Incorporation of high fructose corn syrup with different fructose levels into biscuit: An assessment of physicochemical and textural properties

Abstract This study examined the effects of different concentrations of high‐fructose corn syrup (HFCS, 28%, 44%, 55%) used in biscuit formulation on the hydroxymethyl furfural (HMF) acrylamide content, and textural properties were investigated and compared with invert sugar and sucrose‐incorporated samples. No significant difference in the chemical composition (moisture, fat, protein, and ash) among different samples was noted based on the results. The highest L* was associated with a control sample containing sugar and invert sugar, although an increase in F55 content decreased the L* value significantly (p < .05). The highest hardness value was correlated with control samples (6.5 N), although the sample with 12.5% F42 and 25% F55 demonstrated lower hardness 6.27 N, and the lowest hardness value (3.97 N) was related to the sample containing 12.5% F42 and 25% F28. The amounts of water activity of all samples were in the range of 0.22 to 0.29, with the highest amount related to the control sample. The SEM images showed a uniform surface with several holes for all the biscuits. The highest and lowest (HMF) levels were related to the samples containing 25% F55 (46.04) and 12.5% F42 with 2.36 ppm. The control sample with the acrylamide amount of 28.50 ppb and the sample containing 12.5% F42 and 25% F55 with the acrylamide amount of 27.33 ppb showed the highest acrylamide content.

Among the heated bakery products, including biscuit and bread, acrylamide and hydroxymethyl furfural (HMF) levels have been investigated using different studies. The neurotoxic, mutagenic, and carcinogenic effects of acrylamide have been proved on human and animal cells (Shipp et al., 2006). Genotoxic and mutagenic impacts of HMF on human cells, as well as bacteria, are reported. HMF can promote liver and colon cancers in animals such as mice and rats (Glatt & Sommer, 2006;Monien et al., 2012). However, more robust and clear evidence for genotoxic and carcinogenic impacts of HMF is needed (Capuano & Fogliano, 2011).
Few studies have been performed regarding the effects of HFCS with different fructose levels on biscuit properties. In this study, three HFCS with different fructose levels (28%, 44%, and 55%) were used in biscuit formulations, and their effects on sensorial, physicochemical (especially HMF and acrylamide), textural, and microbial products were investigated.

| Materials
Three HFCSs with different fructose levels, including 28%, 44%, and 55%, were obtained from Zar Co. Other chemical reagents, all in analytical grade, were obtained from Merck and Sigma Aldrich Co.
Oil and sucrose were purchased from a local market in Tehran, Iran.

| Chemical composition
The chemical composition (moisture content by the oven method at 105°C, protein content by the Kjeldahl method, fat content by Soxhlet method, and ash content by the oven method at 600°C) of biscuit samples was estimated based on the standard analysis methods (Pasqualone et al., 2015).

| Color measurement
Colorimetric analysis of the samples' outer surface was performed based on the proposed method by Hosseini et al. (2019). A Canon digital camera (A540, 6 megapixels) was located on the top of a wooden box with 50 cm length, 50 cm width, and 60 cm height to capture images. The distance between the camera lens and the samples in the box was adjusted to 25 cm. The box was lightened with a natural daylight source (6,500 K). Different color parameters, including lightness (L*), blueness-yellowness (b*), and redness-greenness (a*), were calculated using Adobe Photoshop ® CS6 from the taken images.

| Texture measurement
The hardness of the fresh, intact biscuit was determined using a three-point bend test. A texture analyzer (TAXT2) with the threepoint bending and 5 kg load cell and 40 mm span between the supports was used for the test. The test was performed with a pretest speed of 1.0 mm/s, a test speed of 3.0 mm/s, and a posttest speed of 10.0 mm/s, with a 10 mm distance and 500 PPS data acquisition rate. The maximum recorded force was considered as the hardness of samples. The test was repeated three times, and the average hardness and the distance to break were reported (Mamat et al., 2010).

| Water activity measurement
The relative humidity (RH) of the air, equilibrated with the biscuits placed in a sealed container (22 ± 1°C), was reported as water activity (aw). A water activity meter (Aqua Lab Series 3) was used for this evaluation (Laguna et al., 2014).

| Fourier transform infrared spectroscopy
The FT-IR spectra of biscuit powders were recorded in the wavenumber range of 4,000-400 cm− 1 by the Avatar 370 spectrometer (Thermo Nicolet Corp.). Initially, biscuit samples were completely powdered and mixed with KBr  and finally pressed to create a tablet.

| SEM measurement
The Vega3 electron microscope (SEM) (TESCAN) was used to evaluate the biscuit samples' surface microstructure. In this regard, the small parts of the dried biscuits were attached to aluminum stubs by double-sided adhesive tape. Then, a thin layer of gold was coated over the samples using Desk Sputter Coater DSR1, Nanostructural Coating Co. The accelerating voltage and temperature used in the test were 10 kV and 25°C, respectively (Gahruie et al., 2020).

| Hydroxymethylfurfural measurement
To measure HMF in samples, 2 g of crushed biscuit was poured into a Greiner tube, and 3 ml of Mili-Q water was added to it, vortexed for 1 min, followed by incubation 50°C for 1 hr. After centrifuging at 3,000 rpm for 10 min, 1,500 µl of the separated supernatant was removed and centrifuged for 15 min at 5,000 × g. The supernatant was collected and filtered through a 0.22µm filter and evaluated by an HPLC system. The HPLC system had a UV detector (248 nm) and a Polaris 5 C18-A column (150 × 4.6 mm, 5 µm) to determine the amount of HMF. The mobile phase was water: acetonitrile (95:5) with a 1 ml/ min flow rate and a typical run time of 20 min per sample. The analytical method had a limit of detection of 1 mg/kg sample powder.
Measurements were carried out in duplicate. A standard method by a calibration curve was used to quantify HMF (Nguyen et al., 2016).

| Acrylamide measurement
The amount of acrylamide was determined based on Wang et al. (2017). In brief, 1.0 g of completely dried biscuit and 5 ml methanol were mixed in a tube and centrifuged at 10,000 × g at 4°C (10 min). The supernatant was then removed, and 0.1 ml Carrez I and 0.1 ml Carrez II solutions were added to precipitate and centrifuged at 5,000 × g at 10°C (10 min). Then methanol was added to the supernatant until reaching 5 ml volume. Then 1 ml of the solution was evaporated by nitrogen purging at 40°C until reaching dryness ultimately and 1 ml distilled water was added to the residue.
A Thermo carbon column was conditioned with methanol (5 ml) and water (5 ml) sequentially for solid-phase extraction cleanup. In the next step, 1 ml of the extract was injected into the solid-phase extraction column to pass through the sorbent material. Elution of the acrylamide presented in the column was performed using methanol (5 ml), and then the elute was collected and evaporated with a nitrogen stream at 40°C. One milliliter of methanol was added to the residue until dissolving and then filtered through a 0.22µm polyvinylidene fluoride filter.
HPLC with a UV-VIS detector (Shimadzu, Japan) was used to determine the amount of acrylamide. A Hypersil ODS-C18 column (250 mm × 4.6 mm, 5 µm, Thermo Scientific) was used at 40°C for this purpose. The injection volume was 20 µl. The mobile phase was water: acetonitrile (95:5) with a flow rate of 0.6 ml/min. Acrylamide was detected at 210 nm, and its amount was determined using a calibration curve constructed in the concentration range of 0-10 µg/ ml. CLASS-VP Shimadzu automated software was used for data collection and manipulation.

| Statistical analysis
All results were reported based on three replications. One-way analysis of variance (ANOVA) at a significance level of 5% was used for statistical analysis of data, and the significant differences between the values were determined by Duncan's multiple range tests using SAS ® (ver. 9.1, SAS Institute Inc.). Table 2 shows the effect of the fructose/glucose ratio on different chemical properties of samples. It can be observed that the incorporation of different sweeteners did not change the moisture, protein, fat, and ash contents of different biscuit samples significantly. These results were due to moisture, protein, fat, and ash contents related to basic formulation; other ingredients such as solid oil, flavor, and lecithin, milk powder, egg powder, and sweetener had similar moisture, protein, and protein, fat, and ash composition. The ranges of moisture content, protein content, fat, and ash were 2.07%-2.23%, 7.69%-7.79%, 14.35%-14.41%, and 0.96%-1.05%, respectively. According to Hooda and Jood (2005), the protein content, moisture content, and ash content of biscuit were 9%-10%, 3%-4%, and 1%-2%, respectively, which is almost similar to our results. In another work, biscuit's moisture content was 2%-4% Laguna et al. (2014). Akoja and Coker (2018) evaluated the effects of okra powder on the properties of wheat four biscuits. They reported that moisture content, ash, fat, and protein content of samples were 8%-10%, 1%-4%, 17%-20%, and 10%-21%, respectively.

| Color
The samples incorporated with different sweeteners showed different colors (Figure 1). The control sample, which was incorporated

| Water activity
In biscuits and other brittle and dry foods, mechanical signatures and crispness are influenced by water activity, moisture content, and water distribution. There is a sigmoid relationship between water activity and content with crispness (Blanco Canalis et al., 2017;Laguna et al., 2014). Water activity measurement was performed to determine whether the sensory appreciation of crispness was influenced by water activity and moisture content or not. Based on Figure 3, a significant difference (p < .05) between different biscuit samples' water activity and different glucose/fructose ratios. The highest water activity was related to the control sample (0.29) containing sugar and invert sugar and the sample containing 12.5% F42 (0.28).
Samples containing 37.5% F42 with 0.24 and sample containing 12.5% F42 and 25% F28 showed the lowest water activity values (0.24 and 0.22, respectively). This phenomenon is due to the high tendency of sugar (sucrose) to interact with water against glucose and fructose. There are differently reports on the effects of sugar composition on the water activity of systems. Also, the water activity of fructose and glucose solution is similar. The water activity of biscuit samples containing inulin and HPMC was reported to range from 0.12 to 0.25 (Laguna et al. (2014).  Note: Data represent the mean ± standard deviation of three independent batches. Different uppercase letters in each column indicate significant differences (p < .05).

| Microstructure properties
The images of the microstructure of biscuit samples observed by the SEM are shown in Figure 5.

| HMF content
Based on Table 3, different glucose/fructose ratios resulted in different amounts of HMF in biscuit samples. The sample's HMF content containing 25% F55 was 46.04 ppm, which showed the highest value, and the samples incorporated with 37.5% F42, 12.5% F42, and 25% F28 had the lowest amounts of HMF (2.31 ppm and 4.05 ppm, respectively). The presence of invert sugar in the control and higher fructose in the sample containing F55 are the main reasons for higher HMF. Caramelization and a specific amino acid route are the parameters involved in the production of HMF (Nguyen et al., 2016). Nguyen et al. (2016)

| Acrylamide content
As presented in Table 3, different glucose/fructose ratios in biscuit formulation resulted in different acrylamide contents in samples.
The highest amounts were related to the control sample containing sugar and invert sugar with the value of 28.50 ppb and the sample containing 12.5% F42 and 25% F55 with 27.33 ppb, respectively.
Other samples showed no significant differences (p < .05) in the amounts of acrylamide. The primary mechanism of production of acrylamide is related to the fructose content, and acrylamide content significantly increased with increasing fructose content. In a study on biscuits by Nguyen et al. (2016), it was revealed that in all the four biscuit types with different formulations, the production of acrylamide was mainly related to fructose. Similar results were also reported by Robert et al. (2005). Before the Maillard reaction, melting and other physical changes take place at a low-moisture content medium. Because of the lower melting point of fructose than glucose, this sugar causes higher acrylamide production (De Vleeschouwer et al., 2009). Some studies have proved that the limiting factor in acrylamide production in yeast-leavened wheat bread and heated wheat flour was free asparagine (Europe, 2015;Muttucumaru et al., 2006;Surdyk et al., 2004). Mesías et al. (2016)  reported that the acrylamide content of samples was between 151 and 1,187 mg/kg.

| CON CLUS ION
This study aimed to evaluate the effects of different amounts of high fructose corn syrup with different fructose levels on biscuit formulation compared with inverted sugar and sucrose. The results showed no significant differences among moisture, fat, protein, and ash con- The SEM images showed a uniform surface with several holes on the biscuit surface. The highest HMF amount was related to the sample containing 25% F55 (46.04 ppm). The control sample with an acrylamide amount of 28.50 ppb and the sample containing 12.5% F42 and 25% F55 with an acrylamide amount of 27.33 ppb showed the highest acrylamide content. Finally, the results showed that the sample containing F42 was the best sample for replacing sucrose with HFCS.

ACK N OWLED G EM ENT
The authors thank Dr. Amin Mousavi Khaneghah (University of Campinas), a native English expert, for improving the English of the article.