Physicochemical properties of rice bran blended oil in deep frying by principal component analysis

Abstract

This study aimed to obtain a rice bran blended oil with good quality in deep frying. The thermal stability, nutrients and harmful substances of rice bran oil (RBO) and other four oils (palm oil, PO; cottonseed oil, CO; sunflower oil, SuO; soybean oil, SO) were analyzed. Besides, the blended oil formulas were established by the principal component analysis method, and their physicochemical properties, frying characteristic indicators, nutrients, and harmful substances were compared. The results provided that two suitable blended oil formulas (F1: 50% RBO + 40% PO + 10% CO; F2: 60% RBO + 35% PO + 5% CO) of good frying performance were attained by principal component analysis. The acid value (1.19 mg/g), peroxide value (0.09 meq/kg), residual oil rate (8.07%), 3-MCPD ester reduction content (2.33 mg/kg), benzopyrene concentration content (0.95 μg/kg) and vitamin E consumption rate (67.86%) in F2 were lower than that in F1. Moreover, the oryzanol retention rate (87.84%) of F2 was higher than that of F1. In summary, F2 was more conducive to human health and more suitable than F1 in deep-frying. This information had an important directive on the industrial production of rice bran blend oil.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13197-022-05472-7.

Introduction

Rice could provide 20% of the dietary energy to human, which becomes the primary source of daily nutrition. Rice bran (RB), accounting for 8% of rice grains, is an essential by-product of the rice milling process, which is used as animal feed, oilseeds, and medicine ingredients (Mendes et al. 2021). According to the grain production data of the National Bureau of Statistics of China in 2020 (http://data.stats.gov.cn), the average annual production of RB is about 12 million tons, and the production rate is 7%. The Intensive processing of RB can obtain phytic acid, protein, and oil. And the by-products of RB can extract fatty acids, rice bran wax, rice bran fat, oryzanol, and ferulic acid. Moreover, they are widely used in special oils, oleochemicals, cosmetics, and other industries.

Rice bran oil (RBO) is considered a healthy oil, which has positive effects on hypercholesterolemia, hypertension, hyperglycemia, insomnia, and other chronic diseases. It contains rich linoleic acid and some natural antioxidants, such as phospholipids, glycolipids, phytosterols, and tocopherols (Alfaro et al. 2015), these compounds contribute to the shelf life and moisturizing in the skin. Linoleic acid has beneficial health effects in preventing and treating a variety of diseases, such as reducing body fat deposition, reducing the development of atherosclerosis, stimulating immune function, anticancer and lowering blood sugar (Fuke and Nornberg 2017). It is also an excellent alternative materials for biodiesel in industrial applications (Hoang et al. 2021). Most important of all, RBO has good flavor (Latha and Nasirullah 2014) and frying stability (Hamid et al.2014), which is regarded as a potential frying oil.

The sensory characteristics and nutritional properties of oil would be changed during frying (Xu et al. 2021), such as the content of aldehydes, carboxylic acids, hydrocarbons, ketones, esters, lactones, and aromatic compounds would change (Dobargane and Márquez-Ruiz 2015) and triglyceride polymers would form (Xu et al. 2019). the poor oxidation stability and nutrients retention rate always occur in single-product oil in the deep frying process (Torri et al. 2019). Therefore, the blended oil is used as frying oil to solve the problems (Upadhyay et al. 2017). The blending oil of palm oil (PO) and cottonseed oil (CO) have lower polarity and polymer compounds than single oil (Arslan et al. 2017), and the soybean oil (SO) mixing with sunflower oil (SuO) could improve its thermal stability and oxidation resistance (Boukandoul et al. 2019). In addition, the fried fries have better flavor, texture, and color of olive and palm blended oil than that used of single product oil (Koohikamali and Alam 2019). It is essential of great importance that blending oil should be adopted to develop as a frying oil. RBO has a high content of phytosterols, γ-oryzanol, and tocopherol. These natural ingredients provide to increase the thermal antioxidation stability of oils (Ali et al. 2019). However, there is limited researches on frying blended oil of RBO. In the high frying temperature, there is always occurred a series of complex oxidation reactions and oil deterioration (Xu et al. 2020). Therefore, it is necessary to study the oxidation stability and harmful substance content of blended oil.

The purpose of this study was to explore the physicochemical properties, frying characteristic indicators, nutrients, and harmful substances of RBO and other single-product oils (PO, CO, SuO, SO), the nutritional function, sensory evaluation and oxidation stability were considered comprehensively and then the blended oil formulas were obtained by the principal component analysis (PCA) method. Though the frying quality comparison of different blended oils, the best RBO blended oil formula was obtained.

Materials and methods

Materials

The RBO, PO, CO, SuO, and SO were purchased from Yihai Kerry grain and oil (Wuhan) Co., Ltd (Wuhan, China). The French fries were provided by Lutosa (Shanghai, China). The reagents were purchased from Sinopharm Chemical Reagent Co., Ltd (Hubei, China). The standards of tocopherol, fatty acid, oryzanol, phytosterols, 3-MCPD ester, and benzopyrene were purchased from Supelco Chemical Co. Ltd (Shanghai, China).

Frying procedure

The frying procedure was referred from Xu et al. (2019) with some modifications. The oil of 5 L was filled into the fryer and heated to 180 ± 2 °C. After the pre-heating period, the fries were fried as frying cycle (3 min frying time + 12 min waiting time), and the overall frying duration was 30 h. The 80 mL filtering oil and 100 g fries were collected every 3 h, and the oil was transferred directly to 100 mL bottles and stored at 4 °C in the dark until use.

Residual oil rate

The 100 g fries were crushed into granules, wrapped tightly with filter paper, put into a conical flask with a stopper, added in the conical flask and submerged by 400 mL n-hexane. The sample was placed in an SB25-12DTD ultrasonic extraction machine (Cienta, China) for 1 h, set for 24 h, then subjected to rotary evaporation using a YRE2000E rotary evaporator (Yuhua, China), which was order to obtaining residual oil in the fries.

$$w\left(\%\right)=\frac{m}{M}\times 100$$

Among them, w was the frying oil content; m was the frying oil content; M was the mass of the fry content.

Physicochemical properties

The acid value (AV), vitamin E (VE), trans fatty acids (TFA), absorbance, viscosity, peroxide value (POV), polar component (PC), and phytosterol were determined according to the Cd 3d–63, Ce 8–89, Ch 2a–94, Ch 5–91, Tq 1a–64, Cd 8b–90, Cd 20–91, Ce 12–16 official recommended by AOCS. Method for determination of benzopyrene concerning BS EN 16619:2015. The γ-oryzanol content was determined using the method outlined in CODEX.

Fatty acid composition

The fatty acid composition was carried out by the Agilent 7890A gas chromatography (GC) system (Agilent, USA) by Gao et al. (2018) with some modifications. The oil of 0.20 mg was methylated under alkaline conditions and analyzed by an Agilent capillary column (30 m × 0.25 mm, 0.25 μm). The GC adopted a programmed temperature rising mode, with an initial temperature of 180 °C for 5 min and a temperature of 230 °C at a rate of 3 °C/min for 15 min. The carrier gas was high-purity helium with a flow rate of 1.0 mL/min, the inlet temperature was 250 °C, the injection volume was 1 μL, and the split ratio was 20:1. The fatty acid methyl ester peaks were identified by comparison to the retention times of known standards and the percentage of each peak area to the sum of all peak areas were quantified.

Oxidation stability index (OSI)

The continuous bubbling of air at a flow rate of 20 L/h through oil samples (3.0 g) held at 120 °C was used to measure OSI with a Rancimat 892 model (Metrohm, Switzerland).

3-Monochloropropane-1, 2-diol ester (3-MCPD ester)

The pre-processing methods were referred to Küsters et al. (2010) with some modifications. The content of 3-MCPD ester was measured by a gas chromatography-mass spectrometer (GC–MS, Trace1310-ISQ, Thermo Fisher, USA). The GC with HP-5MS capillary column (30.0 m × 0.25 mm, 0.25 μm, Thermo Fisher, USA) adopted the programmed temperature rise mode. The initial temperature was 85 °C, and it was held for 0.5 min. It raised to 150 °C at a rate of 3 °C/min, then it was up to 180 °C at a rate of 12 °C/min, finally heated up to 280 °C at 25 °C/min and hold for 7 min. The carrier gas was helium (99.999%), and the flow rate was 1.0 mL/min. The injection volume was 1 μL, and the split ratio was 20:1. The electron energy of MS was 70 eV, the ion source and transmission line temperature was 250 °C and 280 °C, respectively.

Sensory evaluation

A 10-member analytical sensory panel trained and experienced in evaluating fried foods, rated the French fries including color, appearance, crispness, flavor, taste and overall acceptability by using 5-point hedonic scale with 1 = dislike extremely and 5 = like extremely. All evaluations were conducted in a panel room with individual booths, temperature control, and natural lighting. The scores received by each samples were then averaged. Panelists were instructed to rinse their mouth with water between samples.

Statistical analysis

Correlation analysis was performed on the 12 indexes (X₁-OSI, X₂-acid value, X₃-absorbance, X₄-viscosity, X₅-PC, X₆-saturated fatty acid (SFA), X₇-monounsaturated fatty acid (MUFA), X₈- polyunsaturated fatty acid (PUFA), X₉-TFA, X₁₀-VE loss rate, X₁₁-sensory evaluation, and X₁₂-residual oil rate, respectively) of the five kinds of oil (1-RBO, 2-PO, 3-CO, 4-SuO, and 5-SO, respectively). The weight of each indexes was the same. The correlation matrix comprised the Pearson correlation coefficient between all pairs, which included duplicate information between different factors with a strong positive correlation. Therefore, the representative evaluation indicators were selected by PCA method to make the new variables irrelevant, which could reduce the probability of information overlap. The correlation coefficient can describe the relationship between the principal component and the original variable.

$$r_{C_i{X}_j}=Corr\left(C_i,X_j\right)=a_{\mathit{ij}}\sqrt{Var{C}_j}=a_{\mathit{ij}}\sqrt{{\lambda }_i}$$

where $a_{\mathit{ij}}$ represents the amount of $X_j$ information extracted by the ith principal component.

The results were characterized by one-way ANOVA analysis using SPSS 17.0 (IBM, USA) and Origin 8.5 (Originlab, USA), then conducted a multi-range test of Duncan and made a significant test of the difference between product quality attributes. PCA of the data was performed using 95% confidence (p < 0.05).

Results and discussion

Physicochemical properties of single-product oils

Figure 1 showed the physicochemical properties of five single-product oils. The residual oil rate of oils after frying was depicted in Fig. 1a. The rate of samples from high to low was SO (13.75%), PO (10.85%), SuO (9.45%), RBO (9.07%), and CO (8.03%). Residual oil could increase the risk of cardiovascular disease through raising the LDL-cholesterol concentration and decreasing the HDL-cholesterol concentration (Ascherio and Willett 1997), therefore, CO was better for health.

Fig. 1.

a Changes in residual oil rate of RBO and other single products during the frying. b Changes in color of RBO and other single oils during frying. c Changes in acid value of RBO and other single oils during frying. d Changes in the content of polar components of RBO and other single products during frying (color figure online)

As shown in Fig. 1b, the absorbance value of five oils was rapid rise during the frying process. The absorbance was sorted by decreasing order from RBO (0.74), PO (0.59), SuO (0.50), CO (0.50) to SO (0.36) after frying 30 h. The results provided that the absorbance of the oils was gradually darkened with the frying process, and the oil oxidative degradation reaction may occur in contact with oxygen and moisture at high temperatures, forming carbonyl compounds and epoxy acids, which increased the absorption of visible light (Tatani et al. 2006). It may be also due to the presence of phospholipids during heating and auto-oxidation, leading the color to darker.

Figure 1c provided the AV change of oil samples during frying, which had an upward trend. The AV of RBO, PO, CO, SuO, and SO increased by 1.01 mg/g, 1.33 mg/g, 1.27 mg/g, 2.50 mg/g, and 1.21 mg/g, respectively. The AV of SuO was raised slowly in the early stage (0–9 h), while the oxidative rancidity was accelerated after 9 h. RBO had the slowest rise, which was relatively stable in the frying process.

Figure 1d showed the PC of five oils during the frying process. The PC 27% of RBO, CO, SuO, and SO was 10 h, 18 h, 13 h, and 20 h, respectively, while PO was not exceeded the limit during frying scope. Research shows that the PC level and SFA profile had not correlation (Li et al. 2017). Although the SFA content of RBO was higher than that of SuO and SO, while the time reached PC 27% was earlier. The PC in edible oil would promote the production of odorous substances and the early fat (Ghosh 2007). Although the initial PC of RBO was the highest, the increase was slower than SuO.

Fatty acid composition and VE content of single-product oils

Table 1 showed the changes of the fatty acid composition and VE content in five single-product oils during frying, and the minor fatty acids, such as C16:1, C17:0, C17:1, C20:0, C22:0, C22:1, and C24:1 were constituted less than 1.00%. After frying, the C14:0, C16:0, C18:0, and C18:1 contents of all samples were increased, while the contents of C18:2 and C20:1 were decreased. Due to the oxidation reaction of unsaturated double bonds under high-temperature conditions, the contents of SFA and MUFA were increased, and the content of PUFA was decreased. Some double bonds of PUFA were oxidized to form saturated structures, such as C18:2 and C18:3 to C18:0 or C18:1. In addition, MUFA became short carbon chains after oxidative degradation, resulting in C14:0 and C16:0. Given the loss of linoleic acid, its oxidative degradation led to secondary oxidation products and polar compounds (Nagarajan et al. 2017). Except for CO, the content of C18:1T was increased after frying, which proved that frying would produce TFA which were not good for human health.

Table 1.

Changes of fatty acid composition and VE content in RBO and other single oils during frying

FAC (%)	RBO			PO			CO			SuO			SO
	0 h	12 h	24 h	0 h	12 h	24 h	0 h	12 h	24 h	0 h	12 h	24 h	0 h	12 h	24 h
C14:0	0.27	0.30	0.33	1.11	1.14	1.21	1.03	1.20	1.24	0.07	0.24	0.34	0.07	0.06	0.06
C16:0	17.50	18.50	19.03	42.68	43.14	49.56	22.50	23.93	26.04	6.45	6.8	8.05	10.65	11.12	12.01
C16:1	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.11	0.11	0.11	0.08	0.10	0.80
C17:0	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.09	0.10	0.13
C18:0	1.64	1.80	1.85	4.01	4.48	4.72	3.82	3.96	3.99	3.16	4.46	5.44	4.15	4.85	5.24
C18:1 T	0.11	0.19	0.30	N.D	0.04	0.61	0.30	0.05	N.D	N.D	N.D	N.D	N.D	0.24	0.42
C18:1	43.96	44.65	46.04	28.45	30.87	34.72	22.97	23.24	23.56	28.28	29.87	31.73	24.11	26.79	28.58
C18:2 T	0.06	0.08	0.11	0.42	0.43	0.45	0.72	0.85	0.88	0.91	0.95	1.02	0.66	0.70	0.74
C18:2	33.57	31.13	28.77	20.99	19.37	7.10	46.79	43.98	41.98	59.59	55.14	51.28	51.99	49.32	46.62
C18:3 T	0.57	0.60	0.62	N.D	N.D	1.02	0.26	0. 54	0.56	N.D	0.04	0.06	0.34	0.50	0.67
C20:0	0.17	0.60	0.85	0.38	0.40	0.42	0.44	0.50	0.55	0.03	0.05	0.07	0.16	1.20	1.20
C20:1	0.76	0.73	0.71	0.15	0.07	N.D	N.D	N.D	N.D	0.15	0.14	0.14	0.87	0.85	0.85
C18:3	0.64	0.60	0.60	0.23	0.13	0.05	1.06	1.21	1.31	0.09	0.06	0.02	5.82	3.38	1.38
C22:0	0.18	0.19	0.19	0.00	0.00	0.08	0.00	0.00	0.00	0.71	0.71	0.72	0.41	0.66	0.70
C22:1N9	0.58	0.60	0.62	N.D	N.D	N.D	N.D	N.D	N.D	N.D	N.D	N.D	N.D	N.D	0.49
C24:0	N.D	N.D	N.D	0.09	0.09	0.09	N.D	N.D	N.D	0.24	0.25	0.25	N.D	N.D	N.D
SFA	20.11	21.68	22.24	48.93	49.25	56.08	28.34	30.20	32.26	11.58	13.85	16.43	16.39	19.79	19.34
MUFA	45.30	46.31	47.37	28.6	30.94	34.72	22.97	23.24	23.56	28.65	30.27	32.09	25.45	27.86	31.11
PUFA	34.21	31.81	29.37	21.22	19.5	7.15	47.85	45.39	43.29	59.68	55.3	51.30	57.81	42.94	48.42
TFA	0.73	0.86	1.02	0.42	0.47	2.08	1.28	1.34	1.44	0.91	0.99	1.08	1.41	1.20	1.00
VE (mg/100 g)
α-tocopherol	24.39	15.51	4.21	20.98	12.98	0.45	16.37	9.63	4.08	47.05	23.85	11.79	10.57	6.78	3.56
γ-tocopherol	11.24	13.4	N.D	11.56	6.9	0.22	46.98	2.98	0.03	N.D	N.D	N.D	98.43	65.29	34.02
α-tocotrienol	19.11	6.91	N.D	4.28	0.23	N.D	N.D	N.D	N.D	0.93	1.05	0.87	1.97	1.86	1.78
γ-tocotrienol	23.88	7.23	N.D	5.44	0.41	N.D	N.D	N.D	N.D	N.D	N.D	N.D	0.78	0.69	0.66
Total	78.62	43.05	4.21	42.26	20.52	0.67	63.35	12.61	4.11	47.98	24.9	12.66	111.75	74.62	40.02

N.D. is not detected. T is trans fatty acid

A linear decrease in total VE content was observed during frying (Table 1). There were four types (α- tocopherol, γ-tocopherol, α-tocotrienol and γ-tocotrienol) of VE in oils. Before frying, SO had superior VE content (111.75 mg/100 g), followed by RBO (78.6 mg/100 g), while PO had the lowest content (37.98 mg/100 g). After frying for 24 h, the content of SO was obviously retained, while others were declined. The retention rates of VE in descending order were SO (32.63%), SuO (26.39%), CO (6.49%), RBO (5.34%), and PO (1.76%). The tocopherols were found in CO (63.35 mg/100 g), while tocotrienols were not detected. The α-tocotrienol was only found in SuO (0.87 mg/100 g) and SO (1.78 mg/100 g) after frying. The tocopherols and tocotrienols were present in both oil samples (11.78 mg/100 g and 0.87 mg/100 g in SuO, 37.58 mg/100 g and 2.44 mg/100 g in SO). The degradation rates of the different tocopherols were γ-tocopherol > α-tocopherol, thus suggesting a superior antioxidant effect of γ-tocopherol. The VE was usually considered a nutritional supplement for fats and oils.

Sensory evaluation, OSI and viscosity measurements

As depicted in Table 2, the viscosity of frying oils increased as a consequence of oxidation and polymerization reaction. The viscosity change of oils in descending order was CO (43.45 mm²/s), PO (42.84 mm²/s), SuO (40.78 mm²/s), SO (38.95 mm²/s) and RBO (37.6 mm²/s). RBO had the smallest increase in viscosity, which might be the natural active substances inhibit the oxidation reaction in the frying process. In terms of oxidation stability, PO was the most stable (1.90 h). The comprehensive score of sensory evaluation was highest for PO, followed by RBO and CO. The flavor of French fries fried in PO was more acceptable to evaluators, due to the high saturation, fries fried in it tended to feel greasy when the time was extended, which had a certain impact on sensory evaluation. Fries fried with RBO had a certain grain fragrance, less greasy sense, which is more conducive to be accepted by consumers; and the fries fried in CO had a good color. The overall acceptability of fries fried in SuO and SO was lower than other oils.

Table 2.

Changes of sensory evaluation, OSI and viscosity in RBO and other single oils during frying

	RBO	PO	CO	SuO	SO
OSI (h)	1.16 ± 0.02b	1.90 ± 0.03a	0.35 ± 0.02d	1.02 ± 0.01c	1.22 ± 0.01b
Viscosity (mm2/s)	37.60 ± 0.01c	42.84 ± 0.02a	43.45 ± 0.01a	40.78 ± 0.02b	38.95 ± 0.01c
Sensory attributes
Appearance	4.50 ± 0.01c	4.80 ± 0.01a	4.60 ± 0.02b	4.60 ± 0.02b	4.20 ± 0.01d
Color	4.80 ± 0.01a	4.70 ± 0.01b	4.90 ± 0.02a	4.70 ± 0.02b	4.60 ± 0.01c
Crispness	4.20 ± 0.02b	4.90 ± 0.01a	3.60 ± 0.01c	4.20 ± 0.01b	4.30 ± 0.02b
Taste	4.50 ± 0.01b	4.80 ± 0.02a	4.10 ± 0.02c	3.50 ± 0.01d	4.20 ± 0.02c
Overall acceptability	4.70 ± 0.01b	4.90 ± 0.01a	4.70 ± 0.02b	4.50 ± 0.01c	4.30 ± 0.01c

Values are means ± standard deviation. The superscript letters indicate the statistical difference in rows in significant level at 5%

Principal component analysis

Figure 2 depicted the loading schematic diagrams of the principal component. The first principal component (PC1) was contributed to 53.64% of the variance, reflecting the fatty acid composition, natural active ingredients contents, and frying sensory characteristics. Among them, the fatty acid composition was of great significance in the oil stability and played a vital role in guiding the blending of RBO. The change in natural active ingredients contents was also reflected the thermal stability of oils, which could prevent lipid from accelerated oxidation during the deep-frying process. The second principal component (PC2) was contributed to 23.30% of the total change, reflecting the oxidative stability during frying. The third principal component (PC3) was contributed to 14.63% of the variance. It showed the changes in color and viscosity of oils during the frying process. The firstly three principal components accounted for 91.56% of the total variability, and the frying characteristics of five single-product oils could be described by these principal components. By standardizing the initial factor component matrix, the equation for each principal component could be obtained as follows:

$$\begin{array}{cc}\hfill {\text{F}}_1& =-0.05{\text{X}}_1+0.13{\text{X}}_2-0.3{\text{X}}_3-0.17{\text{X}}_4+0.19{\text{X}}_5+0.33{\text{X}}_6+0.39{\text{X}}_7\hfill \\ \hfill & +0.39{\text{X}}_8+0.27{\text{X}}_9-0.39{\text{X}}_{10}-0.37{\text{X}}_{11}+0.26{\text{X}}_{12}\hfill \\ \hfill \end{array}$$

$$\begin{array}{cc}\hfill {\text{F}}_2& =0.44{\text{X}}_1-0.39{\text{X}}_2+0.1{\text{X}}_3-0.28{\text{X}}_4-0.36{\text{X}}_5-0.29{\text{X}}_6+0.17{\text{X}}_7\hfill \\ \hfill & -0.01{\text{X}}_8+0.34{\text{X}}_9-0.09{\text{X}}_{10}+0.18{\text{X}}_{11}+0.4{\text{X}}_{12}\hfill \\ \hfill \end{array}$$

$$\begin{array}{cc}\hfill {\text{F}}_3& =0.31{\text{X}}_1+0.2{\text{X}}_2-0.36{\text{X}}_3+0.55{\text{X}}_4-0.47{\text{X}}_5+0.2{\text{X}}_6-0.04{\text{X}}_7\hfill \\ \hfill & +0.02{\text{X}}_8-0.3{\text{X}}_9-0.1{\text{X}}_{10}+0.06{\text{X}}_{11}+0.27{\text{X}}_{12}\hfill \\ \hfill \end{array}$$

Fig. 2.

Principal component analysis loading plot

Based on the above formula, the comprehensive score formula was:

$\text{F}=0.5858{\text{F}}_1+0.2545{\text{F}}_2+0.1597{\text{F}}_3$ and $\text{F}=0.13{\text{X}}_1+0.01{\text{X}}_2-0.21{\text{X}}_3-0.08{\text{X}}_4-0.06{\text{X}}_5+0.15{\text{X}}_6+0.25{\text{X}}_7+0.23{\text{X}}_8+0.20{\text{X}}_9-0.27{\text{X}}_{10}-0.16{\text{X}}_{11}+0.30{\text{X}}_{12}$.

The total scores comprehensively from descending order were CO, RBO and PO. Therefore, these three oils were formulated into blend oils as two different proportions: F1: 50% RBO + 40% PO + 10% CO and F2: 60% RBO + 35% PO + 5% CO.

Residual oil rate, physicochemical properties of blended oils

Figure 3 showed the physicochemical properties of blended oils. After 10 min of freezing, both two blended oil samples were clear and transparent, and 24 h later, both groups had no flocculation phenomenon (Supplementary material). Figure 3a described AV of blend oils during frying. The AV curves of blend oils were started in a low average level (F1: 0.16 mg/g and F2: 0.13 mg/g), with a rapidly rise, it attained high levels eventually (F1: 1.37 mg/g and F2: 1.32 mg/g). At 30 h, the AV of F1 was higher than F2, while frying fries for a short time (within 27 h), F1 had lower AV. By the end of frying, the AV of F1 and F2 increased by 1.21 mg/g and 1.19 mg/g, respectively. The AV of F2 was lower than that of F1during frying.

Fig. 3.

a Changes in acid value of RBO and its blends during frying. b Changes in peroxide value of RBO and its blends during frying. c Changes in the content of polar components in RBO and its blends during frying. d Changes in residual oil rate of RBO and its blends during the frying

Figure 3b described the POV of blended oils during frying, which indicated the oil chemical stability. With the extension of frying time, all the treatments showed three trends of evolution. The POV raised rapidly at the end of the first frying period, and then decreased, finally with an increasing trend, which was in line with the research results of Guillén et al. (2005). F2 had lower increase than F1, and they peaked around 3 h and 5 h, respectively. With the extension of time, the POV of F1 showed an upward trend after 18 h, which indicated that the frying oil was close to the critical value of use, while F2 had not shown an obviously increase at 30 h, this trend indicated that F2 performed better at the early oxidation stages. Because of the dynamic process of oil oxidation, the POV of oils was fluctuating with the continuously alternating of peroxide decomposition and synthesis. Therefore, the POV change depended on the effect of oxidation and decomposition of blended oils.

Figure 3c showed the PC of blended oils during frying. The PC curves of blended oils were rose, and peroxides decomposed into short-chain acids, aldehydes, ketones, alcohols, and volatile products at high-temperature. The threshold for PC lies between 25 and 27% (Gertz 2000), the PC 27 of F1 and F2 was 16 h and 15 h, respectively. Compared with RBO, F1 and F2 prolonged the PC exceeding duration by 60% and 50%, respectively. Therefore, blended oil had a significant effect on extending the frying duration.

Figure 3d provided the residual oil rate during the frying. The average residual oil rates of F1 and F2 were 9.32% and 8.07%, respectively, and the rate in F2 was relatively lower than others. In 180 °C frying temperature, the frying time was prolonged, and the viscosity of oil will decrease, thus promoting the oil absorption in French fries. With the frying, a hard shell would be formed on the outside of fries, so as to stop the oil absorption in French fries, resulting in a dynamic process of residual oil rate in French fries (Liu et al. 2021).

Nutrients retention and harmful substances of blended oils

Figure 4 showed the nutrients retention and harmful substances in samples. Figure 4 showed the contents of phytosterol and oryzanol during frying. It can be known that the phytosterol content of blended oils were generally decreased during the frying process. The overall downward trend of F1 was gentle, while that of F2 fluctuated wildly. Before 8 h, phytosterol content of F2 decreased sharply, then increased at about 8 h, and decreases at around 19 h, with a retention rate of 79.05%. At the end of frying, F1had a higher retention rate of phytosterol (90%). Oryzanol had potent antioxidant and free radical scavenging capacity (Latha and Nasirullah 2014; Mishra et al. 2012), the oryzanol retention rate of F2 (87.84%) was higher than F1 (79.26%), which had a good effect on the retention nutrients after blending. After frying, the oil retained part of the antioxidant activity, because of the loss of oryzanol was less. Sterol ester in phytosterol were prone to forming free phytosterol, which was occurred heat-accelerated esterification reaction with fatty acid. The ferulic acid triterpene alcohol and phytosterol may convert to oryzanol ester (Latha and Nasirullah 2014). Therefore, phytosterol had a downward trend, while the content change of oryzanol was less.

Fig. 4.

a Changes in the content of phytosterols and oryzanol in RBO and its blended oil during frying. b Changes in 3-MCPD ester content of RBO and its blends during frying. c Changes in the content of benzopyrene of RBO and its blends during frying. d Changes in trans fatty acid content in RBO and its blends during frying

Figure 4b showed the 3-MCPD ester concentration of blended oil during frying. 3-MCPD ester would be produced due to high-temperature treatment, the change of 3-MCPD ester concentration during frying needs to be controlled consequently (Qi et al. 2015). During the frying process, the 3-MCPD ester content showed a decreasing trend as a whole, indicating that there was a degradation reaction of 3-MCPD ester in the process of frying fries (Ilko et al. 2011). The 3- MCPD ester in F1 and F2 was reduced by 2.33 mg/kg and 2.18 mg/kg, respectively.

Figure 4c indicated the benzopyrene content of blended oils during frying. The result showed that the benzopyrene content of oils showed an overall downward trend. Neither of them exceeded the EU limit of 2 μg/kg. The blend oils produced fewer harmful substances. The benzopyrene content of F2 (1.41 μg/kg and 0.95 μg/kg) was lower than that of F1 (2.05 μg/kg and 1.90 μg/kg).

Figure 4d described the TFA content in blended oils during frying. The total TFA content of F1 decreased by 58.23% while F2 increased by 19.09%. Various complex chain reactions would occur during high-temperature frying, causing the double bond to be isomerized. In terms of growth rate, F1 was better than F2. Except for the C18:1T content in F1, C18:2T and C18:3T showed a decreasing trend, whereas the content of C18:1T, C18:2T and C18:3T in F2 showed an upward trend. Besides, the TFA content change of blended oils was also significantly lower than that of RBO, the blended oil had lower TFA content and there was more healthy than RBO.

Fatty acid composition and VE content in blend oils during frying

Table 3 showed the fatty acid composition and VE content in blended oils during frying. The stability of blended oils was relatively better than RBO. The fatty acid change of F1 was more significant than that of F2. The SFA and MUFA contents in F1 increased by 1.77% and 2.31% respectively, while PUFA content decreased by 4.17%. The presence of RBO in the mixture oil can significantly reduce the oxidation and thermal degradation reaction in unsaturated fatty acids (Wai et al. 2009).

Table 3.

Changes of fatty acid composition and VE content in RBO and its blended oil during frying

Frying oil	Frying time/h	α-Tocopherol (mg/100 g)	γ-Tocopherol (mg/100 g)	α-Tocotrienol (mg/100 g)	γ-Tocotrienol (mg/100 g)	Total (mg/100 g)	SFA/%	MUFA/%	PUFA/%
RBO	0	24.39	11.24	19.11	23.88	78.60	19.41	43.96	36.63
	12	15.51	1.34	6.91	7.23	31.00	21.04	43.31	36.63
	24	4.21	N.D	N.D	N.D	4.21	22.76	45.95	31.28
F1	0	24.80	6.80	15.17	17.71	64.50	31.23	41.47	27.30
	12	8.37	0.48	0.66	1.49	11.00	32.88	42.64	24.48
	24	N.D	N.D	N.D	N.D	N.D	33.00	43.78	23.13
F2	0	22.41	6.41	13.78	23.96	66.60	28.50	43.34	28.37
	12	10.84	1.53	3.46	5.61	21.40	28.38	43.89	27.73
	24	N.D	N.D	N.D	N.D	N.D	30.59	44.67	24.19

N.D. is not detected

The VE content of blended oils was not detected after 24 h, the disappearance rate of F2 was significantly faster than F1. This might be due to the higher amount of RBO in F1, and the contents of various natural active substances in F1 were more abundant.

In short, F2 was the better formula in research, which had better physicochemical properties and nutraceuticals retention. Moreover, it had lower residual oil rate and benzopyrene content.

Conclusion

In the present study, a multi-dimensional investigation of the frying performance was used in five single-product oils, then the principal component analysis method was used to provide blended oil, and RBO, CO and PO were selected. Compared with pure RBO, the AV of blended oils were lower, the PC 27% was prolonger in the frying process. After 30 h, the POV of blended oils was lower and the overall frying situation was better than that of RBO. The retention rate of oryzanol and tocopherol in F2 (87.84% and 32.13%) was higher than that in F1 (79.05% and 17.05%), while the retention rate of phytosterol in F1 (89.91%) was higher. F2 had higher nutritional value, better frying performance, and was more suitable for frying. This research guided the development of RBO blended oil as frying oil.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

RBO

Rice bran oil

PO

Palm oil

CO

Cottonseed oil

SuO

Sunflower oil

SO

Soybean oil

AV

Acid value

POV

Peroxide value

OSI

Oxidation stability index

VE

Vitamin E

PC

Polar component

PUFA

Polyunsaturated fatty acid

TFA

Trans fatty acids

SFA

Saturated fatty acid

MUFA

Monounsaturated fatty acid

PCA

Principal component analysis

GC

Gas chromatography

Author contributions

MZ: Conceptualization, Writing—Original Draft; YC: Formal analysis, Investigation; CH: Resources, Writing—Review and Editing, Supervision; DH: Resources, Writing—Review and Editing, Supervision; PG: Conceptualization, Resources, Writing—Original Draft.

Funding

The financial support provided by the project of Key Laboratory of Ministry of Education in Wuhan Polytechnic University (2020JYBQGDKFB03) is greatly appreciated.

Data availability

The data and material set supporting the materials and results of this article are included within the article.

Code availability

Not applicable.

Declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Ethics approval

Not applicable.

Consent to participate

Not applicable.

Consent for publication

The work described has not been published before; it is not under consideration for publication elsewhere; its publication has been approved by all co-authors; its publication has been approved by the responsible authorities at the institution where the work is carried out.