Agricultural and Biosystems Engineering

Agricultural and Biosystems Engineering

Publications

1999

Physical and Nutritional Properties of Corn Masa

By-product Streams
Kurt A. Rosentrater
Iowa State University, karosent@iastate.edu

Rolando A. Flores
United States Department of Agriculture

Thomas L. Richard
Iowa State University

Carl J. Bern
Iowa State University, cjbern@iastate.edu

Follow this and additional works at: http://lib.dr.iastate.edu/abe_eng_pubs

Part of the Agriculture Commons, and the Bioresource and Agricultural Engineering Commons
The complete bibliographic information for this item can be found at http://lib.dr.iastate.edu/
abe_eng_pubs/80. For information on how to cite this item, please visit http://lib.dr.iastate.edu/
howtocite.html.

This Article is brought to you for free and open access by the Agricultural and Biosystems Engineering at Iowa State University Digital Repository. It
has been accepted for inclusion in Agricultural and Biosystems Engineering Publications by an authorized administrator of Iowa State University
Digital Repository. For more information, please contact digirep@iastate.edu.
Physical and Nutritional Properties of Corn Masa By-product Streams
Abstract
Production of corn masa-based products is flourishing in the United States, as is the generation of masa
processing waste. Masa by-products have potential for value-added utilization, an option which could produce
less pollution in the environment and economic benefits for masa processors. Physical and nutritional
properties of these byproducts are needed for the proper design of processing operations and by-product
applications, but information concerning masa byproducts is not currently available. Thus the objective of this
study was to fully characterize typical masa by-product streams. The masa by-products studied had moisture
contents between 88.15% and 89.29% (w.b.), water activity values between 0.999 and 1.000, densities
between 1030.85 and 1047.32 kg/m3 , yield stress values between 1440.04 and 1618.08 N/m2 , pH values
between 6.17 and 6.30, Hunter L values between 35.15 and 49.13, a values between 0.27 and 0.98, and b
values between 6.85 and 9.38. Drying curves were developed to predict drying behavior. The dried by-
products had protein contents between 4.76% and 4.90% (d.b.), crude fat contents from 0.74% to 5.76%
(d.b.), ash contents between 17.41% and 19.09% (d.b.), and carbohydrate contents from 71.93% to 75.41%
(d.b.), which was due primarily to fiber, with hemicellulose levels of 20.82% to 24.06% (d.b.) and cellulose
between 30.55% and 31.83% (d.b.). Dry masa by-products also consisted of 4.68% (d.b.) calcium. Therefore,
dehydrated masa by-products seem very suitable for use as livestock feed additives.

Keywords
Evaluation, Food processing, Food waste, Residue characterization, Residue utilization

Disciplines
Agriculture | Bioresource and Agricultural Engineering

Comments
This article is from Applied Engineering in Agriculture 15, no. 5 (1999): 515–523.

This article is available at Iowa State University Digital Repository: http://lib.dr.iastate.edu/abe_eng_pubs/80

fpe 1510 ms 9/25/01 12:52 PM Page 515

PHYSICAL AND NUTRITIONAL PROPERTIES

OF CORN MASA BY-PRODUCT STREAMS
K. A. Rosentrater, R. A. Flores, T. L. Richard, C. J. Bern

ABSTRACT. Production of corn masa-based products is flourishing in the United States, as is the generation of masa
processing waste. Masa by-products have potential for value-added utilization, an option which could produce less
pollution in the environment and economic benefits for masa processors. Physical and nutritional properties of these by-
products are needed for the proper design of processing operations and by-product applications, but information
concerning masa byproducts is not currently available. Thus the objective of this study was to fully characterize typical
masa by-product streams. The masa by-products studied had moisture contents between 88.15% and 89.29% (w.b.), water
activity values between 0.999 and 1.000, densities between 1030.85 and 1047.32 kg/m3, yield stress values between
1440.04 and 1618.08 N/m2, pH values between 6.17 and 6.30, Hunter L values between 35.15 and 49.13, a values
between 0.27 and 0.98, and b values between 6.85 and 9.38. Drying curves were developed to predict drying behavior.
The dried by-products had protein contents between 4.76% and 4.90% (d.b.), crude fat contents from 0.74% to 5.76%
(d.b.), ash contents between 17.41% and 19.09% (d.b.), and carbohydrate contents from 71.93% to 75.41% (d.b.), which
was due primarily to fiber, with hemicellulose levels of 20.82% to 24.06% (d.b.) and cellulose between 30.55% and
31.83% (d.b.). Dry masa by-products also consisted of 4.68% (d.b.) calcium. Therefore, dehydrated masa by-products
seem very suitable for use as livestock feed additives.
Keywords. Evaluation, Food processing, Food waste, Residue characterization, Residue utilization.

A
gricultural processing industries produce great
quantities of waste materials. From 1991 to 1992
agribusinesses in Kansas alone produced over
72 000 Mg/yr of waste products and processing
residues. Approximately 10% (7173 Mg/yr) of these
wastes were attributable to the grain processing industry
(Nelson and Flores, 1994). Due to heightened economic
and environmental concerns, landfilling has declined as the
waste disposal method of choice, and disposal alternatives
have been more frequently investigated. Current options
include reprocessing / r ecycling, resale, incineration,
composting, biomass energy production, land application
as soil conditioning ingredients, and reuse as livestock feed
ingredients. The last alternative includes both direct
feeding of the waste products and feeding the by-products
Article was submitted for publication in August 1998; reviewed and
approved for publication by the Food & Process Engineering Institute of
ASAE in June 1999. Presented as ASAE Paper No. MC98-134.
Journal Paper No. J-18033 of the Iowa Agriculture and Home
Economics Experiment Station, Ames, Iowa, Project No. 3326, and
supported by Hatch Act and State of Iowa Funds. Contribution No. 99-3-J
from the Kansas Agricultural Experiment Station.
The authors are Kurt A. Rosentrater, ASAE Student Member,
Graduate Student, Agricultural and Biosystems Engineering Dept., Iowa
State University, Ames, Iowa; Rolando A. Flores, ASAE Member
Engineer, Associate Professor, Department of Grain Science and
Industry, Kansas State University, Manhattan, Kansas (formerly with
Agricultural and Biosystems Engineering Dept. and Food Science and
Human Nutrition Dept., Iowa State University, Ames, Iowa); Thomas L.
Richard, ASAE Member Engineer, Assistant Professor, and Carl J.
Bern, ASAE Fellow Engineer, Professor, Agricultural and Biosystems
Engineering Dept., Iowa State University, Ames, Iowa. Corresponding
author: Rolando A. Flores, Kansas State University, 104 Shellenberger
Hall, Manhattan, KS 66506-2201; voice: (913) 532-4064; fax: (913) 532
7010; e-mail: raf@wheat.ksu.edu.
after further processing (i.e., dehydration, extrusion, or
blending with other feed components) (Bohlsen et al.,
1997).
In recent years, many research efforts have been aimed
at the development of livestock feed additives from
agricultural wastes and by-products. One aspect has
included the investigation of residential wastes as potential
feed products (Polanski, 1992). Another area has included
the development of feed ingredients from slaughterhouse
by-products (Luzier and Summerfelt, 1995; Martins and
Guzman, 1994; Wang et al., 1997). The utilization of grain
by-products as livestock feed components has also been
extensively researched (Ham et al., 1994, 1995; Hussein
and Berger, 1995; Larson et al., 1993). Corn masa
production is one facet of the grain processing industry that
generates large quantities of waste materials, but to date,
has received little research attention regarding alternatives
for by-product disposal.
Corn masa is used in the production of corn snack foods
and tortilla chips, and has traditionally been utilized in the
preparation of tortillas, which have been a staple in the diet
of Mexican and Central American peoples for centuries.
Foods made with corn masa include tacos, tostadas,
tamales, quesadillas, panuchos, enchiladas, and others
(Krause et al., 1992; Ortiz, 1985; Serna-Saldivar et al.,
1990). Currently, Mexican foods and corn-based snacks in
the United States are becoming more popular. Over $4
billion worth of Mexican foods were marketed in 1986, and
approximately $2.5 billion worth of tortilla chips were
produced in 1994 (Gomez et al., 1987; Wood, 1994). Corn
masa is produced by simulating ancient Aztec methods on
an industrial-scale. Whole corn is boiled with 120 to 300%
water (original corn weight basis) and 0.1 to 2.0% lime
(original corn weight basis) for 0.5 to 3.0 h at 80 to 100°C,

Applied Engineering in Agriculture

VOL. 15(5): 515-523 © 1999 American Society of Agricultural Engineers 0883-8542 / 99 / 1505-515 515
fpe 1510 ms 9/25/01 12:52 PM Page 516
and is then steeped for up to 24 h. This process, called
nixtamalization, can be either a batch process or a
continuous process, depending on production equipment.
The cooked grain (nixtamal) is then separated from the
steep liquor (nejayote), which is rich in lime and dissolved
corn pericarp tissue, is washed to remove any excess lime
and pericarp, and is then stone ground to produce corn
masa. The masa will then be molded, cut, or extruded and
then baked or fried to make tortillas, corn chips, or tortilla
chips; or it will be dried and milled into flour (Gomez et
al., 1987; Rooney and Serna-Saldivar, 1987; Serna-Saldivar
et al., 1990).
The steep liquor and the rinse water contain between 2
and 6% total (dissolved and suspended) solids. Generally
the suspended solids (50 to 60% of the total solids) are
removed by screening, centrifugation, or decanting, and
disposed of in landfills. The remaining water and dissolved
solids are sent to municipal facilities for treatment. The
total solids in the waste stream, which consist primarily of
fiber-rich pericarp tissues, represent corn dry matter loss
that occurs during processing. Estimates of the original
corn dry matter loss have ranged from 5.0% to 17.0%. The
following were found in the literature: 8.5 to 12.5%
(Pflugfelder et al., 1988), 8.0 to 17.0% (Rooney and Serna-
Saldivar, 1987), 7.0 to 13.0% (Khan et al., 1982), 5.0 to
14.0% (Katz et al., 1974), 11.0 to 12.0% (Bressani et al.,
1958), and 13.3% (Gonzalez de Palacios, 1980). Corn dry
matter loss during nixtamalization is affected by many
processing variables, including corn hybrid and quality,
lime type and concentration, cooking and steeping times
and temperatures, friction during washing and transport,
and process equipment used. These processing losses can
be economically significant due to lost masa yield, waste
processing and disposal costs, and potential environmental
pollution. This waste generation is of particular concern in
areas of substantial masa processing, such as Mexico City,
where over 2400 Mg of corn is processed every day into
corn masa for tortilla production (Gonzalez-Martinez,
1984).
Gonzalez-Martinez (1984) investigated four biological
treatment options for “nejayote”, including activated
sludge, anaerobic contact processing, a submerged aerobic
fixed-film cascade reactor, and an anaerobic packed-bed
reactor. However, little research has been conducted into
alternative disposal methods for corn masa dry matter
losses (i.e., the separated suspended solids), and none could
be found regarding the utilization of corn masa waste solids
in livestock feed rations. To effectively utilize and add
value to these by-products, physical and nutritional
properties for corn masa residual streams must be
quantified. Characterization of by-product materials
provides data that are essential for livestock diet
formulation, design of equipment and processing facilities,
and optimization of unit operations such as spray drying,
extrusion, blending, mixing, separating, heating, freezing,
dehydration, and material flow (Ferris et al., 1995;
Stroshine and Hamann, 1995). Little information has been
gathered concerning the properties of corn masa by-
products, however. Because the demand for corn masa-
based products is increasing in the United States, the
objective of this study was to characterize and quantify the
physical and nutritional properties of typical corn masa by-
product streams so that subsequent by-product applications
516
may effectively be developed for these waste materials.
Physical properties studied included moisture content,
water activity, density, yield stress, pH, color, and drying
analysis. Nutritional properties studied included protein,
crude fat, ash, mineral composition, amino acid
composition, and fiber content.
MATERIALS AND METHODS
SAMPLE COLLECTION AND PREPARATION
Corn masa by-product samples, which were gray, pasty,
semi-solid slurries, were obtained from a local white corn
masa processing facility after the decanting process
(i.e., the samples collected were the separated suspended
solids from the waste stream). Five 3.5 L samples were
collected at random times from random batches and
random production runs, from both batch and continuous
processing lines during the fall of 1996. The 10 samples
were then placed in frozen storage at –10°C until testing
commenced. Prior to analysis, the samples were thawed at
room temperature (25°C ± 2°C) for 24 h. All physical
properties, except moisture content and drying analysis,
were determined at room temperature. All nutritional tests
were conducted as specified by the standard method used
for each property.
EXPERIMENTAL DESIGN
To investigate the properties of typical corn masa by-
products, five samples from each of two processes (a batch
process and a continuous process) were analyzed; thus two
populations were present in this study. All properties were
studied using a completely randomized design. For all
physical property testing (moisture content, water activity,
density, yield stress, pH, color, and drying analysis), five
replicates were taken from each sample, which led to an
overall sample size of 25 experimental units for each
production process (population). This maintained the Type-
I (α) and Type-II (β) error rates at the 0.05 level and
allowed the detectable statistically significant difference
between populations to remain at 1.0 standard deviations.
Protein, crude fat, and ash testing were performed with two
replicates taken from each sample, which led to an overall
sample size of 10 experimental units from each population.
This maintained the Type-I (α) and Type-II (β) error rates
at the 0.05 level, but the detectable difference between
populations increased to 1.8 standard deviations. Fiber
analysis was conducted with three replicates from each
sample, which produced a sample size of 15 experimental
units for each production process. This maintained the
Type-I (α) and Type-II (β) error rates at the 0.05 level, and
the detectable difference between populations was set at
1.4 standard deviations (Nelson, 1985). Formal statistical
analyses on the collected data were performed via
Microsoft Excel (Microsoft, 1993) and SAS (SAS, 1992)
software.
Amino acid composition was studied on the continuous-
process samples only, as was mineral analysis. One
replicate from each of the five samples was used in the
amino acid study; whereas, one replicate from two of the
five samples was randomly selected for mineral analysis.
Neither amino acid composition nor mineral composition
was included in the formal statistical analysis that was
conducted on all other physical and nutritional properties.
APPLIED ENGINEERING IN AGRICULTURE
fpe 1510 ms 9/25/01 12:52 PM Page 517
PHYSICAL PROPERTIES
Moisture Content and Water Activity. Moisture
content (wet basis, w.b.) of the masa by-product samples
was determined using a method similar to AACC Method
44-15A (“Moisture-Air-Oven Methods”) (AACC, 1995).
By-product samples of 2 to 3 g were placed in open tared,
aluminum moisture dishes and dried at 130°C for 3 h in a
forced-convection laboratory oven (Thelco Model 160DM,
Precision Scientific, Inc., Chicago, Ill.). After drying, the
samples were allowed to cool in a desiccator, and sample
mass was then measured with an electronic balance (Model
A-250, Denver Instrument Co., Arvada, Colo.). To
determine water activity of the masa by-products, 4-g
samples were placed in a previously calibrated water
activity meter (Aqualab Model CX-2, Decagon Devices,
Inc., Pullman, Wash.).
Mass Density. The density of each by-product sample
was determined using a specific gravity cup (Model H-
38000-12, Cole-Parmer Instrument Co., Barrington, Ill.)
and an electronic balance. The specific gravity cup has a
known mass and volume; hence, to measure sample
density, the cup was filled completely with sample
material, the lid was placed on the cup, excess sample
material was removed, and the filled cup was then weighed
on the balance.
Yield Stress. The yield stress of the by-product samples
was determined using the vane method with a digital
viscometer (Model HBDV II+, Brookfield Engineering,
Stoughton, Mass.) and a four-finned vane, which was 1.01
cm in height and 1.2 cm in diameter; each fin was 1.4 mm
in width. By-product samples were placed in a 150 mL
beaker which was 8 cm in height and 5.3 cm in diameter;
the vane was then placed 2 cm below the sample surface,
and a controlled constant shear rate of 0.34 s–1 (0.5 rpm)
was applied. Using the vane method with a controlled shear
rate, the yield stress was determined using the following
equation:
Tm = π D
3
H + 1 τy (1)
2 D 3
where Tm is the torque exerted on the vane at yielding, D is
the overall vane diameter, H is the overall vane height, and
τy is the resultant yield stress (Dzuy and Boger, 1985;
Yoo et al., 1995).
pH. To determine the pH of the masa by-product
samples, 10-g samples were placed in a 50-mL glass
beaker. A bench-top pH meter (Model PHB-62, Omega
Engineering, Inc., Stamford, Conn.) was then used to
measure the pH.
Color. The color of each masa by-product sample was
determined using a Hunter Colorimeter (Model LabScan
SN-12414, Hunter Associates Laboratory, Inc., Reston,
Va.), using a view-port and view-area size of 1.27 cm
(0.5 in.), and the L-a-b opposable color scales (Hunter,
1983). The samples were placed in 100 mm × 15 mm
plastic petri dishes and were positioned in the instrument
for color determination after proper instrument calibration.
Drying Analysis. The drying studies that were
performed on the masa by-products were similar to those
conducted by Rosentrater and Flores (1997) on swine
blood components. Masa by-product samples of 2 to 3 g
VOL. 15(5): 515-523
were placed in aluminum moisture dishes that were 8 cm in
diameter and 2.5 cm in height and were placed in a forced-
convection laboratory oven (Thelco Model 160DM,
Precision Scientific, Inc., Chicago, Ill.). Samples were
dried at temperatures of 80, 100, and 120°C for 2 h, and
sample mass over time was measured with an electronic
balance. Sample masses were recorded every 5 min during
the first hour of testing and every 10 min thereafter. Thus,
drying curves were developed on a dry basis (d.b.) and
drying rate prediction equations as a function of moisture
content and drying temperature were developed for the two
by-products.
NUTRITIONAL PROPERTIES
Proximate Composition. Proximate analyses of dry
masa by-product samples included the determination of
protein, crude fat, and ash contents. Protein and crude fat
analyses were conducted by the Meat Laboratory at Iowa
State University, Ames, Iowa. Protein analysis followed the
AACC Standard Method 46-11A (“Crude protein-improved
Kjeldahl method, copper catalyst modification”), using a
factor of N × 6.25. Crude fat was determined using Method
30-25 (“Crude fat in wheat, corn, and soy flour, feeds, and
cooked feeds”) (AACC, 1995). Ash content was
determined using the AACC Standard Method 08-03
[“Ash-rapid (2-hour, 600°) method”, 1995], which utilized
2 g samples placed in a muffle furnace at 600°C for 2 h.
Because protein, fat, ash, and carbohydrate contents sum to
100% in the dry by-product, the carbohydrate contents of
the dry masa by-product samples could be calculated by
difference.
Fiber Composition. Fiber analysis was conducted by
the Forage Quality Laboratory at Iowa State University,
Ames, Iowa, using the Van Soest detergent system to
determine Neutral Detergent Fiber (NDF), Acid Detergent
Fiber (ADF), and lignin fractions of the masa by-product
samples. These methods are widely used for the
determination of fiber composition of biological materials
and potential livestock feed ingredients (Moore and
Hatfield, 1994; Van Soest et al., 1991).
Mineral Composition. Mineral analyses of the masa
by-product samples were conducted by the Analytical
Services Laboratory at Iowa State University, Ames, Iowa,
and included the determination of calcium, potassium,
magnesium, and phosphorous. Prior to analysis, all samples
were dried at 103°C for 24 h, and the portions used for
calcium, potassium, and magnesium were then digested in
nitric acid and hydrogen peroxide. Calcium and
magnesium determinations were performed using Method
SMEWW 3111 B, which utilized flame atomic absorption
spectrophotometry, and potassium analysis was conducted
using Method SMEWW 3500-K D, which utilized flame
photometry (APHA, 1995). Phosphorous content was
determined by dry ashing the samples using Method AOAC
2.020c and then performing spectrophotometric
measurements using the molybdovanadophosphate method
prescribed in Method AOAC 2.026 (AOAC, 1980).
Amino Acid Composition. Amino acid profiles of the
masa by-product samples were conducted by the Protein
Structure Core Facility at the University of Nebraska
Medical Center, Omaha, Nebr. The samples were dried,
subjected to 6 N HCl hydrolysis for 20 h at 110°C, and
517
fpe 1510 ms 9/25/01 12:52 PM Page 518

then analyzed in an automated amino acid analyzer (Model mean values were significantly different (p < 0.05),
6300, Beckman Instruments, Inc., Fullerton, Calif.). however, which may reflect differences in the resulting
material structures due to differences in the batch and
RESULTS AND DISCUSSION continuous processing operations.
PHYSICAL PROPERTIES Yield Stress. Masa by-products, as shown in table 1,
Moisture Content and Water Activity. The moisture have very large yield stresses that are similar in magnitude
content results are shown in table 1. Masa by-products are to those of minced fish paste (1600-2300 N / m 2 )
very wet materials, even after the decanting process, and (Nakayama et al., 1980). Yield stress is a rheological
have moisture contents near 90% (w.b.). These moisture parameter that quantifies the shear stress required to initiate
levels preclude the direct shipping of the by-product flow in a liquid or semi-solid material, results from the
materials for incorporation into disposal alternatives; existence of particle structures, and is vital for the design of
instead, a drying mechanism will be required to reduce the transport processes and operations (Steffe, 1992). Because
high cost which is associated with the transportation of masa by-products have such large yield stress values,
water. Additionally, the water activity levels for the masa special considerations will need to be given to conveying
by-product samples (table 1) are greater than 0.999; thus operations for these materials during further processing.
the by-products are susceptible to microbial spoilage. pH. The pH results are shown in table 1. It was expected
Water activity quantifies the amount of “free” water that pH values would be fairly high for the masa by-
(i.e., unbound water) available in materials for use by product samples because nixtamalization generally
microorganisms and chemical agents, and hence is a employs a pH greater than 10.0 during the lime cooking
measure of susceptibility to spoilage and deterioration. process (Serna-Saldivar et al., 1990). Instead, the resulting
Products with no free water (aW = 0.0) are not at risk for pH values were fairly low (between 6.17 and 6.30). This
spoilage; whereas, materials with 100% free water (aW = may have occurred due to a large proportion of the lime
1.0) are at risk for rapid spoilage, which is the case for washing away with the rinse water during processing, or
masa by-products. Materials become safe from yeast possibly due to microbial activity during transit between
growth below water activities of approximately 0.9, safe the processing facility and the laboratory. Investigating the
from bacteria growth below approximately 0.8, and safe cause of the low pH values would be a valuable future
from mold growth below approximately 0.6 (Barbosa- study.
Canovas and Vega-Mercado, 1996). Color. The Hunter L, a, and b results are shown in
In order to effectively utilize masa by-products, table 1, and show some differences in color between the
dehydration is necessary to prevent microbial spoilage, and batch-processed and the continuous-processed masa by-
can be accomplished through drying, blending, or extrusion products, even though both processes used the same source
processing to an acceptable moisture level of and hybrid of raw white corn. Essentially, these results
approximately 12% (w.b.), which is typically describe masa by-products as gray material with a small
recommended for feed products because it substantially degree of yellowness.
reduces transportation costs and is microbiologically safe Drying Analysis. Drying analysis of biological
(Beauchat, 1981; Wang et al., 1997). materials is important for the proper design of drying
Mass Density. As shown in table 1, masa by-products processes and equipment. For the masa by-products in this
had density values slightly greater than that of water study, sample mass over time was monitored and used to
(approx. 999 kg/m 3). This was not unexpected due to the determine moisture contents and resulting drying rates at
high moisture contents of the by-product samples. The drying temperatures of 80, 100, and 120°C. A multiple
regression procedure with STEPWISE model selection
Table 1. Physical and nutritional properties of corn masa by-products
(SAS, 1992), selecting only significant terms, was then
used to determine polynomial regression prediction
Batch Continuous
Process Process
equations for drying rates as a function of moisture content
and drying temperature. The regression equation for the
Sample C.V. C.V.
Property Size (n) Mean (%) Mean (%) drying of batch-processed masa by-products, which had a
coefficient of determination (R2) of 0.93 and a sample
Moisture content (%, w.b.)* 25 89.29 0.68 88.15 1.13
Water activity (-)* 25 1.000 0.32 0.999 0.12 variance of 0.099, was determined to be:
Density (kg/m3)* 25 1030.85 1.23 1047.32 1.25
Yield stress (N/m2)* 25 1618.08 9.99 1440.04 19.69 DR = (0.0058) – (0.5719)M + (0.7625)M 2
pH (-) 25 6.30 2.12 6.17 9.35
Color: Hunter L value (-)* 25 49.13 4.96 35.15 6.25 – (0.5061)M 3 +
Color: Hunter a value (-)* 25 0.27 16.34 0.98 24.39 (0.1191)M 4 – (0.0117)M 5
Color: Hunter b value (-)* 25 9.38 11.63 6.85 7.15
Protein (%, d.b.) 10 4.76 8.71 4.90 7.29 + (0.0004)M 6 + (0.0111)(M)(T) (2)
Fat (%, d.b.)* 10 0.74 15.66 5.76 21.35
Ash (%, d.b.) 10 19.09 10.11 17.41 18.36
Carbohydrates (%, d.b.)* 10 75.41 2.30 71.93 3.18 where DR is the predicted by-product drying rate (g H2O/g
NDF (%, d.b.) 15 54.97 4.73 53.32 3.69 sample-h), M is the material moisture content (decimal dry
ADF (%, d.b.) 15 30.91 9.93 32.50 2.81 basis, d.b.), and T is the oven drying temperature (°C). The
Lignin (%, d.b.)* 15 0.36 68.07 0.67 55.89
Calcium (%, d.b.) 2 --- --- 4.68 17.39 graph of this regression function is shown in figure 1. A
Potassium (%, d.b.) 2 --- --- 0.07 15.43 similar regression equation for the drying of continuous-
Magnesium (%, d.b.) 2 --- --- 0.13 13.14 processed masa by-products, which had a coefficient of
Phosphorous (%, d.b.) 2 --- --- 0.57 9.68
determination (R2) of 0.90 and a sample variance of 0.060,
* Denotes a significant difference between batch - and continuous-process was determined to be:
means at the 0.05 level.

518 APPLIED ENGINEERING IN AGRICULTURE

fpe 1510 ms 9/25/01 12:52 PM Page 519
Figure 1–Drying curves for batch-processed corn masa by-products at drying temperatures of 80, 100, and 120°C.
DR = (0.0017) – (0.6537)M 2 + (0.2719)M 3
– (0.0330)M 4 + (0.0001)M 6 + (0.0118)(M)(T)
– (0.0003)(M 4)(T) + (0.0001)(M 5)(T) (3)
The graph of this prediction equation is shown in figure 2.
When a material is dried, water is first removed within a
“constant drying rate” period, during which both surface
and unbound water evaporate at a constant rate. After the
unbound water has been removed, the drying rate begins to
fall, and the process enters a “falling rate” period, of which
there may be more than one. At this point, drying is no
longer governed by surface evaporation but by mass
transfer within the material. This moisture transfer may
occur due to several processes, including evaporation
within the solid material, diffusion, and capillary
movement, and may be adversely affected by gravity,
material pore structures, and case hardening at the material
surface (Barbosa-Canovas and Vega-Mercado, 1996).
Second derivative analysis was used to determine the
inflection points in the regression curves, and thus
determine where the critical moisture contents occurred
(i.e., where the falling rate periods began). Batch-processed
masa by-products at 80°C had falling rate periods that
began at moisture contents of 6.08, 3.25, and 0.82 (d.b.); at
100°C had falling rate periods that began at moisture
contents of 5.97, 3.27, and 0.82 (d.b.); and at 120°C had
falling rate periods that began at 5.95, 3.28, and 0.82 (d.b.).
Continuous-processed masa by-products at 80°C had
falling rate periods that began at moisture contents of 7.81,
5.27, 2.61, and 2.00 (d.b.); at 100°C had falling rate periods
VOL. 15(5): 515-523
that began at moisture contents of 7.80 and 4.41 (d.b.); and
at 120°C had falling rate periods that began at moisture
contents of 7.79 and 4.17 (d.b.). For continuous-processed
by-products, critical moisture contents decreased with
increasing drying temperature, but drying temperature
seemed to have little effect on critical moisture content for
batch-processed samples. Also, it appears that, at least
initially, drying rates for the batch-processed by-product
are higher than those for the continuous-processed by-
product. The differences in drying behavior between the
two processing techniques may be due to differences in the
material structures which result from the cooking process
used (i.e., continuous vs batch processing).
Thermal Properties. Although thermal analyses were
not actually conducted, enough information was collected
to estimate thermal conductivity, specific heat, and thermal
diffusivity. Thermal conductivity can be estimated using
two prediction equations (Anderson, 1950; Spells 1961):
k = kW XW + kS(1 – XW) (4)
k = 0.056 + 0.57XW (5)
where k is the estimated thermal conductivity (W/m·K),
kW is the thermal conductivity of water (0.602 W/m·K),
XW is the mass fraction of water (-), and k S is the thermal
conductivity of organic solids (approx. 0.259 W/m·K).
Using equation 4, predicted thermal conductivities of
batch- and continuous-processed masa by-products were,
respectively, 0.565 and 0.561 W/m·K, and with equation 5,
0.565 and 0.558 W / m ·K. Specific heat can also be
predicted using two equations (Stroshine and Hamann,
1995; Choi and Okos, 1986):
519
fpe 1510 ms 9/25/01 12:52 PM Page 520
Figure 2–Drying curves for continuous-processed corn masa by-products at drying temperatures of 80, 100, and 120°C.
cp = 0.837 + 3.348M (6)
cp = 4.180XW + 1.711Xp + 1.928Xf
+ 1.547Xc + 0.908Xa (7)
where cp is the estimated specific heat (kJ/kg·K), M is the
moisture content (decimal wet basis — w.b.), Xp is the
mass fraction of protein (-), Xf is the mass fraction of fat
(-), Xc is the mass fraction of carbohydrates (-), and Xa is
the mass fraction of ash (-). Using equation 6, the predicted
values of specific heat for the batch- and continuous-
processed masa by-products were, respectively, 3.826 and
3.788 kJ / k g·K, and using equation 7, 3.848 and
3.817 kJ/kg·K. Thermal diffusivity can then be determined
using:
α= k (8)
ρc p
where α is the thermal diffusivity (m2 /s), and ρ is the mass
density (kg/m 3 ). Using equations 4 and 6 to estimate
thermal conductivity and specific heat, the predicted
thermal diffusivities for batch- and continuous-processed
masa by-products were, respectively, 1.43 × 10–7 and 1.41
× 10 –7 m2 /s. These values were all very similar to those for
water, which was not unexpected, due to the high moisture
contents of the masa by-product samples. Still, a future
investigation of measured thermal properties would be very
beneficial to verify the predicted estimates.
520
NUTRITIONAL PROPERTIES
Proximate Composition. The proximate compositions
of dry masa by-products are shown in table 1, and are vital
for the assessment of the suitability of masa by-products as
livestock feed additives. The results show that for both
processes, the by-products are low in both protein and fat,
moderate in ash content, and high in carbohydrate content.
Jurgens (1988) has outlined nutritional requirements for
various livestock diets. Growing and finishing swine (18 to
109 kg in body mass) require 13 to 15% crude protein in
their rations; growing and finishing beef cattle (300 to
400 kg) require 10.6 to 13.5% crude protein; growing and
finishing sheep (25 to 50 kg) require 12.9 to 17.2% crude
protein; and poultry require 14.0 to 23.2% crude protein.
Consequently, to effectively utilize masa by-products as a
livestock feed source, a high-protein supplemental material,
such as soybean meal, should be used in addition to the
masa by-products for the development of a feed ingredient.
Fiber Composition. The fiber composition results are
shown in table 1. Carbohydrates in organic materials
consist of nitrogen-free extract and crude fiber. Nitrogen-
free extract contains soluble sugars and starches, which are
easily digested by livestock. Crude fiber, on the other hand,
consists of cellulose and hemicellulose (which can be
digested only by ruminant animals due to microflora in the
rumen) and lignin, which is essentially indigestible.
Ruminant microflora produce cellulase, the enzyme
necessary to digest cellulose. Neutral Detergent Fiber
(NDF) consists of cell wall materials, provides a measure
of the total cellulose, hemicellulose, and lignin content of a
material, and can be used as an indicator of voluntary feed
intake for livestock. Acid Detergent Fiber (ADF) provides a
measure of total cellulose and lignin content of a material
APPLIED ENGINEERING IN AGRICULTURE
fpe 1510 ms 9/25/01 12:52 PM Page 521

(Jurgens, 1988). Utilizing the above relationships, it was
determined that the hemicellulose contents of batch- and
continuous-processed masa by-products, respectively, were
24.06 and 20.82% (d.b.), and the cellulose contents,
respectively, were 30.55 and 31.83% (d.b.). Masa by-
products are high in cellulose and hemicellulose and are
low in lignin, and therefore would be well suited for use as
potential feed ingredients for ruminant animals, which can
digest these fibrous materials.
Mineral Composition. The mineral compositions of
dry continuous-processed masa by-products are shown in
table 1. Of all minerals studied, calcium occurred in the
greatest quantity, with a mean value of 4.68% (d.b.), which
accounted for 26.9% of all ash material in the dry masa by-
product. The remaining minerals studied—potassium,
magnesium, and phosphorous—combined, accounted for
only 4.4% of all ash in the dry by-product. These results
suggest that masa by-products are potentially a source of
both calcium and fiber for livestock diets. However,
because substantial variability occurred in the data, it
would be beneficial to conduct a further study of mineral
content.
Amino Acid Composition. The amino acid
compositions of dry continuous-processed masa by-
products are shown in table 2. Even though corn masa by-
products are low in protein content, the protein which is
present is rich in the essential amino acids leucine and
proline, and in the nonessential amino acids alanine,
asparagine, glutamine, and glycine. However, masa by-
products are low in the essential amino acids arginine,
histidine, and methionine, and in the nonessential amino
acid tyrosine.
PROPERTY RELATIONSHIPS
The relationships between the 15 measured physical and
nutritional properties in the study were investigated using a
correlation analysis of the mean property values for each
sample. Fifteen of the resulting Pearson product-moment
correlations (Speigel, 1994) were significant (p < 0.05)
(table 3). The remainder of the correlations were not. The
correlation coefficient (r) quantifies the strength of the
linear relationship between two variables, and as shown in
the table, eight of the variable combinations in the study
had resulting correlation coefficients greater than 0.80 and
thus exhibited fairly strong linear relationships. The most
Table 2. Amino acid composition of the protein in dry continuous-
process corn masa by-products (amino acid g/100 g protein, d.b.)
Amino Acid Mean (n = 5) C.V. (%)
Alanine 11.35
Arginine 2.22
Asparagine 8.06
Glutamine 13.85
Glycine 10.18
Histidine 2.58
Isoleucine 3.58
Leucine 10.13
Lysine 4.62
Methionine 1.10
Phenylalanine 3.72
Proline 9.38
Serine 6.44
Threonine 6.10
Tyrosine 0.62 17.67
Valine 6.03
Table 3. Statistically significant correlation coefficients
(p < 0.05) for masa by-product properties
Correlation
Variable Associations Coefficient (r)
a b –0.767
a Density 0.744
a Fat 0.819
a Moisture content –0.704
b Density –0.690
b Fat –0.905
Carbohydrate a –0.808
Carbohydrate Density – 0.718
Carbohydrate Moisture content 0.778
L a –0.828
L b 0.978
L Density –0.700
L Fat –0.944
Lignin Yield stress –0.865
Moisture content Density –0.886
highly significant correlations occurred between L and b
(r = 0.978), L and Fat Content (r = –0.944), and b and Fat
Content (r = –0.905), and thus warrant further exploration.
To further investigate the relationships and interactions
between the masa by-product properties, a principal
components analysis was conducted on the 15 measured
properties using the mean property values for each sample.
Principal components analysis is used to reduce the
dimensionality of multivariate data by summarizing the
variance in the data and projecting it into a set of
uncorrelated orthogonal linear combinations of the original
variables. For this study, these linear combinations, or
principal components, have the form:
yP.C. = a1X1 + a2X2 + . . . + a15X15 (9)
where yP.C. is a principal component value, or score, a1
through a 15 are the principal component coefficients
(i.e., eigenvectors), and X1 through X15 are the original
property variables that were measured in the study (Everitt
and Dunn, 1991). The results for the principal components
analysis are presented in table 4, which shows the
eigenvectors and eigenvalues for the first five principal
components. The first five principal components accounted
for 92.6% of the total variability in the data, and thus
provide both a convenient and comprehensive summary of
the information contained in the original 15 property
variables, utilizing a reduced dimensionality of only five
variables. Although the interpretation of principal
components is very subjective, the first principal
component appears to be an indication of material
2.99
structure, while the second principal component appears to
5.87 be an indication of chemical composition. Another
2.25 advantage in using principal components analysis to
2.14 summarize data is the ability to easily identify outliers,
4.88 through examination of scatterplots of the calculated
3.24
1.25 principal component scores. Using this approach, it was
3.62 determined that an outlier did occur in the continuous-
5.17 process by-product data. This outlier was produced by two
9.09 yield stress readings that were somewhat lower than the
5.51
0.89
typical yield stress values for the by-product materials, and
1.77 might indicate a possible difference in material structure
2.01 due to the continuous-cooking process. Even though the
current study has been fairly extensive, it would be useful
0.74 to further investigate these relationships and interactions.

VOL. 15(5): 515-523 521

fpe 1510 ms 9/25/01 12:52 PM Page 522

Table 4. Eigenvectors of the first five principal components Barbosa-Canovas, G. V., and H. Vega-Mercado. 1996. Dehydration of
for masa by-product properties* Foods. New York, N.Y.: Int. Thomson Publishing.
Property P.C. 1 P.C. 2 P.C. 3 P.C. 4 P.C. 5 Beauchat, L. R. 1981. Microbial stability as affected by water
activity. Cereal Foods World 26(7): 345-349.
Moisture content (%, w.b.) 0.316 0.251 0.096 –0.280 –0.015
Bohlsen, D., S. Weeder, and S. Wang. 1997. Tracking organic
Water activity (-) 0.136 –0.043 0.539 –0.394 0.143
Density (kg/m3) –0.344 –0.109 –0.094 0.248 0.020 waste. Resource 4(9): 13-14.
Yield stress (N/m2) 0.267 –0.152 –0.412 –0.132 –0.049 Bressani, R., R. Paz y Paz, and N. S. Scrimshaw. 1958. Corn
pH (-) 0.037 –0.333 0.378 –0.281 –0.134 nutrient losses: Chemical changes in corn during preparation of
L (-) 0.353 –0.157 0.122 0.205 0.112 tortillas. J. Agric. Food Chem. 6: 770.
a (-) –0.351 –0.078 0.073 –0.216 –0.141 Choi, Y., and M. Okos. 1986. Thermal properties of liquid foods-
b (-) 0.345 –0.133 0.132 0.173 0.208
review. In Physical and Chemical Properties of Food, ed. M.
Protein (%, d.b.) –0.047 0.312 –0.043 –0.067 0.711
Fat (%, d.b.) –0.326 0.195 0.006 –0.363 –0.075 Okos, 35-77. St. Joseph, Mich.: ASAE.
Ash (%, d.b.) 0.034 –0.564 0.101 0.217 0.044 Dzuy, N. Q., and D. V. Boger. 1985. Direct yield stress measurement
Carbohydrate (%, d.b.) 0.302 0.338 –0.102 0.148 –0.071 with the vane method. J. Rheology 29(3): 335-347.
NDF (%, d.b.) 0.163 0.355 0.360 0.195 –0.352 Everitt, B. S., and G. Dunn. 1991. Applied Multivariate Data
ADF (%, d.b.) –0.186 0.216 0.292 0.434 –0.232 Analysis. New York, N.Y.: Halsted Press.
Lignin (%, d.b.) –0.248 0.024 0.323 0.240 0.433
Ferris, D. A., R. A. Flores, C. W. Shanklin, and M. K. Whitworth.
Eigenvalue 6.592 2.834 2.141 1.288 1.035 1995. Proximate analysis of food service wastes. Applied
Proportion of 43.94 18.90 14.27 8.59 6.90 Engineering in Agriculture 11(4): 567-572.
variation explained (%) Gomez, M. H., L. W. Rooney, R. D. Waniska, and R. L.
Total variation 43.94 62.84 77.11 85.70 92.60
explained (%) Pflugfelder. 1987. Dry corn masa flours for tortilla and snack
food production. Cereal Foods World 32(5): 372-377.
* P.C. denotes principal component eigenvector. Gonzalez de Palacios, M. 1980. Characteristics of corn and
sorghum for tortilla processing. Unpub. M.S. thesis. College
Station, Tex.: Texas A&M University.
Gonzalez-Martinez, S. 1984. Biological treatability of the
wastewaters from the alkaline cooking of maize (Indian corn).
SUMMARY AND CONCLUSIONS Environ. Technol. Letters 5(8): 365-372.
Characterization of the residual streams from masa Ham, G. A., R. A. Stock, T. J. Kopfenstein, E. M. Larson, D. J.
processing plants provides physical and nutritional data Shain, and R. P. Huffman. 1994. Wet corn distillers by-products
necessary for effective utilization of these by-product compared with dried corn distillers grains with solubles as a
materials. Masa by-products (i.e., suspended solids source of protein and energy for ruminants. J. Animal Science
removed from masa processing waste water) are very 72(12): 3246.
Ham, G. A., R. A. Stock, T. J. Klopfenstein, and R. P. Huffman.
suitable for use as livestock feed additives, or components
1995. Determining the net energy value of wet and dry corn
thereof. Because these by-products are high in fiber gluten feed in beef growing and finishing diets. J. Animal
(especially cellulose and hemicellulose), they would be Science 73(2): 353-359.
best suited for ruminant diets. Additionally, the products Hunter Associates Laboratory, Inc. 1983. Instruction Manual:
have a substantial calcium content, so there is also potential Hunterlab LabScan Spectro Colorimeter. Reston, Va.: Hunter
for use as a calcium source for livestock diets. No matter Associates Laboratory, Inc.
how masa by-products are eventually utilized, they must be Hussein, H. S., and L. L. Berger. 1995. Effects of feed intake and
dehydrated to reduce transportation costs, to decrease dietary level of wet corn gluten feed on feedlot performance,
microbial activity, and to increase shelf life. This can be digestibility of nutrients, and carcass characteristics of
accomplished through drying, blending, or extrusion growing-finishing beef heifers. J. Animal Science 73(11): 3246.
Jurgens, M. H. 1988. Animal Feeding and Nutrition. Dubuque,
processing, all of which would be appropriate areas for
Iowa: Kendall/Hunt Publishing Co.
future investigations. Katz, S. H., M. L. Hediger, and L. A. Valleroy. 1974. Traditional
maize processing techniques in the New World. Science 184: 765.
ACKNOWLEDGMENTS. The authors would like to thank the Khan, M. N., M. C. Des Rosiers, L. W. Rooney, R. G. Morgan, and
Recycling and Reuse Technology Transfer Center V. E. Sweat. 1982. Corn tortillas: Evaluation of corn cooking
(RRTTC) at the University of Northern Iowa, Cedar Falls, procedures. Cereal Chemistry 59(4): 279-284.
Iowa, for financial support and the Center for Crops Krause, V. M., K. L. Tucker, H. V. Kuhnlein, C. Y. Lopez-Palacios,
Utilization Research (CCUR) at Iowa State University, M. Ruz, and N. W. Solomons. 1992. Rural-urban variation in
Ames, Iowa, for use of equipment and facilities. They limed maize use and tortilla consumption by women in
would also like to acknowledge the assistance of Dr. Guatemala. Ecol. Food & Nutrit. 28(4): 279-288.
Larson, E. M., R. A. Stock, T. J. Klopfenstein, J. H. Sindt, and R.
Hyesun Park during initial stages of nutritional analysis.
P. Huffman. 1993. Feeding value of wet distillers by-products
for finishing ruminants. J. Animal Science 71(8): 2228.
REFERENCES Luzier, J. M., and R. C. Summerfelt. 1995. Partial replacement of
American Association of Cereal Chemists (AACC). 1995. fish meal with spray-dried blood powder to reduce
Approved Methods of the American Association of Cereal phosphorous concentrations in diets for juvenile rainbow trout,
Chemists. St. Paul, Minn.: Am. Society of Cereal Chemists. Oncorhynchus mykiss (Walbaum). Aquaculture Res. 26(8): 577.
American Public Health Association (APHA). 1995. 19th Ed. Martins, S. N., and E. C. Guzman. 1994. Effect of drying method
Standard Methods for the Examination of Water and of bovine blood on the performance of growing for tambaqui
Wastewater. New York, N.Y.: Am. Public Health Assoc. (Colossoma macropomum, Curier 1818) in experimental
Anderson, S. A. 1950. Automatic Refrigeration. Norberg, culture tanks. Aquaculture 124(1-4): 335-341.
Denmark: MacLaren & Son Ltd. for Don Foss. Microsoft Excel v. 5.0a. 1993. Redmond, Wash.: Microsoft Corp.
Association of Official Analytical Chemists (AOAC). 1980. 13th Moore, K. J., and R. D. Hatfield. 1994. Carbohydrates and forage
Ed. Official Methods of Analysis of the AOAC International. quality. In Forage Quality, Evaluation, and Utilization, 229-
Arlington, Va.: Assoc. Official Analytical Chemists Int.

522 APPLIED ENGINEERING IN AGRICULTURE

fpe 1510 ms 9/25/01 12:52 PM Page 523

280, eds. G. C. Fahey Jr., M. Collins, D. R. Martens, and L. E. Steffe, J. F. 1992. Rheological Methods in Food Process
Moser. Madison, Wis.: American Society of Agronomy. Engineering. East Lansing, Mich.: Freeman Press.
Nakayama, T., E. Niwa, and I. Hamada. 1980. Pipe transportation Stroshine, R., and D. Hamann. 1995. Physical Properties of
of minced fish paste. J. Food Sci. 45(4): 844-847. Agricultural Materials and Food Products. West Lafayette,
Nelson, L. S. 1985. Sample size tables for analysis of variance. J. Ind.: Copy Cat.
Quality Technol. 17(3): 167-168. Van Soest, P. J., J. B. Robertson, and B. A. Lewis. 1991. Methods
Nelson, R. G., and R. A. Flores. 1994. Survey of processing for dietary fiber, neutral detergent fiber, and nonstarch
residues generated by Kansas agribusinesses. Applied polysaccharides in relation to animal nutrition. J. Dairy Sci.
Engineering in Agriculture 10(5): 703-708. 74(10): 3583-3597.
Ortiz, E. L. 1985. The cuisine of Mexico: Corn tortilla dishes. Wang, L., R. A. Flores, and L. A. Johnson. 1997. Processing feed
Gourmet 45(May): 62-63. ingredients from blends of soybean meal, whole blood, and red
Pflugfelder, R. L., L. W. Rooney, and R. D. Waniska. 1988. Dry blood cells. Transactions of the ASAE 40(3): 691-697.
matter losses in commercial corn masa production. Cereal Wood, M. 1994. What makes tortillas and tacos taste so good?
Chem. 65(2): 127-132. Agric. Res. 42(11): 17.
Polanski, J. 1992. Legalizing the feeding of nonmeat food wastes to Yoo, B., M. A. Rao, and J. F. Steffe. 1995. Yield stress of food
livestock. Applied Engineering in Agriculture 11(1): 115-119. dispersions with the vane method at controlled shear rate and
Rooney, L. W., and S. O. Serna-Saldivar. 1987. Food uses of whole shear stress. J. Texture Studies 26(1): 1-10.
corn and dry-milled fractions. In Corn Chemistry and
Technology, 399-430, eds. S. A. Watson, and P. E. Ramstad. St.
Paul, Minn.: Am. Assoc. Cereal Chemists.
Rosentrater, K., and R. Flores. 1997. Physical and rheological
properties of slaughterhouse swine blood and blood
components. Transactions of the ASAE 40(3): 683-689.
SAS v. 6.08. 1992. Cary, N.C.: SAS Institute, Inc.
Serna-Saldivar, S. O., M. H. Gomez, and L. M. Rooney. 1990.
Technology, chemistry, and nutritional value of alkaline-cooked
corn products. Adv. Cereal Sci. & Technol. 10: 243-306.
Speigel, M. R. 1994. Statistics. New York, N.Y.: McGraw-Hill, Inc.
Spells, K. E. 1961. Thermal conductivity of some biological fluids.
Physics in Med. & Biol. 5(2): 139-153.

VOL. 15(5): 515-523 523

fpe 1510 ms 9/25/01 12:52 PM Page 524