The effect of modified (preheated) soybean concentrate powder on high protein biscuit

Protein is a key structural component in many foods. Soybeans are a plant-based protein source, that is used in food. Soy concentrate is generally high in nutrients, particularly protein, with an essential amino acid profile that is immense in plant products and closely resembles animal protein. In increasing the characteristics of processed soybean products, such as physical modification, namely the preheated process. The preheated process is a physical modification that requires heating the product at a specific temperature and time, with the aim of reaching the protein denaturation point for improving the functional characteristic of protein contents. Greater public interest in healthy diets, leading to contributes to the development of protein-enriched foods, one of which is biscuits. The aim of this study was to determine the effect of modified preheated soybean concentrate powder in the physical characteristic of high protein biscuits (11 - 15%). Texture characteristics, macrostructure, and sensory quality of biscuits were studied. In this study, soy protein was preheated for 20 minutes at a temperature of 80°C. The results showed that preheated treatment affected a decrease in texture quality compared to control biscuits, also it induced a decrease in porosity of biscuits, compared to control. The preheated treatment that reaches the denaturation point simulate improving the physical characteristic of biscuit, one of which is not give an excessive textural effect in the formulation of high-protein food. The best-preheated treatment was found in an 11% soy biscuit with a great texture. ImageJ analysis of macrostructural test results revealed that the porous biscuits were found in 11% preheated soy biscuits. According to the comparison sensory, triangle test, the panelists cannot defined a pair of control and 11% soy biscuit.


Introduction
Protein is an essential macronutrient [1].Besides that, proteins may have other non-nutritional uses in food; the most prevalent one is to provide the distinctive structure of the individual food.This structure also plays a significant role in determining the cooked protein's texture [2].Therefore, it plays an essential role in the critical structure of foods [1,2].In this decade, there has been much interest in plantbased nutrition for developing high-protein foods enriched with foreign protein sources, one of which is by adding soybean products [3,4].Soybean contains 40% protein with the highest quality and quantity of other plant-based protein sources, 20% cholesterol-free fat, including good polyunsaturated fatty acids like linoleic acid (50% of total fat content), vitamins such as vitamin A, B, and C and D, also minerals content like calcium (Ca) and phosphorus (P) [3,5,6].One of the soybean products used in developing baked goods is soy concentrate, a processed soybean product with high nutritional value and quality.1230 (2023) 012166 IOP Publishing doi:10.1088/1755-1315/1230/1/012166 2 Soy concentrate is generally high in nutrients, particularly protein.Soy protein contains much essential amino acid such as lysine, leucine, isoleucine, threonine, valine, phenylalanine, and tryptophan and lack contents of Sulphur such as methionine and cysteine [7,8].The primary components of soy protein that are largely stored (up to 80%) in globulins include glycinin (11S globulin) and β-conglycinin (7S globulin) [9,10].Protein in soybeans is generally stored in globulins.β-conglycinin, which has a trimeric glycoprotein structure, binds through hydrophobic interactions, that undergo complex association-dissociation depending on pH and ionic strength [9,11].Glycinin has a structure consisting of 5 subunits consisting two polypeptides, namely acidic (A) and basic (B) polypeptides connected by covalent, disulfide (SS) bonds [12,13].
In increasing the functional value and characteristics of processed soybean products, one of which is a physical modification, namely the preheated process.The preheated process is a physical modification that requires heating the product at a specific temperature and time, with the aim of reaching the protein denaturation point in the form of aggregate, for improving the functional characteristic of protein contents [14].The heating process in soy protein induces irreversible unfolding and disrupts the quaternary structures of β-conglycinin (7S) and glycinin (11S).The disulfide bonds between the broken chains can initiate disulfide (S-S)-sulfhydryl (S-H) exchange reactions that play a role in protein crosslinking, forming aggregates.The exposure of hydrophobic groups results in hydrophobic interactions between protein molecules that can promote protein aggregation and cross-linking in the process of food structure formation [15].The transition of denatured protein formation in soy protein occurs at ±60℃ [12].β-conglycinin denatures at 75 ℃, while glycinin protein denatures at 90 ℃.The denaturation mechanism is controlled by subunit composition, interactions between subunits, ionic strength, and reducing agents that will cause aggregation and polymerization reactions [16].The aggregation process by heating the two main protein components in soybeans results in the opening of the hydrophobic groups on β-conglycinin and interaction (via disulfide bonds) with glycinin, leaving the N-terminal region with hydrophilic glycans exposed to the outside, so the aggregation process is stopped and soluble aggregates are formed [13].The heating of soy protein components of soy concentrate, namely βconglycinin (7S) and glycinin (11S), has different aggregation formation characteristics.According to J. Guo et al., [12] heating soy protein clusters simultaneously shows the interaction between βconglycinin and glycine-the presence of β-conglycinin groups covering the surface of glycine aggregates, becoming non-hydrophobic residues [12].The following is a scheme of the formation of soy protein aggregates of βconglycinin and glycinin, according to J. Guo et al [12].In order to create bakery products enriched with foreign protein sources, there has been much interest in plant-based nutrition, one of which is the development of biscuits [3].Based on United States Agency for International Development (USAID) and WFP/WHO (2011), the protein content in high-energy biscuits is 10 g or 10 -15% per 100 g protein packaged products [17].Based on Indonesia's Recommended Dietary Allowance (RDA), consuming high-protein biscuits can fulfill 10 -15% of daily protein needs [18].Generally, the higher addition of protein concentration in biscuits increases the hardness of the biscuit texture compared to general biscuits [19].The addition of protein concentration in high-protein biscuits with the addition of soybean concentrate (native) results in an increase in texture hardness by the thermal process (baking), with an increase in the concentration of protein added resulting in changes in the characteristics of the biscuit structure [5,20].The aim of this study was to determine the effect of modified preheated or denatured soybean concentrate powder in the physical characteristic of high protein biscuits (11 -15%).Texture characteristics, image analysis from macrostructure, and sensory quality of control and treatment biscuits (non-preheated and preheated) were studied.

Materials and Methods
The raw materials used for making 11 -15% high protein biscuits were 12% of denatured soy concentrate powder mixed with amount of water, all purposed flour, sweet potato flour, skim milk powder, egg yolk, 0.2 -0.35% mineral powder and syrup (magnesium and zinc supplement, ferro syrup), sugar, margarine, vanilla powder, and leavening agent such as baking powder and baking soda.

Preparation of denatured soy concentrate powder
Denatured soy concentrate powder was prepared by calculating, and diluting 12% soy protein content of soy concentrate powder in amount of water for heating in waterbath by holding the temperature at 80℃ for 20 minutes.As the concentration of high protein biscuit increased, the amount of denatured soy protein increased in dough.The ratio of the addition of soy concentrate powder and water is shown in the data table below:

Formulation of biscuit
The amount of denatured soy protein was varied according to percentage of protein content in high protein biscuits (11%, 13% and 15%), meanwhile the other ingredients were fixed.In this study, the formulation of biscuits were divided into 2 groups experimental sample that is undenatured (native) soy concentrate (NPH) and denatured soy concentrate (PH).Formulation of biscuits used in this study showed as follow:

Baking procedure
The biscuit was baked using the creaming method according to Misra and Tiwari [21], Zaker Md and TR [22], with some modification process.Ingredient was used for preparation biscuit according to formula above.Sieve dry ingredients and mixed to obtain a uniform blend.Fat or margarine and sugar were mixed thoroughly to obtain shortening cream.The shortening mix was mixed with mixer, further for 3 minutes at 860 rpm.Both dry and wet mixture were mixed separately.Then mixed dry ingredients was added to shortening cream, and mixed for 1 minute at 860 rpm.The mixture of soy protein (undenatured and denatured) and water was added to the dough bowl and mixed thoroughly.When the dough was ready, it was kept for 10 -15 minutes as it is and then used for molding with customized frame of mould cookies (thickness : 3 cm ; height : 5 cm) on the baking tray.The shape dough of biscuit was baked at a temperature of 110 -135℃ for 20 -30 minutes.The baked biscuits were allowed to cool at room temperature, then packaged in an airtight container.

Texture (hardness) of biscuit
Texture analysis of biscuit was conducted to determine hardness.Hardness of biscuits was determined using TA.XTExpress (Stable Micro Systems) equipped with a 5 kg loading cell.Exponent Lite Express software was used to measure and acquire data.The procedure for texture analysis was according to the method of Di Cairano et al. [23], Filipčev et al. [24], with a modification.Biscuit hardness was determined using a 6 mm cylindrical probe (P/6).The apparatus configuration and setting were: pre-test speed 1 mm/s, test speed 4 mm/s and post-test speed 10 mm/s with travel distance 10 mm.The sample was placed on the two holders of the adapter and the cutting probe was lowered until it met the sample.The probe acts as a third point of contact, exerting increasing pressure until the biscuit structure breaks.Measurements of the hardness of biscuits were made at deform biscuits, when the maximum peak force shown and form the area under the curve.That maximum peak force was used to calculate the hardness value.The measurement of each sample was done in three replication, each replication conducting three biscuit sample for each formulation were analyzed.

Image analysis of biscuit
Image analysis of the macrostructure from biscuits is conducted to determine biscuit porosity based on descriptive and fractal dimension analyses.The image sample was taken with a camera, where the samples were placed inside a black box, adjusting the light and angle as desired.A descriptive analysis of biscuits by describing the visible image that showed the porous structure of biscuits from each formulation of biscuits and the form shown by 1200x120 pixels.Fractal analysis of biscuit pores was determined using ImageJ Analysis software to measure and acquire dimension fractal data from each sample.Images were transformed into 8-bit grey scale binary images with the same pixel size (1200 x 120 pixels) and were threshold.The procedure of dimension fractal contains two significant steps, thresholding the image and determining the fractal dimension value from box counting.The procedure for biscuits macrostructure analysis was conducted according to Kuhn et al. [25] with some modifications.The thresholding of the image was using auto-threshold settings by the Otsu method in ImageJ software.Thresholding by Otsu method is shown where the thresholding or image segmentation process is based on a discriminant analysis to find the maximum separation point of two classes (highest and lowest) [26,27].Following the steps, the fractal dimension (Df) values of the biscuit (pores) structure are calculated using the box-counting method.The fractal dimension found, based on the calculation of the scaling rule shown below : N ℇ stand for number of boxes at a certain scale that contains part of the image (foreground pixels), while ℇ is the corresponding scale (in pixels), and D stand for the fractal dimension on a two-dimensional (2D) form.Since image analysis determines Df value in a two-dimensional space, it is necessary to add an extra dimension for calculated D value that represent the three-dimensional features of the image structure, according to the equation below.
Df = D + 1 [25] (2) In this study, there were seven samples from each biscuit treatment, and each sample consisted of at least five macroscopic images recorded.

Sensory evaluation
To assess the differences between control and soy treatment biscuits, a comparison sensory evaluation, namely a triangle test, was conducted by an untrained panelist with no replication.A panel of fifteen judges was recruited from staff and students at the Food Industrial Technology Department, Faculty of Agro-industrial Technology, University of Padjadjaran.This number of panelists is considered adequate for indicating an odd sample from three samples in each set (different from the other two), in which testing samples consist of six pair serving or six sets of control biscuit and treatment (undenatured and denatured) biscuit (11%, 13%, and 15%).Tasters asses all three samples, then pick the sample with different characteristics (color, taste, texture, or hardness) from each set sample or the odd one out.The significance of the triangle test requires a number of the correct answer from the total response was 10% (α = 0,10).The minimum correct answer for fifteen panelists was eight out of fifteen (8/15) from the total responses according to the critical value number of the triangle test.

Statistical analysis
In this study, all experiments were repeated at least three times, and the data were expressed as mean.The experimental design used was descriptive descriptive explanatory research, then the calculated results of the experiments were subjected to statistical analysis using paired t-test, to determine significant difference between control sample and experimental sample that had a treatment (preheated) (undenatured and denatured) using IBM SPSS software (version 25.0, SPSS Institute Inc., Chicago, IL, USA).Significant difference was defined as p < 0.05.

Texture (hardness) of biscuit
The texture property of biscuits is specifically related to their eating quality.In terms of breaking strength, the peak force needed to snap biscuits is a measure of hardness, one of the many textural parameters for biscuits, that is regarded as a crucial characteristic [28].The hardness of native (undenatured) and preheated (denatured) soy protein biscuits was determined by force (gram) required in compressing the sample calculated from a force (g)time (s) graph.The effect of adding native and modified (preheated) soy concentrate powder on the hardness properties of the cookies is presented in   formulations significantly changed the biscuits textural (hardness) quality.According to the peak force graph, the control biscuits showed the lowest peak force (818.49g) compared to the undenatured and denatured soy concentrate powder biscuits.The results from group A showed that the higher incorporation of the undenatured soy concentrate powder decreased the peak force of the soy biscuits from 1701.77 g (NPH 11) to 1254.69 g (NPH 13), then increased to 1472.34 g (NPH 15). the results from group B showed that the higher incorporation of the denatured soy concentrate powder decreased the peak force of the soy biscuits from 1512.17 g (PH 11) to 1378.86 g (PH 13), then increased to 1750.16 g (PH 15).The difference in the speed of time in reaching the peak force in biscuit hardness characteristics is influenced by the composition of the biscuit and the interaction between the ingredients, as well as the quality of the interaction between the ingredients.In the formulation of high-protein biscuits based on commercial soy concentrate, increasing the protein concentration in the biscuit formulation resulted in a decrease in the hardness of the biscuit texture.Increasing the amount of denatured soy concentrate with preheated process in 11%, 13%, and 15% biscuit formulations led to an increase in biscuit hardness as the protein concentration in the biscuits increased.The result of decreasing the highest point of biscuit-breaking strength is opposed to the results of previous research by Yang et al., where the addition of black soybean flour increased the hardness (N) of biscuit texture as the amount of protein in the biscuit increased [29].Similar results were found by Ghoshal & Kaushik, where the addition of 15-25% soymeal flour concentration increased the texture hardness of biscuits compared to the control biscuits [5].The increase in the peak breaking point or hardness of the biscuits is due to the increase in protein content in the biscuits, where the addition of soy flour results in a decrease in the gluten content of the biscuits, causing a reduction in the formation of gluten networks that provide a crunchy texture [30,31].
Preheated treatment of soy protein changes the structure of the native form, which is tightly folded into a denatured structure [32].Under native conditions, soy protein has a greater water-holding capacity (more hydrophilic site) than denatured protein (increased hydrophobic site), so the denaturation process on soy protein is expected to minimize the effect of extreme changes on biscuits.The increase in hardness as soy concentration increases in biscuits is due to the interaction of protein and starch through hydrogen bonding during the dough development [28].In Xie et al., lower water absorption is associated with higher texture hardness [33].The decrease in texture hardness of group A biscuit was influenced by increased moisture content due to the baking process at a less-than-optimal temperature and time [34].Increasing the concentration of soybean causes a decrease in the starch content of the dough.It increases soybean protein's water absorption ability, reducing the formation of biscuit structure [35].The increase in the breaking force peak point of the biscuit compression hardness graph generally shows a greater compressive force with increasing time.The graph of group B results is by the above statement.The NPH 13 sample of group A showed that the time required to reach the peak of maximum force took longer with the lowest level of hardness.This indicates a difference in the internal structure and hardness of the biscuits produced [34].As shown on a diagram of the average biscuits hardness (Figure 4), the undenatured and denatured soy protein biscuits significantly changed the hardness quality.The results from both groups showed that the incorporation of the soy concentrate powder resulting a higher hardness level than the control biscuit (1256.34g).From the results of the undenatured soy powder concentrate (Group A), it can be seen that the hardness of the biscuits decreased with the higher addition level of undenatured (native) soy concentrate powder, from 2498.83 g to 1551.71 g.On the otherwise, it can be seen that the hardness of the denatured soy biscuits (Group B) increased with the higher addition level of denatured soy concentrate powder, from 2155.08 g to 2226.15 g, except for the hardness decrease of PH 13 sample (1859.66g).
The increase in biscuit texture hardness is influenced by the increase in protein content and the interaction of the protein with sugar during dough handling and baking [19].The interaction of protein and starch during the dough handling process until the baking process has a role in increasing the hardness of the biscuit texture.The results of the research on the hardness texture of group A biscuits are in line with the results of Tang & Liu, where the addition of soy protein resulted in a decrease in biscuit hardness lower than the control, which affected the changes in the physical characteristics of the biscuits [3].Contrary to the above results, group B had similar results obtained by the results of Yang et al.'s research, where the hardness of the biscuit texture increased the hardness value of the biscuits in the black soybean flour-blend base biscuit compared to the control biscuit of 6.81 N (wheat flour base) [29].These results also align with the study by Olakanmi et

Hardness (Average) of Biscuits from Group A and Group B
Group A Group B wheat protein with legume products and legumes affected the quality characteristics of the bread produced, resulted in a decrease in bread volume along with an increase in the hardness of the bread crumb texture during an increase in the level of soy protein substitution [36].Increasing protein substitution in high bread will slow gluten formation, decreasing the gas retention ability during the proofing to the baking process.Making biscuits with high-protein ingredients contributes to the strong binding of protein and starch through hydrogen bonding during dough development and baking [33].
The increased protein content of biscuits results in increased protein interactions with other components, such as protein to protein, protein to carbohydrate, sugar, and other component interactions during the dough development and baking process, which is also associated with increased biscuit texture hardness [19,28,37,38].
Increasing the addition of soybean concentrate increases the free sulfhydryl (SH-free) and disulfide (S-S) content in the dough.Generally, disulfide bonds (S-S) intrinsically in gliadin and glutenin will be disrupted and recombined during gluten formation.The free sulfhydryl (S-H) combine to form disulfide bonds (S-S), contributing to the gluten network, interacting with wheat proteins to produce weak dough due to the inhibition of gluten network formation [39,40].The disruption of gluten formation in the presence of soy protein is related to the degree of denaturation of soy protein.Soy protein denatured during the dough handling process can form gluten due to the breakdown of molecular protein chains, increasing its interaction with wheat proteins [39].Preliminary heat treatment of soy protein can improve the baking properties of wheat flour-composite (soy) based products [40].The heating process of soy protein in biscuit making with the preheating process results in irreversible denaturation of soy protein, with exposure of hydrophobic and sulfhydryl groups that enhance the formation of protein aggregates, resulting in gelation when protein concentration is high [41].The denaturation of soy protein in the preheated process is expected to reduce the interaction of soy protein with the dough during the dough handling process (mixing-baking), to produce a reduction in the interaction of soy-wheat protein forming the biscuit matrix, to produce a decrease in the texture hardness of soy biscuits compared to native soy.
In the baking process, there is an interaction between starch and protein.During the baking process, protein interaction and protein interaction with starch (gelatinization) are found, affecting the product's structural properties.Gelatinized starch and protein can decompose, rearrange, hydrolyze, and physically crosslink, improving the texture of the resulting product [42].In Ma et al., protein denaturation leads to aggregation during the preheated treatment process.The second aggregation inhibition of the particles occurs because the less reactive groups are allowed to crosslink between the protein particles due to the denaturation and aggregation process in the preheated treatment [41], according to the results of J. Wang et al. research in the reheating process, particles that are stable to heating can inhibit aggregation due to more significant conformational changes with a higher degree of denaturation and have a more compact aggregate structure.On the other hand, aggregation of soy protein at 85℃ results in partial denaturation of soy protein due to the lower preheated temperature than glycinin (11S) denaturation temperature.
Based on the results of the hardness level of the biscuits produced, group B biscuits have an increased hardness level as soy protein is added to the biscuit formulation.The results between groups A and B showed that at a concentration of 11%, preheated biscuits produced a lower texture than biscuits with the addition of native soy protein.In groups 13 and 15%, preheated biscuits had a higher texture than biscuits with the addition of native soy protein.It is due to the interaction between soy protein and wheat protein during dough handling, where there is undenatured soy protein.The proteins not denatured in the preheated process further interact during baking.In conclusion, a significant difference (p<0.05) was found between the mean hardness value of control and undenatured or denatured soy concentrate biscuits (Table 4.).The enhancement of the biscuit's hardness is significantly more massive with the incorporation of denatured soy concentrate at 15% level biscuits formulation with the increased hardness average value was calculated, therefor further biscuits formulation is required to resulting not such excessive changes in biscuits product.

Image analysis of biscuit
The macrostructure of native (undenatured) and preheated (denatured) soy protein biscuits was determined by descriptive analysis and Fractal dimension (DF value) determination.A descriptive analysis of the biscuit was accomplished by describing the image of the biscuit's pores structure that proof by the total area of porosity from each samples.Determining fractal dimension value is required to measure and acquire dimension fractal data from each sample.In order to understand the effect of adding native and modified soy concentrate powder in high protein biscuits formulation, the images of the biscuits macrostructure with 1200 x 120 pixels were used.The effect of adding native and modified (preheated) soy concentrate powder on the macrostructure by descriptive analysis properties of the cookies is presented in Figure 4.The average of porosity area and porosity size from biscuit are presented in Figure 5 and 6.The thresholding image of the biscuit pores structure are presented in Figure 7.The average fractal dimension values from each treatment sample are presented in Figure 8.

Porosity Area of Group A and Group B Biscuits
Group A Group B The picture above (Figure 5.) shows that the control biscuits have a big porous with some more pores compared to treatment (undenatured and denatured soy protein) biscuits.The picture also showed that the incorporation of the soy concentrate powder decreased the number of pores and size.From the results of the undenatured soy powder concentrate (Group A), it can be seen that the NPH 13 and NPH 15 biscuits have tiny pores that construct all over the biscuits samples, with the NPH 11 biscuits having a similar structure of pores according to the picture above.This is in agreement with the calculation of the total area from biscuits in Figure 5, where there is an increase in the total area of the pore structure in the 11% (35224.80pixels) and 15% (35142.80pixels) NPH biscuits compared to the control biscuits (34820.53pixels), and there is a decrease in the total area in the 13% NPH biscuits (34611.87).The variations in the size of the porosity formed in the biscuits are a significant decrease from the 11%, 13% to 15% NPH samples (242.43 to 200.21, and increased to 204.27 pixels) compared to the control biscuits (443.20 pixels), as shown at the figure 7.
The same results showed in denatured soy concentrate (Group B) that PH 15 biscuits have a small size of pores and decreased the number of pores that construct a denser biscuits samples, with the PH 11 and PH 13 biscuits having similar pores structure with the control sample.This is in agreement with the calculation of the total area from biscuits in Figure 6, where there is a decrease in the total area of

Average Size of Biscuits Porosity
Group A Group B the pore structure as the increase of the soy protein contents in biscuits PH 11% (32557.53pixels), slightly increased in PH 13% biscuits (33916.27pixels) and decrease at PH 15% biscuits (29028.67pixels).The variations in porosity size formed in the biscuits (Figure 7) show a slight decrease in pore size in 11% of biscuits (417.43 pixels) compared to the control.In comparison, 13% and 15% of biscuits have a significant decrease in pore size (229.89and 242.94 pixels) compared to the control biscuits (443.20 pixels).Based on the above figures, it can be concluded that NPH 11, 13, and 15 biscuits have a more porous structure than PH 11, PH 13, and PH 15 biscuits, with a significant decrease in the average size of porosity in the biscuits, compared to the control, which has the highest area structure and porosity size.
The formation of pore structure in soybean addition biscuits affects the shape and variation of pore size with an even distribution of pores in the cross-section of the biscuit.The pore structure of biscuits represents a product quality attribute produced by air cell formation and expansion, as well as volume development in the baking process.Large-sized pores form due to the increased coalescence of air bubbles before the gelatinization of starch in the baking process.The greater air-holding capacity in the dough system leads to a slower air rise in the early baking process, with a lower gas loss rate and a larger specific volume of the baked product [43].Based on result above, the decrease in the pore structure of biscuits is influenced by the increase in the addition of soy concentrate, so biscuits with higher protein form a pore structure with a smaller size and is dispersed.This study's results align with Nicole et al., where the addition of soy protein isolate (SPI) to biscuits resulted in a denser biscuit structure [44].This occurs due to the inhibition of gluten network formation by SPI, by disrupting the gluten structure and limiting the availability of water in gluten network formation.In dough mixing, soy proteins can aggregate and interact with wheat proteins through disulfide bonds [44].It is confirmed by the pore structure image and the quantification of the total area above, where the pore structure formed in the treated biscuits has less pore structure or has small pores size compared to the control.In addition, soy protein can affect pore uniformity, where the resulting pores are not homogeneous and tend to separate with aggregated parts.A decrease in pore structure in the presence of denatured soy protein can form aggregates and inhibit the gluten network, causing a weakening of the gluten network so that it cannot hold air in the heating or baking process.In addition, other factors, such as the rate of water evaporation, can affect the decrease in pore structure in biscuits [44][45][46].In particular, macroscopic variations in pore structure may occur due to variations in moisture content across products, which may indicate differences in the degree of starch gelatinization [46].
In conclusion, there was no significant difference (p>0.05) between the mean total porosity area values of control and treated soy concentrate biscuits, based on the paired t-test results (Table 5).The increase in the pore structure of the biscuits decreased significantly with the addition of undenatured and denatured soy concentrate at the 13% and 15% soy protein addition levels, based on the total area and average area of porosity.Descriptive analysis of the macrostructure of the biscuit samples showed that biscuits containing 11% protein, mainly distributed by soy protein, may not have exaggerated changes in the biscuit samples, as proven by the total porosity area of the biscuits and the average size distribution of the pore structure.

Dimension fractal of biscuits.
In fractal dimension analysis, two main steps are carried out: the thresholding process and the calculation of the DF value by the Box Counting Method.In this study, Otsu thresholding method was applied.The resulting thresholding image is separated into two parts: the pore structure marked with black or image background and the biscuit structure marked with white.Otsu method can evaluate the image's threshold value well within each gray level formed in the image.This approach is also widely used in pore structure segmentation and can classify pore structures, grains, and high-density inclusions [47].The following image is a digital (2D) picture of the results of the biscuit Thresholding process through the Otsu method (1200 x 120 pixels) : The threshold image above (Figure 8) is followed by calculating the DF value using the Box Counting Method.The Box Counting Method is a method of calculating the DF value based on 2D images of the pore structure from the slope of the least squares linear regression on a logarithmic plot (number of boxes) per-log (box size/length) [46].The average value of DF in each control and treatment biscuit sample (group A and group B) was obtained as follows: Figure 9. Fractal dimension (value) of biscuits pores (size = 1200 x 120 pixels) data from group a (non-preheated/NPH) and group b (preheated/PH) soy concentrate on high protein biscuits.Based on a diagram of the average biscuits fractal dimension (Df) value (Figure 9.), the undenatured and denatured soy protein biscuits had slightly significant changed the fractal dimension value.It is shown, the DF values varies in a close range from each samples (2.50 -2.58).The results from both groups showed, the incorporation of the soy concentrate powder resulting a decreased of DF values than the control biscuit (2.58).From the results of the undenatured soy powder concentrate (Group A), it can be seen that the value of fractal dimension increased insignificantly with the higher addition level of undenatured (native) soy concentrate powder, from 2.55 to 2.56.On the other hands, result showed at the group B that the DF value of the denatured soy biscuits (Group B) decreased with the higher addition level of denatured (preheated) soy concentrate powder (except increased in 11% to 13%), from 2.54 to 2.50.
The macrostructure of the biscuit porosity in this research also described and measured experimentally using the fractal concept.In this study, the fractal concepts used in characterizing the morphological structure of biscuits, in an irregular pore structure [48].Based on the DF value method, namely Fractal Brownian Motion (FBM) method, the particle network in the formation of aggregates is divided into two cluster models, i.e.Diffusion-Limited Cluster Aggregation (DLCA) and Reaction-Limited Cluster Aggregation (RLCA).To explain DLCA and RLCA fractal aggregation clusters by quantifying the aggregated particle structure using DF calculations [48,49].Aggregation in DLCA clusters is rapid, with particles of the structural network bonding at first contact and inter-particle collisions limited by Brownian motion.The aggregation in DLCA clusters in the three-dimensional model is shown to vary with DF values of 1.7 -1.8.RLCA clusters generally proceed more slowly due to electrostatic repulsion between particles approaching each other.Aggregation in RLCA clusters is presented in three-dimensional form, with DF values varying from 1.9 -2.1 or 2.0 -2.2 [49][50][51].
Based on the explanation above, the DF value that represents the porosity formed in the biscuits shows that RLCA aggregation clusters are formed in the biscuits, with the results of the DF value greater than 2.5.The research results generated fractal dimension values in measuring pore irregularity or coarseness throughout the sample structure.In Rahimi et al., an increase in higher DF values resulted in pores with more random patterns with irregular geometry shapes [52].Caused by the diffusion of carbon dioxide (CO2) during the handling process, the expansion resulted from the influence of the thermal process, resulting in thinning of the gas cell wall from the pressure gradients, which depends on its thickness.The presence of gluten from wheat flour in the dough affects the expansion of the gas cells, which results in a different pore geometry compared to the reduced gluten content in the dough.This is in line with the results of the research, where the control biscuits had the highest DF value.Similar findings were found by Rahimi et al., where baked products with wheat flour had the highest value, followed by chickpea, millet, tiff, and amaranth [52].
It was explained by Kouhsari et al. and Rathnayake et al. that higher DF values represent more complex structural images or crumb images with coarser gray levels, while lower DF values have simpler and finer structural [46,53].In the results of this study, it was found that the DF value of each treatment was not significantly different, so the pore structure formed tended to be complex, with a decrease in the complexity of the pore structure as the protein concentration of the biscuits increased.
This study also relates to the texture characteristics produced, where in this study, the increase in protein concentration increased the level of hardness of the biscuits.It follows that there is a decrease in the pore structure formed in the biscuits.The texture of biscuits is influenced by the cell structure, with smooth, thin-walled cells having a uniform size, producing a softer texture compared to rough, open, thick-walled cell structures.
In conclusion, a significant difference (p<0.05) was found between the mean fractal dimension value of control and 15% PH denatured soy concentrate biscuits, according to paired t-test results (Table 6).The enhancement of the biscuit's fractal dimension value is slightly increased for group A and decreased for group B, with the incorporation of denatured soy concentrate at 11% to 15% level biscuits formulation.The addition of soy protein in the biscuit formulation did not significantly change the biscuit's structure.The fractal dimension value average was calculated, and the result revealed that suitable high protein biscuits with less excessive change characteristics are 11% and 13% level soy protein supplement.Further biscuits formulation is required to resulting not such excessive changes to resulting high protein biscuits using soy concentrate powder.

Sensory evaluation
The biscuit prepared by incorporation of undenatured soy concentrate and modified (denatured) soy concentrate for different concentration of high protein biscuits (11%, 13% and 15%) were subjected to the sensory evaluation by fifteen panelist using triangle test, with six paired set of biscuits.The data was analyzed, and the compiled data was presented in figure 10.Each treatment soy (undenatured and denatured) was evaluated against control biscuit to asses different sample characteristic in paired sample set.Different characteristic of biscuits choose according to their color, taste and texture or hardness of biscuits samples.Based on the sensory attributes displayed in figure 10 showed that the biscuits were affected by the addition of soy concentrate powder in some parameters.The average scores indicate an increase in taste (flavor) and texture (hardness) with increasing levels of soy concentrate powder incorporated in the formulation of high-protein biscuits with the addition of soy protein.According to the data above, the parameters of color did not exhibit predictable patterns scores with variation treatment, as evidenced by the panelist score (correct answer <8) indicating that panelist cannot distinguish between control biscuit and treatment biscuits.Furthermore, the taste and texture parameters shown that panelist could still point differences between control biscuit and treatment biscuit with amounts of soy concentrate powder added to biscuits at NPH 11% -NPH 15% and PH 15% samples.
Color or appearance is an important quality indicator of fine food that influence consumer acceptance, perceived by the human sense [54].Based on the panelist's perception, the score of the color showed that the addition of soy concentrate, non-preheated (undenatured), and modified or preheated (denatured) cannot differ from the control biscuit, as evidenced by the panelist score (correct answer <8) so that explains the addition of soy concentrate was not affecting the color characteristic of the biscuit, except for the PH 15% biscuit paired samples.Hence, adding soy protein to the biscuit formulation could have well influenced the biscuit's quality.Based on the report from the panelist's, the PH 15% biscuit paired has darker color of biscuit, compared to control.This happened caused of the Maillard reaction which occurs due to several factors that affect the heating process, such as baking, including temperature, humidity, and baking time.The Maillard reaction occurs in the browning nonenzymatic reaction in the baking process between sugar molecules and undenatured amino acids, such as lysine from soy protein [55][56][57].
Taste is the primary determinant of product's acceptability, and has the greatest impact on a product's commercial success.Food identification, acceptance and appreciation are all assisted by taste [54,58].Based on the panelist's perception, the score of the taste showed that the addition of soy concentrate, non-preheated (undenatured), and modified or preheated (denatured) can differ from the control biscuit, as evidenced by the panelist score (correct answer >8), so that explains the addition of soy concentrate affecting the taste characteristic of the biscuit, except for the PH 11% biscuit paired samples.Based on the report from the panelists, all paired samples (11%, 13%, 15%) biscuit has beany taste from soya bean, except PH 11% paired sample.This indicated that the taste of the beany flavor in the biscuit increased with the increased level of adding soy concentrate, this result agree with the work of Dhingra and Sudesh [59], and Mishra and Ramesh [55].The variations in volatiles produced during baking process, was related to the differences in flavor of soy biscuit compared to control biscuit [56].Soybean has fiber, flavonoids, and bioactive peptides contents [60].While soy protein products are made, lipoxygenase enzymes oxidize unsaturated fatty acids, which cause an unpleasant beany taste.This kind of off-flavor is mainly caused by volatile substances like n-hexanal and n-hexanol, which result from the oxidation of the lipid process [61,62].The beany flavor from soybean also happens during the irreversible and reversible interaction between soy protein and volatile content during soybean processing [63].In this study, the PH 11% biscuits show unaffected biscuit taste made from denatured soy concentrate, which indicated the objectional of beany flavor due to elimination while preheating processing.
Touch, mouthfeel, sight, and hearing affect humans perceived texture.The texture is a prerequisite in the acceptance of foods [54].Texture also represents the product's internal structure regarding how it responds to stress.Mechanical properties (such as hardness, adhesiveness, cohesiveness, gumminess, springiness, Etc.) are measured by a kinesthetic sense in the tongue, jaw or lips, fingers, and hands [64].Based on the panelist's perception, the score of the biscuit texture (hardness) showed the addition of soy concentrate, non-preheated (undenatured), and modified or preheated (denatured) can differ from the control biscuit, as evidenced by the panelist score (correct answer >8) so that explains the addition of soy concentrate affecting to increased hardness characteristic of the biscuit, except for the PH 11% and 13% biscuit paired samples.According to the panelist report, all paired samples (11%, 13%, 15%) of biscuit have a more rigid texture with the increased level of the addition of soy concentrate, except the PH 11%, 13% paired sample, which has slightly difference hardness of the biscuit.Hardness, as well as sensory hardness trial, were the most noticeable differences between control biscuits and increased with soy concentrate addition.The sensory panelists cannot defined the control biscuits with either PH 11%, 13% soy concentrate.These results imply that the soy biscuit's acceptability, whether it had 11%, 13% soy concentrate was comparable to the control biscuits.The texture data above (Figure .3) shows that adding soy protein with the increased soy protein desired in the higher protein biscuit leads to a decrease in the texture quality of biscuits.Higher protein contents from soy concentrate, and thus the interaction of protein-protein sources, protein-carbohydrates or starch and other components that interact during dough development and baking, may attributed to the increase in the hardness of biscuits [28,38].The result agree according to Kulkarni et al., [65] Cappa et al., [30] and Mishra & Chandra [55].
Following the sensory comparison triangle test, according to the results, the panelists could not define differences biscuits characteristics (color, taste, texture) between pairs of 11% PH and control biscuits.So appropriate high protein biscuits with soy protein concentrate discover in the denatured 11% level soy protein supplement, with less excessive characteristic change than the control biscuits.Furthermore, it is necessary to reformulate the biscuits formulation to produce biscuits with higher protein contents with similar physical characteristics (color, taste, texture) found in available biscuits.

Conclusion
The effect of adding undenatured and denatured soy concentrate on the formulation of high-protein biscuits compared to the control biscuit in this work.The hardness of biscuits increased by the increased on the addition of soy concentrated under two circumstances toward the higher concentration of biscuits.The preheated treatment of the soy concentrate was not optimal due to the partially changed protein structure.Another protein structure still interacted with other protein sources in the heat process, leading to a hardener biscuit.By the experimental descriptive method, the macrostructure analysis showed that increasing the soy protein addition to the biscuits decreased the pore's structure, leading to dense biscuits.According to the Df value of biscuit pores, adding soy protein to the biscuit decreased the Df value.The non-preheated and preheated soy concentrate (11 -15%) used in the biscuit formulation did not significantly affect the biscuit color.However, they significantly affected the taste and texture of the biscuit since panelists found differences between the pair of biscuits-the best-preheated treatment found in 11% soy resulting porous structure with great texture.According to the triangle test, the panelists could not define the pair of biscuits at 11% PH and control.Generally, from this study, further research of preheating treatment on soy protein added to biscuit formulation is required to reach better quality biscuits.

Figure 3 .
Figure 3. Graphic of peak force (hardness) from group b (modified (preheated/PH)) soy concentrate on high protein biscuits.

Figure 4 .
Figure 4. Diagram of average biscuits hardness data from group a (nonpreheated/NPH) and group b (modified preheated/PH) soy concentrate on high protein biscuits.

Figure 5 .
Macrostructure of Biscuits Pores (size = 1200 x 120 pixels) from Non preheated (NPH) group and Modified, Preheated (PH) Soy Concentrate on High Protein Biscuits where dark grey or black pixels represent the porous structure and the light grey pixels represent the biscuit structure.

Figure 6 .
Figure 6.Diagram of average biscuits porosity area from group a (non-preheated/NPH) and group b (modified preheated/PH) soy concentrate on high protein biscuits.

Figure 8 .
Figure 8. Image of biscuit structure (2D) before and after thresholding.top image = control biscuit.left side = non-preheated (NPH) soy concentrate on high protein biscuits 11%, 13%, and 15%.right side = pre-heated (PH) soy concentrate on high protein biscuits.Dark grey or black pixels represent the porous structure and the light grey pixels represent the biscuit structure.

Figure 10 .
Figure 10.Comparison of sensory attributes from non-preheated (NPH) and modified by preheated (PH) soy concentrate on high protein biscuits.
[1] Loveday S M 2019 Food Proteins: Technological, Nutritional, and Sustainability Attributes of Traditional and Emerging Proteins Annu.Rev. Food Sci.Technol. 10 [2] Foegeding E A and Davis J P 2011 Food protein functionality: A comprehensive approach Food Hydrocoll.25 [3] Tang X and Liu J 2017 A comparative study of partial replacement of wheat flour with whey and soy protein on rheological properties of dough and cookie quality J. Food Qual.[4] Lonnie M, Hooker E, Brunstrom J M, Corfe B M, Green M A, Watson A W, Williams E A, Stevenson E J, Penson S and Johnstone A M 2018 Protein for life: Review of optimal protein intake, sustainable dietary sources and the effect on appetite in ageing adults Nutrients 10 [5] Ghoshal G and Kaushik P 2020 Development of soymeal fortified cookies to combat malnutrition Legum.Sci. 2 [6] Mohajan S, Orchy T N and Farzana T 2018 Effect of incorporation of soy flour on functional, nutritional, and sensory properties of mushroom-moringa-supplemented healthy soup Food Sci.Nutr.6 [7] Ermosh L G, Prisuhina N V., Melnikova E V. and Safronova T N 2021 New types of biscuit products with improved amino acid composition IOP Conference Series: Earth and Environmental Science [8] Nogueira A de C, Aquino R K do N de and Steel C J 2022 Empirical rheology of wheat flour doughs with pea, soybean and whey protein isolates Food Sci.Technol.42 1-8 [9] Barac M, Stanojevic S, Jovanovic S and Pesic M 2004 Soy protein modification: A review Acta Period.Technol.3-16 [10] Wang Z, Li Y, Jiang L, Qi B and Zhou L 2014 Relationship between secondary structure and surface hydrophobicity of soybean protein isolate subjected to heat treatment J. Chem.[11] Deak N A, Johnson L A, Lusas E W and Rhee K C 2008 Soy Protein Products, Processing, and

Table 1 .
Ratio of soy concentrate powder and water.

Table 2 .
Formulation of high protein biscuit.

Table 3 .
Peak force (hardness) from group A (non-preheated/NPH) and group b (modified (preheated/PH) soy concentrate on high protein biscuits.
Based on a force-time graph (Figure2, 3 and Table3) shows the maximum peak force of the undenatured and denatured soy protein biscuits.The addition of soy concentrate powder in biscuit

Table 4 .
Average of biscuits hardness data (undenatured or non-preheated/NPH) and (modified, preheated/PH or denatured) soy concentrate on high protein biscuits.
Noted : Data are expressed as mean ± standard deviation of the mean.P value was calculated using paired t test.P< 0.05 (compared with control biscuit).Statistically significant difference is marked with (*) al., which observed that the substitution of

Table 5 .
Average of biscuits porosity area from group a (non-preheated/NPH) and group b (preheated/PH) soy concentrate on high protein biscuits.
Figure 7. Diagram of average size of biscuits porosity from group a (nonpreheated/NPH) and group b (modified preheated/PH) soy concentrate on high protein biscuits.

Table 6 .
Average of fractal dimension (Df) of non-preheated (NPH) and modified, preheated (PH) soy concentrate on high protein biscuits.Data are expressed as mean ± standard deviation of the mean.P value was calculated using paired t test.P< 0.05 (compared with control biscuit).Statistically significant difference is marked with (*).