Formulation and characterization of cookies prepared from the composite flour of germinated kidney bean, chickpea, and wheat

The objective of this study was to optimize the formulation of cookies from the composite flour of germinated kidney bean, chickpea, and wheat using response surface methodology (RSM). Snap force, spread ratio, and overall acceptability served as responses for the optimization of cookie formulation. Optimization and validation of central composite rotatable design (CCRD) of RSM concluded the feasibility of using 19.11 g of germinated kidney bean flour, 31.19 g of germinated chickpea flour, and 50.00 g of germinated wheat flour, per 100 g of flour composition for cookie preparation. Characterization of the novel formulated product was done by analyzing various attributes of flour and cookies. Optimized composite flour formulation (OCFF) exhibited appropriate functional and pasting characteristics required for cookie preparation. Optimized composite flour cookies (OCFCs) had a higher amount of protein (12.32 ± 0.11), fats (22.57 ± 0.23), and crude fiber (5.64 ± 0.02) content as compared with ungerminated wheat flour cookies (UWFCs). in vitro digestibility (carbohydrate and protein) was significantly higher in OCFCs owing to the utilization of geminated grain's flour. Amino acid content of germinated grains enhanced the total essential amino acids in OCFCs. Shelf life of the formulated product was acceptable for up to 90 days when stored at 25°C in aluminum‐laminated sealed bags.


| INTRODUCTION
Wheat flour is an ideal ingredient for various food product formulations; therefore, it is widely used for the production of bakery and confectionery products (Diana, Mirela, & Jianu, 2007). The suitability of wheat flour in bakery and confectionery products could be attributed to its gluten content. Out of few problems associated with the use of wheat, scarcity of wheat in certain regions and gluten-related allergies are common.
In the present research, supplementation with germinated legume flours was adopted to formulate a product with high nutrition and digestibility. Final formulation was characterized for its nutritional, in vitro digestibility, amino acid profile, texture profile, and shelf life.

| Raw material
Wheat (PBW-550), kidney bean (light speckled kidney beans), and chickpea GPF-2 (GF-89-36) were procured from certified seed agency. Selected grains were germinated as per the method described by Arora, Jood, Khetarpaul, and Goyal (2009). The sprouts were rinsed in water after germination, dried initially at 80 C for 15 min to arrest the enzyme activity, and then finally dried at 55-60 C to moisture content of 8.00 ± 1% (db). Dried sprouts and ungerminated wheat grains (PBW-550) were ground separately in lab grinder to fine powder in the form of flour and passed through 60 mesh sieves (US size 60 mesh = 250 μm).

| Experimental design 2.2.1 | Selection of independent variables and responses
Selected germinated legume flours other than germinated wheat flour were selected as independent variables. Response surface methodology (RSM) was used as optimization technique, which has been proven as appropriate tool in the optimization of various bakery products like sponge cakes (Chaiya, Pongsawatmanit, & Prinyawiwatkul, 2014), gluten-free bread (Sanchez, Osella, & Torre, 2004), cassava cake (Gan et al., 2007), and chocolate cake (Moscatto, Borsato, Bona, Oliveira, & Hauly, 2006). Responses like spread ratio, snap force, and overall acceptability were selected by putting a hypothesis that with the incorporation of germinated grains, these functions could be related to specific composition to fit regression equation, which describes the quality composition responses. To serve the purpose of optimization and to evaluate the effect of selected variables on the responses, central composite rotatable design (CCRD) was chosen. Preliminary studies have shown that replacing wheat more than 50% with germinated legumes was not feasible for production of cookies from composite flour containing germinated legumes. Based on trial runs, the upper and lower range for germinated kidney bean flour was selected between 15% and 25%, whereas the range for germinated chickpea flour was kept between 25% and 35%. CCRD was segregated into three different parts, that is, factorial design (two levels), axial points (outside core), and center points. RSM statistical software (design expert 10) distributed the design into 13 experiments with four factorial points, five replicates (center points), and four axial points. The CCRD variables and responses for different formulations are as shown in Table 1. Experiments were carried out in random order to minimize the effect of unexplained variability owing to extraneous factors (Myers & Montgomery, 2002).
Result of responses was analyzed using regression analysis by fitting a suitable model, that is, second-order polynomial equation: where β 0 is the value of response at the center points (0, 0); x i and x ij are the variables of design; β i , β ii , and β ij are regression coefficients; and n is the number of variables.

| Preparation of cookies
Cookies were prepared using a traditional method as described by Chauhan, Saxena, and Singh (2015). Ingredients used were composite flour (proportions were kept according to values given by experimental design), 100 g; grounded sugar, 40 g; sodium bicarbonate, 1.0 g; sodium chloride, 1.0 g; bakery shortening, 50 g; skim milk powder, 20 g; and water, 20 ml. A premix of flour, milk powder, and sodium bicarbonate was prepared separately. Dough prepared was formed into sheet of approximately 0.5-cm thickness, and then, a circular mold was used to cut the dough sheets. Baking was carried out in baking oven (conventional baking oven, Continental India) at 170 C for around 15 min. After 15 min, cookies were allowed to cool down at room temperature. Cookies were packed in aluminum-laminated sealed bags for further analysis.

| Spread ratio
Spread ratio was measured as ratio of diameter to thickness of the cookies (Zoulias, Piknis, & Oreopoulou, 2000). Average value of triplicate measurements was reported, and ratio was observed as Spread ratio = Diameter of cookie mm ð Þ Thickness of cookie mm ð Þ :

| Snap force
A three-point bend test was used to estimate the snap force required to break down the cookie (Chauhan et al., 2015). For this purpose, TA.
XT2i Stable Micro System texture profile analyzer (Stable microsystem, England) was used with three-point bend rig arrangements. Mode of measurement was in compression with units in newton (N). The measurements were carried out under following test conditions. Pretest speed was 1.5 mm/s, test speed was kept at 2.0 mm/s, posttest speeds were kept at 10.0 mm/s, distance of probe was maintained at 30 mm, automatic trigger type force was 20 g, and method settings were maintained by adjusting data acquisition rate at 200 pps. Graph was observed between force (N) and time (T). The peak force required to break down the cookies was reported as snap force (Sindhuja, Sudha, & Rahim, 2005). Results were observed in triplicates. foaming capacity, and emulsification capacity. Water absorption capacity was done by the method of Yamazaki (1953). Oil absorption capacity was estimated by the method of Lin, Humbert, and Sosulski (1974). Sedimentation value was estimated by calculating as the swelling power of flour by using ICC 116/1 standard method (Zeleny's method). Foaming capacity (%) was estimated as per the method described by Mizubuti, Junior, Souza, da Silva, and Ida (2000).

| Overall acceptability
Emulsification capacity was calculated by the method of Naczk, Diosady, and Rubin (1985). The emulsification capacity (EC) was expressed as ml of oil emulsified by 1.0 g of the sample.

| Pasting properties of flours
Pasting properties of different flours were estimated using Rapid Visco Analyzer (RVA Tecmaster, Perten, Australia), using standard testing profile-1. Appropriate sample around 5 g as described in T A B L E 1 Central composite design arrangement and responses for optimization of composite flour with germination legume substitution Experiment no.
Variables (coded) Variables (actual) Responses Note: x 1 , kidney bean; x 2 , chickpea; Y 1 , spread ratio; Y 2 , snap force; Y 3 , overall acceptability. manual was taken and mixed with 25 ml of water. The mixture was stirred for 10 s at 960 and then 160 rpm. Initially, temp-time combination of 50 C for 1 min was employed, and then, after equilibrium, temperature raised to 95 C for 5 min. After the procedure, cooling cycle was carried out by decreasing temperature to 50 C in 3 min.

| Proximate analysis and in vitro digestibility of cookies
Moisture content, crude fat content, crude fiber, and ash content of cookies prepared from optimized formulation were analyzed using standard AOAC methods (AOAC, 2005). Total carbohydrate content was estimated by anthrone method (Ludwig & Goldberg, 1956). Protein content of cookies was estimated by the method of Lowry, Rosebrough, Farr, and Randall (1951). To evaluate the protein digestibility, in vitro method suggested by Chavan, Chavan, and Kadam (1988) was used. This method is based on the digestion of sample by pepsin and pancreatic enzymes. Digested and nondigested sample were analyzed for protein by using method of Lowry et al. (1951). in vitro protein digestibility was calculated as proportion of protein digested using following relationship: In vitro protein digestibility % ð Þ= Total protein− Residual protein Total protein × 100: In vitro carbohydrate digestibility was analyzed using the modified method of Modi and Kulkarni (1976

| Color characteristics of cookies
Color value of different cookies was estimated by using Hunter Lab

| Amino acid content of cookies
Amino acid content of cookies was analyzed using the physiological kits of gas chromatography-flame ionization detection (Phenomenex, USA). Grounded samples were defatted and then hydrolyzed with concentrated HCl. Analysis was performed as instructed in the kit's manual. The GC column used was the ZB-AAA-GC column, which was provided in the kits, and standard analysis conditions were used, as described in the kit's manual. The results obtained were expressed as amino acid (g/16 g N). responses such as spread ratio, snap force, and overall acceptability were reported as follows:

| Optimization of responses
Germination enhanced the disruption of polysaccharides that led to F I G U R E 1 Effect of variables on (a) spread ratio, (b) snap force, and (c) overall acceptability more damaged starch and thus retained more water. Another reason could be attributed to higher protein content in legumes, which tended to absorb more water (Chauhan et al., 2015). Oil absorption capacity of OCFF was also observed higher than in UWF. The observed values for oil absorption capacity of OCFF were 1.38 ± 0.03 g/g ,whereas UWF was reported to have 1.16 ± 0.02 g/g of oil absorbed. The increase in the protein content and hydrophobic interaction of resultant protein molecules of blended flours could be attributed to the variation in oil absorption capacities (Chiemela, Olufemi, & Joseph, 2009). Sedimentation value of raw wheat flour was 60.01 ± 0.01 ml. With blending of different flours and germination, there was reduction in the total gluten fraction that led to a lowered value of sedimentation. Germination and blending of flour affected the foaming capacity of composite flour. Foaming capacity of OCFF (21.43 ± 0.03) was reported than that of UWF (12.76 ± 0.01). Foaming capacity is a function of surface properties of some proteins (Sibian, Saxena, & Riar, 2017). Emulsification capacity of OCFF was also reported to be higher (24.45 ± 0.04).
The variation in the emulsification capacities could be attributed to the protein-lipid interactions (Millward & Rivers, 1988;Sibian et al., 2017).
Rheology of blended doughs was compared by using RVA, to observe the pasting properties related to flours. RVA gave multiparameter values like peak viscosity, trough viscosity, breakdown, final viscosity, setback viscosity, and pasting temperature to observe the behavior of viscosity-related molecules (Table 3). Owing to germination, there was degradation of starch molecule, which leads to lower value of peak viscosity in OCFF. Peak viscosity (cP) of UWF was observed to be 928.00 ± 0.34, which was higher than that of OCFF (519.00 ± 0.47). Trough viscosity is the measure of extreme attainable viscosity at given temperature. Owing to degradation in starch, the trough viscosity of OCFF was reported to be lower  Table 4 indicates the difference in compositional characteristics of optimized composite flour cookies (OCFCs) and UWF cookies (UWFCs). Changes in the proximate composition of flours were previously reported in wheat and chickpea (Sibian et al., 2017;Sibian, Saxena, & Riar, 2016).

T A B L E 3
Functional and pasting properties of optimized flour formulation and ungerminated wheat flour

| Texture profile analysis and color characteristics
Texture profile of cookies was analyzed for the components like hardness, fracturability, and cohesiveness (Table 5) Composite flour cookies were slightly darker than wheat flour cookies (Table 5). Color is one of the quality characteristics, which reflects the cookie acceptability by the consumer. Color characteristics reflect the starch dextrinization, caramelization, and Maillard reaction, which are induced by cooking of product (Chung, Cho, & Lim, 2014). Lightness of OCFCs was observed lower as compared with that of UWFCs, which ranged from 52.37 ± 0.04 to 67.54 ± 0.07. The observations are in accordance with the results obtained by Bolarinwa, Lim, and Muhammad (2018)   T A B L E 5 Texture profile analysis (TPA) and color characteristics of cookies from optimized flour formulation and ungerminated wheat with the processing (germination and substitution) of flour, the formulated cookies were observed with good methionine and improved lysine content. The reason for the good nutritional attribute of legume-blended cookies could be attributed to the amino acid profile of legume flours (Imran et al., 2011). Table 6 represents the total amino acid profile of UWFCs and OCFCs. Total aromatic amino acid content (%) was found lower in OCFCs (8.27 ± 0.04) as compared with UWFCs (8.52 ± 0.06). Aromatic amino acid is more readily available for utilization in chemical changes during baking (Oupadissakoon & Young, 1984). Composite flour thus proved to be beneficial, owing to the supplementation of deficient essential amino acids (Ikumola, Otutu, & Oluniran, 2017).

| Storage studies
Multiple factors like moisture content, peroxide value, free fatty acids, and overall sensory score were observed for the storage studies of cookies (Table 7). Moisture content of cookies sample increased slightly during storage. With the increase in the moisture content, other parameters like peroxide value, free fatty acid, and total plate count increased to some extent. The increase in the moisture content could be attributed to the hygroscopic nature of cookies. The reason for legumes blended cookies to absorb high moisture could also be attributed to higher content of protein and crude fibers. Nagi, Kaur, Dar, and Sharma (2012) reported similar findings in cereal bran-fortified cookies at the end of 90 days. Peroxide value indicates the initial rancidity and degree to which lipid undergoes primary oxidation.
Owing to higher moisture and fat content of OCFCs, the peroxide value was also reported to be higher. Cereals contain a lower amount of unsaturated fats; therefore, a lower amount of oxidation took place in UWFCs (2.12 ± 0.01 to 5.16 ± 0.03). Peroxide value of OCFCs was higher even during the initial day of study (4.12 ± 0.04), which increased to 8.53 ± 0.03 mEq/kg on the 90th day. The results corroborated the studies by Divyashree (2014) in buckwheat-chia seed-fortified cookies, which occurred owing to auto oxidation during storage.
Ease of oxidation depends on the amount of unsaturated fats, storage conditions, and antioxidant activity of food (Akhter, Haider, Muzamil, Zia, & Salahuddin, 2016). Free fatty acid content was higher in OCFCs at initial day (0.41 ± 0.03) owing to its lipid and moisture content. During the storage, the increase in free fatty acid content of both OCFCs and UWFCs was observed. The values of free fatty acid (mg KOH/g) at the 90th day for UWFCs and OCFCs were observed as 0.68 ± 0.05 and 0.89 ± 0.02, respectively. A similar trend of free fatty acids was also reported in soy-fortified cookies (Singh, Singh, & Chauhan, 2000) and composite pasta (Yadav, Sharma, Chikara, Anand, & Bansal, 2014) with the increase in the storage time. Availability of moisture and other nutritive components promote the growth of microbes in cookies. A higher moisture content was observed in OCFCs, which increased during storage. In addition, composite flour cookies were reported to be high in nutrients, which facilitated the microbial growth. Germination converts the complex molecules into simpler molecules like sugar and amino acids, which might be readily available to the microorganism. Therefore, as a result, the microbial load of cookies increased. According to Gilbert (2000), the permissible limit for total aerobic count of ready-to-eat foods should be less than 10 4 -10 6 cfu/g. The microbial count of UWFCs was reported in the range of 1.5 × 10 1 to 1.03 × 10 2 cfu/g during the 90-day storage period.
The initial microbial count in OCFCs was 2.0 × 10 1 and increased up to 1.07 × 10 2 on the 90th day of storage. Although there was an increment in microbial count, the total plate count during the storage period of preservation did not exceed the permissible limit. Yusufu, Netala, and Opega (2016) also observed an increase in the total viable count (cfu/g) in composite flour cookies prepared from maize, yam bean, and plantain during storage. Organoleptic properties depend on various intrinsic and extrinsic factors (Mudgil, Barak, & Khatkar, 2017

| CONCLUSION
The two-way approach of formulating a novel product for the prepa- Textural properties of OCFCs were softer with slightly dark color. All age groups prefer these kinds of cookies with soft texture and high digestibility. The shelf life of OCFCs was found quite comparable with that of UWFCs, whereas sensory score has shown the acceptability of product on a 5-point scale. Shelf life could be extended to a longer period by using specialized packaging techniques.

T A B L E 7
Partial shelf life studies of cookies from optimized formulation and ungerminated wheat