Content uploaded by Tong Wang
Author content
All content in this area was uploaded by Tong Wang on Jan 07, 2015
Content may be subject to copyright.
ABSTRACT: In recent years there has been widespread growth
in extruding-expelling (E-E) facilities for small-scale processing
of soybeans. To compete in a highly competitive market, these
E-E operations are looking for ways to optimize production of
their oil and meal products for values to their customers. The
objective of this study was to determine the ranges of residual
oil contents and protein dispersibility indices (PDI) possible
with E-E processing of soybeans. We also characterized the par-
tially defatted meal for other factors important in food and feed
applications. Residual oil and PDI values ranged from 4.7 to
12.7% and 12.5 to 69.1%, respectively. E-E conditions signifi-
cantly influenced residual lipase, lipoxygenase (L1–L3), and
trypsin inhibitor activities. Chemical compositions were differ-
ent for whole, dehulled, and reduced-moisture soybeans, with
dehulled soybeans tending to produce meals having higher
residual oil contents at higher PDI values. It was possible to
process soybeans with different characteristics (e.g., moisture
content, whole, dehulled) to produce meals and flours with
wide ranges of properties, providing E-E operators with oppor-
tunities to market value-added products.
Paper no. J9700 in JAOCS 78, 775–779 (August 2001).
KEY WORDS: Expelling, extrusion, oil extraction, partially de-
fatted soy flour, PDI, screw pressing, soybean meal, soybean
processing.
Extruding-expelling (E-E) is a relatively new process devel-
oped by Nelson et al. (1) to mechanically recover oil from
soybeans. This process eliminates the need for costly steam
dryers and conditioners and associated steam generation, en-
hances oil extraction over simple screw pressing, and elimi-
nates the use of organic solvents. Small-scale E-E facilities,
also known as mini-mills, are increasing in popularity be-
cause of the low capital investment required and ability to
process identity-preserved and organic products. The low-fat,
high-protein, high-energy meals are desirable products for use
as animal feeds, especially dairy cattle feed (2). E-E soybean
meal reportedly has higher digestible energy and amino acid
availability compared with solvent-extracted meal (3,4). In
addition, the nonuse of organic solvents in E-E meal produc-
tion makes partially defatted soy flour particularly attractive
to producers of natural foods.
To develop value-added products from E-E soybean meal,
it is important to understand the ranges of protein solubilities,
oil contents, enzyme activities, and protease-inhibitor activi-
ties that are possible with this new processing technology. Soy
flours with high protein dispersibility indices (PDI) and low
oil contents are generally considered to be required to produce
food-grade soy flour and high-quality texturized proteins with
fewer processing difficulties, although the activities of certain
enzymes, often associated with high PDI, could contribute to
off-flavor development or antinutritional effects (5). However,
increasing the range of PDI values for partially defatted soy
flour that can be produced by E-E soybean mills could enable
using these products in a wide variety of food applications.
The objective of this study was to determine the ranges of
residual oil contents and PDI values of partially defatted soy
flours that are achievable by changing extruder and expeller
(screw press) conditions within practical confines of a com-
mercial E-E mini-mill operation. These partially defatted soy
flours were characterized to determine their suitabilities for
human food and animal feed applications.
EXPERIMENTAL PROCEDURES
Experimental design. This experiment was designed to use E-E
to produce partially defatted soy flours with the widest possi-
ble ranges of residual oil contents and PDI values. The tar-
geted PDI and residual oil values were selected to represent
the widest range believed, a priori, to be possible and useful
using different processing conditions that are easily attainable
or commonly used at E-E mini-mills. Both whole and de-
hulled soybeans were used.
Raw materials. Whole soybeans (Latham 610) at 9.5%
moisture were obtained from Iowa Soy Specialties (Vinton,
IA). Some of the beans were dried to 6.7% moisture using
ambient temperature (22°C) air; the remainder were used as
is. The beans were dehulled using traditional methods of
cracking the soybeans into 6–8 pieces with a corrugated roller
mill (Ferrell-Ross, Oklahoma City, OK), and then aspirating
the hulls with a Multi-Aspirator (Kice, Wichita, KS).
Extruding and expelling. An Insta-Pro 2500 dry extruder
(Triple “F”; Insta-Pro, Des Moines, IA) was used to dry-
extrude whole and dehulled soybeans. Oil expression was car-
ried out with an Insta-Pro 1500 screw press. The extruder was
capable of varying barrel temperature and mechanical input
by manipulating the screw design and shear lock configura-
tion, as well as via die (nose cone) restriction and design. Ad-
ditionally, the feed rate to the extruder could be changed. Res-
idence time within the extruder was ca. 20–25 s. Processing
Copyright © 2001 by AOCS Press 775 JAOCS, Vol. 78, no. 8 (2001)
*To whom correspondence should be addressed at 1041 Food Science Build-
ing, Iowa State University, Ames, IA, 50011. E-mail: ljohnson@iastate.edu
Characterization of Extruded-Expelled Soybean Flours
Troy W. Crowe
a
, Lawrence A. Johnson
a,b,
*, and Tong Wang
a
a
Department of Food Science and Human Nutrition and
b
Center for Crops
Utilization Research, Iowa State University, Ames, IA 50011
parameters used to obtain different residual oil contents and
PDI values are shown in Table 1. The extruder barrel is di-
vided into three equal sections with temperature gauges
mounted in the middle of each (Zone 1 being close to the die
and Zone 3 being close to the feed).
Three samples were screw-pressed twice after one pass
through the extruder in attempts to produce very low residual
oil contents. After the first pass through the screw press, sam-
ples were collected into large tubs and held until sufficient
material was produced to be refed into the screw press.
Each E-E processing trial was carried out in duplicate. Fol-
lowing E-E processing, the press cake (both single- and twice-
expelled) was placed into plastic-lined paper bags and allowed
to cool to ambient temperature in the open bag until sealing
for transport. Samples were stored at −20°C until milled.
Flour milling. The soybean meal press cake was milled
(94.7% <100 mesh) by first passing it through a set of crack-
ing rolls and then through a Fitzmill (The Fitzpatrick Com-
pany, Elmhurst, IL). The Fitzmill was operated at 7,000 rpm
using the blades in a blunt hammermill configuration, fed at
30 rpm feed screw speed, and fitted with a 1536–0060 screen.
Milled samples were stored at −20°C until analyzed.
Meal characterization. Moisture contents of soy flours
were determined according to the 2-h oven-drying method
(6). Crude fat content was determined by Goldfisch extrac-
tion (7). Nitrogen content was measured by using a Perkin-
Elmer Series II Nitrogen Analyzer 2410 (PerkinElmer Corp.,
Norwalk, CT). Nitrogen content was multiplied by a factor of
6.25 for estimating crude protein content. Lipase activity was
measured in duplicate as outlined by Moscowitz et al. (8)
with the modifications of Guzman et al. (9). Lipoxygenase
activity was measured in duplicate as outlined by Zhu et al.
(10). Trypsin inhibitor (TI) activity and PDI values were ana-
lyzed according to AOCS official methods at Woodson-
Tenent Laboratories (Des Moines, IA). Moisture content,
crude protein, and crude fat were analyzed in triplicate.
Statistical analysis. Statistical analyses were performed
using the General Linear Model procedures of SAS 6.06 (11).
Significance was established at P < 0.05.
RESULTS AND DISCUSSION
E-E equipment performance. Whole soybeans generally pro-
duced higher extruder barrel temperatures compared with de-
hulled soybeans (Table 1). Jin et al. (12) reported that fiber
addition caused extruder torque, die pressure, and specific en-
ergy to increase, which they attributed to increased dough
mass viscosity. Total dietary fiber content (not measured) is
significantly higher for E-E meal from whole soybeans than
from dehulled soybeans. Given the reported health benefits
associated with dietary fiber (13), the use of whole soybeans
might be attractive in food applications, if the fiber is not
detrimental to the performance, taste, and texture of foods in
which the flour is incorporated.
We also observed higher foots contents during oil collec-
tion when dehulled soybeans were processed. This is an im-
portant consideration for processors because more oil settling
capacity will be required when dehulling soybeans prior to E-E
processing.
Proximate analyses. Results from the compositional analy-
ses of the E-E soybean meal samples are presented in Table
2. Partially defatted soy flours with a wide range of PDI val-
ues (12.5–69.1) and residual oil contents (4.73–12.65%) were
produced by E-E. Highest and lowest oil recoveries were
76.0% (PDI/residual oil content/times expelled, 13/5/1) and
35.8% (63/13/1), respectively. Dehulled soybeans tended (not
significant at P = 0.05) to have increased PDI values and
higher residual oil contents compared with whole soybeans
776 T.W. CROWE ET AL.
JAOCS, Vol. 78, no. 8 (2001)
TABLE 1
Extruder and Expeller Operating Conditions for Production of Extruded-Expelled Soybean Flour
Extruder Nose cone Choke setting Feed rate
Current (amps) Barrel temp. (°C)
Sample code
a
configuration
b
(cm) (cm) (kg/h) Extruder Expeller Zone 1 Zone 2 Zone 3
13/5/1-W 11-6-6-6, DF 0.8 1.0 590 128 28 162 147 107
26/5/1-W 11-11-6-6, SF 0.8 1.0 615 119 28 138 88 56
20/5/1-W 11-6-6-6, DF 1.0 1.1 590 112 25 144 107 89
14/7/1 11-6-6-6, DF 1.0 1.1 590 105 22 144 102 76
43/6/1 11-11-6-6, SF 1.0 1.9 730 107 21 129 80 48
38/8/1 11R-11R-11R-11, SF Tight 0.9 590 94 21 132 72 28
45/7/1 11R-11R-11R-11, SF 0.8 1.1 590 95 22 126 57 31
61/10/1 11R-11R-11R-11R, SF 1.0 0.9 590 81 21 117 42 27
63/13/1 11R-11R-11R-11R, SF 1.6 Tight 950 72 25 86 55 27
54/12/1 11R-11R-11R-11R, SF 1.6 0.9 590 81 21 89 34 24
69/12/1 11R-11R-11R-11R, SF 1.6 1.1 590 74 20 88 27 23
35/5/2 11R-11R-11R-11, SF 0.8 1.0 730 109 24 129 99 41
43/7/1-L 11R-11R-11R-11, SF 0.8 1.0 730 119 34 137 76 46
67/10/2 11R-11R-11R-11R, SF 1.6 1.1 590 72 22 85 54 27
58/8/1 11R-11R-11R-11R, SF 1.6 1.0 730 107 28 119 64 29
55/6/2 11R-11R-11R-11R, SF 1.6 1.0 730 107 28 119 64 29
54/8/1-L 11R-11R-11R-11R, SF 1.6 1.0 730 98 28 129 56 37
a
Denotes protein dispersibility index (PDI)/residual oil content/times expelled; W indicates whole beans; L indicates low moisture.
b
Numbers and R denote shear lock type used from feed end to die end of the extruder; DF denotes double flighting of the screw; SF denotes single flighting
of the screw.
processed under identical E-E conditions, as in the case for
sample 14/7/1 (dehulled) vs. 20/5/1 (whole). These results are
contrary to those of Nelson et al. (1), who reported signifi-
cantly higher oil yield when using dehulled soybeans, al-
though that difference diminished following removal of oil
fines or foots.
As expected, twice-screw-pressed flour samples had sig-
nificantly, but modestly, lower residual oil contents compared
with single-screw-pressed flours processed under identical
operating conditions (Tables 1 and 2). Single-screw-pressed
meal was ca. 2 percentage points higher in residual oil con-
tent than twice-screw-pressed meal. Nelson et al. (1), using a
different type of screw press, found PDI was ~2 percentage
points lower in single-screw-pressed flours. In the present
study, no significant changes in the range of PDI values were
observed in twice-screw-pressed flours vs. single-screw-
pressed flours. Thus, screw pressing in series can modestly
decrease the residual oil content while maintaining protein
functionality. This may be significant for use in lower-fat
flours for food applications.
The E-E processed meals produced from reduced-moisture
(6.7%) soybeans did not differ significantly from soybeans
with higher moisture content (9.5%) in compositional proper-
ties (Table 2). Drying did not improve oil recovery. The rela-
tionship between drying and PDI is unclear. There was a 5
percentage point decrease in PDI associated with dried sam-
ples 58/8/1 vs. 54/8/1. In addition, increased barrel tempera-
tures were observed during extrusion of the dried soybeans
(Table 1). Zhu et al. (10) found that PDI significantly de-
creased during dry extrusion with increasing extrusion tem-
perature and moisture content.
PDI was directly correlated with residual oil content (R =
0.824, P < 0.0001; Fig. 1). Comparison of low (10–40),
medium (40–60), and high (>60) PDI samples revealed sig-
nificantly higher mean residual oil content for high- compared
with low-PDI flours (high PDI = 10.9%, low PDI = 5.9%, P
= 0.05; Fig. 2). Temperatures in the three extruder zones were
the most important factors affecting PDI and residual oil of
E-E processed soy flour. As the temperature of extruder Zone
1 increased, both PDI and residual oil content decreased (R =
−0.861 (PDI), R = −0.946 (residual oil), P < 0.05; Fig. 3).
Similar correlations were found with respect to the tempera-
tures of extruder Zones 2 and 3. These data indicate that al-
tering the final PDI and residual oil content of E-E partially
defatted soy flour is possible by adjusting the feed rate and
screw and shear lock configurations, thereby changing the ex-
trusion zone temperatures.
EXTRUDED-EXPELLED SOYBEAN FLOURS 777
JAOCS, Vol. 78, no. 8 (2001)
FIG. 1. Distributions of residual oil vs. protein dispersibility index (PDI)
of extruded-expelled soy flours.
FIG. 2. Residual oil contents of extruded-expelled soy flours for low,
medium, and high PDI ranges. Error bars, ± 1 SD. See Figure 1 for ab-
breviation.
TABLE 2
Chemical Analyses of Extruded-Expelled Soybean Flour
a
Sample Dry matter Crude protein Residual oil
code
b
(%) (% mfb)
c
PDI (% mfb)
13/5/1-W 96.1
g,h
50.4
c,d
12.5
a
4.7
a
26/5/1-W 94.5
e
48.1
b
25.6
b
5.3
a,b
20/5/1-W 95.6
f,g
49.4
b,c
20.0
a,b
5.2
a,b
14/7/1 95.9
g
50.2
c,d
14.3
a
6.8
b,c
43/6/1 94.1
d,e
51.1
d
42.9
c,d
6.3
b
38/8/1 95.2
f
51.4
d
37.8
c
7.8
c
45/7/1 94.8
e,f
51.2
d
45.2
c,d
7.6
c
61/10/1 94.2
d,e
50.6
c,d
61.4
e,f,g
9.6
d
63/13/1 93.8
d
49.6
c
63.0
e,f,g
12.7
e
54/12/1 92.8
c
48.6
b,c
54.0
d,e,f
11.6
e
69/12/1 91.8
b
49.6
c
69.1
g,h
11.7
e
35/5/2 94.3
d,e
51.6
d
35.4
b,c
5.4
a,b
43/7/1-L 96.5
h
50.9
d
43.0
c,d
6.6
b,c
67/10/2 94.2
d,e
50.6
c,d
66.7
f,g
9.9
d
58/8/1 93.7
d
50.9
c,d
58.1
e,f,g
7.8
c
55/6/2 94.0
d,e
52.4
d
55.4
d,e,f
5.7
a,b
54/8/1-L 96.0
g,h
50.4
c,d
53.8
d,e,f
8.1
c
Control 91.3
a
39.7
a
98.7
i
19.7
f
a
Means within each column with different roman superscripts are signifi-
cantly different at P < 0.05.
b
Denotes PDI/residual oil content/times expelled; W indicates whole beans;
L indicates low moisture. See Table 1 for other abbreviation.
c
mfb, moisture-free basis.
The low extrusion temperatures necessary to produce high
PDI generally are less efficient in rupturing soybean sphero-
somes and therefore do not facilitate oil extraction as evi-
denced by the high residual oil contents. This study was de-
signed to produce wide ranges of PDI and residual oil values
(e.g., high PDI, low residual oil), and the results indicate that
partially defatted soy flour with optimal properties (high PDI
and low residual oil content) can be produced by altering feed
materials, extruder configuration, and processing conditions
from those typically used today.
TI inhibitor and enzyme activities of E-E processed soy
flours. TI activities (Table 3) ranged from 4.5 to 97.5% of the
activity of raw soybeans and decreased with increasing extruder
barrel temperature (Zone 1) (R = −0.816, P < 0.05; Fig. 4). Guz-
man et al. (9) varied extrusion temperatures from 127 to 160°C
and reported corresponding residual TI activities in nonexpelled
samples between 31 and 2% of the original activity. Eweedah et
al. (3) and Nelson et al. (1) used similar extrusion systems at
temperatures of 150 and 135–141°C, respectively. In both stud-
ies, TI was reduced to ~6% of its original activity.
Lipase activities were not significantly different among
samples and were not correlated with extruder barrel temper-
ature. These data are in agreement with those previously re-
ported by Guzman et al. (9), who found no trend for lipase
activity in extrusion processed soybean-corn mixtures.
The activities of all three lipoxygenase isozymes (L1, L2,
and L3) decreased with higher Zone 1 temperature (P < 0.05)
and were not detectable in the partially defatted soy flour
samples when extruded at Zone 1 barrel temperatures of
117°C or higher (Table 4). This was expected following E-E
processing because of the high temperatures and long hold
times in both the extruder and expeller. Activity levels of L3,
778 T.W. CROWE ET AL.
JAOCS, Vol. 78, no. 8 (2001)
TABLE 3
Lipase and Trypsin Inhibitor Activities of Extruded-Expelled Soybean
Flours
Lipase Trypsin inhibitor,
Sample code
a
(mM H
+
/min/g) trypsin inhibitor units
13/5/1-W 18.6 ≤2,000
26/5/1-W 21.0 5,200
20/5/1-W 16.2 N/A
b
14/7/1 15.8 5,000
43/6/1 11.8 N/A
38/8/1 28.0 N/A
45/7/1 15.4 13,500
61/10/1 18.8 N/A
63/13/1 15.1 N/A
54/12/1 17.9 26,900
69/12/1 10.7 36,500
35/5/2 20.9 10,200
43/7/1-L 13.8 N/A
67/10/2 10.1 43,500
58/8/1 19.2 N/A
55/6/2 17.5 27,275
54/8/1-L 13.2 N/A
Control 19.4 44,600
a
Denotes PDI/residual oil content/times expelled; W indicates whole beans;
L indicates low moisture.
b
N/A denotes not applicable. See Table 1 for abbreviation.
FIG. 4. Relationship between extruder barrel temperature (Zone 1, °C)
and trypsin inhibitor activity of extruded-expelled soy flour.
FIG. 3. Relationships between PDI (A) and residual oil content (B) of
extruded-expelled soy flour, extruder barrel temperature (Zone 1, °C).
See Figure 1 for abbreviation.
the most heat-labile isozyme, were much lower than those ob-
served for the L1 and L2 isozymes (Table 4). No lipoxyge-
nase activity was detected in partially defatted soy flours ex-
truded at 117°C and higher (Zone 1 barrel temperature).
These results are consistent with those reported by Zhu et al.
(10) and Guzman et al. (9), who detected no lipoxygenase ac-
tivity at temperatures greater than 107 and 127°C, respec-
tively. These data suggest that only those partially defatted
soy flours produced using low temperatures to achieve a high
PDI may contain appreciable lipoxygenase activity. This may
be important in food applications of E-E processed partially
defatted soy flour because these enzymes may significantly
affect the colors and flavors of foods in which the flours are
incorporated.
ACKNOWLEDGMENTS
Journal Paper No. 18834 of the Iowa Agriculture and Home Eco-
nomics Experiment Station, Ames, Iowa, Project No. 3507, and sup-
ported by Hatch Act and State of Iowa funds, Department of Food
Science and Human Nutrition, Center for Crops Utilization Re-
search, and the Iowa Agriculture and Home Economics Experimen-
tal Station, Iowa State University, Ames, IA 50011.
REFERENCES
1. Nelson, A.I., W.B. Wijeratne, S.W. Yeh, T.M. Wei, and L.S.
Wei, Dry Extrusion as an Aid to Mechanical Expelling of Oil
from Soybeans, J. Am. Oil Chem. Soc. 64:1341–1347 (1987).
2. Said, N.W., Dry Extrusion—Mechanical Expelling, INFORM
9:139, 141–144 (1998).
3. Eweedah, N., J. Gundel, and T. Matrai, Protein Degradability,
Amino Acid Composition, Trypsin Inhibitor and Urease Activ-
ity of Raw and Heat-Treated Fullfat Soybean, Arch. Anim. Nutr.
50:361–367 (1997).
4. Zhang, Y., and C.M. Parsons, Effects of Extrusion and Ex-
pelling on the Nutritional Quality of Conventional and Kunitz
Trypsin Inhibitor-Free Soybeans, Poult. Sci. 72:2299–2308
(1993).
5. Schwimmer, S., Source Book of Food Enzymology, AVI Pub-
lishing, Westport, CT, 1981.
6. American Oil Chemists’ Society, Official Methods and Recom-
mended Practices of the American Oil Chemists’ Society, 4th
edn., AOCS Press, Champaign, 1993, Method Ba-38.
7. Approved Methods of the American Association of Cereal
Chemists, 8th edn., American Association of Cereal Chemists,
St. Paul, MN, 1983, Method 30-25.
8. Moscowitz, G.J., R. Cassaigne, I.R. West, T. Shen, and L.I.
Feldman, Hydrolysis of Animal Fat and Vegetable Oil with
Mucor miehei Esterase: Properties of the Enzyme, J. Agric.
Food Chem. 25:1146–1150 (1977).
9. Guzman, G.J., P.A. Murphy, and L.A. Johnson, Properties of
Soybean–Corn Mixtures Processed by Low-Cost Extrusion, J.
Food Sci. 54:1590–1593 (1989).
10. Zhu, S., M.N. Riaz, and E.W. Lusas, Effect of Different Extru-
sion Temperatures and Moisture Content on Lipoxygenase In-
activation and Protein Solubility in Soybeans, J. Agric. Food
Chem. 44:3315–3318 (1996).
11. Statistical Analysis System, SAS Institute, Cary, NC, 1991.
12. Jin, Z., F. Hsieh, and H.E. Huff, Extrusion Cooking of Corn
Meal with Soy Fiber, Salt, and Sugar, Cereal Chem. 71:227–234
(1994).
13. Gordon, D.T., Functional Properties vs. Physiological Action of
Total Dietary Fiber, Cereal Foods World 34:517–525 (1989).
[Received July 13, 2000; accepted July 25, 2001]
EXTRUDED-EXPELLED SOYBEAN FLOURS 779
JAOCS, Vol. 78, no. 8 (2001)
TABLE 4
Lipoxygenase Isozyme (L1, L2, L3) Activities of Extruded-Expelled
Soybean Flours
a
Barrel temperature (°C)
Sample code
b
L1 L2 L3
13/5/1-W ND
c
ND ND
26/5/1-W ND ND ND
20/5/1-W ND ND ND
14/7/1 ND ND ND
43/6/1 ND ND ND
38/8/1 ND ND ND
45/7/1 ND ND ND
61/10/1 ND ND ND
63/13/1 16.4
b
12.9
b
4.4
a
54/12/1 9.3
a,b
8.1
a
3.9
a
69/12/1 10.7 14.1
b
5.1
a
35/5/2 ND ND ND
43/7/1-L ND ND ND
67/10/2 7.8
a
12.1
a,b
2.1
b
58/8/1 ND ND ND
55/6/2 ND ND ND
54/8/1-L ND ND ND
Control 100.0
c
100.0
c
100.0
c
a
Means within each column with different nonitalic superscripts are signifi-
cantly different (P < 0.05).
b
Denotes PDI/residual oil content/times expelled; W indicates whole beans;
L indicates low moisture.
c
ND, not detectable. See Table 1 for other abbreviation.