J Food Sci Technol

https://doi.org/10.1007/s13197-018-3363-z

ORIGINAL ARTICLE

Optimization and characterization of an extruded snack based

on taro flour (Colocasia esculenta L.) enriched with mango pulp
(Mangifera indica L.)
C. A. Pensamiento-Niño1 • C. A. Gómez-Aldapa2 • B. Hernández-Santos1 •

J. M. Juárez-Barrientos1 • E. Herman-Lara1 • C. E. Martı́nez-Sánchez1 •

J. G. Torruco-Uco1 • J. Rodrı́guez-Miranda1

Revised: 21 June 2018 / Accepted: 30 July 2018

Ó Association of Food Scientists & Technologists (India) 2018

Abstract The aim of this study was to optimize and Keywords b-Carotene
Expansion index
Mango
Taro
characterize an extruded snack made with taro flour and
mango pulp. A central experimental design composed of Abbreviations
the following three variables was used: mango pulp pro- BD Bulk density (g/cm3)
portion (MPP = 0–10 g/100 g) in taro flour, feed moisture EI Expansion index (–)
content (FMC = 16–30 g/100 g) and extrusion temperature FMC Feed moisture content (X2) (g/100 g)
(zone 4 of extruder) (T = 80–150 °C) using a single-screw MPP Mango pulp proportion (X1) (g/100 g)
extruder with a compression screw ratio of 3:1. Increasing SME Specific mechanical energy (J/g)
FMC values decreased the torque, pressure, specific T Extrusion temperature (X3) (°C)
mechanical energy (SME), expansion index (EI), water WAI Water absorption index (g/g)
solubility index and pH values and increased the residence WSI Water solubility index (%)
time, bulk density (BD), hardness and total colour differ- DE Total colour difference (–)
ence. Increasing T values led to a decrease in the torque,
pressure, BD and hardness values, while increasing MPP
values only caused a significant increase in the hardness Introduction
values and b-carotene content and a decrease in the pres-
sure value. The optimal extrusion conditions were Currently, snack foods are being redesigned to increase
T = 135.81 °C, FMC = 18.84 g/100 g and MPP = 7.97 g/ their nutritional value by adding micro or macronutrients,
100 g, with a desirability value of 0.772, to obtain a snack phytochemical components, vitamins, and antioxidants,
with EI = 1.52, BD = 0.66 g/cm3, hardness = 24.48 N, b- among other ingredients, to make the snack foods attractive
carotene content = 99.1 lg/g and SME = 428.54 J/g. The to consumers by the nutraceutical properties of these
mango pulp is an available and economical source of b- compounds (Honi et al. 2018). The bioactive compounds
carotene for the enrichment of extruded expanded taro typically found in small amounts in foods are currently
snacks. added as ingredients in the development of new products
not only for their bioactive potency but also for their
colouring properties (Falfán-Cortés et al. 2014; Emin et al.
2012), in this sense, some research has been carried out to
& J. Rodrı́guez-Miranda enrich snacks with antioxidant compounds such as b-car-
[email protected] otene. Guzman-Tello and Cheftel (1990) reported a
1
reduction from 73 to 38% of the initial synthetic trans-b-
Tecnológico Nacional de México, Instituto Tecnológico de
carotene content mixed with wheat flour when extruded at
Tuxtepec, Av. Dr. Vı́ctor Bravo Ahuja S/N., Col. 5 de Mayo,
C.P. 68350 Tuxtepec, Oaxaca, Mexico 125 and 200 °C, respectively. Falfán-Cortés et al. (2014)
2 evaluated the effect of the addition of passion fruit pulp
Área Académica de Quı́mica, Ciudad del Conocimiento,
ICBI-UAEH, Carr. Pachuca-Tulancingo Km 4.5, (0–7% (w/w)), extrusion temperature (80–140 °C) and feed
C.P. 42184 Mineral De La Reforma, Hidalgo, Mexico moisture content (16–30% (w/w)) in a blend of corn starch,

123
 J Food Sci Technol

reporting that at a passion fruit pulp concentration of Materials and methods

1.42% and barrel temperature of 127 °C and & 27% of
feed moisture content were the best conditions to preserve Raw materials
the b-carotene in extruded products (0.007 mg/100 g of
dry base (db) and the highest expansion index (1.8)). The Taro (Colocasia esculenta L.) and mango of the Manila
b-carotene retained during the extrusion process are variety (Mangifera indica L.) were purchased at the local
attributable to the raw material (source of b-carotene), the market in the city of San Juan Bautista Tuxtepec, Oaxaca,
type of extruder and the extrusion conditions, with thermal Mexico. The taro was washed, peeled and cut into 0.5 cm
degradation being the main factor contributing to the losses thick slices and then dried at 65 °C for 25 h (Rodrı́guez-
of b-carotene during extrusion (Falfán-Cortés et al. (2014). Miranda et al. 2011a). The dried slices were ground and
Extrusion cooking is a high-temperature short-time process sieved until a particle size of 0.59 mm (# 30 mesh) was
in which food materials are plasticized and cooked by a obtained. Mango pulp was obtained according to the pro-
combination of temperature, pressure and mechanical cedure described by Falfán-Cortés et al. (2014). The mango
shear, resulting in molecular transformation and chemical was peeled, pulped, and crushed and then centrifuged at
reactions. This technology uses a continuous process with 1027 9 g for 10 min at 15 °C (Hettich D-78532 1706-01
high productivity, significant retention of nutritional qual- centrifuge, Model Rotina 380R, Germany). The precipitate
ity, and natural colour and flavour of food (Navarro-Cortez (pulp) was blended with taro flour in different concentra-
et al. 2016). This method presents multiple advantages, the tions according to the experimental design (Fig. 1,
main advantage being that the ingredients undergo a series Table 1). The samples was placed in ZiplocÒ bags and
of transformations, which would ordinarily be carried out stored at 4 °C until use. The chemical compositions (% db)
in several unit operations (mixing, shearing, cooking, of the taro flour and mango pulp were: 6.12 and 4.33
drying and texturing), in a single fast and efficient step protein, 0.25 and 0.33 lipids, 1.82 and 8.83 fibre, 3.09 and
(Rodrı́guez-Miranda et al. 2011b). Therefore, this method 9.19 ash, 88.72 and 77.32 carbohydrates, as well as a b-
is important for the production of snacks with higher carotene content of 68.45 and 108.85 lg/g, respectively.
nutritional content and the use of non-conventional and
perishable raw materials to give an added value, as in the b-carotene content
case of taro and magician and with the above decrease the
losses after harvest. Taro (Colocasia esculenta (L.) Schott) The b-carotene content was determined in a Spectropho-
is a tuber with edible starch belonging to the Araceae tometer (Cary 60 UV–VIS MY11510011, ÓAgilent
family. Taro cultivation in developing countries has Technologies, Inc., USA.) according to the procedure
become important in recent years due to the high fibre described by Ying et al. (2015) in raw materials and
content (0.6–0.8 g/100 g); protein content (2–6 g/100 g); extruded products. Four millilitres of dimethylsulfoxide
and mucilage, vitamin, phosphorus, calcium and starch (DMSO) were added to 80 mg of sample in a water bath at
content (70–80 g/100 g db) of taro (Rodrı́guez-Miranda 75 °C, and the mixture was stirred at 150 rpm for 50 min.
et al. 2011a). In contrast, mango (Mangı́fera indica L.) is a Subsequently, the mixture was cooled to 25 °C, and 4 mL
fruit that belongs to the Anacardiaceae family, and the of butyl-hydroxy-toluene (BHT) at 0.1% (w/v) in hexane
edible part of fruit is between 60 and 75% of the whole (C6H14) was added, and the mixture was vortexed for 10 s
fruit. Mango has high carotenoid content when mature and and left to stand for 30 min. Two to three drops of pure
is a good source of provitamin A (Liu et al. 2013). The ethanol were added to precipitate any protein present in the
major component of mango is water, which is present at hexane phase. The hexane layer was recovered and placed
84%. The sugar content varies from 10 to 20%, and the in a 10 mL beaker; 0.5 g of anhydrous sodium sulfate was
protein content is 0.5%. The predominant acid is citric added (Na2SO4 anhydride); and finally, the absorbance of
acid, although malic, succinic, uronic, tartaric and oxalic the recovered hexane was analysed at a wavelength of
acids are also found in smaller amounts (Ribeiro et al. 450 nm.
2007). Therefore, the aim of the study was to optimize the
process to develop an extruded snack based on taro flour Colour and pH
(Colocasia esculenta L.) enriched with mango pulp
(Mangifera indica L.) prepared using a single screw The colour of the extruded products was determined by
extruder and to evaluate the effect of extrusion tempera- using a Hunter Laboratory tristimulus colorimeter (Ul-
ture, feed moisture content and the proportion of mango traScanÒ VIS, Hunter Laboratory, Model USVIS1347,
pulp in taro flour on some process parameters, physical, Hunter Associates Laboratory, Inc., Reston, Virginia,
functional properties and b-carotene content of the extru- USA). The values L*, a* and b* were obtained and used to
ded snacks.

123
J Food Sci Technol

Fig. 1 General schematic diagram of obtaining and evaluating extruded snacks based on taro flour enriched with mango pulp

calculate the total colour difference (DE). The pH was Specific mechanical energy (SME)
measured in a dispersion of 1 g of flour in 10 mL of dis-
tilled water at 25 °C. The SME (J/g), defined, as the total mechanical energy
input required obtaining 1 g of extruded product, was
Extrusion process calculated as described by Rodrı́guez-Miranda et al.
(2012a).
Extrusion was conducted in a single-screw laboratory
extruder (Extruder 19/25DN, Model 832005.007, Braben- Characterization of extruded products
derÒ GmbH & Co. KG, Germany) with a length of
428 mm, diameter of 19 mm, compression ratio of 3:1. A Expansion index (EI) and bulk density (BD)
3 mm circular exit die was used. The extruder had four
heating zones (Zone 1 = 40 °C, Zone 2 = 60 °C, Zone The EI was calculated in 20 samples with a Vernier caliper
3 = 80 °C and Zone 4 = 80–150 °C). Before extruding, the (Vernier, Science Purchase, 0604CAL6, USA), dividing
formulations were blended (Fig. 1), and the moisture the diameter of the extruded products by the internal
content of each was adjusted from 16 to 30 g/100 g diameter of the exit orifice of the extruder (Rodrı́guez-
according to the experimental design (Table 1). Miranda et al. 2014a). The BD was determined as descri-
bed by Navarro-Cortez et al. (2016).
Variables of the extrusion process

Residence time and torque

The residence time and torque was measured according to

procedures described by Rodrı́guez-Miranda et al. (2012a).

123
123
Table 1 Experimental data of extruded snack for response surface analysis
Run Extrusion process variables Responses
MPP (g/100 g) FMC (g/100 g) T (°C) (X3) RT (g/ To (Nm/ P (N/ SME (J/ EI (– BD (g/ WAI (g/ WSI H pH b-C (lg/ DE
(X1) (X2) min) s) cm2) g) ) cm3) g) (%) (N) g)

1 2 (- 1) 19 (- 1) 94 (- 1) 37.01 27.13 1740 460.67 1.13 1.46 6.35 10.77 3.40 6.27 64.28 32.89
2 8 (1) 19 (- 1) 94 (- 1) 36.32 32.93 1193 569.79 1.10 1.33 6.34 10.41 94.08 6.25 49.73 33.39
3 2 (- 1) 19 (- 1) 136 (1) 37.15 18.10 1041 306.10 1.26 0.84 6.10 14.09 16.14 6.25 78.27 33.47
4 8 (1) 19 (- 1) 136 (1) 37.74 15.87 600.5 264.16 1.47 0.70 5.83 15.89 8.47 6.20 109.98 34.53
5 2 (- 1) 27 (1) 94 (- 1) 44.17 23.20 510 330.04 1.01 1.26 5.36 10.58 65.44 6.14 62.59 35.02
6 8 (1) 27 (1) 94 (- 1) 43.21 14.33 426.8 208.44 0.92 1.35 5.24 13.23 36.85 6.16 30.27 33.16
7 2 (- 1) 27 (1) 136 (1) 43.44 10.77 250.3 155.74 1.47 0.91 5.75 9.56 20.69 6.14 102.05 33.18
8 8 (1) 27 (1) 136 (1) 42.13 11.70 241.5 174.51 1.35 0.96 6.34 12.51 95.73 6.11 138.15 36.19
9 0 (- 1.682) 23 (0) 115 (0) 41.12 31.77 808.1 485.45 1.03 1.34 5.51 8.97 67.65 6.23 49.91 33.81
10 10 (1.682) 23 (0) 115 (0) 33.14 25.23 575.9 478.45 1.05 1.30 5.28 11.46 77.16 6.19 142.50 36.89
11 5 (0) 23 (0) 80 (- 1.682) 32.80 28.10 1408 538.31 1.12 1.39 6.24 10.09 96.54 6.30 115.59 32.24
12 5 (0) 23 (0) 150 (1.682) 34.03 18.07 255.3 333.55 1.65 0.74 6.84 11.88 1.13 6.27 52.55 32.09
13 5 (0) 16 (- 1.682) 115 (0) 33.92 47.80 1435 885.37 1.63 0.69 5.68 24.14 0.80 6.22 6.26 31.09
14 5 (0) 30 (1.682) 115 (0) 41.95 16.83 390.2 252.15 1.01 1.31 5.20 11.72 3.51 5.92 6.78 38.50
15 5 (0) 23 (0) 115 (0) 37.54 23.50 673.2 393.34 1.02 1.39 6.74 11.99 3.20 6.15 38.40 34.00
16 5 (0) 23 (0) 115 (0) 35.12 20.83 617.3 372.66 1.02 1.38 5.57 9.60 3.52 6.19 35.38 36.16
17 5 (0) 23 (0) 115 (0) 35.06 23.67 583.1 424.14 1.03 1.39 6.58 15.98 3.45 6.23 38.42 37.08
18 5 (0) 23 (0) 115 (0) 35.01 21.23 601.5 381.08 1.05 1.37 5.51 14.27 3.40 6.24 39.28 35.21
19 5 (0) 23 (0) 115 (0) 34.94 15.93 591.1 286.53 1.02 1.43 5.60 11.28 3.27 6.11 38.43 37.10
20 5 (0) 23 (0) 115 (0) 34.23 22.97 559.3 421.53 1.02 1.40 5.93 13.47 3.22 6.21 38.61 37.39
MPP mango pulp proportion, FMC feed moisture content, T temperature, RT residence time, To torque, P pressure, SME specific mechanical energy, EI expansion index, BD bulk density, WAI
water absorption index, WSI water solubility index, H hardness, b-C b-carotene content, DE total color difference
J Food Sci Technol
J Food Sci Technol

Functional properties DE within range, established in the Design Expert 7.0

program (State-Ease Inc., Minneapolis, MN, USA) as
Water absorption index (WAI) and water solubility index optimal values. To predict the dependent variables at the
(WSI) optimum value, the complete models were considered in
desirability function.
These indexes were determined according to the method-
ology described by Rodrı́guez-Miranda et al. (2012a).
Results and discussion
Hardness
Effect of mango pulp proportion (MPP), feed
The maximum fracture strength in compression (Rodrı́- moisture content (FMC) and extrusion temperature
guez-Miranda et al. 2014b) of the extrudates was deter- (T) on extrudates
mined on a TA-XT2 texture analyser (Texture
Technologies Corp. Scarsdale, NY/Stable Micro Systems, Regression coefficients
Haslemere, Surrey, UK) using a Warner–Bratzler cutting
blade with a test speed of 5 mm/s. The hardness was Table 2 shows the regression coefficients for all the
determined in Newton (N). Fifteen repetitions were per- responses analysed. The MPP (X1) has significant linear
formed per treatment. effect (p \ 0.05) on pressure, hardness and b-carotene
content and a quadratic effect (X21) on hardness and b-
Experimental design and data analysis carotene content. On the other hand, the FMC (X2) exhibits
a significant linear effect (p \ 0.05) on residence time,
Central experiments were designed (Table 1) with three torque, pressure, hardness, SME, EI, BD, WSI, pH and DE
independent variables by using a commercial statistical and quadratic effect (X22) on residence time, pressure, EI,
package (Design-Expert 7.0.0 Statease Inc., Minneapolis, BD, WSI and pH, while T (X3) has a significant linear
MN, USA). The independent variables were as follows: effect (p \ 0.05) on torque, pressure, EI, BD and hardness
mango pulp proportion (MPP = 0–10 g/100 g), feed and a quadratic effect (X23) on pressure, hardness, EI, BD,
moisture content (FMC = 16–30 g/100 g) and extrusion pH, DE and b-carotene content. The X1X2 and X2X3 inter-
temperature (zone 4 of extruder) (T = 80–150 °C). The actions have significant effects (p \ 0.05) on pressure and
limits established for the independent variable MPP were hardness, and the X1X3 interaction has a significant effect
defined by preliminary tests. The levels higher than 10 g/ (p \ 0.05) on b-carotene content. The quadratic regression
100 g of MPP can hinder the expansion of the product model adjusted to the experimental data show that the
during extrusion, resulting in a more compact and hard model is accurate for the evaluated responses (p \ 0.05),
product, unwanted for the consumer. The feed volume with the exception of the WAI (p [ 0.05). The models had
(20 g/min) and screw speed (100 rpm) were kept constant. an R2 [ 0.653.
The experimental data were evaluated using the response
surface methodology (RSM) to investigate the effect of the Process variables
extrusion process (FMC and T) and the MPP on the
response variables. The statistical significance of the terms The residence time values varied from 32.80 to 44.17 g/
of regression was examined by analysis of variance min (Table 1). Figure 2a shows that the residence time was
(ANOVA) for each response. significantly influenced (p \ 0.05) by the FMC. It is
observed that the increase in FMC causes an increase in the
Process optimization residence time; high residence time values were observed
when the FMC values ranged from 27 to 30 g/100 g, while
The objective in the optimization process was to find the at low FMC values (16–23 g/100 g), the residence time
process conditions and proportion of mango pulp with the value decreased. This phenomenon occurs because the
maximum content of b-carotene, greater expansion as well FMC causes a decrease in the viscosity of the mixture,
as lower energy expenditure in development of a snack having an effect similar to that of high lipid levels in the
enriched in b-carotene based on taro flour. Numerical system (Rodrı́guez-Miranda et al. 2012a).
optimization was carried out by superposition of the dif- The torque value varies from 10.77 to 47.80 Nm
ferent response surfaces based on the maxima and minima (Table 1). Figure 2b shows that at low FMC (16–23 g/
observed, i.e., maximum EI, minimum BD, minimum 100 g) and low T (80–115 °C) values, the torque value
hardness, maximum b-carotene content, minimum SME increases, and increasing the FMC (18–27 g/100 g) and T
and residence time, torque, pressure, WAI, WSI, pH and (94–135 °C) value reduces the torque value. The decrease

123
123
Table 2 Coefficients estimated by multiple linear regression of the physicochemical characterization of extruded snacks using taro flour and mango pulp blends
Coefficients Response
RT To P SME EI BD WAI WSI H pH b-C DE

Intercepto 35.190* 21.650* 605.961* 386.142* 1.060* 1.396* 5.983* 12.794* 3.423 6.186* 37.414* 36.160*
Linear
X1 - 1.157 - 1.124 - 108.184* - 3.472 0.000 - 0.014 - 0.014 0.822 10.503* - 0.011 12.936* 0.578
X2 2.798* - 6.305* - 352.163* - 131.576* - 0.093* 0.087* - 0.200 - 1.917* 8.087* - 0.068* 2.320* 1.152*
X3 0.134 - 4.250* - 269.828* - 74.162 0.165* - 0.226* 0.128 0.737 - 14.097* - 0.012 8.461 0.194
Quadratic
X21 1.470 0.603 16.680 - 4.827 - 0.017 - 0.027 - 0.183 - 1.097 24.363* 0.008 23.914 - 0.287
X22 1.755* 1.952 88.367* 25.865 0.083* - 0.142* - 0.164 1.629* - 2.417 - 0.040* - 7.794 0.481
X23 0.158 - 1.312 66.954* - 21.097 0.107* - 0.120* 0.224 - 0.825 14.166* 0.034* 19.622* - 1.411*
Interactions
X1 X2 - 0.271 - 1.438 111.938* - 21.252 - 0.048 0.050 0.093 0.521 - 11.117* 0.008 - 1.672 - 0.051
X1 X3 0.116 0.221 22.612 - 1.337 0.025 - 0.006 0.057 0.309 0.001 - 0.010 14.336* 0.679
X2 X3 - 0.423 1.379 105.813* 31.497 0.049 0.064 0.283 - 1.318 8.315* 0.001 9.139 - 0.068
R2 0.730 0.674 0.964 0.660 0.855 0.929 0.653 0.722 0.684 0.869 0.656 0.711
P of F (model) 0.0506 0.0024 \ 0.0001 0.0017 0.0350 0.001 0.1324 0.0571 0.0699 0.0022 0.1163 0.0671
Lack of Fit 0.0116 0.0289 0.0057 0.0222 0.1106 0.004 0.9590 0.3852 \ 0.0001 0.8444 \ 0.0001 0.2953
*Bold numbers indicate estimates of significant parameters (p \ 0.05)
X1 mango pulp proportion, X2 feed moisture content, X3 temperature, RT residence time, To torque, P pressure, SME specific mechanical energy, EI expansion index, BD bulk density, WAI
water absorption index, WSI water solubility index, H hardness, b-C b-carotene content, DE total color difference
J Food Sci Technol
J Food Sci Technol

Fig. 2 Response surface plots of process variables a residence time, b torque, c pressure, and d specific mechanical energy (SME)

in the To values with the increase in the FMC value could were at MPP = 0–7.6 g/100 g, FMC = 16–23 g/100 g and
be due to the unavailability of water for gelatinization of at T = 80–135 °C. The increase in the T value with
the starch, which results in a decrease in the apparent increasing FMC values affects the viscosity of the mass
viscosity; the decrease in the torque value can be attributed throughout the extruder, which leads to a decrease in the
to the reduction of friction in the extruder as a result of the die pressure. Kaur et al. (2000) have reported that
increase in the FMC value, which minimizes the difficulty increasing temperatures decrease the viscosity of the
of processing the mixtures. In general, the torque value materials. The interaction of FMC-T increases the pressure,
tends to decrease with increasing T values due to a because with the increase of the FMC it causes an exces-
decrease in viscosity (Kannadhason et al. 2009). sive viscosity of the molten material that leads to a low
The pressure value obtained was 241.5–1740 N/cm2 torque of the motor, while the increase of the T increases
(Table 1). Figure 2c shows that the pressure value increa- the pressure of the saturated vapors exceeds the melting
ses with decreasing MPP and FMC values. The increase in pressure towards the exit of the extruder. In the other hand,
the T value during the process causes a decrease in the in the interaction of MPP-FMC, the increase in MPP
pressure value, and this effect can be attributed to the increases the concentration of sugars in the blend, Fan et al.
decrease in the viscosity of the molten mixture caused by (1996) reported that sugars reduced the expansion of the
the increase in the T value. The highest-pressure values section of corn extrudates, monosaccharide’s have a more

123
 J Food Sci Technol

pronounced effect than disaccharides. This result in a softer in FMC can reduce the elasticity of the mass by plasticizing
matrix that collapses under the high vapor pressure during the melt and, therefore, reduces gelatinization, decreases
expansion, reducing the final degree of expansion, expansion and increases density (Korkerd et al. 2016).
accordingly there is a higher pressure exerted during the Paraman et al. (2013), mention that moisture content of
extrusion process. feed higher than 28% significantly decreases the expansion
SME values ranged between 155.74 and 885.37 J/g and increased the hardness of the crispy proteins, as well as
(Table 1). SME during extrusion plays a direct role in the the density of the product increases, this drunk to an
macromolecular transformations and in the interactions inadequate denaturation of the proteins.
among the different components in the materials. The BD is considered as an index of the extent of puffing and
macromolecular transformations can be the conversion of ranged from 0.69–1.46 (g/cm3) (Table 1). Figure 3b shows
starch or changes in the structure of the protein that, con- that at high FMC values and low T values, BD increased.
sequently, determine the rheological properties of the However, when the FMC and T value were 16–27 g/100 g
materials (Aguilar-Palazuelos et al. 2012). Figure 2d shows and 115–150 °C respectively, the BD value decreased.
that when the FMC value decreases the SME value This decrease was a consequence of the expansion of the
increases; this effect was observed at MPP values of extrudes. Some authors have reported that the expansion of
0–7.6 g/100 g and FMC values of 16–23 g/100 g, while at extruded products depends on the degree of starch gela-
MPP values of 2–7.6 g/100 g and FMC values of 18–30 g/ tinization, the amount of starch and the processing condi-
100 g, the lowest values of the SME were observed. The tions (Rodrı́guez-Miranda et al. 2014b). At comparatively
viscosity of the food increases when the FMC value higher temperatures melt viscosity get reduced hence
decreases, and therefore, the SME value increases (Rodrı́- bubble walls become too thin to contain the vapour pres-
guez-Miranda et al. 2012a). It was also observed that upon sure, resulting in more bubble fracture, thus increasing rate
increasing the T value, the SME value decreases. High of collapse and overall expansion decreased, increasing the
temperatures are associated with a decrease in the viscosity BD (Borah et al. 2016). The BD value of the snacks is an
of the melt within the extruder, which in turn reduces the indirect measure of the EI, since the BD value has a neg-
input of energy to the extruder. Upon increasing the FMC ative correlation with the expansion (Navarro-Cortez et al.
value, the viscosity decreases, which eventually leads to an 2016).
SME minimum (Rodrı́guez-Miranda et al. 2012a).
Water absorption index (WAI) and water solubility index
Extrudate properties (WSI)

Expansion index (EI) and bulk density (BD) WAI values ranged between 5.20 and 6.84 g/g (Table 1).
WAI is associated with the amount of water absorbed by
EI of the extruded products ranged from 0.92 to 1.65 the starch granules after swelling and can be used as an
(Table 1). Increasing the FMC value decreased the EI index of the degree of gelatinization of the starch.
value, and increasing the T value favoured the expansion of (Rodrı́guez-Miranda et al. 2011b). Moreover, the WAI is
the snacks. This phenomenon can be attributed to the fact associated with the hydrophilic balance of the proteins
that expansion occurs at high temperatures and low mois- present in the mixture, which changes based on the degree
ture content, causing transitions, phase transformations and of denaturation of the proteins, whereby the extrusion
structural changes in the biopolymers, leading to the for- process modifies the solubility index of the proteins
mation of air bubbles inside the starch and retaining the air (Rodrı́guez-Miranda et al. 2012a). The variables (MPP, T
bubbles after the exit of the material (Moraru and Kokini and FMC) did not show significant effect (p [ 0.05) on
2003). Korkerd et al. (2016), mention that beyond the WAI.
critical temperature of extrusion there is a decrease in the WSI oscillated between 8.97 and 24.14% (Table 1).
EI, which depends on the type of starch and moisture Figure 3c shows that at low FMC values (18–27 g/100 g)
content, which could be the result of increased dex- and high T values (94–150 °C), the WSI increased, which
trinization, excessive softening and possible structural may be because, under these processing conditions, a
degradation of starch, in which with stands the high vapour greater amount of gelatinized material causes greater sol-
pressure and collapse. Figure 3a shows that at MPP values ubilization. Solubility is used in the extrusion process as an
of 2–8 g/100 g and T values of 115–150 °C, an increase in indicator of the degradation of molecular components such
the EI value was obtained. This increase in the EI value as starch, fibres and proteins (Seth and Rajamanickam
with an increase in the T value and decrease in the FMC 2012). WSI indicates the total degradation of starch gran-
value has also been reported by other authors (Falfán- ules. The increase in solubility of the extruded products is
Cortés et al. 2014; Ruiz-Armenta et al. 2018). The increase attributed to the dispersion of amylose and amylopectin

123
J Food Sci Technol

123
 J Food Sci Technol

b Fig. 3 Response surface plots of process variables a expansion index pH

(EI), b bulk density (BD), c water solubility index (WSI), d hardness,
e pH, f b-carotene content and g total color difference
The pH values of the extrudates ranged between 5.92 and
6.30 (Table 1). Figure 3e shows that increasing FMC and T
values decreased the pH values of the extruded products,
molecules followed by gelatinization when the processing which may be because the increase in the FMC and T
conditions are mild and to the formation of low-molecular- values during the extrusion process helps release some
weight components under harsh conditions (Colonna et al. organic acids (e.g., fatty acids) present, leading to low pH
1984). The relatively low FMC and T resulted in low mass values (more acidic) of the samples (Navarro-Cortez et al.
shear with the consequent lower macromolecular degra- 2016). Bowen et al. (2006) reported that the presence of
dation. The higher the dextrinization, attributed to higher free fatty acids may cause changes in the pH of the system
and degradation of the starch to small more soluble parti- and result in hydrolysis of the starch. Sriburi and Hill
cles, which increases the WSI (Carvalho et al. 2013). (2000) reported an increase in pH in extruded products,
with initial values between 1 to 6.5 increasing to 2.15 and
Hardness 6.63, respectively.

The hardness of the extruded products ranged from 0.80 to b-carotene content
96.54 N (Table 1). Notably, the lowest value of hardness
was obtained at MPP = 5 g/100 g, FMC = 23 g/100 g and The b-carotene content ranged from 6.26 to 142.50 lg/g
T = 150 °C. Figure 3d shows that the increase in the FMC (Table 1), and the highest b-carotene content was found at
and MPP values directly affected the increase in the MPP 10 g/100 g, FMC 23 g/100 g and T 115 °C (Table 1).
hardness values of the extruded products; the same beha- Figure 3f shows that increasing MPP and T values helped
viour was reported by Liu et al. (2013). Low hardness in the retention and final preservation of the b-carotene
values of the extrudates were obtained at low MPP content in the product. The highest b-carotene content in
(5–7.97 g/100 g) and FMC (16–23 g/100 g) values. Ruiz- the products was obtained at high MPP values and high
Armenta et al. (2018) reported that at higher extrusion extrusion T values, because the mango pulp is high b-
temperatures, greater expansion is obtained, and the hard- carotene content (108.85 lg/g). On the other hand, the T
ness of the extrudates is reduced. Moraru and Kokini value did not have a significant effect (p [ 0.05) on this
(2003) reported that with high values of EI and low values variable, and the conservation of this component can pos-
of texture, extruded products are generated at high tem- sibly be attributed to the presence of a protein-carbohydrate
peratures, leading to several events, such as the structural matrix that protects this bioactive component during the
transformation of biopolymers; phase transition and extrusion process (Basto et al. 2016). An additional aspect
nucleation; and swelling, growth and collapse of air bub- to consider is that extrusion is a short-time cooking pro-
bles. The negative effect of the MPP-FMC interaction is cess, so it is possible that there was not enough time to
due to the fact that the increase in MPP decreases the starch cause effective damage to low molecular weight com-
content present in the mixture and by increasing the FMC pounds; however, the heating and shearing could have
that provides the motive power for the expansion is ruptured the cell wall, which may have helped to release
affected, therefore the expansion decreases and porosity more carotenoids, such as b-carotene (Rojas-Garbanzo
probably due to excessive softening by excess water et al. 2011), therefore the interaction of MPP-T helps the
thereby decreasing the hardness of the final product (Mo- release of the b-carotene contained in the blend. The results
raru and Kokini 2003). While the interaction FMC-T are consistent with those reported by other authors (Basto
increased the hardness of the final product, because the et al. 2016; Falfán-Cortés et al. 2014).
increase of the FMC and increase of T allow the raw
material to undergo a glass transition during the extrusion Total colour difference (DE)
process and thus facilitates the deformation of the matrix
and its expansion, however there is a temperature range There are many reactions that occur during extrusion
where the expansion reaches a maximum, and this opti- cooking that affect colour (Falfán-Cortés et al. 2014). The
mum temperature depends on the type of starch, and if not DE value of the products varied from 31.09 to 38.50
reaching that temperature, the excess water increases the (Table 2), with the highest DE value observed at MPP 5 g/
final product stiffness and hardness (Moraru and Kokini 100 g, FMC 30 g/100 g and T 115 °C (Table 2). Figure 3g
2003). shows that increasing FMC values increased the DE value,
while at high T values, the DE value decreased. The
increase in the T value during the extrusion process leads to

123
J Food Sci Technol

Table 3 Optimum values of extrusion process parameters and responses for extruded snack
Process parameters Importance Target Experimental value Optimum value Desirability
Min Max

MPP (g/100 g) 3 Range 0 10 7.9 0.772

FMC (g/100 g) 3 Range 16 30 18.84
T (°C) 3 Range 80 150 135.81
Responses Predicted values
Residence time (g/min) 3 Range 32.80 44.17 35.56
Torque (Nm) 3 Range 10.77 47.80 23.40
Pressure (N/cm2) 3 Range 241.5 1740 567.8
SME (J/g) 4 Minimize 155.74 885.37 428.54
EI 5 Maximize 0.92 1.65 1.52
BD (g/cm3) 5 Minimize 0.69 1.46 0.66
WAI (g/g) 3 Range 5.20 6.84 5.85
WSI (%) 3 Range 8.97 24.14 17.08
Hardness (N) 5 Minimize 0.80 96.54 24.48
pH 3 Range 5.92 6.30 6.21
b-carotene content (lg/g) 5 Maximize 6.26 142.50 99.10
DE 3 Range 31.09 38.50 34.39
MPP mango pulp proportion, FMC feed moisture content, T temperature, SME specific mechanical energy, EI expansion index, BD bulk density,
WAI water absorption index, WSI water solubility index, DE total color difference

products with less luminosity and tends to darken the increased the residence time, apparent density, hardness
products. It is known that the reduction of sugars and and total colour difference. Increasing the temperature led
proteins (amino acids) in food can occur at high processing to a decrease in torque, pressure, bulk density and hardness,
temperatures and can promote non-enzymatic browning while an increase in the proportion of mango pulp caused a
(Maillard reaction), which results in the darkening of the significant increase in the hardness and b-carotene content
final product (Navarro-Cortez et al. 2016). and a decrease in pressure. Therefore, mango pulp is an
available and economical source of b-carotene for the
Optimization enrichment of extruded expanded taro snacks.

Numerical optimization by superposition of the different Acknowledgements The authors are grateful to CONACYT (Con-
sejo Nacional de Ciencia y Tecnologı́a) for the Master Science
response surfaces was used to determine the optimal level scholarship for C. A. Pensamiento-Niño, and the National Techno-
of the three independent variables based on the concept of logical of Mexico (TecNM) for the financing the project 5901.16-P.
convenience and desirability. In Table 3, we can observe
the importance assigned to each of the response variables
as well as the intervals determined. At these conditions References
predicted values were: T = 135.81 °C, FMC = 18.84 g/
100 g and MPP = 7.97 g/100 g with a desirability value of Aguilar-Palazuelos E, Zazueta-Morales JDJ, Harumi EN, Martı́nez-
Bustos F (2012) Optimization of extrusion process for produc-
0.772, to obtain a product with the following characteris- tion of nutritious pellets. Food Sci Technol 32(1):34–42
tics: EI = 1.52, BD = 0.66 g/cm3, hardness = 24.48 N, b- Basto GJ, Carvalho CWP, Soares AG, Costa HTGB, Chávez DWH,
carotene content = 99.1 lg/g and SME = 428.54 J/g. de Oliveira Godoy RL, Pacheco S (2016) Physicochemical
properties and carotenoid content of extruded and non-extruded
corn and peach palm (Bactris gasipaes, Kunth). LWT-Food Sci
Technol 69:312–318
Conclusion Borah A, Mahanta CL, Kalita D (2016) Optimization of process
parameters for extrusion cooking of low amylose rice flour
Taro flour enriched with mango pulp could be used to blended with seeded banana and carambola pomace for devel-
opment of minerals and fiber rich breakfast cereal. J Food Sci
produce snacks. Increasing the moisture content in the feed Technol 53(1):221–232
decreased the torque, pressure, specific mechanical energy, Bowen SE, Gray DA, Giraud C, Majzoobi M, Testa CEM, Perez
expansion index, water solubility index and pH and LAB, Hill SE (2006) Lipid oxidation and amylopectin molecular

123
 J Food Sci Technol

weight changes occurring during storage of extruded starch Paraman I, Supriyadi S, Wagner ME, Rizvi SS (2013) Prebiotic fibre-
samples. J Cereal Sci 43:275–283 incorporated whey protein crisps processed by supercritical fluid
Carvalho AV, Bassinello PZ, Mattietto RDA, Rios ADO, de Lima extrusion. Int J Food Sci Technol 48(10):2193–2199
ACP, Koakuzu SN, Carvalho RN (2013) Physicochemical, Ribeiro SMR, Queiroz JH, de Queiroz MELR, Campos FM, Sant’Ana
technological and sensory characteristics of a rice (Oryza sativa HMP (2007) Antioxidant in mango (Mangifera indica L.) pulp.
L.) and bean (Phaseolus vulgaris L.) soup prepared by extrusion. Plant Foods Hum Nutr 62(1):13–17
Int J Food Sci Tech 48(10):2057–2063 Rodrı́guez-Miranda, Rodrı́guez-Miranda J, Rivadeneyra-Rodrı́guez
Colonna P, Doublier IL, Melcion JP, De Monredon F, Mercier C JM, Ramı́rez-Rivera JE, Juárez-Barrientos JM, Herrera-Torres E,
(1984) Extrusion cooking and drum drying of whcat starch, Navarro-Cortez RO, Hernández-Santos B (2011a) Caracteri-
physical and macromolecular modifications. Cereal Chem zación fisicoquı́mica, funcional y contenido fenólico de harina de
61(6):538–543 malanga (Colocasia esculenta) cultivada en la región de
Emin MA, Mayer-Miebach E, Schuchmann HP (2012) Retention of Tuxtepec, Oaxaca, México. Cienc Mar 15(43):37–47
b-carotene as a model substance for lipophilic phytochemicals Rodrı́guez-Miranda J, Ruiz-López II, Herman-Lara E, Martı́nez-
during extrusion cooking. LWT Food Sci Technol 48:302–307 Sánchez CE, Delgado-Licon E, Vivar-Vera MA (2011b) Devel-
Falfán-Cortés RN, Verdalet-Guzmán I, Martı́nez-Bustos F (2014) opment of extruded snacks using taro (Colocasia esculenta) and
Effects of some extrusion variables on physicochemical charac- nixtamalized maize (Zea mays) flour blends. LWT-Food Sci
teristics of extruded corn starch-passion fruit pulp (Passiflora Technol 44(3):673–680
edulis) snacks. Plant Foods Hum Nutr 69(4):365–371 Rodrı́guez-Miranda J, Delgado-Licon E, Ramı́rez-Wong B, Solı́s-
Fan J, Mitchell JR, Blanshard JMV (1996) The effect of sugars on the Soto A, Vivar-Vera MA, Gómez-Aldapa CA, Medrano-Roldán
extrusion of maize grits. I. The role of the glass transition in H (2012a) Effect of moisture, extrusion temperature and screw
determining product density and shape. Int J Food Sci Technol speed on residence time, specific mechanical energy and
31(1):55–65 Physicochemical properties of bean flour and soy protein
Guzman-Tello R, Cheftel J (1990) Colour loss during extrusion aquaculture feeds. J Anim Prod Adv 2(1):65–73
cooking of b-carotene-wheat flour mixes as an indicator of the Rodrı́guez-Miranda J, Hernández-Santos B, Herman-Lara E, Vivar-
intensity of thermal and oxidative processing. Int J Food Sci Vera MA, Carmona-Garcı́a R, Gómez-Aldapa CA, Martı́nez-
Technol 25:420–434 Sánchez CE (2012b) Physicochemical and functional properties
Honi B, Mukisa IM, Mongi RJ (2018) Proximate composition, of whole and defatted meals from Mexican (Cucurbita pepo)
provitamin A retention, and shelf life of extruded orange-fleshed pumpkin seeds. Int J Food Sci Technol 47(11):2297–2303
sweet potato and bambara groundnut-based snacks. J Food Rodrı́guez-Miranda J, Gomez-Aldapa CA, Castro-Rosas J, Ramı́rez-
Process Preserv 42(1):e13415 Wong B, Vivar-Vera MA, Morales-Rosas I, Medrano-Roldan H,
Kannadhason S, Muthukumarappan K, Rosentrater KA (2009) Effects Delgado E (2014a) Effect of extrusion temperature, moisture
of ingredients and extrusion parameters on aquafeeds containing content and screw speed on the functional properties of
DDGS and tapioca starch. J Aquac Feed Sci Nutr 1(1):6–21 aquaculture balanced feed. Emir J Food Agric 26(8):659–671
Kaur L, Singh N, Kaur K, Singh B (2000) Effect of mustard oil and Rodrı́guez-Miranda J, Ramı́rez-Wong B, Vivar-Vera MA, Solı́s-Soto
process variables on extrusion behaviour of rice grits. J Food Sci A, Gómez-Aldapa CA, Castro-Rosas J, Medrano-Roldan H,
Technol 37(6):656–660 Delgado-Licon E (2014b) Efecto de la concentración de harina
Khoo HE, Prasad KN, Kong KW, Jiang Y, Ismail A (2011) de frijol (Phaseolus vulgaris L.), contenido de humedad y
Carotenoids and their isomers: color pigments in fruits and temperatura de extrusión sobre las propiedades funcionales de
vegetables. Molecules 16(2):1710–1738 alimentos acuı́colas. Rev Mex Ing Quı́m 13(3):649–663
Korkerd S, Wanlapa S, Puttanlek C, Uttapap D, Rungsardthong V Rojas-Garbanzo C, Pérez AM, Bustos-Carmona J, Vaillant F (2011)
(2016) Expansion and functional properties of extruded snacks Identification and quantification of carotenoids by HPLC-DAD
enriched with nutrition sources from food processing by- during the process of peach palm (Bactris gasipaes HBK) flour.
products. J Food Sci Technol 53(1):561–570 Food Res Int 44(7):2377–2384
Liu FX, Fu SF, Bi XF, Chen F, Liao XJ, Hu XS, Wu JH (2013) Ruiz-Armenta XA, Zazueta-Morales JDJ, Aguilar-Palazuelos E,
Physico-chemical and antioxidant properties of four mango Delgado-Nieblas CI, López-Diaz A, Camacho-Hernández IL,
(Mangifera indica L.) cultivars in China. Food Chem Martı́nez-Bustos F (2018) Effect of extrusion on the carotenoid
138(1):396–405 content, physical and sensory properties of snacks added with
Moraru CI, Kokini JL (2003) Nucleation and expansion during bagasse of naranjita fruit: optimization process. CyTA J Food
extrusion and microwave heating of cereal foods. Compr Rev 16(1):172–180
Food Sci Food Saf 2(4):147–165 Seth D, Rajamanickam G (2012) Development of extruded snacks
Navarro-Cortez RO, Hernández-Santos B, Gómez-Aldapa CA, Cas- using soy, sorghum, millet and rice blend: a response surface
tro-Rosas J, Herman-Lara E, Martı́nez-Sánchez CE, Juárez- methodology approach. Int J Food Sci Technol 47(7):1526–1531
Barrientos JM, Antonio-Cisneros CM, Rodrı́guez-Miranda J Sriburi P, Hill SE (2000) Extrusion of cassava starch with either
(2016) Development of extruded ready-to-eat snacks using variations in ascorbic acid concentration or pH. Int J Food Sci
pumpkin seed (Cucurbita pepo) and nixtamalized maize (Zea Technol 35(2):141–154
mays) flour blends. Rev Mex Ing Quim 15(2):409–422 Ying D, Cheng LJ, Chibracq G, Sanguansri L, Oiseth SK, Augustin
MA (2015) The format of b-carotene delivery affects its stability
during extrusion. LWT Food Sci Technol 60(1):1–7

123