EXTRUSION

OF
TEXTURIZED PROTEINS
Joseph P. Kearns
Wenger International, Inc., Kansas City, Missouri, U.S.A.

Galen J. Rokey and Gordon R. Huber

Wenger Manufacturing, Sabetha, Kansas, U.S.A.
Abstract

Extrusion cooking is a continuous process by which many foods are produced on an

industrial basis. Texturized proteins, a unique product made by extrusion, can he produced
from a wide range of raw ingredient specifications, while controlling the functional
properties such as density, rate and time of rehydration, shape, product appearance and
mouthfeel. Raw material specifications for extrusion of texturized proteins have increased
and now include: PDI ranges from 20 to 80; fat levels from 0.5 to 6.5%; fiber levels up to
7%; and particle size up to 8 mesh (2360 micron). Additional benefits of extrusion cooking
are denaturing of the proteins, deactivation of heat liable growth inhibitors, control of bitter
flavors and the homogeneous bonding of ingredients that may include colors, chemicals and
other additives which can have an effect on appearance or textural quality.

Extrusion of texturized proteins is one of many successful applications of this unique cooking
process. There are other methods utilized to produce texturized proteins including spun soy
protein isolates and formed meat analogs. Spun soy protein isolates involve redissolving
precipitated vegetable proteins and passing them through a spinneret into a precipitating bath.
The bundles of fibers resulting from this are compacted, shaped, flavored, cooked and/or
dried and packaged. Formed meat analogs are blends of various protein sources such as
isolates, glutens, albumin, extrusion-cooked vegetable proteins and others which are blended
with oils, flavors and binders before forming them into sheets, patties, strips or disks.

Extrusion-cooked texturized proteins include meat extenders in the form of chunks or small
granular pieces which are wet milled or produced directly off the extruder. Extruders also are
able to produce a meat analog tha t has a remarkable similarity in appearance, texture and
mouthfeel to meats. The utilization of extrusion cooking throughout the food industry has
shown that a variety of products can be made on extrusion equipment. Some of these
products include breakfast cereals, breadings, snacks, instant rice, instant pasta, starch
modifications, animal and aquatic feeds.

This paper, however, will center on the production of texturized proteins produced by
extrusion cooking.

Raw Materials

Traditionally the most popular raw material for production of textured vegetable proteins in
an extrusion system has been slightly toasted defatted soy flour. This defatted soy flour
usually meets the following characteristics: 50% protein minimum, 3.5% fiber maximum,
1.5% fat maximum and PDI of 60 to 70.

Soy flour with these specifications allows controllable production of textured proteins in
chunk and extended form on single screw extrusion cookers. Other vegetable protein sources
also have been used as raw materials for texturizing, and these include: glandless cottonseed
flour, rape seed or canola concentrates, defatted peanut flour, defatted sesame flour, as well
as soybean grits, flakes, meal, concentrates and isolates.

The use and development of twin screw extrusion cookers in the field of texturized proteins
has increased the raw material specification range to include raw materials that include: PDI
ranges from 20 to 70, fat levels from 0.5 to 6.5%, fiber levels up to 6% and particle sizes up

2
to 8 mesh.

Protein dispersibility index. The Protein Dispersibility Index (PDI) is the percentage of total
protein that is dispersible in water under controlled conditions of extraction. The PDI is now
the preferred measurement with regard to specification of raw ingredients, as it is a more
reliable figure when compared with NSI. Textured soy products have been produced with raw
materials ranging from 20 to 70 PDI.

Fat level. Raw materials have been texturized containing 0.5 to 6.5% fat levels. This higher
range of fat (5.5%) allows mechanically extracted soybean meal to be texturized into meat
extenders and meat analogs. When extruding material with higher fat levels, generally it can
be said that increased shear energy input and higher temperatures are required to maintain a
desired product integrity.

Fiber level. The presence of fiber in extruded soy proteins inhibits or blocks the interaction or
cross- linking of protein molecules necessary for good textural integrity. Changing the
extruder configuration to impart more shearing action into soy proteins containing higher
levels of fiber may achieve a final product with textural properties similar to soy protein with
lower levels of fiber.

Particle size. With regard to successful production using single screw extrusion cookers, the
exact limitations of particle size requirements of defatted soy flour have never been
determined and the range is very wide. The only limitations encountered are as follows: Very
fine flour, below 400 mesh (38 micron), should be limited because of problems in wetting
this very fine powder without lumping; also, very coarse product, over 80 mesh (180 micron),
requires complicated pre- moistening, and sometimes, whole granules are seen in the finished
product. We, therefore, recommend a product grind of 95% through 100 mesh (150 micron),
with a maximum of 50% through 325 mesh (45 micron) for defatted soy flour.

Twin screw extruders can, sometimes use raw material with a particle size range up to 8 mesh
(2360 micron) without affecting the textural properties of the final product.

It is also believed that proper preconditioning has a great effect on the ability of the extrusion
cookers to utilize the larger particle size raw materials for textured protein production.
Preconditioning is a time, temperature and moisture level relationship and controlling
variables allows for the soy flour or grits to be evenly pre-moistened and pre-tempered.

Protein levels. The percentage of protein is normally inversely related to the remaining
constituents of a raw material such as fat and fiber. For example, soybean protein content
goes up as the oil and hulls are removed. Therefore, as a protein level of a raw material
decreases, the textural integrity and water holding capacity decrease and the bulk density of
the final products increases.

Adjustments in pH. Increasing the pH of vegetable proteins before or during the extrusion
process will aid in texturization of the protein. Extreme increases in pH will increase the
solubility and decrease the final product (1) Modifying pH above 8.0 also may result in the
production of harmful lysinoalanines; (2) Lowering the pH has the opposite effect and will
decrease protein solubility, making the protein more difficult to process. Undesirable sour
flavors in the texturized vegetable protein products may be evident, if the pH is adjusted
below 5.0.

3
Modifying the pH to the alkaline side will increase the water absorption. This is generally
done by using calcium hydroxide or the more widely used sodium hydroxide at about 0.1% or
as required.

Calcium chloride. Calcium chloride (CaCl2 ) is very effective in increasing the textural
integrity of extruded vegetable proteins and also aids in smoothing its surface. Dosing levels
for CaCl2 range between 0.5 and 2.0%. With the addition of CaCl2 and small amounts of
sulfur, soybean meal containing 7.0% fiber may be texturized, retorted for one hour at 110°C
and still maintain a strong meat- like texture. Sodium chloride (NaC1) does not appear to add
any benefit to the texture of extruded vegetable proteins. In fact, it tends to weaken the
textural strength.

The addition of sodium alginate will increase chewiness, water- holding capacity and density
of extruded protein products. Sugar will also disrupt the textural development of soy proteins
(3).

Soy lecithin. When added to formulations of vegetable proteins at levels up to 0.4%, lecithin
tends to assist smooth laminar flow in the extruder barrel and die configuration, which allows
the production of increased density soy products. The ability to make dense vegetable protein
products is related to the higher degree of cross-linking that occurs during the extrusion
process.

Sulfur and sulfur containing ingredients. Sulfur is known for its ability to aid in the
cleavage of disulfide bonding, which assists the unraveling of long twisted protein molecules.
This reaction with the protein molecules causes increased expansion, smooth product surface
and adds stability to the extrusion process. These benefits, however, are not without some
undesirable side effects that include off- flavors and aroma.

Sodium metabisulfite, sodium bisulfite, as well as cystine, can be used with effects similar to
those from using sulfur.

The normal dosing levels for sulfur or sulfur derivatives - are in the 0.1 to 0.2% range.
Cystine is used at approximately 0.5 to 1% level.

Color enhancers. When supplementing light colored meats with meat extenders made from
textured vegetable proteins, it is desirable to bleach or lighten the color of the extruded meat
extender. Bleaching age nts such as hydrogen peroxide are often used for this purpose. Dosing
levels for the hydrogen peroxide range from 0.25 to 0.5%. Pigments such as titanium dioxide
are also used at levels between 0.5 and 0.75% to lighten color, but at increased levels will
weaken the textural properties of extruded vegetable proteins.

Extrusion Process
The above raw materials or combinations are generally mixed prior to the extrusion cooker,
except in the cases where small liquid additions can be added directly to the base raw
materials in the extrusion cooker itself.

4
Definition and functions of extrusion of textured proteins. Extrusion cooking has been
defined as “the process by which moistened, expansile, starchy and/or proteinaceous
materials are plasticized in a tube by a combination of moisture, pressure, heat and
mechanical shear. This results in elevated product temperatures within the tube, gelatinization
of starchy components, denaturization of proteins, the stretching or restructuring of tractile
components and the exothermic expansion of the extruder” (4). Extrusion is widely used to
accomplish this restructuring of protein-based foodstuffs to manufacture a variety of textured
convenience foods. When mechanical and thermal energy are applied during the extrusion
process, the macromolecules in the proteinaceous ingredients lose their native, organized
structure and form a continuous, visco-elastic mass. The extruder barrel, screws and die align
the molecules in the direction of flow. This realignment “exposes bonding sites that lead to
crosslinking and a reformed, expandable structure’ that creates the chewy texture in
fabricated foods (5).

In addition to retexturing and restructuring vegetable food proteins, the extrusion cooking
system performs several other important functions: It denatures proteins. Proteins are
effectively denatured during the moist, thermal process of extrusion. Denaturation of protein
“lowers solubility, renders it digestible and destroys the biological activity of enzymes and
toxic proteins” (6). It causes deactivation of residual heat- labile growth inhibitors native to
many vegetable proteins in a raw or partially processed state. Growth inhibitors exert a
deleterious physiological effect on man or animals, as revealed by growth or metabolism
studies. It controls raw or bitter flavors commonly associated with many vegetable food
protein sources. Many of these undesirable flavors are volatile in nature and are eliminated
through the extrusion and decompression of the protein at the extruder die. The use of
preconditioning and atmospheric-venting devices in the design of the extrusion system also
assists in volatilization and removal of off- flavors. It provides a homogeneous, irreversible,
bonded dispersion of all microingredients throughout a protein matrix. This not only insures
uniformity of all ingredients such as dyes throughout the product, but provides a means
whereby minor ingredients can be intimately associated with potential reaction sites
promoting cross-linking or other desirable che mical and physical modifications. And, it
controls the shape and size of the final extruded protein in convenient and transportable
portions for packaging in the retail or institutional marketplace.

Methodology of extruders relating to texturized proteins.

It is convenient to divide the basic components of an extrusion system barrel configuration,
and the die and knife assembly. The design of each of these components is engineered to
accomplish a specific function in the process of texturizing vegetable food proteins. Within
the design features, the operating conditions are adjusted to vary the texture of the finished
product. The effects of each processing step on the product will be addressed specifically.

A typical arrangement of the components of an extrusion system is shown in Figure 1. This

arrangement includes a live bin/feeder, preconditioner and extrusion cooker. The live
bin/feeder provides a means of uniformly metering the raw materials (granular or floury in
nature) into a preconditioner or mixing cylinder and subsequently into the extruder itself.
This flow of raw materials must be uninterrupted and rate-controllable. This component
controls the product rate or throughput of the entire system. The feeder portion also can be
used to initiate the preconditioning of raw materials through the injection of steam to control
raw material temperature and moisture. Not all extruded products require a preconditioning
step. It is particularly useful where the particle size of the raw vegetable protein is large (a
grit or flake). This initial steam preconditioning promotes moisture and heat penetration of

5
the individual particles, resulting in uniform moisture application and elevated raw material
temperature. The raw material then passes through the atmospheric preconditioner or mixing
cylinder where moisture can be uniformly applied in the form of water and/or steam to
achieve a moisture content of 10 to 25%. If steam is applied, it is carefully metered into the
raw material to precondition it at modest temperatures of 65 to lOO0 C. The mixing cylinder is
vented to avoid excess steam and undersirable volatile flavour components found in the raw
vegetable protein. Flavours, coloring agents and other additives may also be introduced at
this phase of the process to insure thorough and continuous mixing of all foodstuffs entering
the extruder barrel. As mentioned earlier, the preconditioning step is also an effective means
of initiating control of growth inhibitors found in many raw vegetable proteins.

Fig. 1 Cutaway view of a high capacity extrusion cooker.

Most preconditioners (Fig. 2) contain one or two mixing! conveying elements which consist
of rotating shafts with radially attached pitched paddles. Atmospheric or pressurized
chambers may be utilized in this preconditioning step. Pressurized preconditioners can
achieve higher discharge temperatures, but have the disadvantages of potential nutrient
destruction and higher operating costs. Without the use of preconditioning, it is difficult to
make good laminar-structured, textured soy protein. Unpreconditioned vegetable proteins
have a strong tendency to expand rather than laminate due to nonuniform moisture
penetration which does not allow uniform alignment of protein molecules. Uniform moisture
penetration of raw ingredients significantly improves the stability of the extruder and final
product quality. This moisture-temperature-time history allows the extrusion of raw materials
having larger particle sizes without sacrificing final product quality.

Fig. 2 Typical preconditioners.

The preconditioner discharges directly into the extruder assembly which consists of the barrel
and screw configuration. Here, the major transformation of the raw or preconditioned
vegetable protein occurs, which ultimately affects the final product texture. Extruders are

6
popularly classified as either being a single or twin screw design. In both designs, the impact
on final product texture is affected by screw and barrel profile, screw speed, processing
conditions such as temperature, moisture, pressure, raw material characteristics and die
selection.

The initial section of the extruder barrel is designed to act as a feeding or metering zone and
simply conveys the raw or preconditioned vegetable protein away from the inlet portion of
the barrel and into the extruder. The product then enters a processing zone where the
amorphous, free- flowing vegetable protein is worked into a colloidal dough. The
compression ratio of the screw profile is increased in this stage to assist in blending water or
steam with the raw material. The temperature of this moist, proteinaceous dough is rapidly
elevated in the last 2-5 seconds of dwell time within the extruder barrel. Most of this heat is
from mechanical energy dissipated through the rotating screw and may be assisted from the
direct injection of steam or from external thermal energy sources. The screw profile can be
altered by the pitch, flight height and angle, and steamlock diameter, which affects the
conveying of this plasticized food material down the screw channel. The net flow patterns of
the product within the screw are quite complicated and difficult to understand and describe.
Retention times of 5-15 seconds, temperatures of 100 to 200°C and moisture levels of 15 to
30% all influence the protein dough quality just behind the die and the final product
expansion. The temperature at this point actually melts the protein into a visco-elastic,
plasticized mass having very high viscosities. This plasticized material is extruded through
the die openings, and expansion occurs as the product is released to ambient pressures. Final
product density has been shown to correlate with temperatures and moistures of extrusion.

Extruded meat analogs. It is possible to produce an extruded meat analog which has a
remarkable similarity to meat in appearance, texture and mouthfeel (7). Although the process
to extrude meat analogs is similar to that used for the production of meat extenders, it often
uses two extruders or a long-barreled, single extruder. Through this configuration, the dwell
time in the extruder is extended, the moisture content increased to 30% and the pressure
maintained below 150 PSIG. The vegetable proteins exposed to this process are stretched,
cooled and formed into uniform layers of parallel orientation and cut into bite-sized portions.
The lower pressures used inboard of the die reduces product expansion and disruptions to the
laminar structure. Due to the dense nature of these analogs, they are able to retain their
integrity and meat- like texture even after retorting them for 2-3 hours.

The extruder barrel. The above discussion is valid for processing texturized protein on both
single and twin screw extruders, however, there are differences between these two extruder
barrel designs. Single screw extruders are the machines that have produced and still produce
the greatest tonnage of textured soy protein products in the world. Over the years of extrusion
development, the process has become more of a science, which has resulted in the ability to
design screws, heads and shearlocks for single screw extruders for specific functions. This
variety or selection of parts includes straight and spiral-ribbed jacketed heads which allow the
control of friction or product shear, internally, while the flow of cooling water or heating
fluids in the external jacket has allowed control of temperatures of the product as it passes
through the various extruder sections. Coupling these heads with the wide range of extruder
screws that are unrestricted in their number of flights, the flight pitches, cut flights, the screw
volumes and the shearlocks between these screws allow for a very versatile machine that has
been successful in producing a wide range of products. In the case of texturized proteins, the
use of a narrow range of raw materials has been the limiting factor in the use of single screw
extruders. Their use requires good uniform raw materials and a maintenance program to keep

7
the extruder barrel parts within tolerances.

The twin screw extruder barrel generally used in the food industry consists of two corotating
intermeshing self-wiping sets of screws. Because the screws intermesh they do have
restrictions in screw design. They are limited by the flight profile and root diameter. The twin
screw extruder has a segmented smooth bore head that is jacketed. The screws are also
segmented so that their arrangement can be modified. The twin screw barrel design does have
the advantage of a more positive feeding characteristic over the single screw barrel design.
The twin screw design, as a result, has been more lenient in raw material specifications.

In either case, single or twin, the combination of preconditioning and the extruder barrel
arrangement results in modification to the raw materials and the result is a texturized protein.
Hydration and heating cause unravelling of the long twisted protein molecules in vegetable
proteins. In the extrusion process, these molecules align themselves along the streamline
flows of the screws and dies. The increase in shear temperature and retention time causes
cross- linking to occur, yielding a textured product that is layered, and the resulting
denaturation or cross- linking can be considered an irreversible endothermic chemical
reaction. The extent of cross- linking seems to be a function of the time, temperature and
moisture history and can be related to changes in apparent viscosity of the extrudate. The
proper exposure to shearing action, as the protein molecules align themselves for cross-
linking during the extrusion process, is important.

The rheology of soy proteins or soy protein doughs will not be discussed in this paper.
Several papers published in the past describe the effects of shear rate, shear stress, adequate
time and temperature history (5,8,9). When a proper screw profile for the extruder is derived,
one can meet final product quality requirements with a much broader range of raw materials.

Die. Dies for texturized vegetable proteins should have smooth streamlined flows that do not
disrupt or cause shearing effects to the already laminated and cross- linked protein molecules
(8). Harper describes in detail the effects of various land lengths and an open area of dies for
extruding texturized vegetable protein.

On-Going Research
Various studies have been performed to confirm and demonstrate the use of a broader range
of raw materials. These studies include the examination of textural properties, and the work
was performed with a Wenger TX-80 twin screw extruder at the Oak B. Smith Extrusion
Research Center in Sabetha, Kansas.

Protein dispersibility index range tests. Commercially available ingredients were used in the
tests. Soy flour (200 W) from Central Soya with a PDI of 70 and soy grits from A.D.M.
(Nutrisoy) with a PDI of 20 and particle size of 20 mesh were mixed together to obtain five
mixes with various PDI levels ranging from 20 to 70. These are shown in Table 1.

8
TABLE 1
Formulas and PDI’s for PDI Range Tests

Test # Formula Calculated PDI Tested PDI

1 100% soy flour 70 79

2 80% soy flour, 20% soy grits 60 68
3 60% soy flour, 40% soy grits 50 58
4 40% soy flour, 60% soy grits 40 58
5 100% soy grits 20 27

All the formulations were run under approximately the same extruder conditions and with the
same final product densities. Samples from the five formulations were tested for water
holding capacity (WHC) and textural integrity (TI). Table 2 shows that for the full range of
PDI levels, there is no significant difference in water holding capacity or textural integrity,
with the exception of Test #2. The increased WHC and textural integrity values in Test #2
can be attributed to the processing conditions of the extruder that yielded a lighter density
product.

TABLE 2
Effects of Raw Material PDI On Final Product Quality

Test # PDI of raw materials Density g/l WHC Textural

integrity
1 (79)100% soy flour 280 2.03 1.19
2 (68)80% soy flour,
20%soygrits 216 3.12 1.31
3 (58)60% soy flour,
40% soy grits 288 1.93 1.03
4 (48)40% soy flour,
60% soy grits 288 2.01 1.22
5 (27)100% soy grits 280 2.29 1.27

During these energy-conscious times we must point out that using raw materials with
decreasing PDI levels also requires additional specific mechanical energy input to maintain
the same quality of texture in the extrudate (Table 3). This is due to the reduction in available
reactive sites for cross- linking in the lower PDI soy protein, thus requiring additional
shearing action to expose those reactive sites.

9
TABLE 3
Required Energy Requirements for Extruding Soy Proteins at Various PDI Levels

Test # Raw Extruder Required WHC Density g/l Textural

material RPM energy integrity
actual PDI KWH/kg
1 79 425 .112 2.03 280 1.19
3 58 425 .147 1.93 288 1.03
4 48 450 .181 2.01 288 1.22

Fat level tests. It is an advantage in certain situations to have the ability to texturize vegetable
proteins having high levels of residual fat. For example, it may be more economically
feasible to use mechanically extracted soy meals which have fat levels between 3 and 6.5% to
produce meat extenders and meat analogues. It also has been documented that texturization of
whole soybeans via extrusion processing improves the feeding quality of soy protein products
(Table 4).

TABLE 4
Effect of Processing on Nutrient Performance

Raw materials Ave TME Kj/gra % Fat Available

GE Kj/gr digestibilityb protein ratio

Texturized whole soybeans .756 98.2 .992

Conventionally processed
full fat soy .754 95.3 .981

Raw soybeans (ground .63 .81.0 .980

through 1/16” screen)
a
Average total metabolizable energy (Ave TME) and gross energy (G.E.) studies performed
by McNab (10).
b
% Fat digestion performed by McNab (10).

Table 5 shows the materials used, and Table 6 shows the processing conditions and product
characteristics of extruded raw materials having various fat levels. In general, increased fat
levels in vegetable proteins require increased shear energy input and higher temperatures to
maintain a desired product integrity.

It can also be noted that the split and whole beans were texturized and they contained 18 to
20% fat. However, these products did not have desirable characteristics after rehydration. The
defatted soy flour with 4.5% fat added into the raw material did produce a good laminar
product with good integrity.

10
TABLE 5

Fat and Fiber Content of Raw Materials

Raw material Total% fat % Fiber
1. Split soybean @ 16% MCWB 18-20 6-7
2. Whole soybeans @ 14% CWB 18-20 6-7
3. Defatted soy flour (w/4.5% add fat} 6 3.0-3.5

TABLE 6
Effect of Raw Materials and Processing Conditions on Rehydrated Product
characteristics

Ingredient % Fat Extruder Mechanical Temp. Extruder Characteristics

energy pressure after
input PSIG rehydration
KWH/kg
1. Split beans 18-20 SX-80 .135 193°C 1500 soft or mushy
2. Whole beans 18-20 SX-110 .191 127°C 1400 soft or mushy
3. Flour w/fat 6 TW-80 .181 154°C 400 firm with
good integrity
& laminated
structure

Fiber levels. Table 7 presents the materials used in these tests. As discussed earlier, fiber
inhibits the cross- linking of protein molecules necessary for good textural integrity.

This is overcome by imparting more shear as the product passes through the extruder barrel.
Results are in Table 8.

TABLE 7
Fiber Content of Commercially Available Soy Proteins

Ingredient Supplier PDI Particle size % Fiber

soy grits ADM Nutrisoy 20 PDI 20 mesh 3.5 max
47% soybean meal Lortscher Agri Service 30 PDI 20 mesh 3.5 max
44% soybean meal Lortscher Agri Service 30 PDI 20 mesh 7.0 max

11
TABLE 8
Effect of Fiber Content on Final Product Quality

Formula Density g/l WHC Textural % Fiber

Integrity
Soy grits 280 2.29 1.27 3.5 max

40% (44% soybean meal) 288 2.03 1.16 5.0 max

60% (47% soybean meal)

Particle size tests. In time past, particle size variations have been important in achieving the
proper textural properties in a textured soy product. Table 9 shows particle size and PDI’s of
three raw ingredients that were extruded yielding the same bulk densities. It appears from
Table 10 that the particle size had no significant effect on the product’s textural properties.
Preconditioning of the raw ingredients may be the highest contributor to the insensitivity to
particle size variation for these test runs.

TABLE 9
Ingredients of Various Particle Size

Soy flour Central Soya 200-W 70 PDI flour

Soy grits ADM Nutrisoy 20 PDI 20 mesh
Soy grits ADM Nutrisoy 20 PDI 8 mesh

TABLE 10

Effect of Raw Material Particle Size on Final Product Quality

Formulation Density g/l WHC Textural

integrity

Soy flour 280 2.03 1.19

Soy grits (20 mesh) 280 2.29 1.27
Soy grits (8 mesh) 280 1.96 1.22

Effects of screw profile and retention time. Table 11 shows the effect of screw profile on
final product characteristics. The table has a section on continuous fighting of the extruder
screws and a section regarding screw fighting that is interrupted or has breaks in the screw
profile.

Three types of products were made of both continuous and interrupted fighting. These are
dense chunks, light chunks and blown product (hamburger style TVP).

12
TABLE 11
Effect of Screw Profile on Final Product Characteristics

Processing w/ Continuous flighting W/Interrupted flighting

conditions Dense Light Blown Dense Light Blown
Chunk Chunk Chunk Chunk
Preconditioning 82.0 82.0 82.0 82.0 82.0 82.0
temperature °C

Extruder RPM 275.0 350.0 300.0 275.0 350.0 300.0

Extrusion 105.5 108.0 113.0 105.5 108.0 113.0

temperature °C

Moisture added 2.1 1.7 1.4 2.1 1.7 1.4

KG/min

S.M. Energy inputa .122 .114 .131 .130 .120 .131

KWH/kg

Product Characteristics
Density g/l 328.0 216.0 192.0 320.0 216.0 192.0
WHC b 1.01 2.24 3.15 1.00 1.57 1.64
a
Specific Mechanical Energy input
b
Water Holding Capacity by Wenger Mfg.

The processing conditions, preconditioning temperature, RPM, extrusion temperature,

moisture added and energy input were kept as close as possible to the same for each case.

Product density was maintained; however, the water holding capacity went down especially
on the light chunk and blown product, as interrupted fighting was used. The test data
demo nstrated that although adequate shear rates are necessary to enhance the cross- linking
effects of protein molecules, over-shearing, once the cross-linking has begun to occur, may
disrupt the layered structure of the protein molecules, resulting in decreased water holding
capacities.

The effect of moisture on the rheology of soy protein doughs is shown in Table 11. Increased
moisture levels increase mobility of the chemically reactive protein molecules, increasing the
potential for reactive sights to come in close proximity, which facilitates cross- linking,
resulting in increased densities of the extrudate.

Table 12 shows a 36% increase of retention time in the extruder barrel. This allows an
approximately equal increase in the maximum capacity of the extruder, while maintaining an
equivalent final product quality. At the same time, the required specific mechanical energy
input per kilogram of extrudate is reduced.

13
TABLE 12
Effect of Retention Time on Maximum Capacity of Extruder

Extruder barrel Max. capacity Specific mechanical

length: dia. ratio to maintain same energy input
quality final product
161/2:1 500 kg/hr. .130 KWH/kg
221/2: 1 800 kg/hr. .090 KWH/kg

Product enhancement concepts. Due to consumer preferences there is a growing demand for
marketing gimmicks. Therefore, product shape and design have become an important issue in
extrusion processing. Recent developments allow increased flexibility in color development
of extruded products.

Dyes or color pigments may be injected with high pressure pumps into the final sections of
the extruder barrel to achieve special effects such as marbling, striping and multiple colored
products. The conical designed final screw and head section that divides the product flow into
separate chambers allows tremendous flexibility of product enhancement, not only with color
schemes, but with the addition of heat-sensitive chemicals for textural development or
reactions.

Post Extrusion Processes

These are processes that are performed on the extrudate prior to the drying and cooling
system. The most predominate process would be particle size reduction. This is normally
referred to as “wet milling,” as the moisture content is high at this point. There are various
devices that can be used for particle size reduction of textured protein and these can include
hammermills, rotating cutting heads and shredders. Wet milling tests were performed on light
and medium density chunks and the results are shown in Table 13.

Grinding devices used must be based on desired final product sizes and intended uses.
Generally speaking the hammermill style device will yield a higher final bulk density as a
result of fines generated. The rotating cutting head and shredder designs will generate less
fines and maintain the approximate bulk density of the original sample. There are adjustments
that can be made in each of these grinding devices to modify the particle size generated.

14
TABLE 13
Comparison of Wet Milling Devices for Textured Soy Protein

Sample Mesh size, % Bulk

4 6 8 10 16 PAN Density
G/L
Run 1.050583 (light density)
Raw chunk -- -- -- -- -- -- 144
Fitz, round ¼”a 0.15 11.09 20.83 11.90 22.81 33.22 227
Fitz, square ¼” .43 20.77 24.67 10.00 22.33 21.8 219
VGR, 5/32” squareb 1.88 9.09 15.16 12.98 30.44 30.44 156
Run 2.50583 (medium density)
Raw chunk -- -- -- -- -- -- 192
Fitz, round ¼”a 0.06 3.99 11.83 8.56 22.00 53.55 284
Fitz, square ¼” 0.39 7.33 14.06 8.96 20.74 48.79 258
VGR, 5/32” squareb 1.07 2.73 8.86 9.65 32.73 44.96 211
Urschel, 120 screenc 0.91 20.18 28.14 13.38 22.74 14.64 201
a
Fitzmill Hammermill
b
Wenger Shredder
c
Urschel Comitrol

Drying and Cooling

In the extrusion processing of textured proteins the moisture content is elevated, and as a
result, the cooked product must have the moisture reduced for safe storage prior to its use.
Proper design and size selection of the dryer cooler is important for making good quality
products, as well as overall cost of production, because this production step is generally the
most costly.

There are many dryer/cooler designs, but for the purpose of discussion, we will review the
tray-type dryer/coole r (Fig. 3), because it is the most popular style for textured soy protein
production. Tray-type dryer/coolers can be built with various lengths, widths, number of
passes, materials, conveyor styles, air flows, sanitary considerations and heating systems.

Fig. 3 Dryer/cooler.

15
The ability to construct dryer/cooler with various lengths, widths and number of passes
allows proper sizing for the extruder output, as well as offering flexibility in equipment
placement in existing or new facilities. The two pass design not only saves floor space, but
allows the product to be turned over in the dryer, improving evenness of drying. Using
varispeed motors on the dryer passes allows for changes in the product bed depth, and thus
retention time of the product in the dryer.

Dryer/cooler cons truction materials include mild steel, stainless steel and combination of
both. In human food production the dryer/coolers are normally equipped with stainless steel,
at least where the product comes in contact with the dryer, or are all of stainless steel
construction. Product-carrying conveyors are perforated mild or stainless steel trays that can
be coated with teflon or silverstone. Also available are wire mesh screens that are carried on
special supports. The wire mesh design has greater open area for even air flow and allows for
smaller particles to be dried and retained on the conveyor.

Air flow is generally up through the product with the heating elements mounted in a single
side plenum chamber above air recirculation fans. These recirculation fans create the static
pressure below the dryer trays resulting in evenness of drying across the product bed. A
percentage of air is exhausted to control the humidity in the dryer, while the balance is
recirculated, which permits maximum efficiency in energy consumption. The heating
elements mounted in the side plenum are generally steam coils or gas burners. This is usually
based on the location of the plant and the cost of the fuel to be used. Steam coils are generally
the most popula r. Full insulation on all sides of the dryer coupled with the heating systems
allow for safe operating temperatures up to 350° (175°).

Sanitary considerations include doors wherever possible for cleaning access and
maintenance. Fines recovery systems remove fines from the bottom of the dryer/cooler and
deposit them in a cross auger that removes them to the side of the dryer/cooler. Dryers are
commonly equipped with integral coolers or extensions of the bottom dryer pass. Internal
airlocks are used to reduce the amount of air leakage between these two zones. Air flow is
generally down through the product and a percentage of this exhausted air is used as makeup
air for the dryer, yielding a more efficient dryer, reducing energy consumption.

Final Produc ts
The products discussed above include texturized proteins in the meat extender form, as well
as chunk style products. The meat extenders, when ground to a particle size simulating
hamburger, are used at a 25 to 30% level in dishes such as pizza, hamburgers, meatloaf, chili,
tacos, enchiladas, sloppy joes and meat sauces. Chunk style products are effectively used in
soups, stews, meat pies, dry soup mixes and oriental dishes.

Recently, a new line of natural flavors has come to our attention, and flavor response in meat
extender products is remarkable. We have tested beef, ham, chicken, lobster, clam and shrimp
flavors and have been quite impressed with their utility, flavor survival through the extrusion
cooking process and lack of residual soya flavor in the end product. When blended with a
system for producing soups, chowders or gravies, for example, their similarity to meat has
proven to be excellent.

16
Particularly interesting is the idea of using natural flavors as opposed to synthetic flavors.
Because the alcohol, fat or water bases of most synthetic flavors are quite volatile, most of
the synthetic flavor’s effectiveness is lost at the high temperatures necessary for structuring
and moisture removal. Additionally, synthetic flavors are often leached out of the product and
are often dispersed throughout the liquid phase during the rehydration of textured soy protein
in the canning operation or in the making of soups or stews. Conversely, the natural flavor
agents we have utilized seem to be enhanced by cooking, drying and rehydration steps and
react in much the same manner as natural meats, in that thermal processing is required to
bring out their flavor. When included in a flavor system such as soup stock or meat gravy, the
product may be frozen or canned to provide stability. Additionally, systems may be designed
for convenience type products which could include the textured vegetable protein extender in
a gravy or soup stock to be finished by the end user with fresh vegetables, noodles or other
constituents to develop nutritious, inexpensive and convenient complete meals, entrees or
soups.

Discussions above on textural integrity and water- holding capacity help food scientists
determine how texturized proteins will function in the food systems utilizing these products.
The following test procedures are the ones used in the studies.

Procedure for determining water holding capacity.

(Equipment Needed: Beaker, sieve screen and scales)

Weigh out small amount of sample and record weight. Soak in room temperature water for 20
minutes. Let drain on screen for 5 minutes. Reweigh sample. (Rehydrated wt. - Original
wt.)/Original wt. = Water absorption.

Test for measuring integrity of textured vegetable proteins.

(Equipment needed U.S. 20 mesh screen; gram scale; 600 ml. beaker; aluminium foil;
autoclave, capable of holding 15 psi; food grinder with 1.8 in. holes in plate; and water bath.
Ingredients needed: At least 50 g of sample and tap water.)

Begin test by soaking at least 50 g of sample in a 600 ml beaker with sufficient water to fully
cover all of sample and allow for absorption. Let hydrate one hour, stirring occasionally.
After hydration, cover beaker with aluminium foil and place in a preheated retort. Pressure
cook for 30 mm under pressure at 15 psi (250°). Cool sample to 70-80° by putting beaker in
running tap-water bath. Do not pour off excess water. Grind hydrated sample in food grinder
fitted with plate with 1/8 inch diam. holes. If water comes out of grinder, keep with sample.
Weigh out 50 g of the hydrated, cooked and ground sample and spread evenly on a U.S 20
mesh screen. Spray rinse the sample on the screen for one minute with tap water at ambient
temperature under 12 psi. Spray nozzle should be approximately 9 in. from the screen and
slowly moved back and forth and rotated so that the spray water contacts the entire screen
surface during the rinsing operation. Flow of water should be about 3 gal in 1 mm. Shake off
excess water. Drain sample on the screen for 4 mm. During the draining period, blot the
bottom of the screen with paper towels. Weigh the residue. (This is most easily done by
having the screen tared out on the scales before the test begins. Then you can weigh out the
residue with the tared screen). To calculate integrity index, divide the weight of residue by
the weight of hydrated sample.

17
Example:
If 50 g of sample were spray rinsed, and 55 g of sample was left on the screen after rinsing,
55/50 - 1. 1 integrity index.

Discussion

Many tests have been performed with a wide variety of raw materials. Soybean-based raw
materials such as defatted soy flour, soy concentrates, soy isolates and soy grits with various
particle sizes, fat levels and fiber levels can all be used to produce texturized proteins. The
new developments in processing techniques will allow work to continue with raw materials
such as glandless cottonseed flour, rapeseed proteins or canola concentrates, defatted peanut
flour and defatted sesame flour in order to allow a higher utilization of these protein sources.
Initial work with these raw materials on single screw extruders resulted in texturized
products, however, the marketability of these products are questionable. The use of the twin
screw extrusion cookers should revitalize the interest in these raw materials.

Texturized proteins have been successfully marketed and produced throughout the world as a
result of food scientists’ ability to apply the extrusion process to produce products that fit into
the individual society’s normal eating habits. Interactions of equipment manufacturers, end
users and food scientists will result in further improvements and new developments in the art
of producing and utilizing texturized proteins.

Acknowledgement
J. McNab of the Agricultural Research Council, Roslin, Scotland, performed metabolizable
energy and fat digestion studies used to prepare Table 4.

References

1. Simonsky, R.W., and D.W. Stanley, Can. Inst. Food Sci. Technol. J. 15:294 (1982).

2. DeGroot, A.P., P. Slump, V.J. Feron and L. Van Beek, J. Nutr. 106:1527 (1976).

3. Boison, G., M.V. Taranto and M. Cheryan, J. Food Technol. 18:719 (1983).

4. Smith, O.B., “Textured Vegetable Proteins” presented at the World Soybean Research
Conference, University of Illinois, Aug. 3-8, 1975.

5. Harper, Judson, Food Technol. 40:70 (1986).

6. Altschul, A.M., Basic Books, New York (1965).

7. Smith, O.B., “Products of Textured Soy Proteins,” 1st Latin America Soy Protein
Conference, Mexico City (Nov. 1975).

8. Holay, S.H., and J.M. Harper, J. Food Sci. 47:1869 (1982).

9. Hagaiv, R.C., S.R. Dahl and R. Villota, J. Food Sci. 51:367 (1986).

18