carbonate in

the melter liquor by adding a slurry of calcium hydroxide and then bubbling
carbon dioxide through it under controlled conditions of alkalinity and
temperature. The object of carbonatation is to precipitate the insoluble
calcium salts of certain ionic impurities, e.g. sulphates, and to remove as
much as possible of the insoluble and semi-colloidal impurities, such as
starch, gums, protein and some colored substances. These impurities are
both absorbed by, and enmeshed in, the conglomerated particles of the
calcium carbonate precipitated by the reaction of the carbon dioxide and
calcium hydroxide. It has been shown experimentally (ref. 4) that
comparatively little of the lime is directly involved in the purification
process as such, which is not particularly sensitive to temperature, pH,
residence time, and so on. Most of the lime is needed to precipitate
sufficient carbonate to act as an efficient filter aid; both the quantity of
this precipitate and its effectiveness as a filter aid are dependent upon
the reaction conditions. If the impurities are to be removed efficiently
the reaction conditions must ensure that the physical nature of the
precipitate formed is such that it can be easily removed by filtration.
The quantity of lime needed and added as a slurry containing 12-15Z w/w
CaO, varies with the characteristics of the particular melter liquor being
processed, but for normal conditions the optimum dose rate is generally in
the range of 0.6 - 1.0Z CaO on melter liquor solids.
The carbon dioxide is obtained from washed boiler flue gas, typically
containing 8.5 - 10Z carbon dioxide, compressed in rotary compressors and
pumped to the saturation tanks. The introduction of the CO* is usually a
two-stage operation with the major part of the gassing taking place in the
first saturator at a temperature of 80 - 85°C. The amount of lime added and
the pH of the fully carbonated liquor must be carefully controlled if the
main requirements of good filterability and color removal, together with a
useful drop in ash content, are to be met.
Separation of the clarified liquor from the calcium carbonate
precipitate is by filtration on vertical leaf filters, preferably of the
rotating leaf type (ref. 5). In recent years it has been found more
economical to use synthetic filter cloth, which can be acid treated and
reused, rather than cotton cloths which can be used only for a limited
period. Because of the quantity and nature of the precipitated calcium
carbonate, no external filter aid is normally required for precoating the
filter, or for body aid during the filtration cycle.
At the end of the filter cycle, the residual cake is sluiced from the
plates with hot water and run from the filter body as a sludge. These
sludges are refiltered, usually on plate and frame presses, and residual
sugar washed from the cake before disposal.
Decolorization
The sugar liquor clarified either by carbonatation or phosphatation
still contains colored components and a large proportion of these (80-90Z)
must be removed prior to the crystallization process in order to produce
good quality refined sugar. The decolorization process is therefore very
important and since it also represents a major investment for the refiner it
needs to be selected with care.
There are three traditional systems of decolorization, all of which use
carbonaceous adsorbents, namely, bone char, granular activated carbon and
powdered vegetable carbon. These processes are still widely used, but in
recent years ion exchange resins have become increasingly popular as
decolorizing agents.
Bone char. Bone char, made by carbonizing specially selected sun-dried
cattle bones, is an adsorbent comprising about 10Z carbon and 902 basic
calcium phosphate (hydroxyapatite). The carbon is distributed over about
502 of the hydroxyapatite surface (ref. 6).
As well as being an excellent adsorbent for ionic color components and
color precursors found in sugar liquor, bone char also removes certain
inorganic ions, so that treatment of sugar liquor with bone char generally
results in the removal of about 85Z of the color and about 202 of the ash.
Bone char is also intrinsically alkaline and has a natural buffering action
on the liquor treated; this has the desirable effect of maintaining the pH
above 7. Bone char is also structurally quite hard and hence requires
little new char make up due to attrition (generally less than 0.12 on total
sugar solids treated).
The sugar syrup to be treated is passed through the bone char
contained in cisterns or cells, at an average rate which will give about 4-5
hours contact time.
270
There is no doubt that bone char is unrivalled as a single adsorbent,

but it is costly to use (ref. 7). The regeneration rate is high, being
10-15Z on sugar solids treated, and this leads to high energy costs for
regeneration as well as high maintenance costs on extensive conveying,
elevating, de-watering and kilning equipment. All this associated plant and
the treatment vessels themselves are very costly and, as a result, very few
new bone char decolorization plants are being installed. As existing bone
char plants need expansion or renovation, they tend to be replaced by
alternative decolorization systems.
Granular activated carbon. Activated carbons are prepared from a
variety of raw materials of vegetable origin, such as wood and paper mill
products, or of mineral origin, such as coal. They are activated by various
methods, including thermally and chemically with acids, various salts, etc.
The method of activation determines the effectiveness of the carbon as a
decolorizer of sugar syrups. Different types of carbon also have varying
degrees of affinity for the different color constituents in sugar syrup.
The granular activated carbons usually used in the sugar industry are
prepared from coal. In common with all activated carbons, they have a high
carbon content (about 90Z), but the residual inorganic constituents do not
give this material any of the buffering or de-ashing properties so useful in
bone char. Consequently, a small addition of magnesite is made to the
granular carbon during service to prevent a drop in pH of the treated
liquor. Granular carbon decolorization is sometimes followed by ion-exchange
deashing.
The granular carbon decolorizing system is operated in the same way as
the bone char system, but the service time is very much longer and the
regeneration rate is only about 0.5 - 0.8Z on sugar solids treated.
Consequently, the physical size of both the operating and the regeneration
plant is much less than for a bone char plant of a similar capacity and the
lower capital and operating costs can make this decolorizing system
economically more attractive.
There is a decolorization system, patented under the name of CANESORB,
in which a granular carbon is added to, and acts in conjunction with, bone
char. The adsorbent is essentially a mixture of granular carbon and bone
char in a ratio of about 1 to 2 by volume; its properties are said to be
such as to give it the decolorizing characteristics of very good quality
271
granular carbon. The mixture is regenerated at bone char regeneration

temperatures of 500-600°C, but in practice a reduced burn rate of about 70Z
of that using bone char alone is claimed. The most likely application for
this material would appear to be as a means of increasing the effective
capacity of an existing bone char installation by blending in a proportion
of the more highly adsorbing CANESORB. However, success of this application
depends on the practicability of maintaining the condition of mixed
adsorbents when regenerated under compromise conditions.
Powdered activated (vegetable) carbon. Since coloring matter is
adsorbed onto the surface of the carbon, the finer the carbon particles are
the more coloring matter they can adsorb per unit weight. Powdered
activated carbon, generally prepared from raw materials of vegetable origin,
can only be used on a throw-away basis and hence cannot be considered as an
equivalent alternative to bone char or granular carbon. Nevertheless, it
can, at economic levels of use, provide the means for a very useful color
reduction in certain applications.
For example, it can be a very suitable decolorizing agent for a small
refinery, perhaps attached to a raw sugar factory, where higher variable
operating costs are preferable to a high fixed capital investment. Quite
small quantities of carbon can be used, 0.1 - 0.5Z on the weight of sugar
solids treated is usually sufficient, and no costly regeneration facilities
are required. Furthermore, the process is flexible in that the amount of
carbon used can be varied depending upon the quality of the sugar liquor
being processed. At economic levels of use it will reduce liquor color by
about 20Z but, as with granular carbon treatment, no ash is removed.
Powdered carbon is also used quite successfully in batch installations
operated by commercial users of white sugar, such as soft drink
manufacturers, to decolorize syrup made from a less expensive non-refined
sugar to produce a liquid sugar of the required color and clarity for
immediate use.
When using powdered carbon, batch operation is usually more
satisfactory than continuous operation. A slurry of the carbon is added to
the hot syrup (82°C) and the mixture is kept hot for about 20 minutes,
during which time it is slowly stirred to ensure good contact between the
carbon and the sugar liquor. Filter aid is added and the liquor is filtered
through a precoated filter.
272
Powdered carbons tend to be acidic and pH adjustment with sodium

carbonate, either before or after treatment, may be necessary to avoid
inversion.
DEVELOPMENTS IN REFINERY PROCESSES

Clarification
Carbonatation and phosphatation remain the only basic clarification
processes used in sugar refineries. The efficiency of the traditional
phosphatation process has been improved quite significantly over recent
years, but the carbonatation process has remained practically unchanged for
many decades.
Phosphatation. The improvements in the phosphatation process have
included the introduction of flocculating agents (polyacrylamides) to assist
flocculation and removal of impurities. More significantly, the removal of
color in this process has been more than doubled by the introduction of the
TALOFLOC process. The basis of this process is the addition of a cationic
surfactant (TALOFLOC) which has the property of interacting with negatively-
charged colorant molecules to form a color/TALOFLOC precipitate (ref. 8 ) .
The colorant precipitate is scavenged with other insoluble impurities by the
conventional clarification process. The level of addition of the color
precipitant can be varied depending upon the color of the input liquor, but
dose rates are usually in the range of 300-400 ppm on sugar solids treated.
Although TALOFLOC can be used with both carbonatation and phosphatation
processes, its use is more effective in the latter process. This is because
the presence of the cationic surfactant in the calcium phosphate floes
formed during phosphatation reduces the interfacial tension between liquor
and air at the floe surface.
Consequently, air bubbles tend to adhere physiochemically to the floe
surface, and, in the presence of an adequate dispersion of air bubbles,
aeration of the color precipitate-phosphate floe material takes place
spontaneously. This aeration is very easily achieved by means of a
centrifugal aeration pump. Although the pump tends to break up the primary
floes, flocculation can be reestablished by the addition of about 10 parts
per million of a suitable polyacrylamide coagulant, such as TALOFLOTE.
Practical operational experience and the effects of the TALOFLOC
process on an existing refining operation are discussed in some detail by
Simoneaux (ref. 9).
The development of color précipitants and improvements in scum
flocculation have led to significant improvements in phosphatation clarifier
design and scum separation. The scum flotation rate in aerated, TALOFLOC
treated liquors is extremely rapid and more efficient, hence short retention
time, circular clarifiers operating at 85°C have been designed for use with
this clarification/decolorization process. The scum produced is more
compact than from the traditional clarifiers and can be desweetened more
efficiently in a series of circular flotation clarifiers, operated counter
currently and using polyacrylamide flocculant to reflocculate for each
stage. These improvements have transformed the traditional phosphatation
process and led to much clearer syrups with lower color. Also, because the
operating temperature is lower, clarifier residence times are shorter and
less scum is produced, sucrose loss and color formation across the modified
process are significantly less.
The process flow diagram for a typical TALOFLOC process is shown in
Figure 1, and a three stage scum desweetening system is shown in Figure 2.
Because the TALOFLOC process can give up to 60Z color removal, compared
with a maximum of about 30Z with the traditional phosphatation process, it
is very attractive to refineries attached to raw sugar factories. The
TALOFLOC- treated melt liquor from a well washed raw sugar often needs only
a very small amount of additional decolorization to make it suitable for
boiling a good quality white sugar. The subnatant clarifier liquor does in
any case require polish filtration and this can be conveniently combined
with vegetable carbon to provide an extra reduction in color.
Alternatively, should an even better quality sugar be required, an ion
exchange process can be used.
Another FDA approved color precipitant with the trade name of TALOCARB
has been developed which is more effective than TALOFLOC in carbonatation.
This new product is, like TALOFLOC, a cationic surfactant but it has several
positive charges per molecule and this tends to make it more active per unit
weight than TALOFLOC. Commercial trials so far carried out have shown that
TALOCARB, used at a dose rate of about 200 ppm on liquor solids, can
increase the decolorization in a carbonatation process by 25-30Z.
Sì
1 1
1
1 111 I Cavitation
L Aerator \y
Toloftote Lime Sue rate
Preparation Preparation
Ton* Tank rfJL
Phosphoric Tatofloc ci-, JL
Aad Tank Drum Heater
Reaction
Tank Clarifier
Takrflote Line Sucrate

Aeration
Tank
ffl
HokJng Holding "fank Έ
Tank (
L^J L-φ-* Clarified Liquor
Toloftote Lime Socrate

^mr^ Phosphoric Acid Talofloc
Dosing Pump Do&ng Pump Dosing Pump Dosng Pump
Untreated Luiquor Scum
Melt Liquor
In, Heater
a?
Melt Liquor Melt Lquor
"fank Pump
Fig. 1. Talofloc process flow diagram.

Fig. 2, Three-stage scum desweetening process flow diagram.
276
It is envisaged that TALOCARB will find its main application in

providing the carbonatation process with more flexibility to accomodate
variable color levels in the input liquor.
The average percentage decolorizations normally achieved by the
traditional and improved clarification processes are compared in Table 1.
TABLE 1
Color and ash removal by alternative clarification processes.
Average Average
Color Ash
Process Removal (Z) Removal (Z)
Carbonatation 40-50 10
Carbonatation/TALOCARB 60 10
Phosphatation 20-30 Nil
Phosphatation/TALOFLOC 50-60 10
Decolorization
Ion exchange resins. The use of synthetic quaternary resins for the
decolorization of sugar syrups dates from the 1950's. However, the
microporous, highly crosslinked polymeric matrix structure of the gelular
synthetic resins available at that time led to rapid fouling if the color of
the liquor being treated was more than about 150 I.U. These resins can be
used successfully for treating very light color liquors for special
purposes, but great care must be taken to ensure that the maximum allowable
color level is not exceeded.
Developments in resin manufacture in the following decade resulted in
the production of exchange resins based on a styrene-divinylbenzene
copolymer containing macroreticular pores. The practical capacity of these
resins for color was greatly enhanced and this led to their commercial use
as polishing agents following the then normal decolorization process using
bone char or granular carbon. These resins are ideal for polishing duties
because of their selectivity and ultimate capacity for color, including the
larger color molecules, but, although less easily fouled than the gel
resins, they can be irreversibly fouled if regularly exposed to color levels
above 400-500 I.U.
277
A third, and even more significant, advance was made during the 1970's
with the development of acrylic based quaternary resins. This type of resin
has a high decolorization efficiency coupled with the advantage that it can
cope with high color levels in the syrup being treated and the color
adsorbed can be removed quite readily from the resin by normal regeneration
with a 10Z solution of common salt (ref. 10). Consequently, the combination
of an acrylic resin in the primary position to remove the dark colors
followed by a styrene based resin for polishing to remove the final color
has proved ideal. Employing the acrylic resin in this way makes it possible
to use an all ion exchange decolorization system in cane sugar refineries
without the use of bone char or activiated carbon (refs. 11-13).
The basic flow diagrams for a two-stage ion exchange decolorization are
shown in Figures 3 and 4.
As is the case for most natural plant extracts, practically all the
colored bodies and color precursors found in sugar juices and syrups are
ionic in nature, being negatively charged in neutral or basic media. In
view of this, strong base anion exchange resins are employed for the
decolorization of natural products in many applications apart from sugar
refining. To allow the strong base resin to be operated at elevated
temperatures, as is necessary when treating sugar syrup, the exchanger is
used in the chloride form. Some of the colored fraction is removed in
exchange for chloride ions but a significant portion of the color is removed
without being exchanged for another anionic species.
Such sorptive processes involve primarily the aromatic matrix of the
anion exchange resin rather than the ionic functionality. In other words,
color is bound to the anion exchanger by a combination of electrostatic and
hydrophobic bonding. Many of the color constituents in sugar syrups have
some aromatic character and as such would be expected to be more selectively
held by bonding with the matrix of the aromatic styrene resins than with
that of the acrylic resins. The selectivity of the acrylic backbone is,
however, sufficiently high to reduce the color by 60-80Z in practical
treatment, but, because of the lessened selectivity, the color adsorbed is
efficiently removed with 10Z aqueous brine.
The emergence of resins as new decolorizing media has given a more
compact, less labor intensive process which uses less fuel and is lower in
23
CO
PRIMARY SECONOARY
OECOLORISING O E C O L O R I S I NG
COLUMNS COLUMNS
JL, JL J ^ #*■ "*S ^" "S (* '"N
FROM
FILTER r\ wr
INPUT LIQUOR
w TO FINE
HEAT EXCHANGER LIQUOR
STORAGE
Vv/
PRIMARY
LIQUOR OECOLORIZED FINE
SUPPLY LIQUOR LIQUOR
TANK TANK TANK
■a
FINE LIQUOR
LIQUOR SUPPLY PRIMARY DECOLORISEO
PUMP LIQUOR PUMP PUMP
Fig. 3. Two-stage ion exchange decolorization--service fl

PRIMARY
J SECONDARY
DECOLORIZING DECOLORIZING
COLUMNS COLUMNS
TO
REMELTER
ORINE
MEASURING AC IO
TANK MEASURING
I 1 T A NK
SWEETWATER RECYCLEO REGENERATION D

TANK WATER WATER
TANK TANK
Γα
BRINE
SANO
HLTER
& ■Q zr* L·^ zrS

SWfET WATER
PUMPS RECYCLED
WATER
3
REGENERATION BRINE TRANSFER BRiNf ACiD
WATER PUMP PUMP REGENERATION REGENERATION
SPENT PUMP
PUMP PUMP
CHEMICAL
TANK
SOFTENEO
WATER
ORAIN
Fig. A. Two-stage ion exchange decolorization--regeneration flow.
3
VO
280
capital costs than a comparable decolorization operation using carbonaceous

adsorbents requiring thermal regeneration.
The design of resin cells, their performance with various types of
resins and under different operating conditions and the efficiency of
various regeneration methods have been the subject of many publications in
the last decade.
Some of the important differences between resins and traditional
adsorbents are summarized in Table 2.
TABLE 2
Summary of differences between resins and traditional adsorbents.
Resins Bone Char Granular carbon
Effect on Anions such as Ash decreases No effect.

syrup ash sulphate are across char, by
exchanged for about 201.
chloride; no
overall change
in ash.
Selectivity of Removal of low Broad spectrum Broad spectrum

decolorization molecular weight adsorbent. absorbent. Color
and low charge Color reduced by reduced by 80-85Z.
colorants, and up to 85Z.
flavonoids
varies with type
of resin. Color
reduced by 75-85Z.
Regeneration Gradual decline in Thermal regenera- Thermal regenera-

resin performance tion and separation tion brings stock
with age, due to of dense char, carbon to standard
resin degradation brings stock char condition.*
and irreversible to standard
fouling. condition.*
Effluent Colored brine No significant No effluent

solution to be effluent problem, problem, other than
disposed of. other than waste waste waters.
waters.
* Treating poorly filtered liquor and/or badly controlled regeneration

can result in a rapid and serious fall-off in the performance of the
entire stock of adsorbent.
281
Powdered resins. The process of syrup decolorization by ion exchange

resins entails diffusion of colorants from the syrup to the surface of the
resin beads, followed by diffusion through the beads themselves. The latter
process is by far the slower and this means that a large proportion of the
resin inside the beads remains unused. Reducing the bead size is not a
practical solution since this would restrict the flow of syrup through the
resin bed.
These considerations have led to the development of a powdered resin
(Ecodex-R) which can be used as an adjunct to filter aid to provide a
decolorizing precoat for suitable filters. This product is typically used
in a precoat of 7-14 mm thickness to give a contact time of some 1-2
minutes; this is compared with a contact time of about 20 minutes with a
deep bed resin system.
Early results with this powdered resin are promising and it is reported
that a higher decolorization efficiency can be obtained in practice.
Powdered adsorbent/filter aid adjuncts in the Ecodex range have found use in
polishing duties. The low capital costs and simplicity of operation could
make this process economically attractive to small refineries attached to
raw sugar factories, although the high costs of resin replacement would be
likely to make its use too costly for a large traditional refinery (refs.
14-16).
SURE. The problems of diffusion resistance within conventional resin
particles has also recently led to development work on a new decolorization
process called 'SURE' (ref. 17).
The adsorbent being developed for this process is based on a support
polymer of polypropylene, which has a very regular microporous structure
best described as a packed porous bed of minute beads. Physically, this
polymer, with the tradename Accurel, resembles resin but has a much higher
porosity. A long chain quaternary ammonium compound is adsorbed onto the
Accurel by circulating an aqueous solution of the quaternary salt through
the dry column of Accurel particles.
After rinsing out excess quaternary salt, the column can be used for
decolorizing hot (60-75°C) sugar syrup at 65 Bx. Because of the high
porosity of the polymer, space velocities of up to ten times those
attainable with decolorizing resins are claimed, depending upon the color of
the feed liquor. If these short contact times are proved feasible in a
282
production size plant, the size of absorbent units would, of course, be very
much smaller than the resin columns for a similar duty.
The amount of color (BV X IU) which can be removed during each cycle in
the SURE process is said to be comparable to that for conventional resins
(30-40,000 BV-IU's).
As with anionic decolorizing resins, caustic brine is used for
regeneration of the Accurel based adsorbent. All the quaternary compound is
removed from the polymer by the regeneration process, so the polymer support
material has to be reloaded with fresh solution of quaternary salt for the
following cycle. Indications are that the removal and replacement of the
active decolorant in this way will reduce the tendency for the
decolorization efficiency of the adsorbent to decrease with time as compared
with resin and also cause less problems of irreversible fouling by very high
color levels in feed liquor.
The SURE process is still under development and is not yet operating
commercially. No information is currently available on likely operating
costs and how these will compare with ion-exchange resin systems as a result
of the quaternary having to be reloaded every cycle.
TRADITIONAL CANE SUGAR FACTORY PROCESSES

In most sugar factories, juice clarification is the only process which
specifically removes impurities. No additional decolorization is normally
necessary, the remaining purification needed for raw sugar production being
taken care of by the final crystallization process. However, if the factory
wishes to produce a mill white sugar for direct consumption, it is necessary
to introduce some degree of decolorization of the mill juice and perhaps of
the syrup also. The decolorization is usually combined with the
clarification process to produce a much cleaner, paler treated juice, so
that subsequent evaporation and crystallization directly from the raw syrup
yields an off-white sugar locally acceptable for commercial and domestic
use.
Juice Clarification
The purification process in a cane factory starts with the removal of
the larger particles of suspended matter from the mill juice by straining
through coarse conventional mesh strainers or wedge-wire screens. This
still leaves significant amounts of suspended solids, including soil, sand
and fine fibrous particles.
Juice clarification in a factory producing raw sugar is effected by a
combinaton of lime addition and heat. This is the oldest, simplest and
least expensive clarification process for cane juice, but its effectiveness
is such that it is still very widely used after some 200 years. Basically,
the process consists of adding a thin slurry of hydrated lime (sp. gr.
1.03-1.09) to the juice, followed by heating to a temperature a few degrees
above its boiling point. The lime neutralizes the organic acids in the
juice and reacts with the soluble phosphates to form calcium phosphates of
variable composition, while the heat treatment coagulates the protein
present and also de-aerates the juice. In spite of intensive study over
many years, the reactions taking place are still not fully understood, but
the net result is the production of a flocculant precipitate of complex
composition, heavy enough to be separated by settling in suitable subsiders
(clarifiers). The precipitate traps most of the finely
suspended material in the juice, so that the supernatant juice is clear.
Numerous modifications in the application of this basic process have
inevitably been devised and introduced over the years, often because of the
need for special treatment of juices from certain varieties of cane or from
cane grown under certain agronomic or climatic conditions. A comprehensive
summary of the main variations in juice treatment and in settling equipment
which have proved useful is given by Meade and Chen (ref. 18).
To achieve efficient clarification, the main criteria are careful
control of the pH of the limed juice and the temperature after heating.
Because the phosphate precipitate plays such an important role in the
clarification process, the phosphate content of the mill juice is critical;
consequently, phosphoric acid or soluble phosphates are often added to the
juice prior to liming. A minimum level of phosphate of about 350 ppm (as
P2O5) is usually recommended for good clarification.
The thin mud which settles out in the clarifiers is usually mixed with
a quantity of fine particles of a bagasse (bagacillo) and filtered on rotary
vacuum filters; the resultant cake is desweetened by washing with hot water.
Neither the filter pick-up nor the wash runnings from this process are clear
enough to go forward with the clarified juice and both are normally recycled
with the mill juice for re-clarification.
28^
Clarification for mill white sugar production. The juice clarification

processes traditionally used in cane sugar factories for the production of
mill white sugars are carbonatation, sulphitation or a combination of the
two.
The cane juice carbonatation process is similar to that used in the
beet sugar industry, from which it was in fact copied by the cane sugar
industry in Java in about 1880. The process was used somewhat empirically
and improvements were made on the basis of experience; it is still used in
one or two countries in Asia. Basically the process involves the
neutralization with kiln gas to produce calcium carbonate. Single
carbonatation, in which the lime is neutralized in one stage, has been
replaced by double carbonatation, which involves two-stage carbonatation
with filtration after each stage.
At the turn of the century, sulphitation was introduced to further
reduce the pH of the filtered juice (the minimum pH of the carbonatated
juice after the second gassing is determined by the carbonate-bicarbonate
equilibrium at about 8.2). Sulphitation results in a much paler juice for
evaporation and crystallization because of the chemical decolorization
effect of sulphite as a reducing agent as well as the indicator effect of
the lower juice pH. Later the use of sulphur dioxide was extended to the
syrup produced after evaporation, to give the so-called NDouble
carbonatation, double sulphitation' process, which, carefully operated, is
capable of producing a good quality mill white sugar. However, with the
ever increasing cost of producing large quantities of lime at the raw sugar
factory, the carbonatation process is getting very expensive to operate and
is being replaced by alternative processes.
Juice sulphitation combined with simple lime defecation is another
process which has been used in the production of mill white sugar. This
process is also operated somewhat empirically, but whether the juice is
sulphited hot or cold, or in acid, alkaline or neutral conditions, the
ultimate objective is to utilize the reducing properties of the sulphite ion
to inhibit the subsequent formation of colorants during the evaporation and
crystallization processes. The combination of liming, sulphitation and
heating produces a heavy precipitate which can be separated by settling in
conventional juice clarifiers. Double sulphitation, in which the evaporator
syrup is also sulphited to reduce the pH to about 6.5 before
285
crystallization, is also practiced in some sulphitation factories. The

sugar produced by this process, essentially an off-white raw sugar, is often
not very attractive and has a distinct tendency to develop a yellow
coloration during storage.
DEVELOPMENTS IN CANE SUGAR FACTORY PROCESSING

Juice clarification in raw cane sugar factories is still based on the
traditional lime defecation process. Developments in clarification and
decolorization technology in the refining process are, unfortunately, not
applicable to raw factory juices. Nevertheless, there have been
improvements in equipment and the availability of synthetic flocculants
particularly, which have enabled the clarification process to make a
contribution to the general improvement in raw sugar quality over recent
years.
As the economies of cane sugar countries improve, the local demand for
good quality white sugar for both commercial and domestic use increases.
Neither washed raw sugar nor sugar produced by sulphitation alone is
generally acceptable and the capital and operating cost of the double
carbonatation/double sulphitation process is usually much too high to
contemplate. In an effort to meet the demand for good quality white sugar,
refineries can be built on to the end of sugar mills, and this has been done
in some countries. Often, however, the quality requirements of the
consumers cannot justify the cost of fully refined sugar and it is in such
circumstances that recent developments in clarification technology
applicable to raw sugar factories are most useful.
Juice clarification
The most significant improvement made to the traditional clarification
of cane juice in recent years has resulted from the introduction of
synthetic flocculants. Most of the flocculants used in juice clarification
are generically water soluble, partially hydrolyzed polyacrylamides, and a
number of products are available under various trade names. The precise
molecular structure, polymer chain length and degree of hydrolysis differ
among products and the flocculating efficiency of a given polymer can vary
from one area to another depending on the particular juice characteristics
and local operating conditions. Flocculants in dilute aqueous solution
286
(typically 0.5Z) are usually added to the juice at a rate of 1-3 ppm on
juice; dose rates in excess of 3 ppm rarely give any further improvement in
settling rate or mud volume.
These flocculants are very widely used throughout the cane sugar
industry and, in cases of marginal subsider capacity, dirty mill juice
(perhaps from mechanically harvested cane), refractory juices, etc., these
flocculants have proved invaluable since their development and introduction
into the sugar industry, some 25 years ago.
Batch subsiders have long been replaced by efficient continuous
clarifiers in which the residence time for settling of 2-2 1/2 hours is
normal. Much smaller, high capacity clarifiers with residence times of only
20-30 minutes have recently been introduced. These clarifiers cannot be
operated successfully without the use of flocculants and the liming, heating
and flocculant addition all have to be controlled very accurately.
Filtering the clarifier mud on rotary vacuum filters produces a
filtrate which is much too turbid to send forward with the clarified juice.
However, the conventional procedure of returning this filtrate to the mixed
juice and recirculating through the mainstream clarification system has a
number of disadvantages: the recycled juice is again subjected to high
temperatures in the juice heaters and clarifier and this results in
additional color generation and sucrose inversion; the volumetric loading
on the clarifiation process is increased by 15-202 and the impurities in the
recycled stream inevitably reduce the efficiency of the mainstream
clarification.
A separate flotation clarification system has recently been introduced
to overcome these problems. (See Figure 5). The dirty filtrate is treated
with a small amount of phosphoric acid (50 ppm as P 2 0 s ) and lime (100 ppm as
CaO) and aerated; additional flocculation is then promoted by the controlled
addition of an appropriate flocculant. The filtrate flows by gravity to a
small flotation clarifier, where the large aerated floes float to the
surface of the liquid, allowing the clarified filtrate to be removed from
the bottom of the clarifier and passed forward directly into the main
clarified juice stream going to the evaporator. The scum from the top of
the clarifier is returned to the thin clarifier mud and eliminated with the
filter cake. This process is particularly useful for filtrates produced in
, Milk Of Urne
-2?
F bei:liant
Preparahon fptn
lin*
Clarifier <>
Clear Filtrate To
Evaporators
Re act o n pH Sample
Floccukint Tank Pump
Hokltng
Tank Scum Return To
Rotary Vacuum
■â! Acid ■s Fitter Supply l a n k
Flocculant Phosphoric
Carboy
Dosing ftjmp
Acid Dosng
Pump
FUtrate From Rotary
Vocuum Fdttr
V
Aeration
ftjmp
Untreated Filtrate Untreated

Tank Filtrate
Pump
Fig. 5. Filtrate clarification process flow diagram. co

288
factories milling mechanically harvested or mechanically loaded cane

containing field soil.
Clarification for white sugar production

Particulate impurities are present in all clarified juices and these
are concentrated by a factor of four during evaporation. In addition,
sparingly soluble salts are precipitated during the evaporation process and
these add to the turbidity of the final syrup. Particles range from
colloidal to hundreds of microns in size and are virtually impossible to
remove by filtration. During crystallization some of this suspended matter
becomes occluded in the growing sugar crystals, increasing their intrinsic
color, while the remainder of the insoluble material is concentrated in the
molasses, increasing its viscosity, and making efficient separation in the
centrifugal machines more difficult.
Consequently, processes which improve the clarity of the raw syrup can
make a significant contribution to the quality and appearance of the product
sugar. Syrup clarification processes, such as FABFROTH and TALODURA
processes, have recently been introduced into the cane sugar industry with
considerable success.
A flow diagram of the latter process is shown in Figure 6. In this
process raw syrup is heated to a temperature of about 80°C and the hot syrup
is passed to a reaction tank where controlled additions of phosphoric acid
and lime slurry are made to produce a syrup of predetermined controlled pH
containing floes of calcium phosphate. After aeration and reflocculation,
the syrup is run to a flotation clarifier, in which the clear syrup is
separated from the insoluble matter.
A syrup clarification process using phosphoric acid and lime in this
way will remove some 90Z of the suspended solids and about 102 of the
soluble color.
Combining the syrup and filtrate clarification processes described
above with juice sulphitation yields a clear, lightly colored syrup from
which a very good quality white sugar can be directly crystallized.
Syrup produced using this process combination is characterized by a
particularly low turbidity which gives the crystals a sparkling appearance.
The juice sulphitation and syrup clarification processes can be controlled
TalodurQ
Rocculant
Holdng
lank
Λ Clarified
Syrup
Untreated
Syrup
Or Talofiltrate Process
Fig. 6. Syrup clarification process flow diagram.

vo
290
and arranged such that the residual sulphite in the product sugar is
extremely low (generally less than 2 ppm as S0 2 ).
The general quality of * BLANCO DIRECTO* sugar produced commercially
using these modern processes is compared with that of typical fully refined
and mill white sugars in Table 3.
TABLE 3
Blanco Directo sugar compared with refined and mill white sugars.
Property Refined Blanco directo Mill white
Polarization 99.9 99.8 99.6

Color (ICUMSA 4) 10-80 100-200 250-200
Turbidity (ICUMSA 4) 10-30 20-50 100-500
Ash (Z) 0.01-0.04 0.05 0.1
Invert (Z) 0.01-0.04 0.02 0.1
S0 2 (ppm) 1-5 20-50
Acknowledgement
The assistance and information provided by Dr. Christine Goodacre of

Tate and Lyle Group Research on powdered resin and the SURE process in
particular is gratefully acknowledged.
REFERENCES
1 M. A. Clarke, The future of raw sugar quality. Sugar y Azucar,
80 (1985) 32-50.
2 V. E. Baikow, Manufacture and Refining of Raw Cane Sugar. 1st Ed.
1968, Elsevier, Amsterdam.
3 A. P. Saranin, Technology of phosflotation of sugar melt. Sugar Tech.
Rev. 2(1) (1972) 1-72.
4 M. C. Bennett, Liquor carbonatation, Int. Sugar J., 69 (1967) 101-104,
198-202.
5 G. Gaudfrin, Sugar refinery filtration and clarification of carbonated
syrups, Proc. Sugar Ind. Technol, 45 (1967) 97-125.
6 A. I. Macdonald, Bone charcoal: some pointers on supply, chemical
properties and refinery practice. Proc. Sugar Ind. Technol., 34 (1975)
58-66.
7 L. G. Sansaricq, Bone char. Proc. Sugar Ind. Technol., 43 (1984) 86-91.
8 M. C. Bennett, New industrial process for decolorizing sugar. Chem. &
Ind. (London), 22 (1974) 886-91.
9 W. J. Simoneaux, The Taloflote-Talofloc process at Supreme. Proc. Sugar
Ind. Technol., 37 (1978) 24-66.
291
10 C. Loker, Factory and pilot-plant experiences with ion-exchange resin

employed in gross decolorization in a carbonation refinery. Proc. Sugar
Ind. Technol., 42 (1983) 211-221.
11 W. Fries, Cane sugar decolorization by ion exchange resins. Int.
Sugar J., 84 (1982) 325-328.
12 N. G. H. Hindefelt and K. A. Lilja, Int. Sugar J., 88 (1986) 148-153.
13 L. Ramm-Schmidt and G. Hyoky, New resin decolorization station at
Porkkala refinery. Proc. Sugar Ind. Technol., 43 (1984) 241-259.
14 A. Tavares, A. Roberts and R. Kunin, A progress report on the use of
Ecosorb precoat technology for cane syrup clarification and decolori
zation. Proc. Sugar Ind. Technol., 41 (1982) 45-55.
15 A. Tavares, R. Forman and R. Kunin, Refinery and laboratory studies on
the use of Ecosorb precoat formulations for the clarification and
decolorization of cane and beet sugar syrups. Proc. Sugar Ind.
Technol., 42 (1983) 192-210.
16 B. C. Goodacre, Performance of powder resin in conjunction with the
Talofloc process.Proc. Sugar Ind. Technol., 41 (1982) 14-44.
17 D. Frank, D. L. Metcalfe and J. Park, A new sugar decolorization
process. Int. Sugar J., 89 (1987) 46-54.
18 G. P. Meade and J. C. P. Chen, Cane Sugar Handbook, 10th Ed., 1977,
112-147.
Chemistry and Processing of Sugarbeet and Sugarcane, edited by M. A. Clarke and M. A. Godshall
292
Chapter 18
CHEMISTRY OF SUGAR IN FOOD PROCESSING
W.S.C. TSANG AND M. A . CLARKE
SUMMARY
The physical and functional properties of sugars, particularly sucrose,

in foods and beverages are reviewed. Primary functional properties of
sugars include sweetener activity, activity as a flavor enhancer and
sustainer, antioxidant and preservative, and interaction with water to
affect water activity. Chemical reactions include Maillard reactions or
browning compound formation, caramelization, and fermentation.
In beverages, properties of sweetness and flavor are important.
Solubility is especially important in alcoholic beverages. Levels and
characteristics of non-sugars present in the sweetener can affect the
appearance, flavor, and stability of a beverage.
INTRODUCTION
Sucrose, the common table sugar, is a disaccharide composed of a

glucose unit and a fructose unit. Commercial white sugar represents the
highest volume organic chemical produced worldwide in pure form (> 99.5Z).
It is obtained mainly from sugarcane and sugarbeet.
Generally recognized as a source of energy and sweetness, sucrose is
used widely in food as a preservative, flavor enhancer, bulking and body
agent, and texturizer (ref. 1). This chapter will attempt to highlight the
physical and functional properties associated with sucrose which make it
attractive in food applications.
PHYSICAL PROPERTIES OF SUCROSE

The general physical properties of sugars that are important in food
applications are now described (ref. 2).
Crystalline sucrose
Sucrose crystals are monoclinic and form more readily than glucose and
fructose. Crystals have a prismatic habit which is strongly affected by
impurities: the presence of raffinose or dextran causes sucrose to
crystallize in long needle-shapes. Sucrose by its sparkling crystallinity,
or its decorative effect in icings and on bakery products makes food more
attractive.
293
Sucrose solutions
Sucrose is one of the more soluble of the simple food carbohydrates as
shown in Figure 1 (ref. 2 ) . The solubility of sucrose in water, especially
in the presence of water-soluble impurities, is of great commercial
significance in sugar manufacture. The high degree of solubility is
essential in preparation of flavored syrups, jams, jellies and preserves,
and canned fruits.
An important commercial property of sucrose in solution is its
polarization. Sucrose is purchased, and the various stages of manufacturing
processes are monitored, by polarization measurements.
100
80
« 60
<
o
D
tf 40
20
0 20 40 60 80 100
TEMPERATURE
Fig. 1. Approximate solubility of various sugars at different

temperatures (°C).
Viscosity
Viscosity in a solution is a measure of its resistance to flow. The
viscosity of sucrose solutions varies directly with the sugar concentration
and inversely with temperature (refs. 1, 2 ) . Sucrose is intermediate in
viscosity between the extremes of glucose syrups as indicated in Figure 2.
Viscosity is an important property of sugar solutions in foods since it
imparts thickening effects and improved body and mouthfeel, as in puddings,
fruit preparations, and preserves.
29^
Tempero tu re (*C)
Fig. 2 . V i s c o s i t y of c a r b o h y d r a t e sweetener solutions at 70Z s o l i d s .
Osmotic pressure
The p h e n o m e n o n of osmotic pressure has important a p p l i c a t i o n s in the
field of food sugar chemistry, since sugars in solutions h a v e a n intrinsic
osmotic p r e s s u r e . Sugar functions as a p r e s e r v a t i v e by i n c r e a s i n g osmotic
p r e s s u r e to a level w h e r e m i c r o o r g a n i s m s cannot s u r v i v e . A n example of this
effect is seen in jams and j e l l i e s , w h i c h can be stored at a t m o s p h e r i c
temperature and pressure w i t h o u t g r o w t h of m i c r o o r g a n i s m s (unless a layer of
w a t e r c o n d e n s e s on the s u r f a c e ) .
Water activity
The a b i l i t y of the sugars to lower the w a t e r a c t i v i t y (Aw) of the
s o l u t i o n , w h i c h is related to their osmotic p r o p e r t i e s , is a l s o a n important
intrinsic food p r o p e r t y . A closely related d e s i g n a t i o n is the e q u i l i b r i u m
r e l a t i v e h u m i d i t y (ERH) w h i c h m a y be defined as the relative h u m i d i t y at
295
which the sugar solution neither gains nor loses moisture in its surrounding
atmosphere. Sugars are used extensively to control Aw and ERH values in
food to improve microbial stability.
Sugars in solution lower the ERH value and increase the osmotic
pressure, thus making the product less liable to microbial contamination.
The preservative action of sugars in high concentrations is related to the
fact that different organisms have different ERH requirements. The ERH
values below which certain microorganisms will not grow are shown in Table 1
(ref. 1).
TABLE 1
Minimum ERH allowing microorganism growth.
Type of microorganism ERH Z
Common bacteria 91
Common yeasts 88
Common moulds 80
Halophilic bacteria 75
Xerophilic moulds 65
Osmotolerant yeasts 60
Figure 3 shows the percentage moisture content of sucrose crystals at

various humidities when the moisture content has come to equilibrium. There
is a broad band of relative humidity over which the crystal is stable. If
the relative humidity of the air in contact with the sugar is kept under
70Z, little or no caking will occur.
Humectancv
Another property of sugar molecules is their water binding capacity or
humectancy. A humectant product has the ability to resist change in
moisture content, which means that products made with sugar will not dry out
prematurely—they stay softer or fresher longer. The addition of sucrose to
sweet breads and cakes demonstrates this property.
296
Wottr (%)
>.1
O.OI
J I I I I I L
20 40 60 80
Relative humidity (%)
Fig. 3. Relative humidity isotherm for sucrose crystals.
FUNCTIONAL PROPERTIES OF SUCROSE

Sucrose plays a multifunctional role in food technology. The nature of
these functional uses is always related to the physical properties (ref. 1).
Energy source
Sucrose serves as a source of energy in human metabolism. On a weight
basis, both protein and fat exceed sucrose in calorific value (ref. 1 ) .
Sweetness
The primary purpose of sucrose in food products is usually as a
sweetener, since its sweet taste has no undesirable overtones of bitterness,
sourness or saltiness. It makes many foods palatable (refs. 3, 4).
The sensation of sweetness is affected by several factors such as total
acidity, pH, viscosity, temperature, and presence of other constituents
(refs. 2, 5, 6). The sweetness of sucrose is compared to that of some
common sugars in Table 2 (refs. 7, 8).
297
TABLE 2
Relative sweetness of some common sugars.
Sugar Relative sweetness
Fructose 114
Sucrose 100
Glucose 69
Galactose 63
Maltose 46
Lactose 39
Flavor enhancer
Sucrose enhances the natural flavor of foods by creating a better

balance among acidity, bitterness, and sourness, when used in concentrations
in which the sense of sweetness will not override the flavors which are
being accentuated. Sugar itself can also have flavors: flavor in sugar
comes from thermal degradation, as when it is caramelized, and from Maillard
reaction products and plant flavonoids, especially in brown sugars, as
discussed in the chapter by Godshall.
Bulking agent
Sucrose serves as a bulking agent in a variety of formulated food as,
for example, in dry mixes of various types. It also serves along with other
ingredients, to give bulk to many confectionery products, and to baked
goods.
Texturizer
Sugar confers desirable textural properties to foodstuffs varying from
the hard brittle texture of boiled sweets, the viscosity of ice cream, the
gel in preserves, to the body imparted to soft drinks. The variety of
textures available from one ingredient, which has other additional
functionalities, is unusually wide.
Preservatives
A traditional use for sucrose is based on its preservative action and
it is essential in manufacturing fruit conserves, glazed and preserved
298
fruits and in sweetened condensed milk. Sugar in solution functions by

reducing the water activity and increasing the osmotic pressure to a level
where preservation against microorganism is ensured. Water activity is
reduced below the level necessary for microorganisms, even most osmophilics,
to survive.
Fermentation
Sucrose is readily fermented by many microorganisms. Probably the most
important fermentation, and historically one of the oldest processes in food
technology, is the production of ethanol, now in the many forms of
industrial alcohol, rums, beers and wine.
Considerable research has been conducted on the manufacture and flavor
control of rums, as is discussed in the chapter by West and Harris. Flavors
in molasses-based rum differ from cane juice-based rum. Sucrose is inverted
to glucose and fructose during the fermentation process and the fermentable
solids are actually 105Z of the sucrose solids.
Sucrose is widely used as a fermentable carbohydrate as, for example,
in bread baking. In addition, it will enter into Maillard reactions, and
also caramelize to produce a desirable crust color.
Color
There are several pathways along which sucrose reacts to form colored
compounds. Thermal degradation of sucrose leads to caramelization with its
characteristic flavor and brown coloration. When sucrose is inverted and
subjected to heat, in the presence of primary, or secondary amines, or amino
acids, the well-known Maillard reaction is also observed (ref. 2).
The mechanistic routes for Maillard reaction include:
(1) Direct, via Amadori compounds.
(2) Indirect, via dicarbonyl (includes Strecker degradation, a source of
flavors).
(3) Indirect, bypassing Amadori, through Schiff base, and
(4) After Strecker degradation, a route that can generate heterocyclics
and other flavor compounds.
299
Antioxidants
Sugar solutions inhibit the formation of rust on ferrous surfaces.
They also prevent the oxidative degeneration of flavors in canned fruit
(ref. 1 ) .
FOOD APPLICATIONS
Sucrose has been incorporated into a variety of foods and beverages
where it serves several functions besides providing sweetness (refs. 1, 2,
9).
Baking
Sucrose is a necessary component in baked goods. It serves as the
source from which yeasts produce the carbon dioxide essential to the rising
and leavening action. It also contributes to crust characteristics, texture
and product stability.
In a comparison of cakes made with and without sugar (ref. 10), the
cake baked without sugar was dry and tasteless and did not rise. In
addition, the sugar-free cake actually contained more calories per slice
than the cake baked with sugar, as outlined in Table 3.
TABLE 3
Comparison of cakes made with and without sugar.
100 g of cake made 100 g of cake made

with sugar without sugar
Butter or margerine 25 g 33 g
Sugar 23 g
Eggs 20 g 26 g
Flour + spices 29 g 37 g
Cream 3 g A g
Total 100 g 100 g
Energy kcal/100 g 408 kcal - 5 slice 414 kcal - 3-4 slices

content same size slices e. 82 kcal 104-138 kcal
Ingredients other than sugar were identical.

300
Beverages
Sucrose not only provides sweetness but also increases viscosity and
thereby the desirable body or mouthfeel of beverages. It also improves the
acceptability of soft drinks because it enhances fruit flavors (ref. 11).
A haze, called acid beverage floe, can come from organic matter in the
sugar, which will become visible upon acidification (refs. 12, 13). Even
though it is harmless, it resembles microbial infection, and so causes the
beverage to be rejected by the consumer. Another type of floe, also
sometimes caused by sugar, has been observed in sweetened alcoholic
beverages: this is caused by dextrans or other polyglucans in the sugar,
which precipitate upon the addition of alcohol (ref. 13).
Preserves
In jam and jelly manufacture, sugar combines with acid and pectin to
thicken or solidify the products, therefore preserving the fruits. The high
osmotic pressure exerted by sugar solutions is a major factor in suppressing
microbiological spoilage in the storage of food.
Dairy products
Sweetened condensed milk is prepared by evaporating a mixture of fresh
milk and added sucrose. Sugar inhibits mold and bacterial growth as a
result of the osmotic pressure of solutions in high concentrations. Sugar
also contributes to the characteristic flavor of condensed milk, through
reaction with the milk components.
The dairy industry's principal use for sucrose is in ice cream. Apart
from its contribution to product acceptability through sweetness, it
improves mouthfeel and texture. Although there has been some replacement of
sucrose in ice cream by cheaper corn sweeteners, premium brands still use
only sucrose as a sweetener.
Meat products
Sugars are important in development of color in the curing of meats,
such as bacon and ham. Sucrose improves color by establishing reducing
conditions and by tending to prevent oxidation of ferrohemoglobin to
ferrihemoglobin during storage (ref. 3).
301
Confectionery products
Hard-boiled sweets, or hard candies, are produced by boiling together
sucrose, glucose syrup, acid, color, flavor and water. On cooling, the mass
sets to a "glass." It is essential that this glass should not crystallize
or "grain." If it does, the sweet will crumble and disintegrate. The high
degree of supersaturation obtainable by sucrose in solution is essential to
obtain a good "glass."
Sucrose derivatives
A wide variety of sucrose-based chemicals has been prepared with a view
towards utilization of sucrose as a chemical feedstock (refs. 14, 15).
Sucrochemicals have applications as baking additives, emulsifiers, and
viscosity modifiers in foods. They are also used as detergents,
confectionery ingredients and in drug and cosmetic formulations.
At the present time, the economic factors for sucrochemicals are
unfavorable when compared to petrochemicals. Most sucrochemicals, however,
possess certain unique properties such as biodegradability. This topic is
fully discussed in the chapter by Khan.
CONCLUSION
Sucrose has been shown for many years to be sweet and versatile.
Because of its general availability, multifunctionality and relatively low
cost, sucrose plays an important role in food processing, where its many
useful physical and chemical properties offer great opportunity to food
technologists.
REFERENCES
1 W. M. Nicol, Sucrose and food technology, in: G. G. Birch and K. J.
Parker, (Ed.), Sugar: Science and Technology, Applied Science
Publishers Ltd., London, 1979, 211-230.
2 R. S. Shallenberger and G. G. Birch, Sugar Chemistry, AVI Publishing
Co., Inc., Westport, Conn., 1975, 60-71 and 147-159.
3 R. M. Pangborn, Sensory perception of sweetness, in: G. E. Inglett
(Ed.), Symposium: Sweeteners, AVI Publishing Co., Inc., Westport,
Conn., 1974, 23-44.
4 M. Brook, Sucrose and the food manufacturer, in: J. Yudkin, J. Edelman
and L. Hough (Eds.), Sugar J., Butterworths, London, 1970, 32.
5 R. S. Shallenberger, Sugar structure and taste, in: Carbohydrates in
solutions, R. G. Gould (Ed.), Advances in Chemistry Series 117, A.C.S.,
Washington, D.C., 1973, 256-263.
6 H. M. Pancoast and W. R. Junk, Handbook of Sugars, 2nd Edition, The
AVI Publishing Co., Inc., Westport, Conn., 1980, 383-394.
7 R. S. Shallenberger, The theory of sweetness, in: G. G. Birch, L. F.
Green and C. B. Coulson (Eds.), Sweetness and Sweeteners, Applied
Science, London, 1971, 43.
8 M. A. Amerine, R. M. Pangborn, and E. B. Roessler, Principles of
sensory evaluation of food, Academic Press, New York, 1965, 95.
9 A. Lachman, The role of sucrose in foods, The International Sugar
Research Found., Washington, D.C., 1975.
10 Communication, Food Technology Division, Suomen Sokeri Oy, Porkkala,
Finland.
11 L. B. Sjostrom and S. E. Cairncross, Role of sweeteners in food flavor,
in: Use of sugars and other carbohydrates in the food industry,
Advances in Chemistry Series No. 12, A.C.S., Washington, D.C., 1955.
12 E. J. Roberts and F. G. Carpenter, Composition of acid beverage floe,
Proc. Tech. Sess. Cane Sugar Refining Res., 1974, 39-50.
13 M. A. Clarke, E. J. Roberts, M. A. Godshall and F. G. Carpenter,
Beverage floe and cane sugar. Proc. Int. Soc. Sugar Cane Technol.,
1977, 2587-2598.
14 J. L. Hickson, The potential for industrial uses of sucrose, in:
G. G. Birch and K. J. Parker (Eds.), Sugar: Science and Technology,
Applied Science Publishers Ltd., London, 1979, 151-180.
15 P. K. Chang, Sucrose chemicals and their industral uses, in: G. E.
Inglett (Ed.), Symposium: Sweeteners, AVI Publishing Co., Inc.,
Westport, Conn., 1974, 74-77.
303
Chapter 19
SWEETENERS: SAFETY, UTILITY AND EFFICACY
G. N. BOLLENBACK
At a session such as this I usually recite the evidence that exonerates

sugar from almost all the allegations that the sweetener is wholly or in
part responsible for various health and nutritrion problems. Today the
pressure to present such a defense is low. In fact, there are several
recent publications that can be used as references in order to neutralize
any such claims. These publications include (1) the American Council on
Science and Health's paper on sugar; (2) the article by Dr. Thomas Jukes in
World Review of Nutrition and Dietetics; (3) the British Nutrition
Foundation's report on Sugars and Syrups; and (4) the FDA's Sugars Task
Force Report. All of these articles should offer a bulwark of defense for
the industry for many years.
With all this reassurance of the safety of sugar it hardly seems
necessary to reiterate the role or lack of role of sugar in health and
nutrition problems. That is not to recommend complacency on our part; there
will always be someone who finds something in the way of a health or
nutrition problem attributable to sugar.
With these two points in mind—the reassurance of the safety of sugar
and the warning against complacency—I would offer to you two commentaries.
The first is short and that is to note to you that the Sugar Association's
sponsored research program, conducted by the scientific section at the
Association, played an important role in developing many of the scientific
facts upon which the "safety reassurance" publications were based. (Copies
of the projects sponsored within this program are available from the
Association offices).
The second commentary intends to be an aid against becoming complacent
regarding the future allegations relating sugar to health/nutrition. It
includes my thoughts and suggestions as to where the industry might look for
added defense and new offense in establishing where and how sugar fits into
our diet as regards health/nutrition and food technology.
3<A>
I was, in fact, adopted by the sugar industry—coming from the ivory

research tower known as the George Moffatt Research Labs of CPC. These
labs, I understand, have also been retired recently.
Being an adopted son of the industry has had its advantages. My
approach to problems has often been somewhat different from the standard for
the industry. Also, I was able to practice the advice of several men whose
insight to problem solving has proved invaluable.
One of these men was Dr. Henry Gilman of Iowa State. Dr. Gilman taught
all his graduate students the immense importance to research of background
literature, to the point of asking on prelims not only what library stack
the French Chemical Journal was on but what the color of the binding was.
"Check the references." said Dr. Gilman, "You may interpret them differently
from the chap who cited them."
Aaron Yohalem, onetime Senior Vice President of CPC, after a meeting at
which I failed to sell a product to the CPC marketing managers, was a
gentleman who advised me that if a project or idea does not survive at Point
A, put it at the bottom of the pile of ideas and let it surface at some
other, perhaps more appropriate, time. Mr. Yohalem might well have included
the advice that if one presents for approval an attractive short term
project he then stands a good chance of promoting a seemingly less
attractive, long term project of potentially great value.
Finally, a chap known to many of you, Pete Hoynak, who was a most
outstanding contributor to the success of the sugar industry in general,
liquid sugar in particular, and was the first to propose and try to execute
the formation of a total sweetener concept, had much advice worth listening
to. He taught me how to behave at conventions, how to treat customers and,
above all, what to do with information once it was in hand—dispense with it
rapidly to become known as one who stays on top of the field and anticipates
future trends.
The remembrance of advice from these men: being fully conversant with
the literature (Gilman); promoting projects with patience and timing
(Yohalem) and keeping information rolling (Hoynak)—and the fact that I was
able to bring to the scene an outsider's viewpoint were highly instrumental,
I believe, in making the Sugar Association's sponsored research program,
designed to clarify the role of sugar in health and nutrition, such a
resounding success.
Of course there were other ingredients contributing to the success of
the research program. Fred Stare collected for the industry a group of
medical experts who fit beautifully with the industry's problems. Dave
Coursin smoothly held these skitterish experts (later known as our Food &
Nutrition Advisory Council--FNAC) in place until they felt comfortable
working with us industrialists. The FNAC members indeed were the guides to
finding sound answers to the questions of sugar and health.
Then too, we had a special group of members we called our Contributing
Research Members (CRMs) that was composed of representatives of large sugar
user companies. For most of the life of the program the CRM reps were very
high caliber scientists who contributed importantly to the selection of
research projects funded through the program.
Importantly the Sugar Association's Board of Directors allocated the
funds necessary to conduct the program.
Add to these elements my department's intense interest in fact finding,
our ability to communicate with all participants and a high degree of
integrity maintained in conducting the program and we can see in great part
what made the sugar-health-nutrition Sugar Association sponsored research
program succeed in helping significantly in putting the role of sugar in
health into proper perspective.
In reviewing the Sugar Association's research program philosophy,
direction and procedures, I came up with a list of practices we evolved that
may help in future endeavors of this sort. As well, I composed a list of
approaches to problems and projects I strongly believed in but had not put
into practice before my retirement. I intend to offer you both lists but in
so doing I must admit to feeling a bit like Polonius advising his son
Laertes as Laertes prepares to go out into the world and sow some wild oats.
Polonius, in the performances I have seen, gives his advice (known as
precepts in Elizabethan times) in great good humor, recognizing that Laertes
will listen respectfully and then go out and do his own thing. Still,
Polonius will always have his offered bits of advice as reminders when
Laertes comes to him in the future with great personal discoveries.
Polonius and I have at least in common age and a deep experience from
which to draw our precepts.
To the lists. First some Polonius-like precepts based on some research
program procedures we evolved.
306
1) Allegations of relationships between sugar and various health and

nutrition problems are best tracked by members of the sugar industry rather
than those persons involved with medical research. The medical researchers
know what is hot in their own fields (thermogenesis in obesity; cell insulin
receptors in diabetes; gum caries in dentistry) and what avenues are most
profitable to follow when searching for funding. These experts keep track
of where sugar fits into their research picture only in special cases;
industry members know best about specific alleged relationships.
2) Applications of sugar in the food industry (especially when compared
with other sweeteners) do not seem to be well understood by members of the
sugar industry unless an "old timer" is available as advisor. I draw this
conclusion based on some rather disappointing experiences over the past
twelve years.
Where should one go for advice, instruction and research on sweeteners
in the foods fields? Quite frankly, one of the best sources I have found is
a chap named Tom Neuhaus who writes a syndicated food column for newspapers.
Further, a logical source would appear to be the many universities in this
country, although my experience with these researchers has been
disappointing. My advice is to look for a commercial laboratory for
reliable research in the food field.
3) Remember the consumer. We generated a wealth of valuable, sound
data throughout the twelve years of our research program. In doing so we
convinced many communities of the safe role of sugar in the diet—medical
researchers, government agencies, food industry members, many news editors,
some consumerists—but not the consumer.
Some ten years ago the Association had a group of twelve part-time
women employees, all registered dieticians, spread throughout the country.
My personal opinion was and remains that that group of women did a marvelous
job of contacting the consumer directly and putting up a good defense for
sugar in the diet when necessary. Critical to the success of the ladies was
that the consumers not only understood what they said but believed them.
They were credible and understandable. Unfortunately, that part of our
program was almost totally abandoned.
Consumers are important because of their attitudes towards sugar, their
understanding of what sugar is and what it does as a food ingredient, their
changes in tastes and their perceptions of value that sometimes guide their
307
food choices—unfounded as some of those perceptions might be (the value of

high intensity sweeteners, for instance).
4) Maintain contact with anti-sugar advocates. We found that a good
way to defuse many anti-sugar people was to invite them to our technical
meetings, listen respectfully to their views and counter persuasively with
our contrary opinions.
5) Maintain quality control. Among several tasks at the Association, I
was in charge of quality control although the activity was not called by
that name. In fact, it was my assignment to see to it that the technical
information emanating from the office was accurate. From my experience in
various parts of the food industry I was quite aware of the necessity of
quality control and that quality controllers were not everyone's favorite
persons. As with any other product, information of quality caliber insures
reliability and credibility.
Where information and persuasion are concerned it takes years to
develop credibility. In a much shorter time credibility can be destroyed
and a lowering of quality control can stimulate this destruction.
Thus the precepts I offer to those continuing to research sugar's
relationship to health and nutrition are:
1) Consult with sugar industry members for alleged sugar-health
relationships.
2) Consult with commerical labs for research on food applications for
sugar and other sweeteners.
3) Remember the consumer.
4) Maintain contact with anti-sugar advocates.
5) Maintain quality control.
Now let me pass to some specific projects and approaches to problems I
believe would benefit the industry. All of the following at one time or
another were recommended to the attention of my former colleagues. Some
were a bit too premature and I am bringing them to the top of the idea pile.
Others were seeds planted not too long ago; perhaps they need a little
attention in order to germinate properly.
308
MEDICAL RESEARCH
Dental Caries
Dental caries is the one health problem to which sugar consumption
contributes significantly and, therefore, deserves continuing attention.
Caries is almost inevitably referred to as being multifactorial—more than a
single factor is involved in caries production.
Research on caries, I believe, has suffered from the lack of a
multidisciplinary approach. For instance, the three factors essential to
developing caries--host, bacteria and food--can be looked at as more than
the three ring intersect most often pictured. The three factors can be
considered three reactants of a biochemical process (caries production) and
shown as an equation, the rate at which the reaction takes place being
determined by the variations in the three reactants.
This approach allows host, bacterial and food variables to be neatly
cataloged and, at least to my mind, outlines explicitly their importance,
compared with keeping in mind the many variables that do exist.
One dental project I would like to see carried out is that designed to
test the effect of food texture on the rate of caries development. I
understand that some food companies have evaluated this relationship but
have not published the results. With experts in textural measurement now
available, the project is most feasible.
There is another small project I have often thought worthy of carrying
out in the caries area. This one concerns the use of confectioners* sugar
as the standard for developing caries in animal studies. One of our
colleagues, Dr. Basil Bibby, has more than once suggested that the high rate
of caries production in rats fed confectioners' sugar might be more due to
the starch content of the sugar than to the sugar itself. (In the United
States, starch is added to confectioners' sugar, or powdered sugar, as a
moisture absorbing agent).
Obesity
There are a couple of projects on obesity I believe important enough to
the sugar industry that they should be carried out. In the past the
Association sponsored projects that have yielded results indicating that a
high carbohydrate-low calorie diet will maintain a higher metabolic rate
than a high protein-low calorie diet in the short term. I see the need for
a project carried out long enough to show whether or not the high
carbohydrate-low calorie diet eventually results in a significant weight
loss.
Perhaps even more important to the industry, especially from an
economic point of view, is to show, again at long term, whether or not high
intensity sweeteners are effective in reducing calorie intake and,
subsequently, loss of weight.
It should also be of interest to the industry to realize that even if
the high intensity sweeteners are shown not to be efficacious in this
respect, the high intensity sweeteners will still be marketable products
because the public perceives them to be helpful. This philosophy is evident
in the position paper on saccharin published by the American Medical
Association.
Behavior
Behavior takes many forms—hyperactivity, criminality, addiction,
hypersensitivity, bulimia, anorexia, taste and appetite to name a few. The
association of sugar with many of these conditions has been alleged. Most
allegations appear to be well contained. Taste and appetite might bear some
investigation not necessarily for sugar itself but for food products
containing sugar as compared with counterparts based on alternate
sweeteners.
I was involved with some of the latter types of research and would
suggest recognition that the tastes of consumers change. Further, shopping
for top notch investigators in the field is a must and my considered advice
in this respect is to settle with commercial laboratories such as a Monell
Chemical Senses Center.
Diabetes
Sugar is well divorced from the stigma of being responsible for the
development of diabetes. Sugar, when taken with a meal, has even been shown
to be acceptable in the diets of diabetics. It would therefore be easy to
say that there is no profit to the industry in supporting research on sugar
and diabetes.
I do, however, have a couple of favorite projects of long standing I
would like to recommend to your attention. First, a call for a
310
multidisciplinary approach to diabetes research seems advisable. You've all

heard of the glycémie index of foods (or the effect of a given food on blood
glucose level as compared with glucose). Starches from different sources
have given inconsistent glycémie indices in the hands of different medical
researchers.
For many years I tried to interest diabetologists in adding to their
staffs a starch chemist. Such an expert I thought could bring to bear on
the glycémie index problems the knowlege of the physical characteristics of
the various starches and the effects thereof on enzymic susceptibility. I
was not a good salesman.
This year, however, I noticed a publication relating the effect of
degree of rétrogradation of potato starch, depending on heat treatment, on
the starch's glycémie index.
A review of what has been published on starches vs. glycémie index
might be worthwhile in the light of this paper.
A second area for diabetes research I have found is the measure of
glycoslyated hemoglobin of diabetics. I understand that this measure
indicates the degree of progress of the disease. One wonders whether there
is a difference in degree of glycosylation depending on what reducing sugar
is ingested. Is it only glucose that combines with hemoglobin (or other
blood proteins) or do other reducing carbohydrates, such as fructose,
combine to a different degree?
Finally, we have been told for many years that the refining of sugar
removes chromium from the sweetener as it occurs naturally in cane. Sugar
without chromium is said to add stress to the pancreas. Two items are
bothersome about this claim. One, the glucose tolerance factor (GTF) that
apparently aids in the metabolism of glucose, is said to contain chromium.
But the GTF is an elusive factor to the point where some of our
diabetologist contacts have confirmed an inability to isolate the factor by
published procedures. That procedure should be checked with care.
Further, I understand that until last year there was no accurate and
precise method for analyzing for small amounts of chromium. Indeed, the
assay for the presence of chromium seems to be a biological one and one
wonders what is being measured: chromium content or some other factor.
311
Here is a chance to produce some worthwhile research before the

chromium deficiency-sugar relationship becomes too driving an anti-sugar
force.
Heart Disease
About the only concern for the involvement of sugar with cardiovascular
heart disease (CHD) seems to be the sometimes relationship between sugar
intake and rise of blood triglycérides and/or cholesterol. As I have in the
past, I would urge someone in the industry to comb the literature and list
the qualifications or conditions necessary for sugar to promote such
elevations. They are many.
As corollaries to this listing might be added the proper definition of
a "carbohydrate sensitive" person and perhaps some research to help in the
decision as to whether sugar acts as a disaccharide ("disaccharide effect")
under some conditions and as a mixture of glucose and fructose under other
conditions.
In considering the allegations that sugar may be involved in the
production or in the promotion of the production of both diabetes and CHD I
have always found it of great interest that much of the research aimed in
this direction appears to arise at the Nutrition Institute of the USDA.
Perhaps it would be appropriate to ask the question as to why an agency
thought to be devoted to the protection and oversight of our commodity,
sugar, tolerates research obviously directed towards finding something wrong
with the product.
FOOD TECHNOLOGY
A year or so before I retired, the industry decided to initiate
research in food technology with the express purpose of showing that in any
given application, sugar, as compared with other competitive sweeteners,
made the best, most acceptable food product.
I not only applauded the industry's decision to enter into this type of
research, I recommended it some ten years before. And I would have pursued
the sweetener in foods comparison technique at that time.
However, times change. So does expertise in food technology. So do
the tastes of the consumer. Today, I would emphasize, rather than a
comparison of acceptability of food products based on different sweeteners,
312
a search of processing advantages that sugar has over other sweeteners in

the manufacture of a food product. I would also look to the interaction of
the different sweeteners with other ingredients of a foodstuff. Remember,
some of the high intensity sweeteners (fructose and aspartame, for instance)
are much more reactive in a number of applications when compared to sugar.
NEW SWEETENERS
The search for new sweeteners as replacements for sugar will
undoubtedly continue for years. If I were monitoring this development as I
have in the past, I would keep an eye on the possibility that through
advances in biotechnology, sugar itself will be made less expensively, than
isolating it from sugarcane and sugarbeets.
Let me repeat here as an ending that I have given all of this advice in
the past to various members of the industry. This is a summary of some of
my thoughts that I would like to leave with you as a parting gift although I
realize that my offerings may be looked at with much the same attitude as
that with which Laertes heard those of his father. At least I feel I leave
with you a worthwhile checklist for consideration by the industry for
research sponsorship.
313
Chapter 20
CARIBBEAN RUM: ITS MANUFACTURE AND QUALITY
R. HARRIS AND D . H. WEST
SUMMARY
The historical development of the sugarcane and rum industries of the

Caribbean are outlined showing the early importance of these industries to
the development of the area.
All aspects of the technical production of rum are discussed including
feedstock options, fermentation techniques, distillation practices and the
ageing and blending of rum as is practiced in each of the Caribbean
countries. The characteristic quality of light, heavy, blended and aged rum
is shown by 6LC analysis of a typical sample from each area.
The relationship between amino nitrogen and the production of n-propyl
alcohol is discussed and the results of an investigation that monitored the
levels of amino nitrogen in final molasses and sugarcane are given.
HISTORY
Sugarcane has been cultivated for approximately 2000 years. It

originated in Asia and spread to Spain in the early 15th century via North
Africa and Sicily. It was introduced to the Caribbean in the late 15th
century by Christopher Colombus. By the 16th century the Spanish and
Portuguese were producing sugar in the New World on a large scale. In
Hispaniola (Dominican Republic), Cuba and Puerto Rico the Spanish were
producing sugar and molasses and it is very likely that a distilled spirit
was being produced from fermented cane juice and molasses. The Portuguese
were producing sugar, molasses and a distilled spirit from fermented
molasses in the regions of Pernambuco and Bahia in Brazil.
In the 17th century the English, French and Dutch began to colonize the
smaller Caribbean Islands, with Barbados becoming the second British
Caribbean Colony in 1627. Experimentation with sugarcane in Barbados began
in the mid 1630*s with Dutch assistance brought in by the British from
Pernambuco which the Dutch then held. Soon after 1640 Barbados was
producing acceptable quality sugar and was distilling spirit from skimmings
of the copper taiches and from fermented molasses.
The origins of the word 'Rum* are subject to much speculation. Some
claim that it may have been derived from the Latin for sugar, saccharum.
Others claim origins from the Spanish word 'Ron' (» Rum) arguing that the
Spanish probably had distilleries in the Caribbean before the arrival of the
31Λ
English (réf. 1 ) . The strongest claim comes from Barbados (ref. 2) where
an anonymous 1650 description of the island states that:
"the chief fuddling they make in the Island is Rumbullion alias
Kill Devili, and this is made of sugar cane distilled - a hot,
hellish and terrible liquor."
Soon after Rumbullion was abbreviated to Rum as confirmed by a General Court

of Connecticut Order in 1654 to confiscate:
"whatsoever Barbados liquors, commonly called
rum, Kill Devili or the like."
In these early times Caribbean molasses, sugar and rum were traded in
Connecticut, New England and Virginia in exchange for salted fish, pork and
beef, dairy products, flour, lumber and livestock. Caribbean rum was
introduced into England in the late 17th century and had become well known
by the 18th century. Today Caribbean rum is sold in the USA, Canada, UK,
Europe, Australia and New Zealand, and is a major foreign exchange earner
for each of the producing countries. A list of the major producers and
their annual production capacity is given below (ref. 3 ) .
TABLE 1
Rum production.
Country Imperial gallons Liters of pure alcohol
Antigua 250,000 650,000

Bahamas 4,000,000 10,400,000
Barbados 3,500,000 9,100,000
Guyana 6,500,000 16,900,000
Jamaica 11,424,300 29,703,180
St. Lucia 200,000 520,000
St. Vincent 150,000 390,000
Trinidad & Tobago 6,600,000 17,160,000
TOTAL: 32,624,300 84,823,180
DEFINITION
The definition agreed to and accepted by the major producers is as

follows :
"Rum shall be the spirit obtained only by alcoholic fermentation
and distillation of molasses, syrups or cane sugar obtained during
the extraction of cane sugar or of sugarcane juice. Production
must be carried out in such a way that the product has the aroma
and flavor derived from the natural volatile elements contained
in the above materials or formed during the fermentation or
distillation process of the named raw materials.n (Ref. 4)
RAW MATERIALS
Molasses is the raw material used by the major producers. Small
producers in Dominica and Grenada and the French West Indies use cane juice
or mixtures of cane juice and molasses. Molasses is the preferred raw
material because it contains more sugar per unit volume, lower bacterial
loading, has a much longer shelf life than cane juice and is available year
round.
Molasses quality varies depending on the variety of sugarcane,
climatic conditions, fertilizing practices and processing and storage
conditions. All of these variables have a significant effect on both the
quality of the rum produced and the alcohol yield or fermentation
efficiency. The analysis of an average quality Caribbean molasses is shown
in Table 2.
TABLE 2
Average composition of Caribbean molasses.
Brix (hydrometer) 85.00° Bx

Total sugars (as invert) 52.00Z w.w.
Ash (sulphate) 8.00Z w.w.
Nitrogen 0.75Z w.w.
Gums 2.50Z w.w.
pH 5.5
This analysis may be compared with the criteria and ratios for grading
molasses for rum production developed by Arroyo and given in Table 3 (ref.
6).
316
TABLE 3
Typical molasses grades.
Criteria Good Fair Poor
Brix 87.6 85.4 84.2

Sucrose Z w/w 36.4 31.3 34.6
Reducing sugars Z w/w 19.6 20.0 13.5
Total sugars Z w/w 58.0 52.9 49.9
Ash Z w/w 7.3 9.4 11.6
Nitrogen X w/w 1.1 0.5 0.5
P z 0 5 X w/w 0.2 0.1 0.2
Gums X w/w 2.0 2.6 3.8
pH 5.5 5.7 5.3
Ratios
Total Sugars/ash 7.95 5.63 4.30

PaOs/Nitrogen 0.18 0.17 0.40
Gums/Total sugars 0.03 0.05 0.08
Alcohol/Total sugars X w/w. 46.5 44.0 39.0
Three criteria not shown above are unfermentable sugars expressed as a

percentage by weight, color expressed as International Units and
organoleptic appraisal.
Good quality molasses should have a low level of unfermentable sugars,
a color between 100,000-200,000 international units, and a sweet fruity
smell and taste. Normally molasses that has been subjected to high
processing and storage temperatures will not meet these criteria. The GLC
analyses of an 8 month old Barbados molasses that entered storage at 120° F
is shown in Table 4. Since the F/G ratio would have been approximately
unity prior to thermal decomposition, the loss of glucose is 6.4Z or 12.8Z
of the total sugar (as reducing sugars) originally present in the molasses.
TABLE 4
Analyses of 8 month old molasses.
X Sucrose 29.3Z
X Fructose 10.3Z
Z Glucose 3.9Z
F/G ratio 2.6:1
317
Further, Table 5 below shows sugar losses by thermal decomposition

under controlled conditions at 40 and 60° C, as reported by United Molasses
for samples of Trinidad and Jamaica molasses. The data is given for a 15
day period at 3 day intervals. This study indicated that whereas thermal
decomposition is negligible at 40° C, a 3Z loss is incurred in 15 days at
60° C.
TABLE 5
Change in sugar content of molasses stored at high temperature, as a percent
of the original.
Time elapsed Trinidad molasses Jamaican molasses

40° C 60° C 40° C 60° C
3 days nil -0.83 -0.07 -1.13

6 days +0.04 -1.12 -0.10 -1.84
9 days +0.16 -1.66 -0.09 -2.46
12 days +0.12 -1.97 -0.09 -2.98
15 days +0.16 -2.56 -0.08 -3.46
FEEDSTOCK COST COMPARISON
The cost and analysis of other raw materials available to Caribbean rum
producers are compared with blackstrap molasses in Table 6.
TABLE 6
Comparison of raw materials.
Z Available Relative
X Solids Total Sugars US $ Value/
Feedstock (Brix) (as Invert) ton
Blackstrap molasses 88 55 50.00*

High test molasses 80 65 59.00
Raw sugar 99.5 103 93.50
Sugarcane 17 14 12.50
*Assumed value of US $50.00/ton blackstrap molasses.
Using typical analyses for each feedstock and an assumed value of US

$50.00/ton for blackstrap molasses, the relative value of each feedstock is
calculated in terms of fermentable sugars (all values are in short tons).
318
The table shows that at US $50.00/ton for blackstrap molasses the

relative value of raw sugar is 4.25c/lb which is only a fraction of the
current cost of sugar production in the Caribbean.
Similarly, the relative value of sugarcane is approximately
US$9.00/ton, after deducting processing cost. The current cost of
production of sugarcane, although highly dependent on the relative value of
local currency to the US dollar, is in the range of US$35.00 to US$85.00.
Consequently sugarcane is only used as a feedstock under special
circumstances.
The cost of high test molasses is similar to that given for raw sugar
and consequently is not considered to be a viable option as a fermentation
feedstock.
Figure 1 is a ready reference graph showing the relative values of
molasses, cane and raw sugar.
>00f
Φ IS«
<Ä C
v» o
O i.
le «»
o o
50+
IT" 20 To so î
~sT To" 5δ" 2bo 2$0 2
Ί8Β" "ïJ5 TiïT "ΊΒ5Γ MÔ0 3
U.S. d o l l o r / ton
Fig. 1. Graph of relative feedstock values (1) » sugarcane; (2)

high test molasses; (3) = raw sugars.
MOLASSES TREATMENT
Calcium salts present in the molasses precipitate out as scale in the
distillation units, reducing performance and efficiency. To solve this
problem, chemical clarification can be carried out with alumina and calcium
phosphate or by adding sulphuric acid with subsequent heating followed by
cooling and settling or centrifuging. Unfortunately these methods have very
high capital and operating costs and do incur a significant loss of sugar
(ref. 5).
Bacteria present in the molasses can be destroyed by pasteurization of
diluted molasses in continuous high temperature short time pasteurizers.
However, temperature control must be precise to avoid the destruction of
naturally occurring nutrients or cause molasses degradation via Maillard and
caramelization reactions. Here again capital and operating costs are
considerable (ref. 7).
In general it has been found that molasses pre-treatment promises a lot
more than it actually delivers; and by and large the major producers have
avoided this procedure.
In the Caribbean, scaling problems are controlled by injecting an
approved food grade complexing agent into the feed to the analyzer column.
Although the precise reaction of these scale inhibitors is not fully
understood it is clear that non-crystallizable complexes are formed and
deposited as a sludge which can be washed away with high pressure water.
Similarly the bacterial problem is controlled by the use of a
combination of sterilization and antiseptics, such as ammonium bifluoride in
the yeast propagation stage. This control is necessary to ensure that a
high yeast concentration is developed during the fermentation stage, so that
its predominance will inhibit the growth of other microorganisms which could
be detrimental to the quality and yield of the rum produced. It is also
important to ensure that a sufficiently high concentration of alcohol is
produced in the fermenters to act as a natural antiseptic to bacterial
action.
All of the major rum producers in the Caribbean mix molasses and water
in some type of mechanical automatic or semi-automatic mixing/blending unit
to produce a molasses wash of approximately 17-24° Brix corresponding to
10-15Z total sugar. At this point the pH of the wash is adjusted between
4.8 and 5.0 with sulphuric acid and liquid nutrients, e.g. ammonium sulphate
320
and ammonium phosphate, are added via independent metering pumps. To

illustrate the importance of mixing, I'Anson (ref. 5) reported a 15Z
increase in yield after installing mechanical mixing in a plant where mixing
was previously done by hand.
FERMENTATION
The choice of yeast species and strain has a most important influence
on the quality and the amount of rum produced during fermentation. The
yeast selected should meet certain requirements:
1) maximize conversion of sugar to alcohol,
2) be temperature and alcohol tolerant,
3) produce desirable organic by-products known collectively as
congeners or congenerics, such as acids, aldehydes, esters
and higher alcohols in the desired amounts and proportions,
4) ferment at a suitable rate to reduce the risk of infection by
undesirable bacteria.
In the production of light rums, the main aims are for a low congeneric
level and high alcohol yield, criteria best met by using the quick
fermenting, budding type Saccharomyces yeasts. On the other hand, the
fission type Schizosaccharomyces strains are more suitable for the
production of the heavier types of rum where the primary objective is
congener formation rather than alcohol yield (ref. 8 ) .
As a result of international consumer demand for lighter rum, the
budding type yeasts predominate fermentation flora in Caribbean
distilleries, but fission type yeasts are still used in the production of
the highly flavored heavy rums of Jamaica, Guyana and the French islands.
Once the yeast strain has been selected, propagation from pure yeast is
achieved either in a tropically adapted yeast culture plant or by the use of
semi-aerobic "bub" stages commencing with a sterilized wash in the
laboratory and finishing in stainless steel "bub tanks" at the distillery.
The contents of these tanks are used to inoculate the main fermenters.
For a rapid rate of fermentation the bub, or inoculant, is first pumped to
the fermentation tank which is then filled with molasses wash. Using this
technique, fermentation will have reached optimum activity by the time the
fermenter is filled. For a slower fermentation the inoculant is added after
321
the fermenter has been filled with wash. Slow fermentations are used for
heavy rums.
A Caribbean rum distillery utilizing a rapid fermentation usually
requires three bub stages in the laboratory and two in the distillery before
the final fermentation. Each bub stage takes 18 to 24 hours and the final
fermentation takes 20 to 26 hours. The quantity of inoculant used is about
12Z of the final wash and contains 2-3Z yeast mass.
Advantageous or spontaneous fermentation is still practiced in a few
Caribbean rum distilleries. The fermenters or mixing tanks are simply
filled and left to ferment relying on airborn yeasts, naturally occurring
yeasts in the molasses, or yeasts from previous fermentations attached to
the fermenter walls. These fermentations can take from two to four days.
Most literature on fermentation insists that for maximum alcohol
yields, fermentation temperatures must be held at 31-32° C. Unfortunately
these temperatures are impossible to achieve without expensive refrigeration
in most tropical distilleries. However, most yeast strains used in the
Caribbean are temperature tolerant, giving their best results at 34-36° C
with some capable of producing alcohol at 40° C. These tropical yeasts are
very sluggish alcohol producers at lower temperatures.
Plate-type heat exchangers are used extensively for cooling because
their high coefficient of heat transfer allows adequate heat removal with
modest equipment at temperature differentials of only 5° C. The low
temperature differential is important since the temperature of the cooling
water is unlikely to be less than 28° C and is frequently considerably
higher.
At the end of fermentation the alcoholic concentration in the
fermenters will be between 5 to 10Z ethyl alcohol by volume, depending on
the type of yeast, the original sugar concentration in the fermenter and the
fermentation temperature.
Once fermentation is complete, the vessels are allowed to rest for 6 to
12 hours, allowing suspended solids and dead yeast cells to settle out on
the fermenter bottoms prior to distillation.
Batch fermentations are preferred by the Caribbean rum producers
because this offers some flexibility and opportunity for the tailor-making
of specific types of rum. A few semi-continuous systems are in operation,
but no full scale continuous fermentation systems are used in the Caribbean.
322
A peculiarity to Jamaica and the French Islands is the practice of

using "dunder" in the production of their high ester heavy rums (refs. 5,
7). Dunder is the lees of previous distillations which have been allowed to
age and ripen by bacterial action. The matured dunder is added to the
molasses wash prior to fermentation. Symbiotic fermentation by the yeast
and bacteria produce the precursors for the formation of high ester
concentrations in subsequent distillation in pot stills. This symbiotic
fermentation has been shown by Arroyo to reduce the time required to
complete the fermentation for heavy rums (ref. 8 ) .
DISTILLATION
Caribbean rum distillers utilize both batch and continuous distillation
processes. Batch distillation is carried out in pot stills of varying
sizes, design, materials or construction and design. Continuous
distillation is carried out in two, three or four columns, in stills
constructed of stainless steel and copper.
Barbados, Guyana and Jamaica produce rum in both pot stills and
continuous stills, while the Bahamas, Trinidad and Tobago, Antigua, St.
Lucia and St. Vincent utilize continuous stills only.
Pot Stills
Pot stills in use in Jamaica and Guyana consist essentially of a wash
still, a low wines retort, a high wines retort, a condensor and spirit
receiver, depicted in Figure 2. The fermented wash is pumped to the wash
still and heated by either direct firing or a steam coil. The first
distillate, called low wines, is distilled in the low wines retort to
produce a second distillate, called high wines, which is distilled in the
high wines retort to produce the finished product rum. Heads and tails are
collected as fractions at the start and finish of the distillation
respectively, while strong rum is collected during the middle period of
distillation known as the "middle run." The heads and tails are re-used for
the next distillation in the high wines and low wines retorts respectively.
The desired strength of the rum is about 80Z v.v. alcohol; to achieve this,
it is sometimes necessary to install a small rectifier on top of the high
wines retort.
Fig. 2. Typical Caribbean pot still (from ref. 7).
Pot stills in Jamaica are made entirely of copper while in Guyana the
pots are made of local hardwoods, greenheart and wallaba, with swan necks,
pipework and condensors made of copper (ref. 5). Rum produced in Guyana and
Jamaica are heavy, highly flavored rums. The Jamaica rums have a very rich
fruity aroma as a result of the use of dunder in the fermentation.
A lighter pot still rum is produced in Barbados. This rum is distilled
from a mixture of fermented wash and a medium type continuous still rum.
The distillation is carried out in a copper pot still, fitted with a
rectifier, to produce a product at a strength of about 602 v/v alcohol.
This product is redistilled in another copper pot still fitted with a
rectifier to produce double distilled pot still rum at a strength of about
80Z v/v alcohol.
Continous Stills
Two column continous stills were introduced to the Caribbean in the
late 19th Century. Today the bulk of the rum produced in the Caribbean is
of the continous still type augmented with small amounts of pot still rum to
produce individual brands of international repute. Two column stills have
for the most part been replaced by three and four column stills with the
32^
capability of producing a wide range of spirits, from heavy continuous right

through to neutral spirits. A two column still consists essentially of an
analyzer column and a rectifier column, depicted in Figure 3. An aldehyde
column introduced between the analyzer and rectifier effectively constitutes
a three column still (Figure A) while a hydroselector column added between
the aldehyde column and the rectifier produces a four column still (Figure
5). The analyzer column is made of stainless steel and contains
approximately 20 widely spaced distilling plates specifically designed to
cope with high concentrations of suspended and dissolved solids.
Fig. 3. Two column continuous still for heavy to medium type rums.
(From APV Mitchell Brochure, 30 Thornton Road, Thornton Heath,
Surrey CR4 6XT, England.)
The aldehyde and hydroselector columns are made of either stainless

steel or copper and contain approximately 20 plates and 40 plates
respectively. The rectifier is made of copper and contains approximately 68
plates. The plates in these columns are closely spaced and are highly
efficient units for refining clean alcohol water mixtures. These columns
are heated by steam, the choice of direct or indirect methods being dictated
by the quality of spirit required, the quality of the steam available and
the cost of fuel and water used to generate steam. Some highly energy
efficient arrangements exist where vapor may be taken from one column to
drive another. Reflux is supplied to the top of these columns from
stainless steel or copper condensers. It is inadvisable to supply high
strength reflux to the top of the analyzer column because this encourages
the deposit of calcium sulphate scale which reduces performance and
increases down-time for manual cleaning.
Fig. 4. Three column continuous still for medium/light type rum.

(From APV Mitchell Brochure).
Pre-heated fresh wash is passed through a degasser to remove carbon

dioxide and undesirable odors, and then to the analyzer column, where
stripping of the wash takes place, producing an overhead product at a
strength of 50 to 60Z volume alcohol. A portion of this is passed to the
aldehyde column for further concentration and removal of a small overhead
"heads" fraction. In the case of a three column still, the bottom from the
aldehyde column is passed to the rectifier when it joins the remaining vapor
from the top of the analyser to be rectified to a strength of 94 to 96Z v/v
alcohol. Rum at this strength is withdrawn from plate #52-62. A small heads
fraction is also removed from the rectifier condenser. Higher alcohols
accumulate in a band in the region above the rectifier feed plate, the exact
plate being determined by the reflux ratio at which the column is operated.
The band of accumulation is narrowest at high reflux ratios and widest at
lower reflux ratios, hence more higher alcohols will find their way into the
326
I SPENT WXSH TO EFFLUENT TREATMENT

-®
Fig. 5. Four column continuous still for extra light rum and neutral
spirits. (From APV Mitchell Brochure.)
products at low reflux ratios. A side cut high in higher alcohols is

therefore drawn off from the correct plate and passed to a decanter, where
it is washed continuously with water to remove the insoluble amyl alcohols
and recover alcohol. The recovered alcohol and water is recirculated
through the still.
On occasions there may be an accumulation of water-soluble n-propyl
alcohol, and this may necessitate another side cut higher up the column.
In the case of a four column still, the analyzer and aldehyde column
work as one unit to produce a pre-concentrated spirit free of organic acids,
at approximately 801 v/v alcohol, allowing for the removal of a small heads
fraction. The pre-concentrated spirit is transferred to the hydroselector
column where hot soft water is added at the top and extractive distillation
at low alcoholic concentration of 8 to 122 v/v alcohol takes place. Under
these conditions the majority of aldehydes, esters and higher alcohols are
more volatile than alcohol and hence distill up the column and are removed
in a moderate heads fraction. Clean spirit at 8 to 12Z w/w alcohol is
transferred to the rectifier to be concentrated to 95.5Z to 96.5Z w/w
alcohol. In the rectifier column a heads fraction is removed and
327
recirculated through the still and a side cut is provided for small amounts
of higher alcohols that may have escaped removal in the hydroselector. Most
four column still installations are equipped with suitable valves, blanks
and piping, to allow the unit to operate as either a two or three or four
column still. This flexibility in the configuration of the still allows the
distiller an opportunity to manufacture the range from heavy continuous
still rums right through to neutral spirits. The Bahamas, Barbados, Guyana,
Jamaica and Trinidad and Tobago distilleries all have this capability.
MATURATION
Despite much research and experimental work the only reliable method
for maturing is by ageing in white oak barrels. In the Caribbean, once-used
American whiskey barrels are common for maturing rum.
Barbados is peculiar in using both charred and decharred barrels while
all of the other major producers use charred barrels exclusively.
Pot still rum is stored in barrels directly from the still at 80Z w/w
alcohol. Continuous still rum is reduced in strength from 95Z w/w alcohol
to 80Z w/w alcohol, and in some cases as low as 63Z w/w alcohol, and then
put into barrels.
Heavy pot still rums require several years maturation while light
continuous still rums may need only 12 to 18 months. Jamaica and Guyana pot
still rums require 5 to 6 years maturation in most cases and may be as much
as 10 to 12 years for the more highly flavored types (ref. 5).
Barbados pot still rum is quite acceptable after 3 years maturation.
The average maturation period for light and medium continuous still rums in
the Caribbean is approximately 2 years.
The changes that take place in the rum during ageing result from
chemical reactions between the organic components in the distillate and
reactions between the distillate and extracts from the barrel wood. A wooden
barrel is a porous container and therefore allows the passage of vapors out
of the barrel and of air into the barrel by diffusion. The chemical
reactions that take place are mainly oxidations and condensations that
produce acetaldehyde from ethanol, acetic acid from acetaldehyde, and ethyl
acetate from ethanol and acetic acid. Lignin, aromatic aldehydes and sugars
are extracted from the hemicellulose of the wood. Tannins and color are
also extracted.
328
It appears that a charred barrel is more active than an uncharred

barrel and produces more congeners and a darker rum.
Higher alcohols take little part in any of these reactions and their
levels remain essentially the same as is the original distillate.
Aldehydes and acids increase rapidly during the first year but further
storage produces little change. Esters increase at a constant rate
throughout the period of maturation, and may be used to indicate the period
of maturation if the analysis of the raw rum and the conditions of storage
are known.
Quality of the aged rum is directly influenced by maturation time,
temperature, the strength of the rum, ratio of contact area to
volume of the barrel, age of the barrel and whether the barrel is charred or
decharred.
The rate of ageing is dependent on the contact area and the
temperature. It is reported that the rate of ageing at 35° C is
approximately twice that at 25° C; unfortunately the price paid for this is
an increase in the annual loss rate from 5X to 10Z (ref. 7).
After maturation the various types of rum are blended, married, reduced
in strength with high quality demineralized water, color-adjusted with
caramel, filtered and finally bottled for delivery world wide.
QUALITY
Table 7 shows the major chemical differences of pot still and
continuous still rums from Barbados, Guyana and Jamaica. Table 8 shows the
chemical differences between popular aged blended rums from Barbados,
Guyana, Jamaica and Trinidad and Tobago.
Table 8 reveals the current trends towards lighter rums in the
Caribbean. Among the more popular brands, the Barbados rums are the
lightest in character and the Guyana rums are the heaviest. The data also
show that the Jamaica rums have the highest ester contents and the Guyana
rums the highest alcohols.
With respect to quality, the importance of organoleptic assessment
must never be overlooked since some of the components which are analytically
grouped may contribute good or bad characteristics to the rum.
329
TABLE 7
Comparison of pot still and continuous still rums from various Caribbean
countries.*
Continuous still rums

Pot still
Country Congeners rums Medium Very Light
Barbados Aldehydes 45.0 4.0 2.0

Esters 13.0 4.0 2.0
Higher alcohols 93.0 15.0 5.0
Guyana Aldehydes 18.1 4.0 0.4

Esters 24.3 9.5 1.1
Jamaica Aldehydes 16.0 32.1 0.4

Esters 120.0 49.0 4.0
* Samples are current distillation ex still; results are quoted in mg/100 ml

absolute alcohol from gas Chromatographie analysis.
Amino nitrogen and n-propyl alcohol

In Barbados, molasses is delivered to the distilleries from the sugar
factories during the crop season, which lasts from February to June. From
early March, 1985, it was noticed that during fermentation there was
excessive foaming, shorter fermentation times, higher fermentation
temperatures, and that the level of n-propyl alcohol in the light rum had
increased.
Table 9 shows the result of analyses of amino nitrogen in 1985
molasses, 1984 molasses and a mixture of 1983 and 1984 molasses.
These results indicate that the concentration of amino nitrogen in
final molasses had increased steadily over the period under review.
Further, two analyses of the fermented wash prepared from the 1985 molasses
indicated excessively high levels of n-propyl alcohol as is shown in Table
10.
330
TABLE 8
The analysis of popular aged blended rums from various Caribbean countries.
Type of Rum
#1 #2
Country Congeners White Rum Dark rum Dark Rum
Barbados Aldehydes 8.0 12.0 15.0

Esters 18.0 20.0 14.0
Higher Alcohol 18.5 24.0 33.0
Guyana Aldehydes — 13.0 —

Esters — 12.0 —
Higher Alcohols — 141.0 —
Jamaica Aldehydes 12.0 29.0 —
Esters 21.0 38.0 —
Higher Alcohols 26.0 93.0 --
Trinidad &
Tobago Aldehydes 14.0 15.0 22.0
Esters 12.0 15.0 15.0
Higher Alcohols 61.0 59.0 33.0
* Results are quoted in mg/100 ml alcohol from gas Chromatographie analysis.
TABLE 9
Amino nitrogen concentration in molasses.
Amino nitrogen Amino nitrogen

Year ^g/ml)
1985 3158
1984 2369
1983 + 1984 1119
TABLE 10
Alcohol content of fermented wash, Barbados molasses 1984-1985.
Molasses Ex Molasses Ex 1985 crop

Higher alcohols* 1984 crop A B
n-propyl alcohol 54 111 85

iso-butyl alcohol 16 25 18
iso-amyl alcohol 106 198 219
*Results quoted in mg/100 ml absolute alcohol.

331
n-Propyl alcohol levels of 40 to 50 mg/100 ml absolute alcohol in the

fermented wash can be reduced to 6 to 8 mg/100 ml absolute alcohol by
distillation to 95 - 96Z v.v. alcohol in a three column still; that level is
quite acceptable. However, when the levels in the fermented wash are
approximately doubled, then unacceptably high levels occur in the light rum
product.
The levels of iso-butyl and iso-amyl alcohol encountered here can be
easily reduced to acceptable levels by distillation to 95 - 96Z v.v. alcohol
and counter-current washing with water, separation and decantating of the
properly selected side cut from the rectifying column. Iso-butyl and
iso-amyl alcohol can be completely removed by low strength extractive
distillation in the hydroselector column of a four column still.
Unfortunately n-propyl alcohol is more difficult to remove because it is
soluble in water and hence does not separate out in a decanter. Further,
even at the lowest alcohol concentration encountered in the hydroselector
column, the volatility of n-propyl alcohol is only slightly higher than that
of ethyl alcohol (ref. 10) and consequently separation is very tedious.
Since n-propyl alcohol is so difficult to remove during distillation,
it is very important that its concentration in the fermented wash be kept
within a tolerable range.
During the first 3 months of the 1986 sugar crop, analyses of the
molasses for amino nitrogen and analyses of the fermented wash for n-propyl
alcohol were again performed and these results are presented in Table 11.
TABLE 11
Amino nitrogen and n-propyl alcohol in 1986 crop molasses.
Average total nitrogen Average n-propyl alcohol in

1st 3 months wash from 1st 3 months
1986 crop molasses 1986 crop molasses
Factory ^g/ml at 90° Bx) (mg/100 ml absolute alcohol)
Andrews 2933 52.3

Bulkeley 3623 65.3
Carrington 3399 57.0
Foursquare 2841 58.3
Haymans 3753 63.0
Portvale 3410 68.7
332
The data have also been plotted on Figure 6, giving the linear
regression equation for the 6 pairs of analyses. The correlation
coefficient for the regression equation is 0.65 which indicates a weak
relationship; a larger number of samples would be necessary before any
definite conclusions could be stated. Nevertheless, the data do suggest
that if the amino nitrogen level in molasses was reduced to the low 2000's,
as in the early 1980*s, then it is highly probable that n-propyl alcohol in
the fermented wash would not exceed 40-50 mg/100 ml of absolute alcohol. As
stated previously a 3 column still can reduce this level of n-propyl alcohol
to 6-8 mg/100 ml of absolute alcohol. In addition to the above, experiments
were also conducted to determine if the use of inorganic nitrogen, in the
form of ammonium sulphate, would reduce the level of n-propyl alcohol in the
fermented wash. The results of this study were very similar to those of
Parfait and Jouret (ref. 9) who found a reduction in the level of iso-amyl
alcohol, but an increase in the level of n-propyl alcohol.
According to Ayräpää (ref. 10, 14), the formation of n-propyl alcohol
seems to obey no clear rules, but with limited concentrations of
nitrogeneous compounds, it increases with the content of nitrogen. He also
found that at higher nitrogen concentrations, large amounts of n-propyl
alcohol were produced although the actual quantity was largely independent
of the concentration of nitrogen present.
Amino nitrogen in Barbados molasses

Over the past seven years, Barbados Sugar Industry Limited (BSIL) has
monitored the level of amino nitrogen in the molasses produced at each
factory on a weekly basis. The data have been normalized at 90° Brix and
are shown in Table 12. Amino nitrogen was determined by the ninhydrin
colorimetrie method and the data reported as μg/ml. Analyses were carried
out on the composite sample representative of the factories' total annual
production and consequently does not necessarily represent the molasses used
by the distillery.
333
o
X
o
Ü
© ö «I
1 JÎ
< - 581
M Y = 0 . 0 l 8 7 3 x + 25.08
«- —
* 6 R = 8.6558
z <o
^ soi
TBÖ5 JoT56 52Ô5 5sT3ö 3δόδ »00
Toiol omino ni trogen In mo losses (jug/ml ot 9 8 ° Bx>
Fig. 6. Relationship of n-propyl alcohol in fermented wash and total

amino nitrogen in molasses.
TABLE 12
Total amino nitrogen content of Barbados' blackstrap molasses.*
Factory 1980 1981 1982 1983 1984 1985 1986 1987
Andrews 2707 2246 2112 2934 2968 3640 3063 3325

Bulkeley 2677 2257 2124 2755 3159 3637 3517 3567
Carrington 2398 2152 1798 2736 3056 3313 3434 3567
Foursquare 2121 2058 1949 2670 3681 3265 2922 3439
Haymans 2081 1658 2134 2927 3589 3992 3693 3140
Portvale — — 1261** 2236 2981 3257 3363 3155
Average
(not
weighted) 2397 2074 2023 2710 3239 3517 3332 3366
* Results are reported as μg/ml.

** Not included in averages as Portvale*s production was only for a test
period at the end of crop.
33^
The data in Table 13 show an increase in the amino nitrogen content of

Barbados molasses from about 2200 μg/ml in the early 1980's to over 3300
μg/ml in the past 3 years. This represents a 50Z increase and to date we
are unable to give any satisfactory explanation for the increase.
The weekly value for amino nitrogen levels for each factory is shown in
Figure 7. The plot shows a weekly variation that evidently reflects cane
quality and indicates that as the cane matures there is a reduction in the
level of amino nitrogen. The actual reduction in amino nitrogen content was
from about 3600 μg/ml to about 3200 μg/ml which is hardly significant in
terms of the 2200 μg/ml that existed between 1980 and 1982.
The presence of amino nitrogen in sugar cane

To provide further information on amino nitrogen in sugarcane, two
major commercial varieties were monitored over a 45-day period during the
1986 harvest season.
Sampling. Two varieties of sugarcane, B-62163 which represents about
80Z of the cane harvested annually in Barbados and B-73382 a promising new
variety, were sampled at approximately monthly intervals during the first
half of the harvest period. Each sample of cane was subsampled to represent
5 specific areas of the cane stalk.
The leafy portion of the cane was severed at the natural break-point of
the cane. This weak point, which coincides with the growing point or apex
of the cane, is shown in Figure 8. The portions sampled were as follows:
Tl, the uppermost subsample, represents the leafy sheath material in
the area beyond the growing point.
T2 subsample represents the immature cane in the area of the growing
point together with the associated leaf sheaths.
T3 subsample represents the uppermost fully developed portion of the
cane and is taken at the internode immediately below the attachment of the
lowest green sheath.
M subsample represents the cane at the midpoint of the stalk and the
section is taken at the internode halfway along the length of the mature
stalk.
B subsample represents the most mature portion of the cane and the
section taken is the lowest fully developed internode.
335
o I \
<s> I
Λ
/\
il·
<έ$
.*/
v/
or/
A
?
V V
9 %
\>. # I
Σ* %_%
* * - :
ν^
22
6
V
7 g 9 10 12 ΥΓ m
W««k of Crop
Fig. 7. Amino nitrogen content of final molasses (90° Brix)—weekly

composites for individual factories in Barbados during 1986 crop.
On each occasion the cane portions represented by the subsamples above

were peeled to remove the rind and leafy material and analyzed as follows:
TR, the rind from subsamples Tl, T2, and T3.
MR, the rind from subsample M.
BR, the rind from subsample B.
,4Jß^M Ji£*2£j£i
Fig. 8. Diagram of cane stalk showing portions sampled for amino nitrogen.
336
Juice extraction
The subsamples of cane and rind were prepared by chopping and dry
disi~Lwto_- .«H; the juice was extracted in a hydraulic press at 5000 psig.
Juice analysis. The extracted juice was analyzed for total soluble
solids by refractometric brix, X reducing sugars by Lane & Eynon method, and
amino nitrogen by the ninhydrin reaction.
Results. The results of these analyses are summarized in Tables 13, 14
and 15, which give the average of the analyses for each set of
determinations.
These analyses show the expected increase in X Brix and decrease in X
reducing sugars over the period of maturity together with the profile of the
components along the length of the cane stalk and in the rind.
The data on amino nitrogen clearly indicate a higher level of amino
nitrogen in both the Tl samples and the rind samples. The Tl samples are of
the order of 3-5 times higher in amino nitrogen than is the middle portion
of the cane. The rind samples also are 3-7 times higher in amino nitrogen
than the whole stalk.
TABLE 13
Analysis of juice Brix extracted from portions of cane stalk.
Percentage (X) Brix

Days
Variety Date elapsed *T1 T2 T3 M B TR MR BR
B--62163 02/07 0 9,.47 7 .19 13,.86 21..01 21,.93 10,.73 19,.41 24,.50
03/05 26 11,.67 11 .35 18,.46 22,.66 22,.60 13,.47 20..19 22,.56
03/24 45 11,.99 8 .53 17,.69 23,.32 23,.09 18,.08 25..47 25,.50
B--73382 02/07 0 9,.38 7 .57 16,.80 20,.27 21,.65 13,.57 20..27 24,.43
03/05 26 10,.34 12,.21 19,.48 21,.56 24,.37 14,.52 20..72 25..20
C3/24 45 14,.69 14,.65 23,.27 23,.70 25,.13 16,.72 22,.96 26,.27
*T1 - Cane material beyond the natural break point

T2 - Immature cane in the area for the break point
T3 - Cane below most mature leaf sheath attachment
M - Mid peint of cane stalk
B - Bottom of cane stalk
TR - Rind from leafy material from Portions Tl, T2 and T3
Br - Rind from bottom of cane.
33?
TABLE 14
Analysis of reducing sugars in juice extracted from portions of cane stalk.
Percentage (Z) reducing sugars

Days
Variety Date Elapsed *T1 T2 T3 M B TR MR BR
B-62163 02/07 0 1.39 2.31 2.68 1.88 0.21 1.86 1.70 1.00
03/05 26 0.63 2.91 1.68 0.36 <0.10 1.12 0.49 0.14
03/24 45 0.96 2.21 1.47 0.21 <0.10 0.77 0.10 0.10
B-73382 02/07 0 0.50 1.44 1.21 0.14 <0.10 0.59 0.20

03/05 26 0.42 0.72 0.24 0.29 <0.10 0.55 0.32 0.13
03/24 45 0.90 0.53 <0.10 0.11 <0.10 0.74 0.68 0.25
* Cane material subsampling is described in Table 13.
TABLE 15
Analysis of amino nitrogen in juice extracted from portions of cane stalk.
Percentage (Z) amino nitrogen

Days
Variety Date Elapsed *T1 T2 T3 M B TR MR BR
B-62163 02/07 0 0.32 0.07 0.08 0.12 0.16 0.34 0.45 0.98
03/05 26 0.30 0.06 0.08 0.06 0.06 0.39 0.65 0.80
03/24 45 0.30 0.12 0.08 0.06 0.08 0.47 0.54 0.79
B-73382 02/07 0 0.32 0.07 0.05 0.18 0.67 0.50 0.75 1.16
03/05 26 0.27 0.16 0.18 0.09 0.31 0.48 0.40 0.79
03/24 45 0.44 0.65 0.19 0.10 0.37 0.57 0.58 1.30
* Cane material subsampling is described in Table 13.
Quantitatively the contribution of amino nitrogen by the Tl sample and

the rind samples is considerably less than that implied by the analyses,
since the total quantity of Tl and rind material is unlikely to exceed 20Z
of the total cane stalk except in cases of inadequate removal of tops during
harvesting.
However, since rum manufacturers are primarily concerned with
maintaining the content below a critical value, then it is possible that
current methods of harvesting and juice extraction in Barbados' factories
could possibly have contributed to a higher than acceptable level of amino
338
nitrogen in the molasses. In particular the recent introduction of cane

shredders has intensified cane preparation, increasing juice extraction by
up to 2%. This additional juice has, in the main part, been obtained from
the rind portion of the cane as a result of the disintegration of the hard
fibrous rind by the shredders. The hypothesis that increased extraction
increases the level of nonsugars in juice is supported by Fort and McKaig
(ref. 11) who showed that about one third more nitrogen is in the whole mill
juice than in crusher juice. The same effect has been demonstrated by
Geerligs (ref. 12), who showed that the percentage of nitrogenepus nonsugar
increases in second and third mill juices indicating the considerable
influence of extraction on the level of nonsugars in the juice to be
processed.
The recent introduction of mechanical harvesters and the inevitable
decline in the control of manual cutting operations have also led to an
increase in the percentage of green leafy material processed.
Although the effects of increased juice extraction and the processing
of increased quantities of leafy material are difficult to measure, it would
be safe to assume that the trend has been towards the extraction of
additional quantities of amino nitrogen from leaf material during recent
years. However since comparable data on the amino nitrogen content of
Barbados' molasses is not available for the years prior to the introduction
of mechanical harvesting and cane shredders, it is impossible to reach a
definitive conclusion.
No attempt has been made to identify nor quantify the individual amino
nitrogen compounds present, which are known to have a considerable effect on
the formation of n-propyl alcohol during fermentation.
The levels of amino nitrogen in various portions of the cane stalk
indicate that an increase in the extraneous matter content of the harvested
cane and increased juice extraction will increase the amino nitrogen content
of the molasses. Further work should be carried out (ref. 13) to monitor
the effect of fertilizer applications on amino nitrogen (ref. 13).
REFERENCES
1 D. W. Clutam, Flavour Industry 5, (1974) 286.
2 P. F. Campbell, The Story of Barbados Rum. Barbados Rum Book, 1985.
3 Private Communication, West India Rum and Spirit Producers Association.
4 Draft; Barbadian and Jamaican Standards on Rum.
5 J.A.P. I'Anson, Rum manufacture, Process Biochem., July, 1971.
6 Rafael Arroyo, Manufacture of Rum, Sugar, December, 1941.
7 W.H. Kampen, Technology of the rum industry, Sugar y Azucar, 70 (7)
(1975) 36-43.
8 Rafael Arroyo, Manufacture of Rum, Sugar, January, 1942.
9 A. Parfait and C. Jouret, Formation of higher alcohols in rum, Ann.
Tech. Agricola, 24 (3-4) (1975) 421-436.
10 E.D. Unger and T. R. Coffey, Production of lightbodied rum by an
extraction distillation process.
11 Int. Sugar J., July, 1934, p. 272.
12 C. F. Geerligs and W. S. McKaig, Cane Sugar 2nd Edition, p. 131.
13 F.E. Hance, Proceedings Hawaiian Sugar Planters Association Report 37
(1932), p. 27.
14 T. Ayräpää, Biosynthetic formation of higher alcohols by yeast.
Dependence on the nitrogenous nutrient level of the medium, J. Inst.
Brewing 77 (1971) 266-275.
3^0
C h a p t e r 21
DESALTING OF MOLASSES BY COUNTER DIFFUSION IN THE FERMENTATION INDUSTRY
R. A. JOHNSON AND M. S. LEFEBVRE
The total annual world production of beet and cane molasses is now
approximately 37 million tonnes. Seventy per cent of this is utilized as a
carbohydrate supplement in the animal feed industry while most of the
remainder is used as a fermentation substrate to produce a wide range of
products. The latter are generally of considerably higher value than the
parent molasses, which has an average selling price of U.S. $70 per tonne.
An equivalent amount of raw sugar would cost the manufacturer U.S. $150.
The main products which are produced by fermentation, their values per tonne
of raw material and the microorganisms involved are listed in Table 1.
TABLE 1
Products of molasses fermentation.
Organism Main products Value*
Anaerobic fermentations
Yeast Ethanol 140
Rum 200
Glycerol 120
Bacteria Butanol/acetone 105

Lactic acid 400
L-lysine 2000
Dextran 750
Molds Itaconic acid 460
Aerobic fermentations
Yeast Feed yeast 95

Baker's yeast 270
Bacteria Vinegar (acetic acid) 175

Monosodium glutamate 570
Xanthan gum 475
Molds Citric acid 530
* Approximate value (US $) per tonne of molasses (80° Bx) fermented.

Notwithstanding the economic advantages of using molasses and its ready
availability, one disadvantage associated with its use is the presence of
components such as herbicides, thermal degradation products and various
inorganic salts, which are known to cause inhibition of microorganism growth
and activity (ref. 1 ) . Inorganic salts are usually present in very high
levels, and for Australian molasses are typically in the range of 12 to 16
per cent by weight. A breakdown of the main inorganic components of an
Australian molasses is given in Table 2 (taken in part from ref. 2).
TABLE 2
Main inorganic components of an Australian molasses.
Component Concentration (%)
Potassium 4.42
Sodium 0.09
Calcium 1.23
Magnesium 1.00
Chloride 3.20
Sulfate 2.42
Nitrate 0.18
Phosphate 0.04
Total 12.58
The most abundant component is potassium, which comprises 35 per cent

of the salt content, and is of particular significance in yeast fermentation
since it has been shown that inhibition commences at 400 ppm while the
optimum value is 120 ppm (ref. 3). These compare with a value of ca. 15,000
ppm for molasses diluted to 28° Bx for ethanol production. Accordingly,
Syrinx Research Institute has applied patented counter diffusion technology
to the removal of potassium and other inorganic ions from molasses to
enhance fermentation product yield. To date, this work has been limited to
ethanol and L-lysine production but the potential exists for improvements in
several other fermentations. This chapter is limited to a discussion of
ethanol production.
Counter diffusion technology

Counter diffusion is a process in which salts in the solution to be
treated (molasses) are selectively transferred across a semi-permeable
3^2
membrane into a water stream using their concentration gradients as the only
driving force. Counter diffusion membranes have a backbone of
cupro-ammonium reconstituted cellulose in the form of hollow fibers as shown
in Figure 1. Each fiber has an internal diameter of 200 microns and a wall
thickness of 11 microns. Several thousand of these fibres with a total
surface area of either 0.7 or 1.8 m 2 are sealed together at both ends by a
Polyurethane resin. Molasses is pumped through the hollow fibers while the
stripping water is pumped around the outside in counter current flow.
Molasses Out
Stripping
Water In
Polystyreneacrylonitrile
Tube
Molasses In
Fig. 1. Diagram of counter diffusion tube.
Unlike dialyis, where the selectivity between salts and sugars is low,
sugar transport in counter diffusion is minimal because of effects arising
from the immobilizaton of inorganic crystals on the membrane. Through an
effective reduction in pore size, the crystals increase the osmotic pressure
expressed by the components retained in the molasses stream and thereby
increase the counter flow of water through the membrane. The water flux,
which can be controlled by the shear rate of both the feed and water
streams, is adjusted to effectively oppose the transport of sugars in the
opposite direction. This also reduces the removal rates of potassium and
other ions, but the higher diffusion coefficients of these species allow
suitable removal accompanied by high separation factors. This is
illustrated for potassium and sugars in Table 3.
TABLE 3
Single-pass desalting of molasses by counter diffusion*
Potassium
Membrane Molasses (50° Bx) flow removal Sugar removal Separation
type rate (kg/m2-hr) (X) (X) factor**
CM-F3 (1.8 M 2 ) 8.3 23.9 1.03 23

16.3 17.7 0.12 147
21.4 14.8 0.05 276
* Stripping water flow rate = twice molasses flow rate.

**Separation factor = per cent potassium removal
per cent sugar removal
Effect of desalting on ethanol production
Work on ethanol production from partially desalted molasses has

embraced laboratory, pilot and industrial scale trials. By agreement with
the distillery concerned, results of the industrial trials cannot be
divulged at this time but the advantages of desalting can be adequately
demonstrated by reference to the results of the pilot studies (ref. 4).
These are shown in Table 4 and Figure 2.
In this work, clarified molasses (50°Bx) was partially desalted by
counter diffusion and then diluted to a fermentable sugars level of 150 g/1
(ca. 26° Bx) for fermentation. The potassium removal achieved in this case
was 35 percent. Clarified and untreated molasses controls were diluted to
the same fermentable sugars level.
344
CLARIFIED AND DESALTED ]
Γ
L /
,''
^^
CLARIFIED |
Λ "i
HI
/
\ ^*^^* ^^^^ ···"* 1
N
—
I ^^c. * ^^ ·* * —J
Γ \ X^.. / yS . . . · ' ' * UNTREATED Ί
S
Ns ^^*·· * S^ ··"' 1
L
| N^ ^v**··
'^S/**·. y /.^^
^ # .··**'··*"*" Jj 111
V>k SUGAR
v /' \. '··.. ^r ,.·**'
CONTENT
Vv # II (a/L)
1— f ^^. ^ ^ · * · * ι
x
/ s ,,Χ·*^. **·
L /
>N ^N· ' " " ^^^Sw **·· _J
/
r / >X^*v ^^s. "*"*·. I
y/-' S
N \ . ···... UNTREATED J
&
/ ' ^ytf^' "-< ^^^^^CLARIFIED

21 3
Γ CLARIFIED AND DESALTED
1 1 1 1 1 1
25 HOURS
Fig. 2. Batch fermentation curves for untreated, clarified and desalted

molasses.
Table 4 shows that the rate of ethanol production increased from 2.0
g/l-hr for untreated molasses to 4.1 g/l-hr for the desalted clarified
product. The rate observed for clarified molasses, which is the form
normally fermented, was 2.4 g/l-hr. The latter is in agreement with the
value normally observed on an industrial scale in the distillery at which
this trial was conducted. Also, the fermentation efficiency increased to 88
percent on desalting. The latter value, which is calculated using the Gay-
Lussac formula, expresses the amount of ethanol produced relative to that
corresponding to 100 percent conversion of sugars.
Figure 2 shows that the maximum ethanol concentration was achieved
after 18 hours using desalted clarified molasses, compared with 27 hours for
clarified molasses. The fermentation of untreated molasses remained
incomplete after 30 hours.
TABLE 4
Batch fermentation results for untreated, clarified and desalted
clarified molasses.
Molasses Rate of ethanol Fermentation

treatment production (g/l-hr) efficiency
Untreated 2.0 78
Clarified 2.4 80
(this trial)
Clarified 2.4 80
(distillery industrial
average)
Desalted, clarified 4.1 88
Economic considerations
The economic consequences of desalting prior to fermentation are
summarized in Figure 3. The increased fermentation efficiency on desalting
results in increased revenue for the distiller as shown by the horizontal
dotted lines (percent efficiency increase) and right hand vertical scale
(additional revenue). Fermentation efficiency is based on the sugar content
of raw (80° Bx) molasses and hence the calculated revenue increase
incorporates that lost from any sugar loss which may occur during the
desalting process. The 10 percent efficiency increase observed when
clarified molasses was desalted in the pilot scale trial just discussed
corresponds to a revenue increase of U.S. $12 per tonne of 80° Bx molasses.
The operating cost of the counter diffusion plant depends on the number
and lifetime of the membranes used. This cost can be readily determined by
moving vertically from the plant capacity scale to the continuous line
corresponding to the lifetime of the membranes and then to the left hand
vertical scale on Figure 3. For example, for a membrane life of 200 hours
and an efficiency increase of 10 percent, a counter diffusion plant of 200
tonnes per day capacity will have an operating cost of U.S. $5 per tonne.
This corresponds to a profit of U.S. $7 per tonne.
When operating on an industrial scale it is usual to set the initial
fermentable sugars level of the ferment low enough to ensure a minimal
residual sugar content after 28 to 30 hours. Adherence to this setting
after installation of a counter diffusion plant would result in a
3^6
Additional Additional
Operating Cost revenue
Φ (A$/tonne raw molasses (80°bx)) (A$/tonne 80°bx)
molasses)
Efficiency Increase: 10%
17 + + 17
16 4-16
15 | + 15
14 + 14
13 13
12 + 12
11 ill
10 110
9 I 9
8+ -r 8
4
1600 hrs
3
No. of modules*
50 100 150 200 250 300

Capacity of plant
(tonnes of 80°bx molasses
per 24 hours)
1 1 1 1 1 1 1
5 10 15 20 25 30 35 40 45 50 55 60 65
Capacity of distillery
(tonnes ethanol/24 hrs.)
1 1 1 1
100 200 300 400 500 600 700 800
Capacity of distillery
'Each module contains 60 membrane tubes (hectolitres ethanol/24 hrs.)
Fig. 3. Operating cost and additional revenue for distilleries of various

capacities.
TABLE 5
Comparison between counter diffusion, electrodialysis and ion exchange for
molasses.
Counter Electro- Ion

Parameter Diffusion (CD) dialysis (ED) exchange (IE)
Investment
cost 1/5 ED, 1/5 IE
Power Little consumed. Drives process. Little consumed.

Requires only the Larger plants may Requires cost of
cost of pumping require a special pumping molasses
molasses and water sub-station. and regenerating
through modules. chemicals through
columns.
Water Town water Soft water Soft water required.

acceptable. required.
Membrane Major cost Biggest operating None.

cost component. cost. Frequent
Disposable and disassembly for
therefore no cleaning required.
cleaning required.
Chemicals Some required Some required for Large regeneration

for cleaning of plant and membrane chemical requirement.
plant manifolds, cleaning.
etc.
Separation More efficient More efficient All types of ions

characteris separation of separation of removed.
tics and monovalent than monovalent than
losses di- and trivalent di- and trivalent
ions. Sugar loss ions. Sugar losses
minimal. higher than for CD.
Operating Lower than ED and Lower than IE up to Lower than ED

cost IE for all levels 70-80 percent above 70-80
of desalting. desalting. percent desalting.
substantial increase in plant capacity, rather than a large increase in

alcohol production per batch. This increased capacity is represented by the
33 per cent time saving shown in Figure 2. Alternatively, the distiller may
3^8
choose to increase plant capacity by increasing the initial fermentable

sugars level of the ferment. This method results in both time saving and
increased alcohol production per batch.
Comparison with competing technologies

A comparison with the main competing desalting technologies,
electrodialysis and ion exchange, is given in Table 5. The points listed
show that the overall operating cost and the initial investment cost of a
counter diffusion plant are substantially less than for its competitors.
REFERENCES
1. G. P. Meade and J.C.P. Chen, Cane Sugar Handbook, 10th ed., John
Wiley and Sons, New York, 1977, 359.
2. R. A. Johnson, Ash removal studies, Proc. of Bureau of Sugar
Experiment Stations Seminar, Ayr, 1983, 49.
3. R. P. Jones and P. F. Greenfield, A review of yeast ionic nutrition,
Part 1: Growth and fermentation requirements, Proc. Biochem.
April, 1984, 48-62.
4. Syrinx Research Institute Pty. Ltd., Counter diffusion in the
fermentation industry.
Chemistry and Processing of Sugarbeet and Sugarcane, edited by M. A. Clarke and M. A. Godshall
Chapter 22
SUGARCANE PROCESSING TO ETHANOL FOR FUEL PURPOSES
C. E. V. ROSSELL
INTRODUCTION
The start of a fuel alcohol program, the Proalcool, by the Brazilian

Government in 1975, dramatically changed the production profile of Copersucar
sugarcane mills. Table 1 shows those changes. Milling capacity has more
than doubled, annual sugar production remained practically unaltered during
this period, and ethanol production has increased ten times.
TABLE 1
Changes in annual production of Copersucar mills from the beginning of
Proalcool (1975) up to last milling season (1986).
Milling Season 1975 1986
Cane milled (metric tonnes) 27,527,105 67,700,000

Sugar produced (metric tonnes) 2,556,314 2,805,050
Ethanol produced (m3) 308,288 3,181,000
Yield (kg/TC) 93.4 110.0
Several changes have been introduced in mills to achieve this increase

in ethanol production.
Today, a typical sugarcane mill processes approximately half of the
extracted juice directly to ethanol. The rest of the juice goes to sugar
production, an easy and straightforward process because there is no
exhaustion of molasses. A high quality plantation white sugar for direct
consumption is obtained, and after two strikes the rich molasses are sent to
the fermentation unit. Figure 1 illustrates how cane is processed in a
typical Copersucar mill.
350
6
148 kg of Pol
losses during
2.7 kg « Cane »ashing
washing
137.3 kg
loss of pol
6.3 kg « Juice extraction
on bagasse
131 kg
Sugar processing Ethanol processing
65.5 kg 65.5 kg
Juice treatment Juice treatment
8.8 kg 9.4 kg
64.7 kg 65.1 kg
Pol losses t t Pol losses
Rich
Sugar Molasses Ethanol losses
production 15.17 production
18.44 kg - 7.11
1.5 kg
of sugar l i t e r s of ethanol
Pol losses
f
Sugar Ethanol
48.1 kg 69.8 kg of sugar~47.6 liters

of ethanol
117.9 kg
Efficiency z 84.22'/.
Fig. 1. P o l d i s t r i b u t i o n during the processing of cane to sugar a n d e t h a n o l .

F i g u r e s a r e o n the basis of 1 m e t r i c tonne of cane ( T C ) .
351
In Brazil, ethanol fermentation is done on the largest scale anywhere in

the world. Specific processes for fuel ethanol have been established based on
existing technology for ethanol production for the beverages and chemical
industries. In this paper, those processes for fuel ethanol production, as
well as research and development efforts to optimize them, are discussed.
THE PRODUCTION OF ETHANOL

Raw juice treatment for ethanol production
In the early stages of Proalcool the juice treatment process was the
same as for sugar production.
Experience accumulated during the last years has led to a reassesment of
the juice treatment process (ref. 1). A well designed process has to render
a juice free of suspended matter, gums and colloids and with a low microbial
count. Treatment steps have to be done avoiding sugar losses as much as
possible. Equipment selection is controlled by investment costs and
utilities and power consumption.
Treatment varies from mill to mill. Table 2, summarizing the common
procedures to treat juice for fermentation, shows that juice treatment can be
as simple as removal of suspended matter, using screens to remove bagacillo
and hydrocyclones to separate sand and soil. This is the cheapest method of
juice treatment. Its disadvantages are the deposition of a mixture of
suspended solids and yeast at the bottom of the fermentors and the hazards of
contamination, flocculation and gum production when the cane has a poor
microbiological quality or when good housekeeping is not obeyed in the mill.
Severe foaming during fermentation, resulting in a high consumption of
antifoaming agents is another drawback of this treatment.
In other mills the complexity of treatment is increased by introduction
of a heat treatment step or heat treatment followed by sedimentation in
clarifiers (ref. 2). The most advanced process includes screening,
hydrocyclone separation, liming, heat treatment, sedimentation and cooling.
352
TABLE 2
Juice treatment processes for ethanol production.
Relative Investment
Degree of Treatment Process Description Cost
Physical treatment only Screening 1.00

Hydrocyclonic separation
Physical plus heat shock Screening 3.33

and cooling Hydrocyclonic separation
Heating up to 105°C on
heat exchangers
Cooling to fermentation
temperature on plate heat
exchangers
Screening 7.90
Hydrocyclonic separation
Liming
Heating up to 105°C
Sedimentation in clarifiers
Complete treatment Cooling to fermentation
including physical temperature
operations, liming,
heat shock and Screening 6.64
sedimentation Hydrocyclonic separation
Liming
Heating up to 105°C on multistage
direct contact heat exchangers
Sedimentation in trayless
clarifiers
Partial cooling in flash tanks
Final cooling on plate heat
exchangers
Best results and highest yields are obtained with the multistage
treatment. The hazards of contamination are prevented by the pasteurization
effect of thermal treatment. Gums and colloids are removed, reducing
consumption of antifoaming agent along with their inhibitory effects on
fermentation. The deposition of solids on fermentation vats is controlled
and the wear of disks and the bowl of the centrifugal separators is
minimized.
The introduction of heat regeneration in this last treatment has been an
improvement. In particular, the successful introduction of a multistage
353
direct contact heat exchanger (Figure 2) significantly diminishes the low

pressure steam consumption (ref. 3).
y E - EVAPORATOR VAPOUR (1 IS*C )

MULTIJETS .Γ
r::---h n,
HYDROCYCUONES
r ! FLASH TANKS
|79*c]
SEAL
SEAL ^ηΤΑΝΚ -
TANK
ITU Ü y TO FERMENTATI«
PROCESS
Fig. 2. Juice treatment with direct vapor contact heating.
The investment cost of this complex treatment is counterbalanced by the

higher fermentation yield obtained and the guarantee of a trouble free
operation. Collected data on industrial installations (Table 3) showed
higher productivity and lower fermentation time in units operating with
complete treatment compared to the others. A higher yeast concentration on
wine and a lower consumption of expensive antifoam chemicals are also
obtained.
35^
TABLE 3
Influence of different levels of juice treatment on fermentation parameters.
Level of juice treatment
Fermentation Physical plus

parameter Physical treatment heat shock Complete treatment
Productivity
(kg of ethanol/ 3.68 4.09 4.73
m 3 per hour)
Fermentation
time (hour) 11.6 9.0 7.9
Yeast concentration
in wine (Z in volume) 6.5 9.5 11.7
Antifoaming agents
consumption (defoamer
+ disperser) (kg of
product per m 3 of
ethanol) 0.628 0.568 0.521
Sulphuric acid
consumption (kg per
m 3 of ethanol) 6.344 3.212 6.994
THE FERMENTATION PROCESSES
The Meile Boinot process
This process, formerly introduced by French distilleries to Brazil 25

years ago to ferment molasses, has been improved to work on a large scale, on
musts compounded of a mixture of cane juice, syrups and rich molasses.
Figure 3 shows a flow diagram of this process. Table 4 summarizes the
most important parameters.
fi- F 0 : FERMENTORS
lCANE JUICE PF · PREFERMENTOR
I SYRUP : HOLDING TANK
■■ BLOWER
I MOLASSES
> CENTRIFUGAL SEPARATOR
I H 20 ' HEAT EXCHANGER
\ « ABSORPTION TOWER
• STATIC MIXER
■ PUMPS
P BL
Fig. 3. Meile Boinot fermentation process.
The process is a batch fed fermentation with cell recycle. At the end
of fermentation, the wine is centrifuged to recover most of the yeast. The
yeast cream obtained is submitted to an acid treatment.
This treatment is done in the prefermentors, diluting the cream with
water and dropping the pH with sulphuric acid: the yeast are stirred for 2-3
hours. Some reactions occur such as the killing of bacteria and harmful
yeasts, and the reduction of intracellular activity of ethanol. The aeration
of the prefermentor restores the activity of yeasts. Once ready, this
inoculum is sent to fermentation vats, where once the must is fed, alcoholic
fermentation starts again.
356
TABLE 4
Typical parameters of Meile Boinot Fermentation Process.
Equipment and performance data
Description Quantity Capacity Additional information
Fermentors 8-12 300 m 3 Cylindrical fermentors

constructed of carbon
steel.
Centrifuges 3-6 45 m 3 /h Disk centrifuges, with

nozzle discharge
Heat
exchangers 4-6 1,200,000 kcal/h Plate type
Productivity 4-7 Ethanol kg/m3/h Weight of ethanol (kg)

per volume of fermentation
tanks (m3) per unit
of time (h).
PROCESS DATA
Must;
Total reducing sugars as glucose 12-152
pH 5.5-6.2
Temperature 26-32°C
Wine:
Total reducing sugars below 0.5Z
Ethanol content 6-11° GL (9 6L on average)
Yeast concentration 7-16Z (volume)
Yeast viability (méthylène blue
staining) 50-85Z
Fermentation temperature 34-38°C
Fermentation time 5-9 h
Yield (stoichiometric) 85-92Z
Inoculum (yeast foot):

Ethanol content below 4.0° GL
pH 2.2 - 2.8
Yeast concentration 15-35Z
Acid treatment time 1-4 h
Temperature 28-32°C
By-Products:
Glycerol production 6Z of alcohol production
Acid production 2-6Z of alcohol production
Yeast production 4-10Z of alcohol production
(on dry basis)
THE CONTINUOUS PROCESS
In this process must is fed to the first of a series of continuous
stirred tank reactors (3 to 5). Wine flows from one reactor to another,
content of reducing sugars decreases and ethanol increases in a stepwise
manner. More than 60Z of the conversion is done on the first vat; the last
unit polishes the wine, depleting the sugar content. After the fermentation
step, the wine is centrifuged to recover the yeast and flows to the
distillation unit. The yeast cream is treated continuously in continuous
stirred tank reactors in a manner similar to the Meile Boinot process and is
recycled to the first fermentor. In summary, substrate and end products are
submitted to steady state conditions but yeasts are submitted to conditions
that change with time, going stepwise from low to high ethanol concentrations
and then to recovery conditions at the acid treatment step.
This process is an improvement of the Meile Boinot process.
Theoretically it requires low fermentors volume and running costs. It should
be easy to instrumentate and to automate it.
A flow-sheet of the process is shown in Figure 4.
YEAST INOCULUM
JTO OISTI-N.
n uATipH y
Fig. 4. Multistage continuous fermentation process.

358
Table 5 summarizes typical parameters of the continuous process.
TABLE 5
Typical parameters of the continuous fermentation process.
Equipment and performance data
First fermentor volume 700 m 3

Quantity of fermentors approximately 5
Quantity of heat exchangers 3
Air flow rate in prefermentors and first
reactor for aeration and stirring 0.01 vvm
Productivity 4-8 m 3 kg of ethanol/h
Process data
Must; Same as Meile Boinot process
Wine:
Ethanol content 7-ll°GL (8.3°GL on average)
Yeast concentration 7-12Z
Yeast viability 40-75Z
Fermentation temperature 32-36°C
Residence time 6-9 h
Fermentation time 4-7 h
Yeast inoculum; Same as Meile Boinot process
RECOVERY OF FERMENTATION BY-PRODUCTS
During the fermentation of sugar to ethanol, secondary products such as

yeasts and glycerol are produced.
A fermentation process efficiently operated generates an excess of
yeast. When the plant has a complete juice treatment and enough centrifuge
capacity this excess of yeast can be bled, washed to remove the ethanol
carried, dried and sold as single cell protein for animal feed.
Usual bleeding ratio is 25-50 kg of dried yeast per m 3 of produced
ethanol. This ratio does not disturb the fermentation process.
STILLAGE RECYCLE
Stillage is rich in potassium and in other nutrients such as phosphorus
and organic matter. In the south central region of Brazil it is used as a
fertilizer in the cane fields.
Cost of transportation and dosing on fields can be significantly reduced
when stilläge volume is diminished by reusing it at the fermentation for
diluting the yeast cream instead of using water. Also, syrup, molasses and
pre-evaporated juice can be diluted with stilläge.
Stillage can be recycled up to a ratio of 25Z of total stilläge produced
without any change in typical fermentation parameters, including yield and
productivity. Research is being done to increase the ratio of stilläge
recycle without loss of yield and productivity.
DISTILLATION PROCESS
Distillation of ethanol is a well defined physical operation and the
production of both hydrated and anhydrous ethanol proceeds without any
trouble. Installed equipment is conventional and 99Z or more of the input
alcohol is easily recovered. Wine is distilled in a set of 3 columns and the
flegma obtained is rectified to hydrated fuel alcohol in two more columns.
This hydrated fuel alcohol is dehydrated by azeotropic distillation using
benzene as entraîner in a set of two columns.
Fusel oil is obtained as a by-product of distillation. It is being
fractionated by distillation to render amyl and butyl alcohols.
Today, technical research and development efforts on distillation are
directed to ethanol quality in order to match the severe standards required
to guarantee a product noncorrosive to engine parts and fuel circuits.
In the near future, research in the field of ethanol distillation will
be oriented to energy economy, in order to decrease the low pressure steam
consumption on distillation towers.
PRESENT STATUS OF FERMENTATION PROCESSES OPERATING IN BRAZIL

Most units at present in operation are of the Meile Boinot type but this
situation is changing rapidly. Today there are seven multistage continuous
units running in Copersucar mills and two more will start the next milling
season.
360
After the start-up of the first continuous unit, experience on running

this process was gained, and the need for troubleshooting had practically
disappeared. It is now a safe process, and the number of installed units has
begun to increase.
Continuous fermentation process is attractive compared to Meile Boinot
because it has the potential for higher productivity and lower manpower
needs, and is easier and cheaper to equip and automate. Heat exchangers and
equipment for recovering ethanol vapors in C0 2 are smaller in size for the
same daily capacity because they are designed for steady state conditions,
not for peak loads.
Figure 1 shows the profile of sugar losses during processing of
sugarcane to ethanol. Losses during conversion of sugar to ethanol are the
major losses, approximately 10.5 kg per ton of cane.
Once distillation is an efficient unit operation, the highest loss of
sugar occurs during its biological conversion to ethanol. A lot of effort
has gone into the attempt to diminish the gap between stoichiometric and
actual fermentation yield (Table 6). This parameter increased steadily from
the 77/78 season to the 82/83 season and then stabilized at the range of
85-92Z depending on the technological level of the mills (ref. 4 ) .
TABLE 6
Evolution of typical fermentation parameters during previous milling seasons-
average figures.
Season
Fermentation Actual range of

parameter 77/78 78/79 79/80 81/82 82/83 values
Yield 82.9 84.5 85.7 85.6 86.3 85-92

Fermentation
time (hours) 14.5 14.5 13.08 14.0 13.22 5-8.5
Ethanol content
°GL 7.5 8.09 8.11 9.04 9.18 9-12
Yeast concentration
(Z in volume) 6.2 6.5 6.3 7.6 7.6 8-15
Mills sampled 33 30 16 16 6 --
361
Fermentation losses can be classified as metabolic and process losses.

Metabolic losses result from the synthesis by yeast of secondary products,
associated with the production of ethanol. Glycerol, organic acids and yeast
reproduction are the most important. Their ratios of formation per unit of
ethanol are shown in Table 4. Biosynthesis of the secondary products depends
on must composition and fermentation conditions.
Process losses depend mainly on factors related to the operation of the
fermentation and to the design of the fermentation unit. When the process is
run under efficient management and control, and the engineering of the plant
is good, these losses are minimal. Main process losses are:
Sugar not converted to ethanol, remaining on wine;
Loss of yeast caused by incorrect operation of centrifugual
separators (this yeast has to be replenished at the expenses of
sugar);
- Yeast lost during the bleeding of the mud accumulated at the bottom
of fermentation vats. This loss is lowered when complete juice
treatment is done.
Flocculation of yeast caused by lactic bacteria; or physiological
changes in yeast. This flocculation causes severe losses of yeast at
the centrifuges.
Productivity is another important parameter of fermentation. Higher
productivity means more alcohol produced per unit of installation and time.
Capital costs are lower in a highly productive fermentation. Table 6 shows
that fermentation time has been decreased, and ethanol and yeast
concentration in wine has increased over the last seven seasons, increasing
the productivity of fermentations (ref. 4 ) . Productivity has stabilized at
the range of 4-7 kg of ethanol/m3 h depending on the technological status of
the particular fermentation unit. Well conducted processes done in optimized
fermentation units and with complete juice treatment are more productive than
others.
To improve the fermentation processes existing in Brazil, some
requirements must be fulfilled. A good fermentation process has to be
efficient in substrate conversion and must minimize formation of by-products
such as glycerol and organic acids. It must have a high productivity to
reduce equipment size and cost, must be reliable and run steadily without the
need to troubleshoot problems such as flocculation, contamination and
362
oscillations in yeast viability, and must be capable of supporting changes in

feed composition and quantity. It should also be able to handle periods of
time without juice feed caused by problems that stop the tandems or when
there is no supply of cane to the mill.
Our experience accumulated in running the fermentation has given us a
good knowledge of research and development lines to be followed in order to
improve both processes.
Topics that justify further study are:
Improvement of raw juice treatment in order to have a process capable
of rendering a juice free of suspended matter and colloids, with a low
microbial count and at the lowest cost.
Isolation of new yeast strains specific to our fermentation
conditions. A good yeast strain has to be tolerant of high ethanol
concentration, be efficient at higher fermentation temperatures (34-38°C)
able to withstand thermal shocks (temperature increases up to 42-45°C), and
to minimize by-product formation.
- Development of new strains by genetic manipulation of yeasts.
Improvement of equipment used in fermentation units such as:
redesigned fermentation vats with better mixing characteristics, that will
avoid must stratification and feed short-circuits if cascade continuous
fermentation is employed.
- More efficient heat exchangers with higher overall heat transfer
coefficients have to be developed to overcome the heat load of a high rate
process.
- New development of centrifugal separators and precentrifugation
equipment like screens and hydrocyclones.
- The formulation of a kinetic model for Meile Boinot and the Continous
Process, that describes with precision yeast growth and death and ethanol and
by-product accumulation. This model will be useful for designing
fermentation and prefermentation vats and heat exchangers.
- A better knowledge of yeast physiology related to ethanol fermentation
in order to establish and control process parameters and avoid any trouble
shooting and understand biological and physiocochemical changes during acid
treatment of yeast cream.
- Biochemistry of yeast fermentation of sugar to ethanol in order to
elucidate the true stoichiometry of alcoholic fermentation and to determine
the enthalpy change in reaction.
- Microfloral studies on cane juice, must, wine and yeast inoculum to
determine the influence of bacterial contaminants on yield, flocculation,
formation of gums, and related phenomena.
- The development of new microbiological, physical and chemical methods
to monitor and control the fermentation process.
- The effect of stilläge recycling on fermentation parameters like
yield, productivity, by-product formation, yeast viability, and raw materials
consumption.
- The effect of yeast cream bleeding and recovery (i.e. recycling) as a
feedstuff on fermentation parameters.
During the last seven years the Centro de Tecnologia Copersucar (CTC)
improved considerably the fermentation processes that exist in Brazil. Table
7 summarizes the most important results of these studies. Research and
development continues and we have selected some working lines (Table 8) to
attain the objectives described above.
TABLE 7
Advances in fermentation technology during the last eight seasons at the
distilleries of Copersucar Group.
Season Advances
80-81 First experiment on stilläge recycle was done in a

distillery.
81-82 The system of closed vessels and absorption tower for

recovery of ethanol from C0 2 was installed on a
fermentation unit.
82-83 First multistage continuous fermentation process (200,000

1/day), was launched at a COPERSUCAR mill.
82-83 Microbiological control at juice treatment and fermentation

process was introduced. The first microbiology laboratory
was inaugurated at a mill (ref. 5).
82-83 Experiments on yeast strains selection and protoplast

fusion was begun.
364
83-84 An experiment to compare stoichometric yield of Meile

Boinot system versus continuous system at industrial
level was done (ref. 6).
84-85 Microscopic analysis of wine and yeast inoculum was

established as a routine control at distilleries (ref. 7 ) .
84-85 An algorithm to design plate heat exchangers for cooling

fermentation vats was established.
84-85 A Melle Boinot fermentation with multistage continuous

acid treatment for yeast recovery and inoculum preparation
started up.
85-86 A new project for a Melle Boinot fermentation unit was

developed. A new generation of more advanced distilleries
appeared.
85-86 An experiment to establish the actual stoichiometry of the

fermentation process at industrial scale was carried out.
A new mass balance of fermentation was obtained. The
concept of by-products formation instead of fermentation
yield was introduced (ref. 8 ) .
85-86 New analytical techniques including amino nitrogen, P 2 0 5

and glycerol determinations were established at mills.
The concept of Operational Control of fermentation process
was introduced.
86-87 A mill ran an entire season recycling stilläge at a ratio

of 252.
86-87 The microcomputer was introduced to compile analytical,

microbiological and process data, calculate characteristic
parameters of fermentation, and edit the daily reports
(ref. 9 ) .
86-87 Mills began to recover excess yeast from the fermentation.

Another by-product to be explored by mills appeared.
87-88 A new project for multistage continuous fermentation with

more advanced technology was developed.
TABLE 8
Research and development lines selected by CTC to improve existing
fermentation process.
Research program Information to be obtained
1. Scale down of fermentation processes Mass and energy balances

to 15 1 and 25 1 sizes. Process kinetics data
2. Laboratory development of microbiological, New analytical tools

chemical and physical techniques.
3. Yeast screening and genetic manipulation. New yeast strains
4. Process modeling and simulation. Kinetic models

Process design data
5. Process engineering. Process and equipment

improvement
6. Data acquisition at industrial Actual yield and

fermentation units. productivity. Raw
materials and
utilities consumption.
Equipment investment and
production costs.
CONCLUSIONS
Today, the processing of sugarcane to ethanol for fuel purposes is a

well established industry in Brazil. A new technology specific for this high
tonnage industry is emerging. Juice treatment and the recovery of ethanol by
distillation are well established technologies.
Most of the research and development efforts are devoted to the
understanding and improvement of existing fermentation processes in Brazil.
Because the greatest processing losses appear during the conversion of
sugar to ethanol by the fermentation step, a lot of effort is expended in
this area. Among new developments are the recovery of by-products like
yeasts and recycling of stilläge in order to make the process more
profitable.
366
REFERENCES
1 D. T. Oliveira, W. Pizaia, H.P.H. Ackermann and C.E.V. Rossell,
Alternativas de processo no tratamento de caldo para destilaria.
Boletim Tecnico Copersucar 37(1987) 25-31.
2 H. A. Lucredi, J. Finguerut, K. H. Leimer and C.E.V. Rossell,
Verificacao da esterelidade do sistema de tratamento termico do
caldo de cana-de-acucar para a fermentacao, Boletim Tecnico
Copersucar 27(1984) 25-28.
3 W. Pizaia, D. T. Oliveira and C.E.V. Rossell, Direct contact
heating and flash cooling of cane juice for alcohol production.
STAB-Acucar, Alcool e Subprodutos, 4(6) (1986) 121-123.
4 J. Finguerut, H. A. Lucredi and C.E.V. Rossell, Fermentacao alcoolica em
usinas cooperadas: Evolucao e perspectivas, Boletim Tecnico
Copersucar 23(1983) 8-11.
5 Anonimo, Contrôle microbiologico na fabricacao de acucar e alcool.
Boletim Tecnico Copersucar 22(1983) 1-15.
6 J. Finguerut, K. H. Leimer, H. A. Lucredi and C.E.V. Rossell,
Consideracoes sobre o calculo do rendimento da fermentacao alcoolica
continua e descontinua, Boletim Tecnico Copersucar, 28 (1984) 20-26.
7 J. Finguerut, H.A. Lucredi, K. H. Leimer and C.E.V. Rossell, Contrôle
microscopico do processo de fermentacao alcoolica, Boletim Tecnico
Copersucar 34 (1986) 22-26.
8 J. Finguerut, H. A. Lucredi, K. H. Leimer and C.E.V. Rossell,
Estequiometria da fermentacao alcoolica a partir de caldo de cana,
Boletim Tecnico Copersucar 33 (1985) 45-48.
9 J. Finguerut, H. A. Lucredi, M. A. Furco, R. Altomari and C.E.V. Rossell,
Contrôle operacional da fermentacao, Boletim Tecnico Copersucar
38 (1987) 9.
367
Chapter 23
SUCROSE CHEMISTRY: ITS POSITION AS A RAW MATERIAL FOR THE CHEMICAL INDUSTRY
R. KHAN AND H. F . JONES
SUMMARY
Over the past two decades many fundamental aspects of the chemistry of
sucrose have been studied and a unique profile of chemical reactivity has
been revealed. Some of the characteristics of the sucrose molecule are
described and the importance of sucrose as a feedstock for chemical products
is discussed using examples of sucrose-based products of actual or potential
commercial value.
INTRODUCTION
Sucrose is a major product of photosynthesis and in many plants serves
as the major storage saccharide. In plants such as sugarcane it is the main
storage carbohydrate but in other plants it is converted to starch, inulin
or levan. It is also significant that in many plants the transport of
oligosaccharides and polysaccharides from one part of the plant to another
takes place by conversion to sucrose, translocation and re-synthesis.
Sucrose is one of the main products of agriculture worldwide: its
current production in all forms exceeds one hundred million tonnes a year.
It is principally used in beverage and food products but it finds some use
in the chemicals and processing industries.
A chemical industry based on sucrose has either to convert sucrose into
conventional feedstocks such as ethanol and ethylene or to develop new
technology and products which utilize the particular properties of the
sucrose molecule. To elaborate this theme the chemistry of sucrose and some
of its potential and actual commercial applications are discussed.
PHYSICAL PROPERTIES
Sucrose is a sweet, crystalline solid which melts at 160 - 186°C,
depending on the solvent of crystallization. It is optically active and has
a specific rotation [a] D of +66.3° (c 26, water).
Solubility
The lack of solubility of sucrose in non-polar organic solvents has
been one of the most important limiting factors in the development of its
368
chemistry and technology. Sucrose is soluble in dipolar aprotic solvents

such as pyridine, N, îi-dimethylformamide and dimethylsulphoxide (see Table
1, ref. 1) These solvents are not only expensive but often participate in
reactions of sucrose. Water, in which sucrose is readily soluble, can only
be used in reactions in which it does not compete with the hydroxyl groups
of sucrose. An ideal organic solvent for sucrose is still needed.
TABLE 1
Non-aqueous solvents of sucrose.
Solubility at Solubility at
Solvents 100°C (g/100 g) Solvent 100°C (g/100 g)
Dimethylsulphoxide 58.7 Ethanol 1.1

Morpholine 45.1 Propanol 0.4
Di-n-propylsulphoxide 42.0 Mixed méthylène
glycerols 7.5
îï-Methyl-2-pyrrolidone 33.5 Methyl cellosolve 2.8
ltf, Ν,-Dime thy 1 formamide 29.6 Tetrahydrofurfuryl
alcohol 2.5
Propane-1,2-diol 11.0
Pyridine 6.0
STRUCTURE
Sucrose, systematically named ß-D-fructofuranosyl α-D-glucopyranoside,
is a non-reducing disaccharide. Its structure has been confirmed by X-ray
crystallography (ref. 2), neutron diffraction (ref. 3), X H- and 13
C-n.m.r.
(ref. 4) spectroscopy. In the crystal structure of sucrose (Fig. 1,
A
structure 2) the a-D-glucopyranosyl residue adopts the d conformation and
the fructofuranosyl moiety has the 3 T* conformation. All of the eight
hydroxyl groups except 4-OH are involved in hydrogen bonding, two of them
intramolecularly (6'-0H 0-5 and l'-OH 0-2).
369
* CHaOH
Fig. Structure of sucrose.
CHEMICAL REACTIVITY
The hydroxyl groups of sucrose can be divided into sets of three
primary (6,6* and 1*) and five secondary (2,3,4,3* and 4*) hydroxyl groups.
Towards the majority of chemical reactions the primary OH groups exhibit
greater reactivity than the secondary OH groups (ref. 5 ) . For example,
selective monoacetylation of sucrose with acetic anhydride in pyridine gives
sucrose 6-acetate (~50Z) as the major product. The primary hydroxyl groups
in sucrose can be selectively blocked by tritylation (ref. 6) or t-
butyldiphenylsilylation (ref. 7) reactions using trityl chloride or t:-
butyldiphenylsilyl chloride, respectively, shown in Fig. 2.
370
OH
OH
Ron ôa 1 RO
)—o
HO f <
OH OR
R - C(C e H s )3, R3- - H

3
R - SiPh* C(CHs)s, R - - H R - R3- - C(CeHs),
1
R - R - SiPha C(CH)s
Fig. 2. Blocking groups for sucrose, using reagents: i, TrCl, Pyridine;

ii, t.-BDPhSiCl, Pyridine.
Cyclic acetals are valuable protecting groups used in carbohydrate

chemistry. The hydroxyl groups at 4,6,1* and 2 in sucrose can be protected
by the acetalation reaction. 4,6-,0-Isopropylidenesucrose (Fig. 3, structure
3) and 4,6:1',2-di-_0-isopropylidenesucrose (Fig. 3, structure 4) have been
obtained under kinetic control using 2-methoxypropene, N,N-dimethylformamide
and catalytic amount of toluene-p-sulphonic acid (ref. 8). Conventional
acetylation of 3 and 4 with acetic anhydride and pyridine afforded the
peracetates 5 and 6, respectively. Acid catalyzed deacetalation of 5 and 6
gave access to sucroses with free hydroxyl groups at 4,6,1* and 2 positions
371
for chemical or biochemical manipulations at these positions, as shown in

Fig. 3(ref. 9).
OR
OR »o
I R » H
5 R * Ac M R« H
6 R « Ac
I 4
OR OH
HO
HO
OR
HO OR OH
OR
Fig. 3. Acetalation as protecting group system.
The chlorination reaction of sucrose with sulphuryl chloride and

pyridine has been extensively investigated (ref. 10). 6*-Chloro-(7), 6,6'-
dichloro- (8), 4,6,6'-trichloro- (9), and 4,6,1»,6'-tetrachloro-(10)
derivatives of sucrose have been synthesized under controlled reaction
conditions. (Numbers in parentheses refer to structures shown in Fig. 4.)
It is of interest to note that the substitution of hydroxyl groups in
sucrose by chlorine changes its taste characteristics. For example,
4.1 *.6'-trichloro-4.1'.6'-trideoxygalactosucrose (11) is intensely sweet,
372
o^-dichloro-o^-dideoxysucrose (8) i s t a s t e l e s s and 2 , 1 , , 6 , 6 ' -

tetrachloro-2.1 '. 6.6 '-tetradeoxymannosucrose (12) i s intensely bitter (refs.
5, 11).
OH OH
HO11 Y
<
11
Fig. 4. Chlorinated sucrose structures.

SUCROSE AS FEEDSTOCK FOR CHEMICALS
In the overall economics of a high-value, low-volume product from
sugar, the cost of the raw material is not particularly significant.
However, for low-value, high-volume products, the cost of sugar becomes
critically important. A comparison of the price of a chemical derived from
petroleum with that of the cost of sugar required to produce the substitute
can roughly indicate whether or not the use of sugar as a feedstock would be
cost competitive. From the examples given in Table 2, it is apparent that
the production of ethylene from sugar by way of ethanol is not a commercial
proposition. In the cases of ethanol and polyhydroxybutyrate the room to
maneuvre is limited. Production of the other chemicals from sugar appears
economically feasible. However, factors such as production costs, the scale
of production, and socio-economic and political conditions of the country in
question must also be taken into account when appraising the feasibility of
a project. For instance, production of alcohol as fuel may be uneconomical
in a European country but in Brazil or a similar tropical country where
petroleum is imported and sugar is in abundant supply it can make sense to
produce ethanol from sugar.
TABLE 2
Effect of the cost of sucrose on low-value high-volume chemicals.
Product yield Cost of sugar Current selling

(te) per te per te» of price US$ per
Product of sugar product te b
Ethanol 0.46 219 585

Ethylene 0.30 333 308
Polyhydroxybutyrate 0.33 300 750°
Citric acid 0.83 120 2002
Lactic acid 0.91 110 2265
Glutamic acid 0.77 130 1760
Lysine 0.67 149 3300
» Sugar price assumed at US$100 per te.

b
Chemical Marketing Reporter, June, 1987.
c
Price of polypropylene--for which polyhydroxybutyrate might substitute.
ESTERS OF SUCROSE
A wide variety of sucrose esters of actual and potential commercial
interest has been identified (see Table 3). Some of these esters are
37^
already in the market and some are being developed. Applications of these
esters range from food additives such as emulsifiers and low-calorie fats to
industrial chemicals such as detergents and plasticizers.
Sucrose is identified by high water solubility and hydrophilic and
hydrogen bonding properties. In designing an ester of sucrose for a
particular application the unique structural features of the molecule must
be considered. For example, low molecular weight carboxylic monoester
derivatives of sucrose exhibit humectant properties like those of sorbitol
and glycerol. Substitution of a hydroxyl group in sucrose by a long-chain
fatty acid moiety renders the molecule surface active. When the degree of
the substitution is greater than four the product has properties similar to
those of a fat or an oil, depending on the chain lengths of the fatty acids
used.
TABLE 3
Esters of sucrose and their applications.
Sucrose compounds Application Reference
Monoacetates Humectant 12
Monofatty acid esters Surfactant, 13-17
emulsifier
Tribenzoates Bittering agent 18
Octaacetate Bittering agent
Octabenzoate Plasticizer 19
Diacetate hexaisobutyrate Plasticizer 20
Polyfatty acid ester Low calorie fats and oils 21
Sucrose monoacetate
Sucrose monoacetate prepared by selective monoacetylation of sucrose
with acetic anhydride and pyridine has been shown to possess humectant
properties superior to sorbitol (ref. 12). Humectants are hygroscopic
materials which prevent loss of moisture when incorporated into foodstuffs.
The commonly used humectants are glycerol, mono- and di-glycerides of fatty
acids, propylene glycol and sorbitol. For sucrose monoacetate to compete
with these products an economic advantage will have to be demonstrated. In
addition a food humectant should be stable, tasteless and must be non-toxic.
Sucrose monofatty acid esters
Sucrose monoesters of long-chain fatty acids (ref. 13) (shown in Fig.
5) exhibit surface active properties (ref. 13). They are approved as food
additives in Europe and in the United States. These esters are non-ionic,
non-toxic and 100% biodegradable.
OH
Fig. 5. Sucrose monester of a long chain fatty acid.
There are a number of processes described in the literature for the

production of sucrose monofatty acid esters. The first process, "The
Nebraska-Snell Process," comprises the reaction of sucrose with a
triglycéride in Ν,,Ν-dime thy lformamide in the presence of a basic catalyst at
90°C (ref. 14). The reaction product is a mixture of mono- and di-
glycerides and sucrose esters. With some modification the process is used
by Ryoto Company of Japan. In order to avoid the use of toxic and expensive
Ν,Ν-dimethylformamide, Dai-Ichi Kogyo Seiyaku Company Limited has developed
a water-based process (ref. 15). However, the disadvantages of this process
are that it involves several process steps and requires reduced pressure.
Tate and Lyle pic developed a solventless process which simply involved the
treatment of sucrose with a triglycéride in the presence of a basic
transesterification catalyst, potassium carbonate and/or potassium soap, at
130°C at atmospheric pressure (ref. 16). The product mass contain less
monoester (~25Z) as compared to the previous two processes (~50Z), which
makes the purification of the esters more difficult and consequently more
expensive. The Tate and Lyle process is however very rapid and has
potential for conversion to continuous operation (ref. 16). Recently, a
semi-continuous process, "The Worm-Shaft Process" for transesterification of
sucrose, has been described (ref. 17). According to the process, sodium
stéarate, potassium carbonate, methyl palmitate and sucrose are first fed
through a Worm-Shaft reactor at elevated temperature and pressure. The
376
resultant mass is then extruded into a reactor where it is further reacted

at 150°C under reduced pressure for 90 min.
Purified sucrose esters are mainly produced and marketed in Japan.
They are of utility in processed food for such functions as emulsification
antibacterial effect, crystallization inhibition, wheat flour products
improver and lubrication (ref. 22). Sucrose esters exhibit fruit and
vegetable preservative properties (ref. 23). The current price of sucrose
esters is US$10.50 per kg on dry basis. It is estimated that Mitsubishi
Kasei Food Corporation, Tokyo, and Dai Ichi Kogyo Seiyaku, Kyoto, between
them, produce about 2600 tonnes of sucrose esters per year.
The crude sucrose glyceride mixture produced by a solventless process
can be included in detergent formulations (ref. 16). Although the
surfactant properties of the sucrose esters may not be as superior as those
of the synthetic detergents, the raw material costs are lower than the price
of synthetics (see Table 4 ) . Hence, wherever the raw materials are
available and foreign exchange is limited, the production of sucrose-based
surfactants should be considered. Sucrose esters are biodegradable and
their use should not result in any unusual environmental problems.
TABLE 4
Raw material costs for sucrose surfactants by the solventless process
(ref. 16).
Cost per te Price of

crude surfactant0 Dodecylbenzensulphonate«1
Raw material US$ US$ per te
Sucrose 47 110 1034

Tallow 525 223
Potassium carbonate 71 71
Total 643 404
a Tallow (64.5Z), sucrose (27.4Z), K*C03 (8.12);

b Tallow (27.4Z), sucrose (64.5Z), K 2 C0 3 (8.1Z);
c
Sugar price assumed US$170 per te;
Λ
Chemicals Marketing Reporter. Nov. 2, 1987.
Detergent compositions usually contain, in addition to a detergent

active material (15Z), a detergent builder (approx. 30Z) whose role is to
remove hardness from the wash liquor and to prevent re-deposition of dirt.
Water soluble phosphates are extensively used as detergent builders.
However, for reasons of eutrophication, allegedly caused by sodium
tripolyphosphate, and cost (US$792 per tonne), a search for an alternative
adjuvant has been on. Unilever pic has developed a phosphate-free detergent
adjuvant containing 10 parts by weight (p.b.w.) of sodium carbonate, 20
p.b.w. of calcite, 4 p.b.w. of sucrose, and 4 p.b.w. of an anionic
surfactant (ref. 24). We may speculate that such a detergent builder may
also be compatible with the crude sucrose ester glyceride mixture.
Sucrose itself solvates calcium ion strongly but not magnesium.
6,1*,6*-Tricarboxylic acid sucrose, shown in Fig. 6, prepared by catalytic
oxidation of sucrose, has been reported by Hoechst AG of the Federal
Republic of Germany to be a useful complex former in washing agent
formulations (ref. 25).
COOH COOH
HO I I 'coOH
OH OH
1W
Fig. 6. 6,1',6'-Tricarboxylic acid sucrose.
Sucrose polyester
In the developed world, 40Z of the caloric intake comes from fat. The
market demand for low calorie fats and oils has long been recognized. The
Procter & Gamble Company has identified a group of products to suit that
market. These are fatty acid esters of sucrose and are termed as sucrose
polyester (Fig. 7). The ester is produced by a solventless process
involving three steps: interesterification of sucrose and long-chain fatty
acid methyl esters followed by refining and extraction (ref. 21). Sucrose
polyester contains six to eight fatty acid ester groups per molecule. The
structures of the fatty acids esterified to sucrose determine the physical
properties of the resulting fat varying from a solid to a liquid. The
viscosity of sucrose polyester is slightly higher than the triglycéride with
the same fatty acids. Other properties such as taste, appearance,
378
interfacial tension, aroma, lubricity, and immiscibility in water are

claimed to be indistinguishable from triglycéride. The ester is not
hydrolyzed by pancreatic lipases and therefore passes unmetabolized. The
absence of hydrolysis or absorption makes sucrose polyester a low calorie
fat. It acts as a lipophilic solvent in the intestine to enhance the
excretion of fat soluble compounds like cholesterol. It is claimed that
sucrose polyester is not toxic or detrimental to normal metabolism in any
species tested, including man. However, it has not yet been approved as a
food additive by the FDA.
R = CO(CH2)rx CH 3
15
Fig. 7. Sucrose polyester.
POLYMERS FROM SUCROSE

Sucrose in urethane polymer manufacture
The non-reducing and polyhydroxyl functionalities of sucrose make it
useful for conversion to polyfunctional polyols for polyurethane manufacture
(ref. 26). High molecular weight, low functionality polyols are used for
flexible and elastomeric polyurethanes and low molecular weight high
functionality polyols such as sucrose for rigid and solid polyurethanes.
Sucrose cannot directly be used in polyurethane manufacture because it leads
to brittle products. Thus the polyhydroxypropyl ether of sucrose is used in
the preparation of rigid foams. The polyhydroxypropyl ether group confers
miscibility with the diisocyanate and fluorocarbon blowing agent and
provides the molecular spacing necessary to impart such useful structural
properties as strength and flexibility to the finished foam.
The use of sucrose in polyurethanes provides one of its most important
industrial outlets. It is not known how much sucrose is consumed in
Polyurethane foams throughout the world. In order to derive an estimate of
the likely market for sucrose, Bayer's estimate of the total world market
for rigid foams of 946 million metric tonnes in 1986 may be taken into
account (ref. 27). Potentially up to 20Z of this material could be sourced
by sucrose, provided it can compete in price and performance with those of
petroleum chemicals, starches, sorbitol and others. In Europe the market
(ref. 27) for sucrose in polyurethanes has been estimated as 9,400 tonnes in
1984-85 indicating that only a small proportion of rigid foams employs
sucrose (ref. 28). Changes in the balance of raw materials prices could
alter the demand for sucrose significantly.
Polyhydroxybutyrate, a biopolymer
Polyhydroxybutyrate (PHB, shown in Fig. 8) is a polyester produced by
numerous microorganisms especially when provided with a nitrogen-deficient
feed. Sucrose is converted to PHB in 70Z yield by weight of the organism,
Alcaligenes eutrophus (ref. 29). The product is being developed by ICI pic.
Its applications are both unique and limited. It has many similarities with
polypropylene but is biodegradable. However, the production cost, using the
existing technology, is considerably higher than that of polypropylene. It
takes 3.5 tonnes of sugar to produce 1 tonne of PHB. The sugar price needs
to be reduced to around US$150 per tonne and significant economies of scale
must be envisaged before there is any possibility of this product competing
with the existing, large tonnage plastics (see Table 5 ) . In the near future
PHB may find low-volume high-value medical applications such as in surgical
pins and sutures.
-Γ—CH - CH, - COO — Ί -

L
I «
CHS
16
Fig. 8. Polyhydroxybutyrate.
380
TABLE 5
Selling price (US$ per te) of PHB-.
Medical Polyester«1 Polypropylene«1

Sugar price application replacer replacer
US$ per te (10 te p.a.) (10,000 te p.a.) (50,000 te p.a.)
150* 16,743 1,125 892

506e — 2,679 2,264
————
m
P. Rodgers, 31st Congress Confederation Internationale des Betteraviers
Européen, 26-30 May, 1986, Athens, Greece.
to
Economically desirable sugar prices.
c
Current sugar price.
d
Current price of polyester US$1210 per te and polypropylene US$1100
per te (Chemicals Marketing Reporter, Nov. 2, 1987).
Dextrans and levans

Homopolysaccharides like dextrans and levans are produced from sucrose
by a number of bacterial species. Dextran synthesis is catalyzed by the
extracellular glycosyl transferase, dextransucrase, that transfers the os-D-
glucosyl group from a sucrose molecule to the 6-position of the non-reducing
terminal glucose residue, resulting in a polymer composed of a-D-
glucopyranosyl residues connected by l-»6 linkages (17), shown in Fig. 9.
Levans are produced in an analogous manner from sucrose by the enzyme levan-
sucrase. The fructose moiety of sucrose is utilized resulting in 2->6 linked
a-D-fructofuranose polymer (ref. 30), shown as structure (18) in Fig. 9.
The NRRL B-512 strain of Leuconostoc mesenteroides is used generally
for the industrial production of dextran. The dextran contains ~95Z of a-
(l-»6)- and "51 of a-(l-*3)- linkages and has a molecular weight of 40-50 x
10 e .
Dextran is mainly used as a blood plasma extender. For clinical
purposes the native polymer is depolymerized to give dextrans of molecular
weight 40,000 and 70,000. Dextran can be used as a stabilizer of ice cream,
sugar syrup and other confectionery but is generally too expensive. The.
total world production is in the 100-200 tonnes per year with the
purification stage accounting for a high proportion of costs (ref. 31).
*-*ιο
17
HO
HO
/ MO HÔ O
-■»i
18
Fig. 9. Structures of dextran (17) and levan (18).
SPECIALTY SWEETENERS FROM SUCROSE

Isomaltulose. a non-cariogenic sweetener
Isomaltulose, 6-0-(a-D-glucopyranosyl)-D-fructose (shown in Fig. 10),
is a reducing disaccharide with half the sweetness of sucrose. It is known
in the literature as 'Palatinose' (Mitsui Sugar Company, Tokyo, Japan) and
as 'Lylose' (Tate & Lyle pic, Reading, England). Isomaltulose is produced
commercially by Mitsui Sugar Company by a process of transglucosylation of
sucrose by the immobilized enzyme from Protaminobacter rubrum (ref. 32). It
costs roughly two and half times as much as sucrose per kilogram in the
Japanese market. Isomaltulose crystallizes readily as a monohydrate and has
specific rotation [a] D + 97.3°.
Isomaltulose is completely hydrolyzed in the small intestine by a
disaccharidase to glucose and fructose. No adverse effects have so far been
reported. Isomaltulose is only slowly or not at all fermented by dental
plaque and oral streptococci (ref. 33). However, caries in pits and
fissures have often been encountered in some individuals probably due to
high acid production in the dental plaque.
The quality of sweetness of isomaltulose is claimed to resemble that of
sucrose. Because of its low sweetness and non-cariogenicity it has found
speciality application in candies, chocolate, chewing gum and cookies. Its
high cost as compared to sucrose has been the limiting factor in its
commercialization as a food additive.
382
19
Fig. 10. Isomaltulose.
Palatinit. a low calorie sweetener

Suddeutsche Zucker-AG of the Federal Republic of Germany has developed
a process for the production of Palatinit from sucrose. The process
involves microbial transglucosylation of sucrose to palatinose which is then
hydrogenated to give a mixture of 6-0-(a-D-glucopyranosyl)-D-mannitol (20)
and 6-0-(a-D-glucopyranosyl)-D-sorbitol (21) (ref. 34), shown in Fig. 11.
The sweetness of Palatinit is 0.45 times that of sucrose and the taste
profile is 'pure sweet' without the cooling effect of xylitol or sorbitol.
Its use as a substitute for sucrose has been demonstrated in chocolate
slabs, drops, bars, marzipan products and chewing gum (ref. 34).
Palatinit is claimed to be less cariogenic in rats than sucrose and
lactose. Strains of Streptococcus mutans do not produce much acid or
extracellular polysaccharides from palatinose (ref. 35). It has been shown
that Palatinit is partially absorbed in the small intestine, and the
proportion entering the large intestine is quantitatively fermented to
volatile fatty acids, carbon dioxide, methane and hydrogen. From
experiments with rats, pigs and man it is concluded that Palatinit has an
energy value of 2 KCal/g. Like palatinose, its commercial success will
depend upon its price as compared to other nutritive sweeteners.
r-OH
HO-j
/—°
20 2Λ
Fig. 11. Components of Palatinit: 6-0-(a-D-glucopyranosyl)-D-mannitol (20)

and 6-0-(a-D-glucopyranosyl)-D-sorbitol (21).
383
Neosugar. a non-cariogenic sweetener

Neosugar is a mixture of fructo-oligosaccharides with one, two or three
fructose residues linked by way of ß (l-*2) bonds to the fructosyl moiety of
sucrose, as shown in Fig. 12 (ref. 36). The sweetness intensity of neosugar
depends on the composition of the mixture. For example, when 95Z of the
mixture is comprised of nystose and kestose, its sweetness intensity is ~20Z
of that of sucrose. It is claimed that neosugar is a non-cariogenic and
non-nutritive sweetener. However, the latter claim has not been fully
substantiated. There is little doubt that even if it escapes hydrolysis in
the small intestine it will be degraded in the large bowel. Nystose, a
trisachcaride and a component of neosugar, has been shown not to provide D-
glucose or D-fructose for the body which makes it a potential candidate for
a bulk sweetener in diets for diabetic patients.
Neosugar is produced commercially by Meiji Seika Company of Japan. The
process involves the microbial transformation of sucrose using a fungal
fructosyltransferase enzyme (ref. 36).
fo ik 2}o
H Ή Ή
"Vît γ4 ί»»
Vî Ç*x
• a OH
2X
Fig. 12. Components of neosugar.
38il·
Sucralose. a non-nutritive sweetener

Sucralose (structure 11 in Fig. 4) is a derivative of sucrose. It is
chemically termed as 4-chloro-4-deoxy-o>D-galactopyranosyl l,6-dichloro-l,6-
dideoxy-ß-D-fructofuranoside (refs. 11, 37). It has a pure sweet taste
which is indistinguishable from that of sucrose. Sucralose is "650 times
sweeter than sucrose. It is resistant to enzymic hydrolysis and is sixty
times more stable than sucrose to acid hydrolysis. Sucralose will enter the
high quality intense sweetener market once regulatory approval has been
obtained. Sucralose is being developed by Tate & Lyle pic in collaboration
with Johnson and Johnson of the United States of America.
CONVENTIONAL CHEMICALS FROM SUCROSE--ETHANOL AND CHEMICALS THEREFROM

With present technology, ethanol is a key intermediate in the
conversion of sugar to chemicals. The chemical industry based on ethanol
feedstock utilizes the raw material by way of either ethylene, the
dehydration product, or acetaldehyde, the dehydrogenation product, to yield
a host of organic chemicals currently obtained from petroleum dervived
ethylene, as see Fig. 13 (refs. 38, 39). However, the conversions are
costly in energy terms and are limited economically by the fact that the
conversions are accompanied by a loss of mass inherent in carbohydrates
conversions (ref. 31). Approximately 3.8 kg of sugar is required to produce
1.7 kg of ethanol which yields ~1 kg of ethylene.
Nevertheless, in countries like Brazil, India and Pakistan, where
ethanol is produced from sucrose, conversion of ethanol to chemicals is low
in commercial practice (refs. 31, 39).
MISCELLANEOUS
Important organic acids used in food, detergent, pharmaceutical and
polymer industries can be derived from sucrose using chemical or
fermentation processes. These include citric acid, lactic acid, oxalic
acid, gluconic acid and itaconic acid. World production of citric acid is
~3,000,000 tonnes per year. The market for amino acids such as monosodium
glutamate (220,000 tonnes per year) and L-lysine (40,000 tonnes per year)
has been growing dramatically (ref. 31).
385
<
Ethylene glycol
Polyethylene
Vinyl chloride
..Acetic acid
Ethanol £■ - Acetaldehyde — ^ Ethyl acetate
Butadiene
Acetone Diacetone alcohol
Ethylamine
Fig. 13. Ethanol as a raw material for production of organic chemicals.
Sucrose can be converted to a variety of conventional 1,5-bifunctional

furans by way of 5-hydroxymethylfurfural in high yields, as shown in Fig. 14
(ref. 40). These furans could be used to produce polyether and polyester
fibers, polymers with such intrinsic properties of the furan ring as non-
flammability, thermal resistance and surface active products.
Sucrose
\
M
vOv°"
HOOC ° COOH 4 OWC ° CHO
OH
Fig. 14. Synthesis of furans from sucrose.

386
A VIEW OF THE FUTURE FOR SUGAR

In the developed world, use of sucrose in industrial products such as
Polyurethane foams is likely to increase. Products which are currently
projected to have low-volume markets could develop into high-volume products
if the sucrose price becomes attractive.
Polyhydroxybutyrate, a biopolymer, could become commercially attractive
as a general purpose polymer if sucrose is available cheaply (US$150 per
tonne). High-value, low-volume products such as high intensity sweeteners,
vitamins or antibiotics will find their own markets, because the cost of
sucrose as a raw material contributes little to the cost of the finished
product.
In the developing world, use of sucrose and its by-products for the
production of alcohol for energy and chemical purposes is likely to
increase. For oil importing countries with indigenous supplies.of sucrose
and triglycérides the detergent industries should evaluate sucrose fatty
acid esters and similar products.
CONCLUSION
The potential value of sucrose as a renewable carbon source for the
production of chemicals has been recognized but not yet significantly
realized. The reasons for this are mainly technical and economic. Sucrose
is multifunctional, relatively low molecular weight (342), crystalline
organic molecule. However, it has limited solubility in common organic
solvents and its reactions are difficult to control because of its
multifunctionality. Hence, processing and separation of reaction products
are often difficult and costly. On the economic front sucrose has to
compete widely with fossil carbon sources such as oil and coal. At present,
at least in the Western world, sucrose is unlikely to compete with
petrochemicals. However, in countries where it is cheap and in abundant
supply and petroleum is imported at a vast cost, the economics could favor
sucrose as a chemical feedstock.
REFERENCES
1 0. K. Kononenko and K. M. Herstein, Chem. Eng. Data Ser., 1 (1956) 87;
C. J. Moye and B. M. Smythe, Carbohydr. Res., 1 (1965) 284.
2 C. A. Beevers and W. Cochran, Proc. Roy. Soc. London Ser. A., 190
(1947) 257.
3 G. M. Brown and H. A. Levy, Science, 141 (1963) 921.
4 R. U. Lemieux and K. Bock, Jap. J. Antibiotics, 32 (Suppl.) (1979)
5163; K. Bock and R. U. Lemieux, Carbohydr. Res., 100 (1982) 63.
5 R. Khan, Adv. Carbohydr. Chem. Biochem., 33 (1976) 253; Pure & Appi..
Chem., 56 (1984) 833; L. Hough, Chem. Soc. Rev., 14 (1985) 357.
6 T. Otaka, Bull. Chem. Soc. Jpn., 43 (1970) 3199; 45 (1972) 289;
L. Hough, K. S. Mufti and R. Khan, Carbohydr. Res. 21 (1972) 144.
7 H. Karl, C. K. Lee and R. Khan, Carbohydr. Res., 101 (1982), 31.
8 E. Fenton, J. Gelas, D. Horton, H. Karl, R. Khan, C. K. Lee and
G. Patel, J. Org. Chem., 46 (1981) 4057.
9 R. Khan and H. Lindseth, Carbohydr. Res., 71 (1979) 327.
10 L. Hough, Amer. Chem. Soc. Symp. Ser., 41 (1977) 9.
11 L. Hough and R. Khan, Trends in Biochem. Sci., (1978) 61.
12 0. K. Konenko and I. L. Kestenbaum, J. Appi. Chem., 11 (1961) 7.
13 K. J. Parker, K. James and J. R. Hurford, Amer. Chem. Soc. Symp.
Ser., 41 (1977) 97.
14 H. B. Bass, F. D. Snell, W. L. York and L. I. Osipow, U.S. Pat., 2,
893, 990 (1959); Chem. Abstr., 53 (1959) 19422c.
15 F. Yamagishi, F. Endo, H. Ooi, and Y. Fozuka, U.S. Pat., 3,792,041
(1974); J. C. Colbert in "Sugar esters: preparation and application,11
Noyes Data Corporation, New Jersey, 1974, p. 30.
16 K. J. Parker, R. Khan and K. S. Mufti, Brit Patent 1,399,053 (1975);
U.S. Pat., 3,996,206 (1976); Chem. Abstr. 82 (1975) 100608r;
H. R. Galleymore, H. F. Jones, K. James, J. S. Plant and C. L.
Bhardwaj, Brit. Pat. GB 2,065,734B (1983).
17 H. J. W. Nieuwenhuis and G. M. Vianen, European Patent Application,
190 779 Al (1986).
18 J. B. Sheridan, D. Mehale, G. G. Birch and E. B. Rathbone, U.K. Pat.,
GB 2,048,251 B (1983).
19 E. P. Lira and R. F. Anderson, Amer. Chem. Soc. Symp. Ser., 41 (1977)
223.
20 C. H. Coney, Amer. Chem. Soc. Symp. Ser., 41 (1977) 213.
21 R. W. Boggs, Fette. Seifen. Anstrichmittel, 88 (1986) 154.
22 A. G. Perotti, International Flavours Food Additives, 8 (1977) 149.
23 H. S. Tan and D. A. Smink, British Pat. Specification, 1593856 (1981).
24 J. F. Davies, R. S. Lee, A. W. Travili, and R. J. P. Williams, U.K.
Pat., Application GB 2 174 712 A (1986).
25 W. Fritschela, E. Leupold and M. Schilingman, E. P. 218,150 A (1987),
DE 3535720 (1987).
26 K. C. Frisch and J. E. Kresta, Amer. Chem. Soc. Symp. Ser., 41 (1977)
238; A. R. Meath and L. D. Booth, ibid.. 41 (1977) 257.
27 Anonymous, Rubber and Plastics News, Oct., 19 (1987) 6.
28 Anonymous, Zuckerind., Ill (1986) 514.
29 P. P. King, J. Chem. Tech. Biotech., 32 (1982) 2; E. R. Howelles,
Chemistry and Industry (1982) 508; R. M. Lafferty and G. Braunegg,
DE-OS 3,343,576 A, 1 (1983).
30 P. T. Murphy and R. L. Whistler, in: Industrial Gums, R. L. Whistler
and J. N. BeMiller (Eds.), Academic Press, New York, 1973, pp. 513-542;
R. Khan, J. K. Wold and B. S. Paulsen in: Rodd's Chemistry of Carbon
Compounds, M. F. Ansell (Ed.), Elsevier Science Publishing Company,
Amsterdam 1983, Vol. Supplement IFG, pp. 330-331.
31 A. J. Hacking, in: Economic Aspects of Biotechnology, Cambridge
University Press, Cambridge, 1986, p. 112.
32 K. Suzuki, New Food Industry, Jpn., 26 (1984) 14; C. Bucke and
P. S. J. Cheetham, U.K. Pat. Application, GB 2,063,268 A (1981);
M. Munir, EP 91063 (1983); P. Egerer, W. Haese, H. Perrey and
G. Schmidt-Kastner, EP 160,260 A (1985); P. Egerer, W. Crueger and
G. Schmidt-Kastner, EP 200,069 A (1986).
33 D. Birkhed, I. Takazoe and G. Frosteil, Dtsch. Zahnarzt1 Z., 42
(1987) S 124.
34 G. Siebert and W. Grupp, in: Health and Substitutes, Proc. ERGOB
Conference, Geneva, B. Guggenheim (Ed.), Karger, Basel (1978) 109;
H. Schiweck, ibid.. (1978) 138; H. Schiweck, G. Steinle and L. Haberl,
US Pat., 3,865,957 (1975); T. H. Grenby, in: Developments in
Sweeteners-2, T. H. Grenby, K. J. Parker and M. G. Lindley (Eds.),
Applied Science Publishers, London, (1983) 51.
35 G. Siebert, Dtsch. Zahnarztl. Z., 42 (1987) S 128.
36 S.C. Zieseunitz and G. Siebert, Amer. Inst. Nutr. (1987) 846; T. Oku,
T. Tokunaga and N. Hosoya, J. Nutr., 114 (1984) 1574; Anonymous,
Nutr. Rev., 43 (1985) 155.
37 L. Hough, S. P. Phadnis, R. Khan and M. R. Jenner, Brit. Pat.,
1,543,167 (1979); U.S. Pat., 4,549,013 (1985); R. Khan and A. J. Forage,
Sugar Technol. Rev., 7 (1979/80) 175.
38 T. K. Ghose, CHEMRAWN III, World Conference on Resource Material
Conversion, The Hague, The Netherlands, June 25-29 (1984) Section 3.
III. 3; C. Ratledge, ibid.. Section 3.II.2.
39 M. R. Adams and G. Flynn, in: Fermentation Ethanol, an Industrial
Profile, T.D.R.I., London, 1982, p. 10.
40 T. E. Hajj, A. Masroua, J.-C. Martin and G. Descotes, Bull Soc.
Chim. France, (1987) 855.
INDEX
Accurel polypropylene, 281

Acetic acid, 24, 28, 102, 165, 242, 340
Acid beverage floe, 300
Acidogenesis, 103
Acids, 29
Acids, Kreb's cycle, 24
Adsorbents, 173
Aerobic treatment, 101
Affination, 171
Affinities, anions, 50
Affinities, cations, 50
Alcohol, 169
Alcohol floe, 300
Alcohols, 28
Aldehydes, 143, 241, 245
Algae, 54
Alkaline degradation, 26, 29, 30
Alkaline degradation products, 192, 200, 203
Alkaline oxidation, 29
Alkalinity, 26
Alumina, 21, 249
Amadori, 30, 298
Amadori rearrangement, 30
Amides, 140, 141
Amines, 30
Amino acid, diffusion juice, 25
Amino acids, 111, 120, 140, 141
Amino acids in sugar colorant, 198
Amino N, 249
Amino nitrogen, 23
α-Amino nitrogen, 150, 151
Amino-nitrogen, in molasses, 332
Amino-nitrogen, in sugarcane, 334
Ammonia, 29, 102
Ammonium, 23
Amylase, 216
Amylopectin, 209, 226
Amylose, 209, 226
Anaerobic treatment, 101
Analytical control, 148
ANAMET system, 103
Anthocyanins, in sugarcane, 194
Antioxidants, 299
Apigenin, 195
Araban, 142
Arabic, gum, 142
Arabinose, 149
Aristotle, 46
Aroma, absorption, 247
Aroma, in solution, 248
Ash, 140
390
Ash removal, by clarification, 276

Attenuation index, 187
Automation, 147
Bacteria, 54
Bagacillo, 171
Bagasse, 166, 168, 169
Baking, 299
Barbados sugar, 237
Beet gum, 142
Beet industry, historical origins, 1, 21
Beets, frozen, 97
Beets, rotten, 97
Behavior, 309
Bentonite, 21, 249
Betaine, 25, 88, 114, 140, 141
Beverages, 25, 300
BIO-ACT, 106, 107
Bioaugmentation, 105, 106, 107
Biochemical Oxygen Demand, 97
Biocides, 106
Biotin, 120, 122
Bisulfite, 31
Bisulfite addition product, 31
Blackstrap molasses, 167, 168, 243
Blue number, 150, 151
BOD, 97, 98, 101, 107
Bone char, 172, 202
Bone char process, 269
Bounty hunter, 109
Brewers grain, 118
Brix, decolorization effect, 65
Brown sugar, 171, 237, 243
Brown sugar, color, 188
Browning, enzymatic, 30
Bub stage in rum production, 321
Bulking agent, 297
Butyric acid, 102
Byproduct isolation, 111
Byproducts, 169, 356
Byproducts, beet, 8
Caffeic acid, 194

Cakes, 299
Calcium, 23, 26, 134, 135
Calcium carbonate, 27, 30, 35, 98
Calcium oxalate, 24
Calcium precipitation, 29
Calcium saccharate, 166
Calcium salts, 6
Calcium salts, insoluble, 24, 25
Calcium sucrate, 112
391
California Safe Drinking Water and

Enforcement Act of 1986, 109
Cane, stale, 211
Cane, standover, 212
Cane juice, base for rum, 315
Cane juice, treatment for fermentation, 351
Canesorb process, 270
Caramelization, 298
Caramels, 63, 189, 197
Carb elimination, 34
Carbon, granular activated, 270
Carbon, powdered vegetable, 271
Carbon adsorbents, buffering action, 270, 271
Carbon dioxide, 6, 20, 25, 35
Carbonatation, 20
Carbonatation, cane juice, 284
Carbonatation (carbonation) liquor, 267
Carbonate-bicarbonate equilibrium, 27
Carbonation, 6
Carboxylic acids, by ion Chromatograph, 154
Caries, dental, 308
Cations, in raw beet juice, 23
Cell wall polysaccharide, 215
Cell walls, 5, 22
Cellulose, 140, 142, 176, 208
Ceramic membranes (see membranes)
Chaulk, 21
Chemical profile, 245
Chicory, 127, 134, 137
Chloride, 23
Chloride, behavior on resin, 67
Chlorinated sucrose, 371, 372
Chlorine, 105
Chlorogenic acid, 194
Choline, 25
Chromatographie desugarization, 87
Chromatography, 86
Chromatography, large scale, 113
Chromium and diabetes, 310
Citrate, 29
Citrate, color inhibitor, 200
Citric acid, 6, 24, 141, 340, 385
Clarification, 6, 166, 171
Clarification, juice, 282
Clarification, phosphatation, 266
Clarification, refinery, 266
Clarification process, 265
Clarifier, design, 267
Clarifiers, cane juice, 285
Clarity, 20
Clay soils, 21
Clays, 27
Cloudiness, 21
Cogeneration, 6
392
Condenser water, 98
Consumer, 306
Continuous processing, 36
Conversion, of factory, 69
Cold main liming, 37
Cold preliming, 37
Colloids, 6, 20, 27, 29, 54
Color, 20, 24, 30, 172, 298
Color, filtration in measurement, 188
Color, measurement of, 187
Color, measurement, pH in, 188
Color, removal by clarification, 276
Color, white sugar, 63
Color formation, 30
Color forming reactions, 31, 191, 203
Color measurement, reflectance, 188
Colorant, 170
Colorant, amino acids in, 198
Colorant, in crystal, 203
Colorant, high molecular weight, 188, 190, 197, 203
Colorant, iron and, 201
Colorant, low molecular weight, 189, 191, 198
Colorant, in molasses, 199, 203
Colorant, pH effect, 190, 192, 203
Colorant, phosphate and, 201
Colorant, polarity, 193
Colorant, structure, 191, 195
Colorant, tests for, 202
Colorants, fractionation, 189, 192, 203
Colorants, phenolic, 189, 193
Colorants, raw sugar, 186 193
Colorants, refined sugar, 13-1
Colorants, sugarcane, 186 193
Colors, white sugars, 43
Composition, sugarbeet, 8
Confectionery, 301
Copper complex method, 151
Cow's blood, 21
Coulter counter, 160
Counter diffusion, 341, 347
Counter diffusion of molasses, 340
Crystal distortion and oligosaccharides, 230
Crystal elongation, 24
Crystal elongation, C-axis, 229
Crystal growth, retarded, 24
Crystal size distribution, 159
Crystallization, 167, 168
Dairy, 300
Décalcification, 51, 69
Décalcification, chemistry, 57
Décalcification, conventional process, 56
Décalcification, Gryllus process, 59
Décalcification, N.R.S. process, 61
Decolorization, 63, 172
Decolorization, beet juice, 68, 69, 89
Decolorization, bone char, 269, 202
Decolorization, brix effect, 65
Decolorization, cane juice, 284
Decolorization, comparison, 280
Decolorization, effect of anion concentration,
Decolorization, granular activated carbon, 270
Decolorization, ion-exchange resin, 202
Decolorization, powdered carbon, 271
Decolorization, refinery, 269
Decolorization, resin, 276
Decolorization, SURE process, 281
Decolorization, thick juice, 71
Decomposition of sucrose, 254
Degradation, acid, 253
Degradation, alkaline, 29, 253
Degradation, sucrose, 253
Degradation, thermal, 253
Degradation of sucrose, catalysis, 254
Demerara, manufactured, 237
Demerara sugar, 237
Demineralization, 80, 81, 82
Demineralization, mixed bed, 81
Demineralization, partial, 84, 85, 89
Density, 157, 158
Desalting, 343
Desludging of cane juice, 351
Desugaring, molasses, 112
Desugaring, processes, 111
Desugarization, Chromatographie, 87
Detergent, 376
Deterioration, post-freeze, 182, 183
Deterioration products, sugarcane, 181
Dextran, 24, 28, 141, 181, 220, 292, 340, 380
Dextran, removal in process, 225
Dextrans, effect on polarization, 223
Dextrans, production, 220
Dextrans, structure, 221
Diabetes, 309
Diet, 174
Diffusion, 164
Diffusion, counter current, 5
Diffusion juice, 5
Dihydroxyphenyalanine, 30
Dimethylsulfide, 165, 239
Discharge sources, 97
Distillation, continuous, 323
Distillation, rum, 322
Donnan membrane effect, 87
Donnan membrane theory, 87
Dunder, in rum production, 322
394
Economies, of desugarization, 121

EDTA, 132
Electricity, 169
Electrodialysis, 347
Environmental concerns, 108
Environmental Protection Agency, 96
Enzymatic analysis, 152
Enzymatic browning, 30
Enzymatic browning reactions, 196
Enzymes, in sugarbeet, 22
EPA, 96
Equilibrium relative humidity, 294
Esters, in rum, 320
Esters, of sucrose, 373
Esters, of sucrose, applications, 374
Ethanol, 343
Ethanol, cost of production, 340
Ethanol, production in Brazil, 349
Ethanol, production from cane juice, 351
Ethanol, production from sucrose, 384, 385
Ethylene, 373, 384
Extraction, flavor, 238
False pol, 223

False seed, 20
Feed, animal, 118, 119
Feedstock, sucrose as chemical, 373
Fermentation, 28, 298
Fermentation, aerobic, 340
Fermentation, anaerobic, 340
Fermentation, continuous, 357
Fermentation, continuous process, 354
Fermentation, Meile Boinot process, 354
Fermentation efficiency, increase by desalting, 346
fermentation of molasses, 340
Fermentation technology, recent advances, 363
Fermenting fractions, 120
Ferric ion, in color formation, 201
Ferric oxidation, 54
Ferrous ion, in color formation, 201
Fiber, 142, 176
Fiber, dietary, 142, 143
FIBREX, 143, 144
Filter staton, DDS, 39
Filterability and polysaccharides, 225
Filtrate clarification, cane juice, 287
Filtration, cane mud, 283
Filtration, liquor, 268
Filtration cycles, 25
Filtration rate and polysaccharides, 226
Final molasses, 167
Finnsugar-Pfeifer & Langen process, 113
First carbonation, 32, 37
395
Flame photometry, 150

Flavonoids, 191, 193
Flavor, 171, 236
Flavor, cane juice, 238
Flavor, molasses, 243
Flavor, precursors, 240
Flavor, problems, 242, 248
Flavor, process-derived, 242
Flavor, sources of, 238
Flavor in colorant compounds, 199
Flavor enhancer, 297
Flavors, 165
Flavors, isolation of, 238
Floe in beverages, 300
Flocculant, 166
Flocculating agents, 272, 285
Flocculation, 272
Flocculation, of yeast, 361
Fluoréscarnine, 151
Fluorescence of colorants, 196
Fluorimetrie measurements, 150
Flume water, 97
Flume water disposal system, 99
Fodder, 128
Fodder value, 141
Food technology, 311
Formic acid, 24, 28
Fouling, 54, 129, 131, 137
Fouling, resin, 47, 54
Fructose, 23, 28
Fructosyltransferase, 383
Fuel alcohol, 175, 359, 373
Furans, synthesis from sucrose, 385
Fusel oil, 359
Galactan, 142
Galactomannans, 214, 218, 219
Galacturonate, 29
Galacturonic acid, 24
Gas liquid chromatography of sugars, 212, 217
Gel permeation chromatography, 213, 214, 217
Gels, resin, 47
Genetic variance, sugarbeet, 16
Geosmin, 143, 145
Glucan, Roberts', 216
Gluconic acid, 385
Glucose, 23, 28, 30
Glutamic acid, 141
Glutamine, 25, 30, 141, 153
Gluten, 143
Glycémie index, 310
Glycerol, 361
Glycine, 25, 30
396
Glycosides, 24
Glyphosate, 179, 180
Grain size measurement, 160
Gryllus décalcification, 79
Gryllus process, 59
Gum arabic, 142
Gums, 24, 27, 141, 176
Harvest, sugarbeet, 4
Heart disease, 311
Helianthus tuberosis, 127
Hemicellulose, 140, 142, 208
Herbicide, 3
Heritability estimates, 15, 18
High fructose syrup, 127, 137
Hot main liming, 37
Humectancy, 295
Humic acids, 55
Hybrids, interspecific, 176
Hydrocloning of cane juice, 352
Hydrogen peroxide, 106, 204
Hydrogen sulfide, 102, 105, 106, 107
Hydrolysis, 102
Hydrolysis, rate constants, sucrose, 22
Hydrolysis of sucrose, 22
Hydromethylfurfural, 254
Hydroxy-acids, 143
Hydroxyapatite, 269
Hydroxybenzoic acids, 165, 194
Hydroxymethylfurfural 200, 386
Inclusions, crystal, 20
Indicator value (I.V.)· 190
Indigenous sugarcane polysaccharide, 214
Infrared, 155
Inositol, 120, 122
Insolubles, 140
Interfering nitrogen, 25
Inulin, 127
Inulin, hydrolysis, 129
Interaction, flavor, 246
Inversion, 22, 181
Invert, 168
Invertase, 22, 166
Inversion, 73
Ion chromatography, 152
Ion exchange, 28, 46, 111
Ion exchange, of molasses, 347
Ion exchange properties, beet pulp, 46
Ion exchange resins, 276
Ion exchange resins, properties, 47
Ion exchange resins, stability, 52
397
Ion exclusion, 87
Iron, 54, 201
ISP, structure and occurrence, 214
Isoamylase, 216
Isomalt, 146
Isomaltulose, 381
Jellies, 25
Jerusalem artichoke, 127, 135, 137
Juice, treatment, 353, 354
Kestoses, 256, 383

Kieserite, 79
Krebs-cycle, 153
L-LDH, 154
Lactate, 154
L-Lactate dehydrogenase, 154
Lactic acid, 24, 28, 102, 340, 385
Lag phase, in sucrose degradation, 255, 257
Leuconostoc, 181
Leuconostoc dextranicum. 221
Leuconostoc mesenteroides. 211, 221
Levan, 24, 28, 380
Levoglucosan, 197, 254
Levulinic acid, 200, 257
Licorice flavor, 243
Lignin, 140, 176
Lime, 6, 20, 21
Lime cake, 98
Lime pond, 98
Limerock, 25, 35
Liming dose, 268, 283
Lipids, 140, 141
Lobry de Bruyn-Alberda van Ekenstein rearrangement, 200
Loss, unknown, 167
Luteolin, 195
Lylose, 381
Macroporous resins, 47
Magnesite, 270
Magnesium, 23, 28, 76
Magnesium chloride, 78
Magnesium ions, effect on sucrose decomposition, 261
Magnesium hydroxide, 21, 54
Magnesium salts, deposits, 53
Maillard products, 143
Maillard reaction, 30, 197, 241, 298
Malate, 29
Malate, behavior on resin, 67
398
Malic acid, 24
Maturity classifications, sugarcane varieties, 178
MAZON BIO-ACT, 106
MBP 10, 118
Meat, 300
Melanins, 30, 31, 63
Melanoidins, 30, 63, 196, 199
Melassigenic coefficient, 168
Melassigenic effect, 232
Melassigenic properties, 53
Melassigenicity, 67, 68
Melassigenesis, magnesium, 76
Meile Boinot, continuous process adaptation, 357
Meile Boinot process, 354, 356
Membrane filtration, 127
Membrane technology, 168
Membranes, ceramic, 136, 137, 138
Membranes, counter diffusion, 342
Membranes, polymeric, 136, 137, 138
Membranes, tubular, 129
Metallic taste, 248
Methane, 103
Methanogenesis, 103
Milk of lime, 6, 25, 26
Milk producing capacity, 143
Mill white sugar, 164
Mill sugar production, 284
Mineral content, sugarcane juice, 183
Molassed Beet Pulp 10, 118, 119
Molasses, 20, 243, 341
Molasses, base for rum, 315
Molasses, beet, composition, 114
Molasses, clarification, 319
Molasses, decomposition with heat, 317
Molasses, desalting, 340
Molasses, desluding, 319
Molasses, fermentation, 320
Molasses, fermentation and amino nitrogen, 329
Molasses, foaming, 329
Molasses, glucans in, 213, 217, 219
Molasses, pasteurization, 319
Molasses composition, 341
Molasses as fermentation feedstock, 340
Molecular sieve effect, 86
Monogerm, 3
Monosaccharide, 30
Moses, 46
Muds, 167
Muscovado, 237
NAD, 154
NAOH, 154
Naantali desugaring plant, 111 121, 124, 125
399
National Pollutant Discharge Elimination System, 98

Needle-grain, 181
Neosugar, 383
Nephelometry, 188
New Regeneration System, 61
Nitrate, 23, 28, 30
Nitrate, behavior on resin, 67
Nitrate determination, 150
Nitrite, 23, 28, 30
Nitrogen, 99, 102
Nitrogen reduction, 28
Nitrogen, interfering, 25
Noble varieties, sugarcane, 176
Non-cariogenic sweetener, 381, 383
Non-sucrose, 23, 25, 28, 33
NPDES, 98
N.R.S. process, 61
Nutrient packages, 120, 122
Nystose, 383
Obesity, 308
Odor, 105
Odor, beet pulp, 143, 145
Off-flavor, 145
Oleanoic acid, 141
Oligosaccharides, 230
On-line measurement, 147
On-line purity measurement, 155
Organic acid formation, 102
Organic acids, 140, 166
Organic non-sugars, 140
Osmotic pressure, 295
Osmotic shock, 48
Oxalate, 29, 167
Oxalic acid, 6, 24, 141
Oxidation, alkaline, 29
Painted sugar, 237

Palatinit, 146, 382
Palatinose, 381, 382
Pantothenic acid, 120, 122
Parental population, sugarbeet, 11
Pasteurization of cane juice, 352
Pectic acids, 24
Pectin, 6, 24, 29, 140, 141, 142
Pectin substances, 22
Pectins, 114, 134
pH, 20, 21, 26
pH adjustment, 106
pH control, 22, 74, 167
pH optimum, 27
Phenol oxidase, 30
400
Phenolic acids, 165

Phenolics, 184
Phosphate, 166
Phosphates, 6, 23, 29, 99
Phosphorus, 184
Phosphatation, 267
Photosynthesis, 163
o-Phthalaldehyde, 151
Phytic acid, 143
Plant pigments, 189, 193, 194
Plantation white sugar, 163, 174
pOH, 23, 31
Poisoning, of resin, 54
Polarimetry, 146
Polarization, effect of polysaccharides on, 223
Polyacrylamide flocculant, 166
Polyacrylamides, 272, 285
Polyfructan, 127, 128
Polyhydroxybutyrate, 379, 386
Polymers, from sucrose, 378
Polysaccharide CP, 217
Polysaccharides, 170
Polysaccharides, microbial, in sugar, 208
Polysaccharides, in sugarcane, 208
Polyurethane, 378
Polish test, 67
Potassium, in beets, 23
Potassium, in cane juice, 183, 184
Potassium, removal from molasses, 341
Potassium chloride, melassigenic effect, 168
Potassium determination, 150
Potassium permanganate, 106
Powdered resins, 72
Prechlorination, 105
Precipitation, with calcium, 29
Preliming, 32
Preservatives, 297
Preserves, 300
Pressure drop, 51
Pressure filtration, 268
Proalcool program, 349
Production, refined beet sugar, 1, 3
Production, sugarbeet, 1
Propionic acid, 102
Proposition 65, 109
Protein, 6, 22, 29, 134, 140, 141, 165
Protein degradation, 23
Pullulan, 212
Pullulanase, 216
Pulp, beet, 5, 46, 118, 142
Pulp, beet, pelletized, 119
Pulp, moistures, 5
Purification, beet, 6
Purification, juice, 35
401
Purity, 21
Purity, beet juice, 53
Purity, factors affecting, 9, 10
Purity, molasses target, 67
Pyrazines, 143
2-Pyrrolidone-5-carboxylic acid, 29
Pyrroline-carboxylic acid, 153
Quality, sugarbeet, 9
Quality control, 307
Quentin plant, 77, 78
Quentin process, 75, 89, 112
Quentin process, chemistry of, 76
Rate constants, sucrose hydrolysis, 22

Raffinose, 23, 114, 120, 140, 149, 292
Raw sugar, colorants in, 186, 191
Raw sugar, polysaccharides in, 215, 217
Raw sugar, production, 164
Raw sugar, refining, 170
Raw sugar, starch in, 209
Reactions, ion exchange, 48
Reactions, kinetics of ion exchange, 51
Reactivity, sucrose, 369
Recovery, sucrose, 174, 204
Recycling, 168
Reducing agent, 28
Refinery decolorization, 269
Refractometry, 146
Régénérants, alternative, 78
Regeneration, SURE, 282
Remelt, 174
Residual molasses, 117, 121
Resin, acrylic, 47, 64
Resin, combinations, 80, 81
Resin, fouling, 54, 276
Resin, fouling by ferric oxide, 54
Resin, fouling by magnesium hydroxide, 54
Resin, powdered, 72, 73
Resin, structure, 47, 48
Resin, styrenic, 47, 64
Resin decolorization systems, 276
Resin fouling, irreversible absorption, 55
Resins, amino, 47
Resins, carboxylic acid, 47
Resins, gel-type, 276
Resins, gel, 47, 48
Resins, macroreticular, 276
Resins, powdered ion-exchange, 281
Resins, quaternary, 276
Resins, quaternary acrylic, 277
Resins, quaternary ammonium, 47
402
Resins, styrene-divinylbenzene, 276

Resins, sulphonic acid, 47
Resins, macroporous, 47
Ripeners, chemical, 176, 179
Roberts* glucan, 216
Rotary vacuum filters, 286
Rum, 169, 175
Rum, history, 313
Rum, qualities and characteristics, 328
Rum feedstock, cost comparison, 317
Rum maturation, 327
Rum production in Caribbean, 314
Rumpler, 46
Saccharate complexes of calcium, 27

Saccharates, 27
Saccharum, 163
Saccharum officinarum, 176
Safety, of sucrose, 303
Salts, effect on sucrose decomposition, 257
Saponin, 6, 29
Saponins, 24, 140, 141, 143
Sarkaran, 211
Scaling, 98
S c h i f f s base, 30, 298
Screening of cane juice, 352
Scum desweetening, 267
Second carbonation, 32, 37
Seed, monogerm, 3
Selection, Amino N, 10
Selection, extractable sucrose, 12, 15, 16, 17
Selection, high, 14
Selection, K + , 10
Selection, low, 13
Selection, Na-*-, 10
Selection, non-sucrose, 9, 15
Sensory profile, 243
Sieve analysis, 159
Silicates, 167, 171
Slaking process, 25
Sludge, 32, 40
Sludge separation, 32
Soda ash, 25
Sodium, 28
Sodium, in beets, 23
Sodium chloride, effect on decomposition, 257
Sodium determination, 150
Sodium hypochlorite, 55, 204
Soft sugar, 171, 174, 237
Solids levels, 21
Solubility, calcium salt, 29
Sorbitol, 382
Soy bean meal, 21
403
Stability of ion exchange resins, 52

Stale cane, 211
Standover cane, 212
Starch, 208, 225
Steffen process, 113
Steffen sucrate treatment, 112
Stench, 105
Still, design for rum production, 322
Stillage, disposal, 359
Stillage, recycle, 359, 363
Storage, beets, 4
Strecker aldehydes, 241, 245
Strecker degradation, 298
Strongly acidic cation exchange resin, 47, 48
Strongly basic anion exchange resin, 47
Subsiders, cane juice, 285
Sucralose, 384
Sucrâtes, 27
Sucrose, acetalation, 370
Sucrose acetylation, 369
Sucrose amorphous, 253, 255, 262
Sucrose antioxidant properties, 299
Sucrose bulking agent, 297
Sucrose chemical reactivity, 369
Sucrose chlorination, 371
Sucrose color formation by, 298
Sucrose crystal habit, 292
Sucrose crystal shape, 229
Sucrose crystalline, 292
Sucrose crystallization, 257
Sucrose degradation, 256, 259
Sucrose effect on water activity, 294
Sucrose feedstock for polymers, 377
Sucrose fermentation, 298
Sucrose flavor enhancer, 297
Sucrose functional properties, 296
Sucrose in methyl sulphoxide, 254
Sucrose monofatty acid esters, 375
Sucrose osmotic pressure, 294
Sucrose physical properties, 292, 367
Sucrose polyhydroxypropyl ether, 378
Sucrose preservative effect, 294, 297
Sucrose renewable resource, 367
Sucrose safety of, 303
Sucrose solubility, 293, 367
Sucrose structure, 368
Sucrose sweetness, 296
Sucrose texturizer 297
Sucrose viscosity, 293
Sucrose anions, 27
Sucrose in baking, 299
Sucrose in beverages, 300
Sucrose in chemical feed stock, 373
Sucrose and dental caries, 308
404
Sucrose and equilibrium relative humidity, 295

Sucrose ester based detergent, 376
Sucrose esters, synthesis, 374
Sucrose and health, 303
Sucrose and heart disease, 311
Sucrose and humectancy, 295
Sucrose hydrolysis, 258, 262
Sucrose monoacetate, 374
Sucrose monofatty acid esters, 375
Sucrose and nutrition, 303
Sucrose and obesity, 308
Sucrose polyester, 377, 378
Sucrose in polyurethanes, 378
Sucrose and rum, 313
Sugar and behavior, 309
Sugar and diabetes, 309
Sugar loss, in stored molasses, 317
Sugar losses, ethanol production, 360
Sugarcane, burning, sucrose ion, 253
Sugarcane, colorants in, 186, 191, 193
Sugarcane, composition of parts, 178
Sugarcane, deterioration, 181
Sugarcane, juice, composition, 177
Sugarcane, sarkaran in, 211
Sugarcane, starch in, 209
Sugarcane, tonnage, 165
Sugarcane composition, 162
Sulfate, 24, 29, 30
Sulfate, behavior on resin, 67
Sulfates, 6
Sulfitation, 33
Sulfite, 28, 29, 30
Sulfite decolorization, 202
Sulfur, 102
Sulfur dioxide, 204
Sulphite, behavior on resin, 67
Sulphitation, cane juice, 284
Supersaturation, 155, 157
Surfactants, 272, 285
SURE, 172
SURE decolorization process, 281
Sweetener, low calorie, 382
Sweetener, non-cariogenic, 381, 383
Sweetener, non-nutritive, 384
Sweeteners, based on sucrose, 380
Sweeteners, new, 312
Sweeteners, non-nutritive, 384
Sweetness, 296
Sweetness, enhancement, 247
Sweetness, relative, 297
Sweetness, suppression, 247
Syrup clarification process, 288
405
Talocarb, 273
Talofloc process, 204, 272
Taloflote, 272
Tare house, 148
Target molasses purity, 67, 68
Taste, beet pulp, 143, 145
Temperature, effect on sucrose, 255
Texturizer, 297
Thin juice, 6, 20
Thick juice, 6
Titratable acidity, 182, 183
Total suspended solids, 107
Tricarboxylic acid sucrose, 377
Trichlorogalactosucrose, 371
Tricin, 195
Trimethylglycine, 25, 141
Trisaccharides, 256
Turbidity, 173, 187
Turbidity removal, 288
Turbinado sugar, 237, 243
Tyrosine, 30
Ultrafiltration, 128, 129, 130

Unknown loss, 167
Urethane, 378
Varietal, differences, sugarcane, 176, 179

Viscosity, 181, 293, 294
Viscosity, effect of polysaccharides, 224
Visual color, 187
Volatile acids, 28
Volatile profile, 244, 245
Volatiles, 239
Waste, composition, 100, 101

Waste water, 101
Water, treatment, 75
Water activity, 294
Water binding capacity, 295
Water holding capacity, 143
Wavelength, color measurement, 187
Weakly acidic cation exchange resin, 47, 48
Weakly basic anion exchange resin, 50
Weed control, 3
White sugars, comparison, 290
White-end refinery process, 273, 285
XAD-2, 246, 249

406
Yeast, 112
Yeast, recycling in fermentation, 357
Yeasts, species for rum production, 320

Chemistry and Processing of Sugarbeet and Sugarcane Volume 9 PDF

Uploaded by

Copyright:

Available Formats

You might also like

Chemistry and Processing of Sugarbeet and Sugarcane Volume 9 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chemistry and Processing of Sugarbeet and Sugarcane Volume 9 PDF

Uploaded by

Copyright:

Available Formats

SUGAR SCIENCES

Vol. 1. Standard Fabrication Practices for Cane Sugar Mills (Delden)

Margaret A. Clarke and Mary An Godshall

Amsterdam — Oxford — New York — Tokyo 1988

Distributors for the United States and Canada:

ELSEVIER SCIENCE PUBLISHING COMPANY INC.

© Elsevier Science Publishers B.V., 1988

Printed in The Netherlands

In 1987, the American Chemical Society held its spring meeting in

The overall effect has been an apparent surplus of sugar, now

The purposes of this book, and of the symposia, are:

1. To present recent developments in beet and cane sugar manufacturing

2. To present the chemistry of sugar processing and products

3. To consider trends and future possibilities in sugar production

To these ends, the papers herein, by various international authorities

Several papers—those on sucrose chemistry (Chapters 16, 18, and 23),

The editors wish to express thanks to the Officers of the Agricultural

We offer our gratitude and appreciation to the contributors to this

We extend special thanks to our editorial assistant, Jacqueline E.

February, 1988. Margaret A. Clarke

VISTI ANDERSSON, DDS Engineering, P. 0. Box 2249, DK-1019, Copenhagen K,

MICHEL S. LEFEBVRE, Syrinx Research Institute Pty. Ltd., MLC Centre,

AN OVERVIEW OF THE U.S. BEET SUGAR INDUSTRY

"V IWvl Siifrir I'Uils

. Clin· Su>vw Mills

Fig. 2. Harvested acres and yield/acre vs years.

Fig. 3. Pounds of refined sugar/acre vs crop years.

Fig. 4. Crop year vs cwt refined sugar.

associated with weed control. Sugarbeets are grown on irrigated land

SUGARBEET LONG TERM STORAGE

Sugar loss in stored sugarbeets, caused by live root respiration and

semi-soluble colloidal proteins, pectins, and saponins extracted from the

^WWWW/ NONSUGARS 2.6%

Fig. 5. Harvested sugarbeet composition percent on total weight.

F i g . 6. Sugar r e c o v e r e d , sugar l o s t and by-products produced from

1 R. A. McGinnis, Beet Sugar Technology, published by the Beet Sugar

EFFECTS OF PLANT BREEDING ON SUGARBEET COMPOSITION

Considerable attention in the sugarbeet industry has been focused on

MATERIALS AND METHODS

parental population were field evaluated in replicated tests for three

Low selection High selection

Na 152.75* 120.69c 130.00c 200.73c **

K 126.37x 104.23y 93.17z 169.08w

Amino N 32.98o 22.70p 20.91p 50.00n 63.35mm

-*- Values are in milligrams/100 ml of raw juice.

**Data not obtained.

Effect of selection for Na"*-. K+ and amino N on other characters. The

Parental First cycle selection Second cycle selection

Na 152.75b*** 120.69c 200.73a 130.00c

** Biochemical values are expressed as mg/100 ml. Extractable sucrose X and

LOW AMINO-N SELECTION Ç

effects on X extraction. A different pattern was seen following selection for

Parental First cycle selection Second cycle selection

K 126.37b*** 104.23c 169.08a 93.17d

Biochemical values are expressed as mg/100 ml. Extractable sucrose Z and

sucrose or purity and not on the effects on other important non-sucrose

Parental First cycle selection Second cycle selection

Amino N 32.98c*** 22.70d 50.00b 20.91d 63.35a

Selection group mean* Resulting population mean-"·