Lebensm.-Wiss. n.-Techno!.

, 24,125-129 (1991)

Characterization of Coffee Pectin

R. Garcia, D. Arriola, M. C. de Arriola, E. de Porres and C. Rolz

Central American Research Institute for Industry (ICAITI), P.O. Box 1552, Guatemala (Guatemala)
(Received February 10, 1989; accepted July 17, 1990)

Pectin from three coffee (c. arabica) varieties, namely Bourbon, Caturra and Catimor, was extracted and purified. Its
physico-chemical characteristics were determined. The coffee pectin was of low methoxyl and of relatively low average molecular
weight. It did not form gels with sucrose and acid or with Ca++ salts. Coffee pulp (exocarp) was richer in pectin than coffee
mucilage.

Introduction Materials and Methods

All the coffee varieties produced in Central America are of the
Coffea arabica species. The meso carp is 15 to 20% w/w of the
coffee fruit and comprises a mucilaginous substance (muci-
lage) which is rich in pectin, the amounts of which are variously
reported in the literature (1-5). Utilization of coffee pectin has
interested researchers for many years. Cleves (6), mentions
analytical determinations made in Costa Rica in 1931 in which
pectic substances were identified in coffee mucilage. Carbonell
and Vilanova (7) characterized the coffee mucilage as a hydro-
gel without cellular structure. Coleman et al. (8) identified
galacturonic acid in mucilage removed by diluted alkali from
Guatemalan coffee cherries. Calle (9) reports that extraction
of the mucilage with NaOH solutions gave a substance that
jelled after a long resting time. Firm gels were obtained by
addition of sodium hypochlorite to mechanically recovered
mucilage. Fresh coffee pulp (exocarp) immersed in 0.5-2.0%
HCl solutions and autoclaved at 120°C for 30 min was later
pressed to obtain a transparent red syrup which jellified on
cooling after sugar addition. The same author (10) later
described a process to extract pectin from coffee mucilage with
50% HCl and hexametaphosphate as Ca scavenger, thus
producing a stable gel. Qualitative tests were positive for
methanol in the liquid which separated from this gel by
"
syneresis and the author therefore concluded that it was
'" formed by low-methoxyl pectins. The coffee pulp or exocarp
was also extracted by boiling with dilute HCl and hexameta-
phosphate to yield a liquid which jelled on cooling. Pectins
from the mucilage did not contain any similar pigments or
phenolic compounds like those from the pulp. Other pectin
extraction methods were reported later by Calle (11), consist-
ing of adding calcium gluconate to the mucilage to form a gel,
or separating mechanically the mucilage inside an acid bath,
screened and spread over glass surfaces on which methanol
was sprinkled to jellify the pectin.
However, there was no information in the literature concern-
ing the physico-chemical properties of coffee pectin gels,
which are important in determining suitability as a food addi-
tive, especially in the production of jellies. The aim of this
study is therefore to investigate the physical and chemical
properties of the coffee pectins.
0023-6438/91/020125 + 05 $03.00/0
Coffee samples
Coffee fruits were obtained from Bourbon, Caturra and Cati-
mor varieties of C. arabica from a farm in Sacatepequez,
Guatemala, at an altitude of 1500 m.
Commercial pectin samples
As a comparison a commercial pectin sample was employed,
obtained from H. P. Bulmer Ltd., Ryelands Street, Hereford,
HR40LE U.K.
Pectin extraction methods
Coffee fruit samples of the varieties Bourbon (I), Caturra (II)
and Catimor (III) were brought to ICAITI's pilot plant on the
same day they were harvested. First, the samples were washed
with water and detergent. Before the depulping, immature
fruits and floats were eliminated. The mucilage-covered coffee
beans were then placed in a locally-made batch demucilaginat-
ing machine which has a 3 HP electric motor (Sexilago) and
resembles a big waring blendor. This operation requires the
addition of a certain amount of water, and lasts about 15 min.
All materials were weighed to obtain data for a mass balance.
The mucilage was led directly into a 95% v/v ethanol solution
for obtaining the raw precipitate which contains the pectin.
The raw precipitate was separated in a basket centrifuge, then
washed first with a 95% ethanol solution to eliminate sugars
and some pigments, then with acetone. The product was dried
at an ambient temperature to give samples I, II and III.
Three additional samples were obtained as follows. Coffee
fruits of the Bourbon variety were processed as outlined for
sample (I). The beans in the liquid mucilage were removed
with the aid of a screening machine (Russel 2758 W.G.2). The
mucilage was precipitated with an 81% ethanol solution,
acidified with HCl (750 ml 95% ethanol, 50 ml HCl cone., 200
ml water). The precipitate was separated by decantation,
washed with 75% v/v ethanol, then with 81% v/v ethanol and
finally with acetone. It was then dried for 48 h at 35°C, to
obtain sample IV.
Coffee pulp (exocarp) from Sample I was also pressed in a
locally made continuous screw press to obtain coffee pulp
juice, which was mixed with two volumes of acidified ethanol
@ 1991 Academic Press Limited

125
lwt/vol. 24 (1991) No.2

100 KG COFFEE FRUIT

I. 26 Kg coffee lost 44.84 Kg pulp

56.15 Kg coffee mucilage

DEMUCILAGINATION 25.02 Kg coffee

22.98 Kgwater
(dry parchment)

.:,"
31.14 Kg mucilage-water
(8.17 Kg dry mucilage)

27.74 Kg ethanol (dry) SCREENING and Liquid to

+28.84 g HCI PRECIPITATION ethanol recovery

Liquid to ethanol 2 ETHANOL 19.43 Kg

recovery WASHING5 ethanal (day)

To acetone ACETONE
2.35 Kg acetone
recovery WASHING

0.49 Kg
raw precipitate

1
89.87g pectin
Fig. 1 Extraction of pectin from coffee mucilage: mass balance for sample IV

to precipitate the alcohol-insoluble solids, which were then
purified as described above to obtain the raw precipitate of
sample V.
The solid residue collected after pressing the coffee pulp from
Sample V was extracted by boiling for 1 h, in an aqueous HCI
solution of pH 2. The solution was precipitated with ethanol,
and the raw precipitate was washed with ethanol and acetone
and dried as described above, giving sample VI.
The mass balances for obtaining samples IV, V and VI are
presented in Figs 1 and 2.
Pectin purification methods
The raw precipitates of samples I, II and III were purified by
the method described by Tuerena (12): 50 g of the raw
precipitate was dispersed in 500 ml of phosphate buffer (pH
6.5). The mixture was heated to 80°C and 5 g of sodium
hexametaphosphate added. The hot suspension was
thoroughly mixed by stirring. After cooling, the slurry was
centrifuged and the pellet discarded. The pectin was precipi-
tated from the supernatant with 500 ml of acidified propan-2-01
(90:10, propan-2-01:conc. HCI) then removed by filtration
using Whatman No.4 paper. The pectin was washed twice with
250 ml of 75% propan-2-01 and once with 100% propan-2-01
then dried under vacuum.
The raw precipitates of samples IV and V were suspended in
acidified water at pH 1.86, heated at 80°C for 15 min, and then
the quaternary ammonium salt method described by Rom-
bouts (13), adapted to coffee pectins, was applied. When a
clear pectin solution has been obtained, the pH is adjusted to
4.5 and 11 of this solution is mixed with 5 ml of 1M ammonium
sulfate solution. Then the temperature is raised to 37°C and
250 ml of a 2% quaternary ammonium salt (hexa decyl tri-
methyl ammonium bromide) is added, as recommended by
Scott (14). The sample is kept at 37°C for 1 h before centrifug-
ing. The precipitate is washed three times with distilled water,
followed by suspension in 300 ml of 1% NaCI solution and
adjustment of pH to 3.2. A temperature of 37°C is maintained
until the complex had completely dissolved. Then 900 ml
acidified ethanol are added, and the precipitate is homogen-
ized by stirring and filtered through a cheese-cloth. The pre-
cipitate is washed four times with the following solutions: i)
acidified ethanol (4: 1 ethanol 95%/HCI cone.); ii) ethanol
60%; iii) ethanol 70%; iv) ethanol 95% and then vacuum
filtered through a cloth. Finally it is left to dry at an ambient
temperature.
The raw precipitate from sample VI was treated in the same
way as for samples IV and V then purified additionally by
repeated precipitation in acidified ethanol.

126
 lwt/vol. 24 (1991) No.2

100 KG COFFEE FRUIT

1.26 Kg
coffee lost

44.84 Kg pulp

17.94 Kg juice 26.90 Kg pressed pulp

20 Kg ethanol 51 L water
(pure) (pH 2)
13 Kg HCI Heat
( pure)

28 L Solution
VACUUM
F ILTE R

37 Kg et ha nol
(pure)

Liquid to ethanol 14 Kg ethanol

recovery (pure)

Liquid to acetone 1.7 Kg

recovery acetone
26 Kg ethanol
Liquid to ethanol (pure)
recovery

Liquid to acetone 61 Kg
recovery acetone
0.358Kgraw
precipitate
1.023 Kg raw
precipitate
1
49.28g pectin ~
109.1 g pectin
Fig. 2 Extraction of pectin from coffee pulp: mass balance for samples IV and V

Analytical methods
Anhydrogalacturonic acid. m-Hydroxydiphenyl method de-
scribed by Kinter and van Buren (15); Methoxyl and degree of
esterification: gas chromatographic method of Walter et al.
(16); Moisture: oven method, FAO (17); Total ash: calci-
nation method, (18); Acid insoluble ash: HCl boiling method,
FAO (19); Weight average molecular weight: viscosity method
after Rombouts (13), based on the work by Owens et al. (20);
Setting time: twist method by Joseph and Baier (21); Firmness
of pectin jellies: for low methoxyl samples, the method de-
scribed by Gross et al. (22).
Results and Discussion
Table 1 presents a characterization of the coffee pectin
samples, compared with a commercial unstandardized citrus
pectin (Bulmer Citrus 603).
The percentage of anhydrogalacturonic acid indicates the
amount of pectin in the sample after purification and thus will
be affected by the extraction and purification procedures used.
The values vary between 46.6 and 91.19%.
The coffee pectin samples show low methoxyl contents, with
esterification degrees of 18.97-30.55%.
The low molecular weight of the coffee pectin samples was
probably caused by enzymatic degradation during the extrac-
tion process at neutral pH of the mucilage. Values reported in
the literature for commercial citrus and apple pectins are
higher than 50,000, while a citrus pectin sample degraded by
heating at 119°Cfor 20 min gave a molecular weight of 20,000-
30,000 [Owens etal., (23)]. For this reason, for samples IV and
V, an acidified ethanol solution was used to precipitate the
pectin, as proposed by Joslyn and Deuel (24). Weight average
molecular weights for pectin purified from samples IV and V
were the highest of all coffee samples.
The total ash content for coffee pectin samples is higher than
that of the commercial sample, but it meets the Food Chemi-
cals Codex standard (25) of less than 10%; although not the
Bulgarian standard of less than 5% (26).
Only the pectin from sample II had detectable acid insoluble
ash content, and does not meet the FAO and Food Chemicals
Codex standards that require a value of less than 1% (17,25).
Pectin from samples IV and V formed clear jellies on the
addition of Ca+ + ions, but this could not be replicated. Pectin

127
lwtlvoJ.24(1991)No.2

Table 1 Characterization of pectins from three coffee varieties

Analysis I II III IV V VI
(All determinations mucil. mucil. mucil. mucil. pulp juice pulp solid Bulmer citrus
in dry weight base) Bourb. Catur. Catim. Bourb. Bourb. Bourb. pectin 603

Anhydrogalacturonicacid (%) .
m-Phenyl phenol method 50.79 74.43 46.60 94.71 79.70 91.19 93.20
Methoxyl (%)
Gas chromatography 2.11 2.30 1.82 3.97 3.54 11.65
Degree of esterification
Gas chromatography 25.50 18.97 23.88 30.55 23.81 76.59
Pectin yield
% w/w of coffee fruit 0.055 0.039 0.040 0.025 0.089
Pectin yield
% w/w of raw precipitate 10.86 7.57 8.26 21.57 15.98 10.65
Raw precipitate yield
% w/w of coffee fruit 0.51 0.51 0.49 2.00 3.81

Moisture (%) 16.22 16.67 16.06 26.61 24.15 20.66 8.33

Acid insoluble ash (%)
HCI boiling method n.d. 1.39 n.d. n.d. n.d. n.d. n.d.
Weight average molecular weight
Viscosity method 37,582 39,367 36,638 47,868 42,371 22,359 86,774

n.d.: not detected

from samples I, II, III and VI did not form consistent jellies
after the addition of a CaCh solution, following the method
described by Gross et aZ.(22).
Yields. In the fresh Bourbon coffee of 79% moisture content
the total yields of pectin purified from mucilage, pulp juice and
pulp solids were 0.256% w/w (wet base) and 1.22% w/w of the
fresh fruit. This yield is smaller than the pectin contents
mentioned in the literature: Wilbaux (1) reported pectic sub-
stances in the mucilage of C. arabica as 33% dry base, or about
4.95% of the ripe fruit, dry base. Menchu et al. (3) found 15.9
to 18.2% pectin dry base in mucilage samples from Guatema-
lan coffee (2.3 to 3.6% dry base of the ripe fruit), around 50%
of which was water soluble. Orozco (4) calculated 3.19%
pectin dry base of the ripe fruit, from Costa Rican coffee
samples. Rao (5) reported 2.39% pectin, dry base, of ripe C.
arabica fruits. Cleves (6) presents analytical data from Costa
Rican coffee samples of 1.74% wet base of pectin in the ripe
coffee fruits. However, direct comparisons cannot be made
since most values previously reported in the literature were
calculated from per cent anhydrogalacturonic determinations
directly on the mucilage, without previous purification.
Conclusions
Pectin from C. arabica varieties Bourbon, Caturra and Cati-
mor, cultivated at an altitude of 1500 m, is low-methoxyl.
Weight average molecular weights from the coffee pectins
assayed are relatively low.
Pectin from the three coffee varieties tested did not form firm
gels after the addition of Ca++ ions.
Coffee pulp (exocarp) is about 1.9 times richer in pectin than
coffee mucilage.
Acknowledgements
This research was supported by grant No. 936-5542-6-00-3065-
00 from USAID's Program in Science and Technology Coop-
eration (PSTC). The authors are also grateful for the collabor-
ation received from Dr Franz Rombouts of the Agricultural
University, Wageningen, The Netherlands, and from Mr
David R. Stevens, H. P. Bulmer Ltd., Hereford, U.K.
References
1 WILBAUX,R. Les Cafeiers au Congo Beige. Technologie du Cafe
Arabica et Robusta. Bruxelles: Direction de L'Agriculture des
Forets et de L'E1evage, p. 12 (1956)
2 ROLZ,C., MENCHU,J. F., ESPINOZA,R. ANDGARCIA-PRENDES, A.
Coffee fermentation studies. In: ASIC. Proceedings 5th Inter-
national Colloquium on the Chemistry of Coffee. Lisbon, June 14-
19. Paris, pp. 259--269 (1971)
3 MENCHU,J. F., DE ARRIOLA,M. c., FUENTES,A. ANDROLZ, C.
Posibilidad de recuperacion de la pectina del mucilago de cafe. In:
CAT/E. Primera Reunion Intemacional sobre la Utilizacion de
Subproductos de Cafe en la Alimentacion Animal y Otras Aplica-
ciones Agricolas e Industriales. Turrialba, Costa Rica, June 11-14
(1974)
4 OROZCO,R. A. Obtencion de pectina a partir del mucilago de cafe.
In: CA TIE. Primera Reunion Internacional sobre la Utilizacion de
Subproductos de Cafe en la Alimentacion Animal y Otras Aplica-
ciones Agricolas e Industriales. Turrialba, Costa Rica, June 11-14
(1974)
5 RAo, S. G. Pectins as potential by-products of coffee waste.
Journal of Coffee Research, 5 (142), 29-35 (1973)
6 CLEVES,R. Justificacion de un proyecto para investigar la obten-
cion de pectina a partir del mucilago de cafe. Oficina del Cafe, San
Jose, Costa Rica (1975)
7 CARBONELL, R. ANDVILANOVA, M. Beneficiado nipido y eficiente
del cafe mediante el uso de soda caustica. Boletin Tecnico No. 13,
Centro Nacional de Agronomia, Ministerio de Agricuitura y Gana-
deria, EI Salvador, pp. 65-66 (1952)
8 COLEMAN, R. J., LENNEY,J. F., COSCIA,A. T. ANDDICARLO,F. J.
Pectic acid from the mucilage of coffee cherries. Archives of
Biochemistry and Biophysics, 59, 157-164 (1955)
9 CALLE,H. Tentativas para la preparacion de productos alimenti-
cios del cafeto, distintos a la bebida. CENICAFE (Colombia), 9
(10),222-227 (1958)
10 CALLE,H. Metodos de extraccion de las pectinas del cafe. CENI-
CAFE (Colombia), 13 (2), 69-74 (1962)
11 CALLE,H. Pectinas. In: Subproductos del Cafe. Boletin Tecnico
No.6, CENICAFE (Colombia), pp. 39--44 (1977)
12 TUERENA,C. E. Carboxyl group distribution in pectin. PhD Thesis,
University of Nottingham, U.K., p. 63 (1983)
13 ROMBOUTS, F. Methods for research of pectins and pectic enzymes.

128
 lwt/vol. 24 (1991) No.2

Unpublished paper. Agricultural University, Department of Food Shape and size of pectinic acid molecules deduced from viscometric
Science, De Dreijen 12, 6703 BC Wageningen, The Netherlands measurements. Journal of the American Chemical Society, 68,
(1983) 1628-1632 (1946)
14 SCOTI, J. E. Fractionation by precipitation with quaternary am- 21 JOSEPH, G. H. AND BAIER, W. E. Methods of determining the
monium salts. In: WHISTLER, R. L. (Ed.), Methods in Carbo- firmness and setting time of pectin test jellies. Food Technology, 3,
hydrate Chemistry, Vol. V. New York: Academic Press, pp. 38-44 18-22 (1949)
(1965) . 22 GROSS, M. 0., RAO, V. N. M. AND SMIT, C. J. B. Rheological
15 KINTER, P. ANDVANBUREN, J. P. Carbohydrate interference and its characterization of low methoxyl pectin gel. Journal of Texture
correction in pectin analysis using the m-hydroxydiphenyl method. Studies, 11 (3),271-290 (1980)
Journal of Food Science, 47 (3), 756-759; 764 (1982) 23 OWENS, H. S., LOTZKAR, H., MERRILL, R. C. AND PETERSON, M.
16 WALTER, R. H., SHERMAN, R. M. AND LEE, C. A comparison of Viscosities of pectin solutions. Journal of the American Chemical
methods for polyuronide methoxyl determination. Journal of Food Society, 66, 1178-1182 (1944)
Science, 48 (3), 1006-1007 (1983) 24 JOSLYN, M. A. ANDDEUEL, H. The extraction of pectins from apple
17 FAO. Food and Nutrition Paper No. 19. Rome, Italy (1981) marc preparations. Journal of Food Science, 28, 65-83 (1963)
18 AOAC. Official Methods of Analysis of the Association of Official 25 NATIONAL RESEARCH COUNCIL. Food Chemicals Codex, 3rd edn.
Analytical Chemists, 13th edn. Washington D.C. (1980) Washington, D.C.: National Academy Press, p. 215 (1981)
19 FAO. Food and Nutrition Paper No.5. Rome, Italy (1978) 26 Bulgaria. Bulgarian Standard BDS 2608-80 (1980). (Food Science
20 OWENS, H. S., LOTZKAR,H., SCHULTZ, T. H. AND MACLAY,W. D. & Technology Abstracts, 15, 2U130.)

.,
~

129