You are on page 1of 9

Food Research International 43 (2010) 650–658

Contents lists available at ScienceDirect

Food Research International


journal homepage: www.elsevier.com/locate/foodres

Effects of extrusion conditions on system parameters and physical properties of


a chickpea flour-based snack
X. Meng *, D. Threinen, M. Hansen, D. Driedger
Alberta Agriculture and Rural Development, 301 Horticulture Station Road East, Brooks, Alberta, Canada T1R 1E6

a r t i c l e i n f o a b s t r a c t

Article history: Response surface methodology (RSM) was used to study the effects of feed moisture content (16–18%),
Received 30 January 2009 screw speed (250–320 rpm), and barrel temperature (150–170 °C) on extruder system parameters (prod-
Accepted 23 July 2009 uct temperature, die pressure, motor torque, specific mechanical energy, SME) and physical properties
(expansion, bulk density, hardness) of a chickpea flour-based snack. Second-order polynomials were used
to model the extruder responses and product properties as a function of process variables. Product tem-
Keywords: perature and die pressure were affected by all three process variables, while motor torque and SME were
Chickpeas
only influenced by screw speed and barrel temperature. All three variables affected product responses
Extrusion
System parameters
significantly. Desirable products, characterized by high expansion ratio and low bulk density and hard-
Physical properties ness, were obtained at low feed moisture, high screw speed and medium to high barrel temperature. It
Response surface methodology was demonstrated that chickpeas can be used to produce nutritious snacks with desirable expansion
and texture properties.
Crown Copyright Ó 2009 Published by Elsevier Ltd. All rights reserved.

1. Introduction Chickpeas have a high (40–50%) starch content (Huang et al.,


2007; Wang & Daun, 2004), which may favor an extrusion process
Snack foods have become a significant part of the diet of many to produce directly expanded snack foods. Using chickpeas instead
individuals, particularly children, and can influence overall nutri- of cereals in extruded snacks would improve the nutrient density
tion (Shukla, 1994). The most widely consumed extruded snacks of these foods because chickpeas have more protein. Hence, devel-
are made primarily with cereals/grains due to their good expansion opment of chickpea-based snacks could provide an attractive out-
characteristics; however, they tend to be low in protein and many let for chickpea utilization. However, processing of chickpeas into
other nutrients. Hence, there is an increasing consumer demand extruded snacks is limited. Few reports (Bhattacharya & Prakash,
for more nutritious snacks (Agriculture and Agri-Food Canada, 1994; Shirani & Ganesharanee, 2009) indicated that incorporation
2008). of chickpeas into rice flour decreased product expansion. The
Chickpeas are the third most important legume in the world extrudates exhibited increased density and breaking strength, indi-
after dry beans and dry peas (Singh, Subrahmanyam, & Kumar, cating poor textural effects of chickpea flour inclusion. The above
1991). Chickpeas have one of the highest nutritional compositions studies dealt with relatively simple raw material compositions. A
of any dry edible pulse and do not contain any specific major anti- more complex mixture, involving starch and protein fortification,
nutritional factors (Chavan & Salunkhe, 1986). Research has shown may promote expansion and nutritional quality of a chickpea-
that chickpeas are an excellent source of protein (24.4%), dietary fi- based snack.
ber (9.0%), complex carbohydrates (60.0%), folate, and trace miner- Extrusion cooking technology is the most used technique for the
als such as iron, molybdenum, manganese (Poltronieri, Areas, & production of snack foods (White, 1994). Extrusion is a powerful
Colli, 2000; Wang & Daun, 2004). In addition, chickpeas have been processing operation, which utilizes high temperature, pressure,
reported to reduce the levels of cholesterol and blood glucose and shear force to produce highly expanded, low-density products
(Singh & Singh, 1992). Hence, Chickpeas are increasingly being with unique texture properties (i.e. crispiness, crunchiness). Extru-
used in healthy diets to promote general well-being and to reduce sion of snack foods demands close control of many variables such
the risk of cardiovascular diseases and diabetes (Pulse Canada, as feed moisture, feed composition, feed particle size, feed rate,
2008). barrel temperature, screw speed, screw configuration, and die
geometry. These material and process variables determine the ex-
tent of macromolecular transformations during extrusion, which in
* Corresponding author. Tel.: +1 403 362 1365; fax: +1 403 362 1326. turn influence the rheological properties of the food melt in the ex-
E-mail address: xiangfeng.meng@gov.ab.ca (X. Meng). truder and, consequently, the product characteristics of extrudates.

0963-9969/$ - see front matter Crown Copyright Ó 2009 Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.foodres.2009.07.016
X. Meng et al. / Food Research International 43 (2010) 650–658 651

Physical characteristics such as expansion, density, and hardness plate had two circular dies, each with one hole. The diameter of
are important parameters to evaluate the consumer acceptability each hole was 2.5 mm.
of the final product (Patil, Berrios, Tang, & Swanson, 2007). The extruder was powered by a 36.6 kW motor with speeds var-
The effects of ingredient properties and processing conditions iable from 0 to 350 rpm. The screw configuration is shown in Fig. 1.
on final product quality are also reflected by their influence on pro- The extruder was equipped with a torque indicator, which showed
cess responses or extruder system parameters such as motor tor- percent of torque in proportion to the current drawn by the drive
que, die pressure, product temperature, and specific mechanical motor. Die pressure was measured using a Dynisco pressure trans-
energy (SME). As they are a result of different combinations of ducer (PT415D-3M-6, Dynisco, Sharon, MA, USA).
extrusion conditions such as feed moisture, screw speed, and bar- Raw materials were metered into the extruder with a twin-
rel temperature, system parameters can be used to describe or screw volumetric feeder (T-35, K-Tron Corp., Pitman, NJ, USA).
compare extrusion process under different operating conditions The feed rate was kept constant at 56.8 kg/h for all experiments.
(Moraru & Kokini, 2003). It has been shown that SME correlates The moisture content of the feed was varied by injecting water
well with extrudate properties such as expansion, density and tex- (approximately 50 °C) into the extruder with a triple action piston
ture characteristics (Altan, McCarthy, & Maskan, 2008; Dogan & pump (P33, Bran and Lubbe, Delavan, WI, USA). A variable speed
Karwe, 2003; Ilo, Tomschik, Berghofer, & Mundigler, 1996; Meuser die face cutter with one blade was used to cut the extrudates.
& Van Lengerich, 1992; Onwulata, Konstance, Smith, & Holsinger,
2001a). The objective of this study was to investigate the effects 2.3. Sample collection and determination of process responses
of extrusion conditions, feed moisture content, screw speed, and
barrel temperature, on the extrusion system parameters and phys- Samples were collected for 5 min after the extruder system
ical properties of a chickpea flour-based extruded snack using re- parameters (product temperature, die pressure, and motor torque)
sponse surface methodology. reached a steady-state. Product temperature, die pressure, and mo-
tor torque shown on the extruder control panel were recorded
twice, at the start and the end of product collection. Average values
2. Materials and methods were reported. Extrudate samples were cooled to room tempera-
ture and then packed in sealed plastic bags until measurements
2.1. Materials were taken.
Specific mechanical energy (W h/kg) was calculated from rated
Chickpea flour (desi type) was obtained from Diefenbaker Seed screw speed (350 rpm), motor power rating (36.6 kW), actual
Processors (Elbow, Saskatchewan, CA). The chickpea flour had a screw speed, % motor torque, and mass flow rate (kg/h) as the fol-
protein, fat, and fiber content of 22.0%, 6.7%, and 11.0% (wet basis), lowing formula (Pansawat et al., 2008):
respectively. HilmarÒ 8200 heat stable/gelling whey protein con-
actual screw speed ðrpmÞ % motor toque
centrate (82% protein) was obtained from NutraBlend Foods (Cam- SME ¼ 
bridge, Ontario, CA). Whey protein is a high-quality, complete
rated screw speed ðrpmÞ 100
protein that delivers all the essential amino acids needed by the motor power rating ðkWÞ

human body (Burrington, 2004; Davis, 2004; Onwulata, Smith, mass flow rate ðkg=hÞ
Konstance, & Holsinger, 2001b). Incorporation of whey protein in
the formulation would enhance the nutritional profile of the prod-
2.4. Experimental design
uct. Potato starch and lecithin were obtained from Blends (Edmon-
ton, Alberta, CA). Baking powder was purchased from a retail store.
Response surface methodology was used to investigate the ef-
Moisture content of chickpea flour, whey protein concentrate, and
fects of extrusion conditions on the process and product responses
potato starch was measured prior to the extrusion experiments so
of the chickpea flour-based snack. Results from preliminary trials
as to adjust the feed moisture content according to the experimen-
were used to select suitable extruder operating window. The inde-
tal design. The feed blend was prepared according to the formula-
pendent variables considered for this study were feed moisture
tion given in Table 1.
content (16.0–18.0%, wet basis), screw speed (250–320 rpm), and
barrel temperature of the last zone (150–170 °C). A Rotatable, Cen-
tral Composite Design (Myers, 1971) was employed to determine
2.2. Extruder and processing conditions
the extrusion conditions (Table 2). Experiments were randomized
in order to minimize the systematic bias in the observed responses
Extrusion experiments were performed on a co-rotating and
due to extraneous factors.
intermeshing twin-screw extruder (ZSK-57, Werner & Pfleiderer,
Ramsey, NJ, USA). The barrel diameter and its length to diameter
ratio (L/D) were 57 mm and 24:1, respectively. The extruder had 2.5. Determination of product responses
3 barrel zones. Temperatures of the first two zones were main-
tained at 50 and 90 °C, respectively, throughout the experiments, 2.5.1. Expansion ratio
while the temperature at the last zone (compression and die sec- To determine the expansion ratio (ER), the cross-sectional
tion) was varied according to the experimental design. The die diameter of the extrudates was measured with a Vernier digital
caliper. The expansion ratio was calculated as the cross-sectional
diameter of the extrudate divided by the diameter of the die open-
Table 1 ing (Ding, Ainsworth, Tucker, & Marson, 2005; Onwulata et al.,
Feed formulation used in the extrusion experiments. 2001b). The ER values were obtained from 15 random samples
Ingredient % (w/w, wet basis) for each extrusion condition.
Chickpea flour 60.0
Potato starch 23.0 2.5.2. Bulk density
Whey protein concentrate 15.0 Bulk density (g/cm3) of extrudates was determined by a volu-
Soy lecithin 1.0
metric displacement procedure as described by Hwang and Hayak-
Baking powder 1.0
awa (1980) and Patil et al. (2007). Glass beads, with a diameter of
652 X. Meng et al. / Food Research International 43 (2010) 650–658

Fig. 1. Screw configuration used for chickpea snack extrusion.

Table 2
Experimental design for the extrusion experiments.

Run no. Coded levels Actual levels


X1 X2 X3 Feed moisture content (%) Screw speed (rpm) Barrel temperature (°C)
1 1 1 1 16.0 250 150
2 1 1 1 16.0 250 170
3 1 1 1 16.0 320 150
4 1 1 1 16.0 320 170
5 1 1 1 18.0 250 150
6 1 1 1 18.0 250 170
7 1 1 1 18.0 320 150
8 1 1 1 18.0 320 170
9 1.682 0 0 15.3 285 160
10 1.682 0 0 18.7 285 160
11 0 1.682 0 17.0 226 160
12 0 1.682 0 17.0 344 160
13 0 0 1.682 17.0 285 143
14 0 0 1.682 17.0 285 177
15 0 0 0 17.0 285 160
16 0 0 0 17.0 285 160
17 0 0 0 17.0 285 160
18 0 0 0 17.0 285 160
19 0 0 0 17.0 285 160
20 0 0 0 17.0 285 160

1 mm, were used as a displacement medium. Five measurements hardness in N by measuring the maximum force required to break
were made for each extrusion condition. Twelve pieces of extru- the extruded samples. A 50 kg load cell was used. The test speed
dates were pre-cut into short segments of about 1.5 cm long and was 2 mm/s and the distance between the two supports was
the total weight (Wex) was recorded. Glass beads were first poured 48 mm. A force–time curve was recorded and analyzed by a Tex-
into a 250-mL (Vc) graduate cylinder to form a 0.5–1.0 cm thick ture Exponent 32 software program (version 4.0). Ten randomly
layer. Three pieces of extrudates were placed on top of the glass collected samples were measured for each extrusion condition.
beads and enough glass beads were poured in to cover the extru-
dates. The contents were gently tapped 12 times with a stainless 2.6. Statistical analysis
steel spoon. Three more pieces were placed in the cylinder, covered
with more glass beads and again tapped 12 times. This procedure Process responses (product temperature, die pressure, motor
was repeated until all 12 extrudate pieces were used. Again, glass torque, SME) and product responses (expansion ratio, bulk density,
beads were poured into the cylinder until they reached the 250 mL hardness) obtained as a result of the proposed experimental design
mark. Excess beads were removed with a spatula. The weight of were subjected to regression analysis in order to assess the effects
the glass bead displaced (Wgbd, weight of glass bead and extrudates of feed moisture content, screw speed, and barrel temperature. A
in the cylinder less weight of extrudates) was recorded. The weight second-order polynomial regression model
of glass beads required to fill the cylinder (Wgb) was determined by
X
3 X
3 X
3
following the same procedure as for the samples, except that yi ¼ b0 þ bi X i þ bij X i X j
extrudates were not used. Extrudate density was calculated as i¼1 i¼1 j¼i

W ex ðgÞ for the dependent variables was established to fit the experimental


qb ¼ 
W ðgÞ

Vc  1  WgbdðgÞ
ðcm3 Þ data for each response and was computed by using SAS (version 9.1)
gb
statistical package (SAS, 1997), where Xi, XiXi, and XiXj are linear,
quadratic, and interaction effect of the input variables which influ-
2.5.3. Texture ence the response y, respectively, and b0, bi, and bij are the regres-
The textural property of extrudate rods was measured using a sion coefficients to be determined. The response surface plots for
TA.XTPlus Texture Analyzer (Texture Technologies Corp., Scarsdale, these models were plotted as a function of two variables, while
NY, USA). A 5-blade Kramer Shear Cell was used to determine the keeping the third variable constant at its intermediate value.
X. Meng et al. / Food Research International 43 (2010) 650–658 653

Canonical analysis of the predicted model was performed to locate pressure is positively related to melt viscosity at a constant feed
extrusion conditions where the snack would have a minimum den- rate (Jin et al., 1994).
sity and hardness and maximum expansion ratio. The interaction effect was also reflected on the surface plot
(Fig. 3B). The influence of screw speed on die pressure was most
noticeable at low barrel temperatures; at high barrel temperatures
3. Results and discussion
the effect was minor. Die pressure ranged from 4.83 to
6.41 Pa  106 (Table 3). High die pressure was obtained at low feed
Effects of extrusion conditions on process and product re-
moisture content, screw speed, and barrel temperature combina-
sponses are shown in Table 3. Extrudates of different physical
tion (Fig. 3). Similar results have been reported (Altan et al.,
properties were obtained at different processing conditions (Table
2008; Pan, Zhang, & Jane, 1998; Pansawat et al., 2008).
3). Results of the regression analysis are shown in Table 4. The
regression models for product temperature, die pressure, expan-
3.1.3. Motor torque and SME
sion ratio, bulk density, and hardness were highly significant
Similar regression results were obtained for motor torque and
(P < 0.001), with a high coefficient of determination (R2 = 0.97,
SME (Table 4). Screw speed and barrel temperature had significant
0.97, 0.87, 0.99, and 0.91, respectively). Although relatively lower
linear effects on motor torque (P < 0.01) and SME (P < 0.05). As ex-
R2 was noted for motor torque (0.79) and SME (0.78), their regres-
pected, torque decreased with increasing screw speed and barrel
sion models were significant (P < 0.05). Moreover, none of the
temperature, while SME increased with increasing screw speed
models showed significant lack of fit (P > 0.05), indicating that all
and decreasing barrel temperature (Table 4, Figs. 4 and 5). How-
the second-order polynomial models correlated well with the mea-
ever, feed moisture had no effect on SME and torque. The measured
sured data and were statistically significant (P < 0.05).
torque values ranged from 19 to 24% and calculated SME from 87.9
to 115.8 Wh/kg (Table 3). Low torque was observed at high screw
3.1. Process responses speed and barrel temperature and high SME at high screw speed
and low temperature on the surface plots (Figs. 4 and 5).
3.1.1. Product temperature Any variable affecting the viscosity of the food melt in the ex-
Product temperature values ranged between 164.1 and 182.4 °C truder would correspondingly affect motor torque and SME (Akdo-
(Table 3), which were higher than the barrel temperatures set at gan, 1996). The decreases in motor torque and SME with increasing
the last zone of the extruder (Table 3). This was expected, because barrel temperature at the die could be attributed to the changes in
heat was generated during extrusion through viscous dissipation of melt viscosity. Increasing barrel temperature and screw speed
the mechanical energy, which caused an increase of product would both cause a decline in the melt viscosity and consequently,
temperature. a decrease in SME and torque. The torque required to turn the ex-
The predicted response model (Table 4) indicated that linear ef- truder screw is also related to degree of fill in the extruder barrel
fects of all three variables and quadratic effects of screw speed and (Jin et al., 1994). At a constant feed rate, an increase in screw speed
barrel temperature were the primary determining factors for prod- decreased the length of filled flights. This resulted in a decrease in
uct temperature, and there was no significant interaction between the load on the screw shaft motor. Thus, the extruder torque was
the independent variables. Among the three variables, feed mois- reduced. However, as shown in the surface plot (Fig. 5), SME in-
ture content had the least effect on product temperature. The re- creased proportional to the screw speed, although a decrease in
sponse surface plots (Fig. 2) showed that increased screw speed SME was also expected because melt viscosity would decrease
and barrel temperature led to an increase in product temperature. with increasing screw speed. This result indicated that the effect
However, increased feed moisture content caused a decrease in of screw speed dominated the effect of melt viscosity, as reported
product temperature. High product temperature was observed at in other studies (Akdogan, 1996; Bhattacharya, 1997; Jin et al.,
high screw speed and barrel temperature, and low feed moisture 1994). On the contrary, increasing feed moisture content and
content. hence lowering the melt viscosity did not result in a lower SME
The dependence of shear or mechanical energy on screw speed and torque. This could be due to the relatively narrow moisture
is well known. A higher screw speed generates a greater amount of range (16.0–18.0%) used in the study.
mechanical energy or frictional heat, which leads to an increase of The amount of mechanical energy delivered to the extruded
product temperature. As shown in Fig. 2, increasing the screw material plays an important role in starch conversion. Higher
speed from 226 to 344 rpm increased the product temperature SME usually results in greater degree of starch gelatinization and
approximately 15 °C. Della Valle, Tayeb, and Melcion (1987) and extrudate expansion. Hence, increased SME is desired for expand-
Jin, Hsieh, and Huff (1994) reported similar temperature increases ing products.
during extrusion of corn meal or corn starch based materials. Prod- The SME values were lower than those observed for extrusion
uct temperature or melt temperature plays an important role in cooking of cereals (Ilo et al., 1996; Jin et al., 1994; Singh & Smith,
changing the rheological properties of the extruded melts, which 1997), most likely due to the feed composition. In our study, the
in turn affect the degree of expansion. feed blend had a fat, protein, and starch content of approximately
Inversely, water acts as a plasticizer in the extruder, and 5%, 25%, and 47–53% (calculated values), respectively, while cereal/
increasing feed moisture reduces melt viscosity and the mechani- grain-based direct expanded snacks usually contain much less pro-
cal energy dissipation (Ilo et al., 1996). Thus, product temperature tein and more starch. Onwulata and Konstance (2006) suggested
was lower. that a higher protein content of the feed resulted in a lower melt
viscosity and reduced extrudate expansion. They found a large
3.1.2. Die pressure reduction in paste viscosity, which represented the melt viscosity
The regression analysis results (Table 4) indicated that feed in the extruder, when a whey protein concentrate was added to
moisture content, screw speed, and barrel temperature had nega- a corn meal. Moreover, Della Valle, Colonna, and Patria (1996)
tive linear effects on die pressure (P < 0.001) followed by the inter- found that viscous behavior highly depended on starch content,
action effect of screw speed and barrel temperature (P < 0.05). The and melt viscosity decreased with decreasing starch content.
response surface plot (Fig. 3A) indicated that die pressure de- Therefore, the high protein and low starch content of the feed
creased with increasing feed moisture content. This may be attrib- blend may result in a low melt viscosity, which could account for
uted to the reduced melt viscosity at high moisture levels, as die the low SME observed in our study. This is supported by Onwulata
654 X. Meng et al. / Food Research International 43 (2010) 650–658

Table 3
Effects of extrusion conditions on process and product responses.

Extrusion conditions Process responses Product responses


Feed moisture Screw speed Barrel Product Die pressure Motor SME Expansion Bulk density Hardness
content (%) (rpm) temperature temperature (°C)b (Pa  106)c torque (%)d (W h/kg)e ratiof (g/cm3)g (N)h
(°C)a
16.0 250 150 164.2 6.41 24 102.8 3.43 ± 0.26 0.217 ± 0.004 289 ± 28
16.0 250 170 173.2 5.65 22 94.3 3.80 ± 0.28 0.170 ± 0.004 256 ± 32
16.0 320 150 174.9 5.45 20 109.7 4.42 ± 0.34 0.161 ± 0.003 240 ± 27
16.0 320 170 182.4 5.03 19 104.2 4.99 ± 0.42 0.130 ± 0.003 198 ± 21
18.0 250 150 164.1 5.86 22 92.0 3.06 ± 0.20 0.275 ± 0.012 550 ± 48
18.0 250 170 171.5 4.83 21 87.9 3.19 ± 0.26 0.233 ± 0.005 399 ± 37
18.0 320 150 172.5 4.83 20 107.1 3.25 ± 0.28 0.223 ± 0.007 400 ± 54
18.0 320 170 179.4 4.83 20 107.1 3.28 ± 0.28 0.167 ± 0.006 299 ± 35
15.3 285 160 178.1 6.21 23 113.3 4.45 ± 0.32 0.140 ± 0.005 184 ± 24
18.7 285 160 172.1 5.03 22 104.1 3.25 ± 0.28 0.240 ± 0.008 398 ± 41
17.0 226 160 166.6 6.07 24 91.9 3.14 ± 0.24 0.249 ± 0.003 344 ± 23
17.0 344 160 179.8 4.83 19 110.7 4.64 ± 0.40 0.151 ± 0.003 201 ± 16
17.0 285 143 164.5 6.41 24 115.8 3.06 ± 0.20 0.228 ± 0.007 339 ± 17
17.0 285 177 179.2 4.83 19 91.7 4.21 ± 0.36 0.162 ± 0.004 243 ± 34
17.0 285 160 173.7 5.38 20 96.5 4.19 ± 0.30 0.179 ± 0.003 272 ± 25
17.0 285 160 175.6 5.65 22 106.2 4.27 ± 0.28 0.168 ± 0.002 244 ± 41
17.0 285 160 174.8 5.45 22 106.2 4.38 ± 0.28 0.170 ± 0.006 270 ± 28
17.0 285 160 173.9 5.52 22 106.2 4.05 ± 0.30 0.178 ± 0.003 282 ± 28
17.0 285 160 174.3 5.38 22 106.2 3.98 ± 0.36 0.176 ± 0.003 255 ± 32
17.0 285 160 177.4 5.52 20 96.5 3.93 ± 0.34 0.177 ± 0.002 258 ± 32
a
Barrel temperature of zone 3.
b,c,d
Means from two readings.
e
SME: specific mechanical energy.
f
Means from 15 measurements.
g
Means from 5 measurements.
h
Maximum force to break the extrudate, means from 10 measurements.

Table 4
Results of regression analysis (calculated on coded levels where all independent variables are in the range ±1).

Parameters Regression parameter coefficients


Process responses Product responses
Product temperature Die pressure Motor torque SME Expansion Bulk density Hardness
(°C) (Pa  106) (%) (W h/kg) ratio (g/cm3) (N)
Intercept 174.92 1.701 21.35 103.10 3.97 0.1746 262
Feed moisture content (X1) 1.27** 0.068*** 0.27 2.37 0.43*** 0.0285*** 75***
Screw speed (X2) 4.28*** 0.075*** 1.35*** 6.06** 0.36*** 0.0278*** 44***
Barrel temperature (X3) 4.07*** 0.076*** 0.91** 4.29* 0.22** 0.0210*** 36**
Feed moisture  feed moisture ðX 21 Þ 0.06 0.004 0.23 1.17 0.12 0.0055** 19*
Feed moisture  screw speed (X1X2) 0.45 0.032 0.50 2.19 0.24** 0.0030 18
Feed moisture  barrel temperature (X1X3) 0.28 0.022 0.25 1.24 0.10 0.0025 22
Screw speed  screw speed ðX 22 Þ 0.73* 0.007 0.12 1.44 0.11 0.0090*** 12
Screw speed  barrel temperature (X2X3) 0.25 0.051* 0.25 0.89 0.01 0.0001 5
Barrel temperature  barrel temperature ðX 23 Þ 1.20** 0.002 0.12 0.58 0.20** 0.0074*** 19*
R2 0.97 0.97 0.79 0.78 0.87 0.99 0.91
P-value for model lack of fit 0.83 0.40 0.46 0.47 0.16 0.28 0.07
*
Significant at P < 0.05.
**
Significant at P < 0.01.
***
Significant at P < 0.001.

et al. (2001a), who observed a significant reduction in SME input of ER at low moisture levels, whereas the effect was minor at high
into the extrudates when the protein content of a corn meal was moisture levels.
increased by incorporation of dairy proteins. The significant quadratic effect of barrel temperature was also
reflected on the surface plot (Fig. 6B) with a curved surface. Expan-
3.2. Product responses sion ratio of extrudates increased with barrel temperature before it
reached a critical level, approximately 168 °C, after which ER de-
3.2.1. Expansion ratio clined. Similar findings have also been reported by others (Dogan
The regression analysis results (Table 4) showed that feed mois- & Karwe, 2003; Ilo et al., 1996). They suggested the existence of
ture content had a negative linear effect (P < 0.001), while screw temperature plateau for expansion, between 150 and 170 °C,
speed and barrel temperature had positive linear effects depending on the type of food material. This phenomenon may
(P < 0.01) on ER followed by an interaction effect (P < 0.01) of mois- be caused by excessive structure breakdown and starch degrada-
ture and screw speed and a quadratic effect (P < 0.01) of barrel tion under high temperature, which weakened the extrudate struc-
temperature. It was observed on the response surface plot ture and therefore caused it to collapse. On the other hand, the
(Fig. 6A) that increased screw speed led to a significant increase decreased expansion at high temperature could also be a result
X. Meng et al. / Food Research International 43 (2010) 650–658 655

Die pressure (Pa × 106)


A Product temperature (°C) A
184 6.7
179 6.1
174
5.6
169
5.0
164
4.4

344 18.7 15.3


315 17.9 226 16.1
285 17.0 256 17.0
285
256 16.1 315 17.9
226 15.3 344 18.7
Screw speed (rpm) Feed moisture (%) Screw speed (rpm) Feed moisture (%)
Product temperature (°C)

Die pressure (Pa × 106)


B
183
176 7.1
169 6.5
161 5.9
154 5.3
4.7

177 344 143


169 315 344 152
160 285 315
285 160
152 256 169
143 226 256
226 177
Barrel temperature (°C) Screw speed (rpm)
Screw speed (rpm) Barrel temperature (°C)
Fig. 2. Effect of feed moisture content, screw speed, and barrel temperature on
product temperature. Fig. 3. Effect of feed moisture content, screw speed, and barrel temperature on die
pressure.

of increased axial or longitudinal expansion as high temperature


favors expansion in the axial direction (Ali, Hanna, & Chinnaswamy,
1996; Della Valle, Vergnes, Colonna, & Patria, 1997). Therefore, pre-
Motor torque (%)

cise control of extrusion conditions is necessary to achieve desir-


able extrudate properties. 25
The ER of the extrudates was in the range of 3.06 and 4.99 23
(Table 3), similar to those reported for cereal-based extrudates 21
(Ilo et al., 1996; Singh & Smith, 1997), indicating that desirable 19
expansion can be obtained for a chickpea flour-based extruded 18
product. A similar range of ER (2.52–4.73) has also been reported
for other pulses (peas, lentils) when extruded under similar condi-
tions (Patil et al., 2007). High ER was observed at low feed mois-
ture, high screw speed and barrel temperature (Fig. 6). However, 226
177 256
the canonical analysis indicated that the stationary point was a 169
saddle point, which means that there was no maximum ER ob- 160 285
152 315
tained within the extrusion conditions used in our study.
143 344
Barrel temperature (°C) Screw speed (rpm)
3.2.2. Bulk density
Fig. 4. Effect of feed moisture content, screw speed, and barrel temperature on
Bulk density was significantly affected by the linear (P < 0.001)
motor torque (feed moisture content had no influence on motor torque).
and quadratic (P < 0.01) effects of feed moisture content, screw
speed, and barrel temperature (Table 4). The response surface plots
(Fig. 7) showed that product bulk density increased with increasing
feed moisture, whereas decreased with increasing screw speed and
barrel temperature. The density of extrudates varied between Chinnan, 1997). Low density, a desirable characteristic of expanded
0.130 and 0.275 g/cm3 (Table 3), similar to those reported for products, was obtained at low feed moisture, high screw speed and
maize grits extrudates (0.09–0.32 g/cm3; Ilo et al., 1996) and com- barrel temperature (Fig. 7). The canonical analysis predicted that
mercial snack products (rod-shaped, cheese powder-coated corn the minimum bulk density (0.118 g/cm3) was obtained at 15% feed
puffed extrudates, 0.160–0.238 g/cm3; Suknark, Phillips, & moisture, 327 rpm screw speed, and 171 °C barrel temperature.
656 X. Meng et al. / Food Research International 43 (2010) 650–658

iability than the other types of expansion. This agrees with the re-
sults of our study (Table 3), where smaller variations were reported
SME (Wh/kg)
113 for density compared to ER. Patil et al. (2007) also indicated that 30
104 replicates were required for determining ER of pulse-based extru-
95 dates, while only 5 measurements were sufficient to estimate bulk
86 density.
77
3.2.3. Hardness
Hardness of extrudates was significantly (P < 0.01) affected by
linear effects of all 3 independent variables. Feed moisture content
177 344 and barrel temperature had also significant quadratic effects
169 315 (P < 0.05) on extrudate hardness (Table 4). Response surface plots
160 285 (Fig. 8) showed that hardness decreased with decreasing feed
152 256
143 226 moisture content, whereas with increasing screw speed and barrel
temperature. The hardness of extrudates ranged from 184 to 550 N
Barrel temperature (°C) Screw speed (rpm) (Table 3). Low hardness, which is also a favored property of extru-
dates, was observed at low feed moisture, high screw speed and
Fig. 5. Effect of feed moisture content, screw speed, and barrel temperature on
specific mechanical energy (SME, feed moisture content had no influence on SME). barrel temperature. The canonical analysis results showed that
the minimum hardness (184 N) was predicted at 15% feed
moisture, 299 rpm screw speed, and 157 °C barrel temperature.
The hardness of expanded extrudates is a perception of the hu-
A man being and is associated with expansion and cell structure of
the product. The hardness is the peak force required for a probe
5.6
or parallel blades (e.g. Kramer Shear Cell) to penetrate the extru-
Expansion ratio

4.9
date. The higher the value of maximum peak force required, the
4.1 higher the hardness of the sample. Previous studies also reported
3.4 that the hardness of extrudates increased as the feed moisture con-
2.6

A
Bulk density (g/cm3)

0.31
344 18.7
315 17.9 0.26
285 17.0 0.21
256 16.1
226 15.3 0.16
Screw speed (rpm) Feed moisture (%) 0.11

B
Expansion ratio

4.6 18.7 143


4.0 17.9 152
17.0 160
3.4 16.1 169
15.3 177
2.9
2.3 Feed moisture (%) Barrel temperature (°C)

B
Bulk density (g/cm3)

0.31
177 344
169 315 0.27
160 285 0.22
152 256
143 226 0.17
Barrel temperature (°C) Screw speed (rpm) 0.13

Fig. 6. Effect of feed moisture content, screw speed, and barrel temperature on
expansion ratio.

177 226
169 256
Bulk density also describes the degree of expansion undergone 160 285
by the melt as it exits the extruder. The sectional ER considers only 152 315
143 344
in the direction perpendicular to extrudate flow, while bulk density
considers expansion in all directions (Altan et al., 2008; Ilo et al.,
Barrel temperature (°C) Screw speed (rpm)
1996). Due to an inverse relationship between sectional and longi-
tudinal expansion (Dogan & Karwe, 2003; Ilo et al., 1996), bulk Fig. 7. Effect of feed moisture content, screw speed, and barrel temperature on bulk
density as a measure of volumetric expansion, would have less var- density.
X. Meng et al. / Food Research International 43 (2010) 650–658 657

screw speed and medium to high barrel temperature within the


A
Hardness (N) range of our process variables. It was demonstrated that, by prop-
610 erly selecting extrusion processing conditions and ingredient com-
503 position, chickpeas can be used to produce nutritious snack
395 products that have desirable expansion and texture attributes.
288
Acknowledgement
180

This study was funded by Pulse Canada as part of the Pulse


Innovation Project.

344 18.7
References
315 17.9
285 17.0
256 16.1 Agriculture and Agri-Food Canada (2008). The Canadian snack food industry.
<http://www4.agr.gc.ca/AAFC-AAC/display-afficher.do?id=1172692863066&lang=
226 15.3
eng>. Accessed 22.10.2008.
Screw speed (rpm) Feed moisture (%) Akdogan, H. (1996). Pressure, torque, and energy responses of a twin screw
extruder at high moisture contents. Food Research International, 29(5–6),
423–429.
Ali, Y., Hanna, M. A., & Chinnaswamy, R. (1996). Expansion characteristics of
B extruded corn grits. Lebensmittel-Wissenschaft und-Technologie, 29(8), 702–707.
Altan, A., McCarthy, K. L., & Maskan, M. (2008). Evaluation of snack foods from
barley-tomato pomace blends by extrusion processing. Journal of Food
Engineering, 84(2), 231–242.
Hardness (N)

Bhattacharya, S. (1997). Twin-screw extrusion of rice-green gram blend: Extrusion


560 and extrudate characteristics. Journal of Food Engineering, 32(1), 83–99.
465 Bhattacharya, S., & Prakash, M. (1994). Extrusion of blends of rice and chickpea
flours: A response surface analysis. Journal of Food Engineering, 21(3), 315–330.
370 Burrington, K. J. (2004). Benefits of whey proteins in breakfast and snack foods.
275 Cereal Foods World, 49(1), 334–336.
Chavan, J. K., & Salunkhe, D. K. (1986). Biochemistry and technology of chickpea
180
(Cicer arietinum, L.) seeds. CRC Critical Review in Food Science and Nutrition, 25,
107–112.
Davis, L. (2004). Fortifying grain-based products with whey protein. Cereal Foods
143 World, 49(1), 4–5.
Della Valle, G., Colonna, P., & Patria, A. (1996). Influence of amylose content on the
226 152 viscous behavior of low hydrated molten starches. Journal of Rheology, 40(3),
256 347–362.
160 Della Valle, G., Tayeb, J., & Melcion, J. P. (1987). Relationship of extrusion variables
285 with pressure and temperature during twin screw extrusion cooking of starch.
315 169 Journal of Food Engineering, 6(6), 423–444.
Della Valle, G., Vergnes, B., Colonna, O., & Patria, A. (1997). Relations between
344 177
rheological properties of molten starches and their expansion behavior in
Screw speed (rpm) Barrel temperature (°C) extrusion. Journal of Food Engineering, 31(3), 277–296.
Ding, Q., Ainsworth, P., Plunkett, A., Tucker, G., & Marson, H. (2006). The effect of
extrusion conditions on the functional and physical properties of wheat-based
Fig. 8. Effect of feed moisture content, screw speed, and barrel temperature on expanded snacks. Journal of Food Engineering, 73(2), 142–148.
product hardness. Ding, Q., Ainsworth, P., Tucker, G., & Marson, H. (2005). The effect of extrusion
conditions on the physicochemical properties and sensory characteristics of
rice-based expanded snacks. Journal of Food Engineering, 66(3), 283–289.
tent increased (Altan et al., 2008; Ding et al., 2005). This might be Dogan, H., & Karwe, M. V. (2003). Physicochemical properties of quinoa extrudates.
Food Science and Technology International, 9(2), 101–114.
due to the reduced expansion caused by the increase in feed mois- Huang, J., Schols, H. A., van Soest, J. J. G., Jin, Z., Sulmann, E., & Voragen, A. G. J.
ture. The decrease in hardness with increasing barrel temperature (2007). Physicochemical properties and amylopectin chain profiles of cowpea,
is in line with the results of bulk density where reduced density chickpea and yellow pea starches. Food Chemistry, 101(4), 1338–1345.
Hwang, M. P., & Hayakawa, K. I. (1980). Bulk densities of cookies undergoing
was observed. A low-density product naturally offers low hard- commercial baking processes. Journal of Food Science, 45(5), 1400–1402. 1407.
ness. High correlation between bulk density and hardness has been Ilo, S., Tomschik, U., Berghofer, E., & Mundigler, N. (1996). The effect of extrusion
demonstrated (Altan et al., 2008; Bhattacharya, 1997; Ding, Ains- operating conditions on the apparent viscosity and the properties of extrudates
in twin-screw extrusion cooking of maize grits. Lebensmittel-Wissenschaft und-
worth, Plunkett, Tucker, & Marson, 2006). The effect of screw speed
Technologie, 29(7), 593–598.
on hardness might be through its influence on extrudate expan- Jin, Z., Hsieh, F., & Huff, H. E. (1994). Extrusion of corn meal with soy fiber, salt, and
sion. The decrease in hardness with increasing screw speed was sugar. Cereal Chemistry, 71(3), 227–234.
Liu, Y., Hsieh, E., Heymann, H., & Huff, H. E. (2000). Effect of process conditions on
also observed in corn- and barley-based extrudates (Altan et al.,
the physical and sensory properties of extruded oat-corn puff. Journal of Food
2008; Liu, Hsieh, Heymann, & Huff, 2000). Science, 65(7), 1253–1259.
Meuser, F., & Van Lengerich, B. (1992). System analytical model for the extrusion of
starches. In J. L. Kokini, C. Ho, & M. V. Larwe (Eds.), Food extrusion science and
4. Conclusions technology (pp. 619–630). New York: Marcel Dekker.
Moraru, C. I., & Kokini, J. L. (2003). Nucleation and expansion during extrusion and
microwave heating of cereal foods. Comprehensive Reviews in Food Science and
Change of extrusion conditions, feed moisture content, screw
Food Safety, 2, 120–138.
speed, and barrel temperature, affected the extruder system Myers, R. H. (1971). Response surface methodology. Boston: Allyan and Bacon.
parameters and physical properties of the chickpea flour-based Onwulata, C. I., & Konstance, R. P. (2006). Extruded corn meal and whey protein
concentrate: Effect of particle size. Journal of Food Processing and Preservation,
snack. High SME was observed at high screw speed and low barrel
30(4), 475–487.
temperature. High level of feed moisture alone, or in combination Onwulata, C. I., Konstance, R. P., Smith, P. W., & Holsinger, V. H. (2001a). Co-
with high temperature resulted in dense and hard extrudates with extrusion of dietary fiber and milk proteins in expanded corn products.
poor expansion. The products with high expansion ratio and low Lebensmittel-Wissenschaft und-Technologie, 34(7), 679–687.
Onwulata, C. I., Smith, P. W., Konstance, R. P., & Holsinger, V. H. (2001b).
bulk density and hardness, which are generally good characteris- Incorporation of whey products in extruded corn, potato or rice snacks. Food
tics of an extruded snack, were obtained at low feed moisture, high Research International, 34(8), 679–687.
658 X. Meng et al. / Food Research International 43 (2010) 650–658

Pan, Z., Zhang, S., & Jane, J. (1998). Effect of extrusion variables and chemicals on the Shukla, T. (1994). Future snacks and snack food technology. Cereal Foods World,
properties of starch-based binders and processing conditions. Cereal Chemistry, 39(9), 704–715.
75(4), 541–546. Singh, N., & Smith, A. C. (1997). A comparison of wheat starch, whole wheat meal
Pansawat, N., Jangchud, K., Jangchud, A., Wuttijumnong, P., Saalia, F. K., Eitenmiller, and oat flour in the extrusion cooking process. Journal of Food Engineering, 34(1),
R. R., et al. (2008). Effects of extrusion conditions on secondary extrusion 15–32.
variables and physical properties of fish, rice-based snacks. Lebensmittel- Singh, U., & Singh, B. (1992). Tropical grain legumes as important human foods.
Wissenschaft und-Technologie, 41(4), 632–641. Economic Botany, 46(3), 310–321.
Patil, R. T., Berrios, J. DeJ., Tang, J., & Swanson, B. G. (2007). Evaluation of methods for Singh, U., Subrahmanyam, N., & Kumar, J. (1991). Cooking quality and nutritional
expansion properties of legume extrudates. Applied Engineering in Agriculture, attributes of some newly developed cultivars of chickpea (Cicer arietinum L).
23(6), 777–783. Journal of Science of Food and Agriculture, 55(1), 37–46.
Poltronieri, F., Areas, J. A. G., & Colli, C. (2000). Extrusion and iron bioavailability in Suknark, K., Phillips, R. D., & Chinnan, M. S. (1997). Physical properties of
chickpea (Cicer arietinum L.). Food Chemistry, 70(2), 175–180. directly expanded extrudates formulated from partially defatted peanut
Pulse Canada. (2008). Nutrition and health fact sheets. <http://www.pulsecanada. flour and different type of starch. Food Research International, 30(8),
com/food-and-nutrition/fact-sheets>. 18.12.2008. 575–583.
SAS. (1997). SASÒ User’s Guide: Statistics, Version 9.1. SAS Institute, Cary, NC. Wang, N., Daun, J. K. 2004. The chemical composition and nutritive value of
Shirani, G., & Ganesharanee, R. (2009). Extruded products with fenugreek (Trigonella Canadian pulses. Canadian Grain Commission report. July 28, 2004.
foenum-graecium) chickpea and rice: Physical properties, sensory acceptability White, G. (1994). Defining the true meaning of snacks. Food Technology International
and glycaemic index. Journal of Food Engineering, 90(1), 44–52. Europe, 2, 115–117.

You might also like