Skip to main content
Journal of Food Science and Technology logoLink to Journal of Food Science and Technology
. 2015 Jun 26;52(12):7683–7696. doi: 10.1007/s13197-015-1923-z

The changes in the volatile aldehydes formed during the deep-fat frying process

Qing Zhang 1,, Wen Qin 1, Derong Lin 1, Qun Shen 2, Ahmed S M Saleh 3
PMCID: PMC4648908  PMID: 26604343

Abstract

Volatile aldehydes (VAs) formed during soybean oil (SBO) heating, wheat dough (WD) frying, and chicken breast meat (CBM) frying processes were comparatively investigated by solid-phase micro-extraction-gas chromatography–mass spectrometry (SPME-GC-MS). The results showed that relative amounts (RAs) of the most detected VAs were firstly increased to maximum values in oil samples collected at the second hour of the seventh day and the values were then decreased with the increase in the time of oil heating process (control). However, for food frying processes, the time needed for reaching maximum RAs of VAs was shorter and the values were decreased with the increase in frying time. Significant change in contents of the VAs was observed for oil samples fried with CBM due to the high contents of water, protein, and lipid content compared to oil samples fried with WD. Based on the obtained results, free radical reaction, particularly positional isomerization and cis-trans isomerization, was deduced to occur when WD or CBM was fried in SBO. The relatively high RAs of VAs formed during the deep-fat frying process presented certain invaluable measures for evaluating of frying oil and fried food quality and safety.

Keywords: Soybean oil, Deep-fat frying, Frying oil, Volatile aldehyde, Solid-phase micro-extraction-gas chromatography–mass spectrometry

Introduction

In addition to attractive golden appearance and crispy structure, pleasant flavor is important for the popularity of fried food. However, the fume generated during food frying process is usually harmful to the health due to the generation of detrimental compounds, such as polycyclic aromatic hydrocarbons (Lee and Gany 2013), heterocyclic amines (Cao et al. 2013), and varying types of aldehydes (Dung et al. 2006). Therefore, the characterization of the aldehydes formed during the deep-fat frying process is very important from the viewpoint of food quality and food safety.

Generally, high amount of aldehydes can be formed during food frying process or oil heating process (Jerković et al. 2007; Ramírez et al. 2004). These aldehydes are mainly generated from the thermal oxidation of the polyunsaturated triacylglycerols (TAGs) (Zhang et al. 2012). However, the reaction substrates involved in the complex reactions are not only the constituents of frying oil, but also the components of fried food and the reactive intermediates. The Maillard reaction, for instance, occurred between amino-containing compound and carbonyl-containing product is the primary reaction involving the constituents of the fried food and frying oil (Zamora and Hidalgo 2005). Therefore, studying the composition of VAs during deep-fat frying process is important for understanding the mechanism of aldehydes formation.

Many studies have been carried out to characterize the types of VAs formed during deep-fat frying process (Katsuta et al. 2008; Katragadda et al. 2010). However, few studies have investigated the difference in VA profiles of varying types of food fried with SBO. Furthermore, a small number of systematical research studies have been done on the mechanisms of the typical VAs formation. Therefore, the contents of the volatile VAs formed during the thermal treatment of SBO without or with foods were comparatively investigated by SPME-GC-MS in this work.

Materials and methods

Materials and chemicals

Refined SBO, wheat flour, and chilled CBM were purchased from a local market. The major FAs of refined SBO expressed in mass ratio are as follows: 4.49 % stearic acid, 22.49 % oleic acid (O), 50.15 % linoleic acid (L), and 10.54 % linolenic acid (Ln).

Propyl propionate (99 %) and n-alkanes (C5-C20) standard solvent were purchased from Sigma-Aldrich. Chromatographically pure n-hexane was purchased from MREDA technology Ltd.

Sample preparation

WD was made by mixing the wheat flour and deionized water with a mass ratio of 2:1 in a mixer. WD was placed in a refrigerator at 4 °C overnight and was kneaded to a shape of 10 cm-diameter disc and thickness of 0.6 cm before the frying treatment. Fresh CBM samples was purchased according to the daily use amount and were manually cut into uniform strips with a sectional area of 0.6 × 0.6 cm. The gross chemical composition of the initial food was determined and expressed in mass percentage as follows: water, 39.00 %; protein, 9.11 %; lipid, 0.03 %; carbohydrates, 51.52 %; and ash, 0.34 % for WD; water, 72.20 %; protein, 19.80 %; lipid, 4.59 %; carbohydrates, 2.31 %; and ash, 1.10 % for CBM.

A deep-fat frying process was performed in a domestic scale electric fryer with double cylinder (HY-82EX, Guangzhou Huili Food Machinery Co., Ltd., China). Specifically, 5 L of SBO were put into the cylinder and heated to the set temperature (180 °C). No replenishment of fresh SBO was carried out during the whole deep-fat frying process. Food materials (100 g) were placed in a stainless screen frame and fully immersed into the heated SBO. Frying time was set as 10 min for WD and 5 min for CBM due to the difference in the nature of foods. The time interval between the frying batches in the WD-frying treatment was 5 min and was 10 min for the CBM-frying treatment. The frying process was performed for 8 h/day over a period of 7 consecutive days. SBO was heated under the same conditions as a control. Oil samples were collected in triplicate at 2 h intervals. Cold water was used to cool the oil samples to the room temperature prior to the SPME treatment.

SPME treatment of volatile aldehydes

Propyl propionate and hexane were mixed (1:9, v/v) to be used as the internal standard. Extraction of VAs was performed by using a 50/30 μm polydimethylsiloxane/divinylbenzene/carboxen SPME fiber mounted to an SPME manual holder assembly (Supelco, Bellefonte, PA, USA) and a constant temperature magnetic stirrer (PC-420, Corning Inc., Acton, MA, USA).

Oil sample (5 g) and internal standard solution (0.0060 g) were precisely weighed into a 15-mL SPME vial with a 1-cm rotor. The sample-loaded vial was sealed with a specialized cap and was then placed into a water bath maintained at 60 °C on a stirrer. Then, the manual holder assembly was inserted into the vial to expose the SPME fiber at the position of 1 cm over the oil surface. The sample loading and SPME fiber insertion were completed as quickly as possible. A total time of 50 min was taken to perform the VAs extraction and the oil sample was maintained under stirring at scale 5 on a magnetic stirrer. The headspace volatiles of the fresh SBO were also extracted under the same conditions.

GC-MS analysis of volatile aldehydes

Analysis of VAs was carried out by an Agilent gas chromatography mass spectrometer (7890A-5975C). After SPME extraction, the fiber was inserted into the injection port and desorbed for 2 min at the temperature of 250 °C. Helium (99.999 %) was used as the carrier gas at 1 mL/min. Separation of VAs was performed on a DB-5MS column (30 m × 0.25 mm × 0.25 μm, J&W Scientific) and in splitless mode for the best separation efficiency. The temperature program was set at 40 °C for 2 min and was then raised to 160 °C at ratio of 5 °C/min and held for 2 min, and was then raised to 250 °C at ratio of 10 °C/min and held for 10 min. Temperatures of the transfer line, ion source, and quadrupole mass filter were set at 250, 230, and 150 °C, respectively. The mass spectrum was acquired over a range of m/z30-450 by electron impact mode with ionization energy of 70 eV. The n-alkanes (C5-C20) standard were separated and analyzed under the same conditions in liquid injection mode with an injection volume of 1 μL.

VAs were qualitatively analyzed by consulting both the NIST08.L MS library (match quality >80 %) and the Flavornet and LRI&odour Database according to the calculated relative retention index (RI) (Bianchi et al. 2007). The relative quantification of the VAs was achieved by using the internal standard method.

Statistical analysis

Analysis of variance (ANOVA) was used to study the changes in contents of VAs during the deep-fat frying process. Data processing and analysis were carried out by SPSS 17.0 for Windows (SPSS Corp., Chicago, USA).

Results and discussion

As the major volatile compounds, aldehydes showed a large proportion of the detected volatiles in both the heat-treated oil (control) and the food frying oils (71.25, 66.54, and 71.77 % in the oil sample of 7–8 h of processes for oil heating, WD frying, and CBM frying, respectively). As shown in Table 1, both saturated aldehydes (alkanals and substitued alkanals) and unsaturated aldehydes (n-alkenals and alkadienals) were observed. Moreover, both cis and trans configurations are presented in the detected aldehydes.

Table 1.

The volatile aldehydes detected during the processes of soybean oil heating (control), WD frying, and CBM frying a,b

No. RT (min) Compounds (Systematic name) MQ (%) RI Possible precursor Proposed mechanism
1 6.729 3-Methylbutanal 83 917 Ln oxidation
2 8.180 Trans-2-pentenal 86 771 Ln-9-hydroperoxide further oxidation
3 9.289 Hexanal 91 2203 L-13-hydroperoxide; L-12-hydroperoxide β-homolysis; keto-enol tautomerism
4 10.941 Trans-2-hexenal 96 931 Ln-12-hydroperoxide keto-enol tautomerism; isomerization
5 12.274 Heptanal 87 16126 O-11-hydroperoxide further oxidation
6 14.002 Trans-2-heptenal 93 1023 L-12-hydroperoxide β-homolysis
7 14.051 Cis-2-heptenal 97 1031 trans-2-Heptenal cis-trans isomerization
8 14.456 Benzaldehyde 96 1131 L; 2,4-decadienal oxidation, dehydration; further oxidation
9 15.561 Trans,trans-2,4-Heptadienal 95 1006 Ln-12-hydroperoxide β-homolysis
10 17.206 Trans-2-octenal 82 1149 L-9-hydroperoxide; L-10-hydroperoxide further oxidation
11 18.550 Nonanal 86 1105 O-10-hydroperoxide; O-9-hydroperoxide β-homolysis; keto-enol automerism
12 20.196 Trans-2-nonenal 93 1256 L-9-hydroperoxide; 2-undecenal position isomerization; further oxidation
13 21.466 Decanal 84 1205 O-8-hydroperoxide keto-enol tautomerism
14 21.778 Trans,trans-2,4-nonadienal 95 1220 L-9-hydroperoxide; trans-2-nonenal further oxidation
15 22.645 4-Oxononanal 90 1298 L further oxidation
16 23.058 Trans-2-decenal 91 1387 O-9-hydroperoxide β-homolysis
17 24.315 Trans,trans-2,4-decadienal 96 1315 L-9-hydroperoxide β-homolysis
18 25.663 Trans-2-tridecenal 80 1476 trans-2-undecenal combining reaction
19 25.741 Trans-2-undecenal 80 1500 O-8-hydroperoxide β-homolysis; cis-trans isomerization
20 25.741 Trans-2-dodecenal 82 1500 trans-2-undecenal combining reaction
21 27.260 Trans,trans-2,4-undecadienal 88 1430 trans-2-undecenal further oxidation; desaturation
22 27.260 Trans,trans-2,4-dodecadienal 84 1430 trans-2-dodecenal further oxidation; desaturation
23 28.127 Cyclohexanecarboxaldehyde, 4-(hydroxymethyl)- 80 1519 structure-similar alkanes, alkenes oxidation
24 32.747 Cis,cis-10,12-hexadecadienal 93 1890 unsaturated radicals combination reaction
25 32.976 Trans-14-hexadecenal 84 2191 unsaturated FAs decarboxylic reaction; free radical reaction

a Abbreviations: WD wheat dough, CBM chicken breast meat, RT retention time, MQ matching quality, RI relative retention index, O oleic acid, L linoleic acid, Ln linolenic acid

b MQ was consulted from NIST08.L MS database. RI values calculated in relation to the C5-C20 n-alkanes for a DB-5 MS column

Based on the FA composition of SBO, all VAs are related to the free radical reaction occurred in the unsaturated FAs. Generally, the breakdown of ester bond in TAGs, such as thermal degradation and hydrolysis, easily occurs under frying condition to produce the free FAs (Chung et al. 2004). Therefore, the expression of primary oxidation product is in the form of free FAs. Different types of hydroperoxide are formed in the oxidation reaction of O, L, and Ln according to the varying reaction sites around the carbon-carbon double (C=C) bonds (Yin et al. 2011). Each hydroperoxide would further decompose into low-molecular-mass products, such as hydrocarbons, alcohols, aldehydes, ketones, and acids. Therefore, most of the detected aldehydes are either the direct products of oxidation of unsaturated FAs or the reaction products of the oxidative intermediates.

It is interesting that no volatile aldehyde was observed in the fresh SBO. Therefore, all the detected volatile aldehydes were probably produced during the oil heating or food frying treatments. Due to the continuity of the changes in RA of VAs, RAs of VAs of the oil samples which were collected at the fourth and sixth hour during the 7-days treatments were omitted for the succinct expression of the data.

Changes in alkanals content

As the β-homolysis cleavage product of L-13-hydroperoxide, hexanal can also be formed by keto-enol tautomerism of the combination product between 1-hexenyl radical degraded by β-homolysis of L-12-hydroperoxide and hydroxyl radical. As shown in Table 2, hexanal was first presented in the heat-treated oil sample of 2–8 h, meaning that it had not been formed or the yield was too low to be detected at the initial stage of the oil heating process (control). As the processing time increased, the RA of hexanal was significantly increased and reached a maximum value of 10.28 mg/kg in oil sample of 7–2 h. In contrast, hexanal was early detected in the food-fried oil samples, indicating that the reaction rate for formation of hexanal was higher at the early stage of the food frying process than that of the oil heating process (control). Relatively slow increase in RA of hexanal was observed in the WD-fried oil samples as the processing time increased. However, no obvious change in RA of hexanal was found for oil samples fried with CBM. This may be attributed to the relatively stable reaction environment during the heating process and the involving of more reaction substrates during the frying process. In other words, the formation of hexanal was influenced by the food components during the frying process.

Table 2.

The changes in amounts of alkanals formed during the soybean oil heating process and food frying process a,b

Alkanals Treatments 1–2 h 1–8 h 2–2 h 2–8 h 3–2 h 3–8 h 4–2 h 4–8 h 5–2 h 5–8 h 6–2 h 6–8 h 7–2 h 7–8 h
Hexanal Oil heating ND ND ND 0.48 ± 0.01b 1.23 ± 0.03a 1.25 ± 0.01b 2.28 ± 0.04a 1.93 ± 0.02a 4.45 ± 0.02a 8.54 ± 0.07a 4.02 ± 0.01a 4.45 ± 0.04a 10.28 ± 0.08a 8.10 ± 0.05a
WD frying 0.70 ± 0.38b 0.73 ± 0.22b 0.81 ± 0.19a 1.07 ± 0.23a 1.03 ± 0.32a 1.73 ± 0.13a 1.48 ± 0.23b 1.37 ± 0.19b 1.30 ± 0.24b 1.18 ± 0.11b 1.35 ± 0.22b 1.43 ± 0.22b 1.61 ± 0.36b 1.93 ± 0.20b
CBM frying ND 1.11 ± 0.36a 1.11 ± 0.22a 1.11 ± 0.23a 1.03 ± 0.32a 0.88 ± 0.32c 1.46 ± 0.26b 1.28 ± 0.37b 0.99 ± 0.35b 1.06 ± 0.14c 0.97 ± 0.38b 1.10 ± 0.33b 1.00 ± 0.20c 0.96 ± 0.10c
Heptanal Oil heating ND ND ND ND ND ND ND ND ND ND ND ND ND 3.89 ± 0.080a
WD frying ND ND ND ND ND ND ND ND ND ND ND ND ND 0.07 ± 0.01c
CBM frying ND ND ND ND ND ND ND ND ND ND ND ND ND 1.29 ± 0.45b
Nonanal Oil heating 2.54 ± 0.20b 3.44 ± 0.32a 4.21 ± 0.22a 6.12 ± 0.27a 5.53 ± 0.28a 6.45 ± 0.32a 7.06 ± 0.23a 8.38 ± 0.31a 9.09 ± 0.26a 7.04 ± 0.41a 10.20 ± 0.32a 9.99 ± 0.28a 8.34 ± 0.21a 8.38 ± 0.32a
WD frying 3.11 ± 0.34a 3.56 ± 0.51a 4.59 ± 0.42a 4.27 ± 0.54c 3.95 ± 0.49b 6.02 ± 0.54a 4.87 ± 0.65b 5.06 ± 0.19c 5.54 ± 0.68b 5.05 ± 0.17b 6.29 ± 0.54b 6.24 ± 0.52b 5.80 ± 0.63b 6.52 ± 0.63b
CBM frying 1.89 ± 0.22c 3.64 ± 0.22a 4.30 ± 0.26a 5.25 ± 0.28b 5.21 ± 0.24a 4.71 ± 0.25b 4.61 ± 0.28b 5.57 ± 0.24b 5.65 ± 0.23b 4.45 ± 0.32c 4.66 ± 0.22c 5.15 ± 0.34c 5.75 ± 0.37b 5.46 ± 0.23b
Decanal Oil heating ND ND ND ND 0.41 ± 0.10 0.70 ± 0.07 0.63 ± 0.11 ND ND ND 1.14 ± 0.11 0.95 ± 0.07 0.87 ± 0.08 0.81 ± 0.06

a Abbreviations: WD wheat dough, CBM chicken breast meat, 12 h, oil sample collected at the second hour of the first day, ND not detected

b The relative contents of alkanals were express as average value ± confidence interval. The lowercase (a,b,c) behind the content represents the significant difference among the three thermal treatments (p < 0.05, n = 3)

Heptanal is not the direct product of triolein, trilinolein, and trilinolenin oxidation (Selke and Rohwedder 1983; Selke et al. 1977, 1980). Therefore, heptanal was not observed during both the oil heating process (control) and the food frying processes except the last stage of these treatments (oil sample of 7–8 h). Based on the possible oxidation products of unsaturated FAs, the precursor of heptanal might be heptanol decomposed from O-11-hydroperoxide.

Nonanal is not only the β-homolysis product of O-10-hydroperoxide but also the keto-enol tautomerism product of the combination product between 1-nonenyl radical degraded from O-9-hydroperoxide and hydroxyl radical. The RAs of nonanal during the oil heating process (control) were greater than those during the food frying process, indicating that the formation of nonanal was also influenced by the food components. Therefore, the formation rate of nonanal was firstly increased and was then decreased to be lower than the reaction rate of consequent reactions of nonanal as the processing time increased.

As a keto-enol tautomerism product of the combination product between 1-decenyl radical decomposed from O-8-hydroperoxide and hydroxyl radical, decanal was discontinuously observed in the heat-treated oil samples (control) and was not present in the food-fried oil samples. Therefore, decanal is unstable and can easily contribute to the consequent reactions during the thermal treatment of SBO (particularly the deep-fat frying process).

Changes in n-alkenals content

As presented in Table 3, trans-2-pentenal was only observed in a relatively low amount in few heated oil samples during the oil heating process (control). Trans-2-pentenal was previously considered the decomposition products of Ln (Soheili et al. 2002). However, according to the composition of secondary oxidation products (SOPs) of Ln-9-hydroperoxide, trans-2-pentenal might be the further oxidation product of cis,cis-3,6-nonadienal, which is unstable under conditions of high temperature and oxygen. Trans-2-pentenal was not observed during the frying process, which could be attributed to its high volatility under the increased reaction intensity condition.

Table 3.

The changes in amounts of n-alkenals formed during the soybean oil heating process and food frying process a,b

n-Alkenals Treatments 1–2 h 1–8 h 2–2 h 2–8 h 3–2 h 3–8 h 4–2 h 4–8 h 5–2 h 5–8 h 6–2 h 6–8 h 7–2 h 7–8 h
Trans-2-pentenal Oil heating ND ND ND ND ND ND ND ND ND 0.09 ± 0.01 0.03 ± 0.01 ND ND 0.11 ± 0.02
Trans-2-hexenal Oil heating 0.40 ± 0.09b 0.34 ± 0.11b 0.52 ± 0.10a 0.41 ± 0.14a 0.33 ± 0.11a 0.29 ± 0.10b 0.75 ± 0.13a 1.13 ± 0.34a 0.66 ± 0.44a 0.89 ± 0.13a 0.54 ± 0.11a 0.89 ± 0.24a 1.30 ± 0.31a 0.41 ± 0.12b
WD frying 0.32 ± 0.06b 0.24 ± 0.07b 0.03 ± 0.02c 0.33 ± 0.04a 0.27 ± 0.09a 0.49 ± 0.05a 0.45 ± 0.04b 0.33 ± 0.05b 0.20 ± 0.05c 0.41 ± 0.07b 0.19 ± 0.04b 0.13 ± 0.08b 0.24 ± 0.09b 0.61 ± 0.07a
CBM frying 0.60 ± 0.08a 0.59 ± 0.06a 0.26 ± 0.06b 0.22 ± 0.04b 0.19 ± 0.06a 0.22 ± 0.04b 0.25 ± 0.08c 0.35 ± 0.07b 0.37 ± 0.04b 0.21 ± 0.06c 0.40 ± 0.07a 0.17 ± 0.06b 0.33 ± 0.05b 0.14 ± 0.06c
Cis-2-heptenal Oil heating 12.17 ± 1.31a 14.29 ± 1.09b 15.75 ± 0.93a 16.89 ± 1.39a 17.90 ± 1.23a 21.11 ± 1.94a 24.81 ± 1.26a 23.09 ± 1.48a 23.86 ± 1.33a 27.94 ± 1.29a 22.86 ± 1.40a 22.10 ± 1.46a 25.22 ± 2.01a 23.66 ± 1.36a
WD frying ND ND ND 15.88 ± 0.82a 14.23 ± 0.72b 21.82 ± 1.53a 14.92 ± 1.06b 14.61 ± 0.49b 14.51 ± 0.82b 13.40 ± 0.10b ND ND ND ND
CBM frying 12.64 ± 0.09a 17.14 ± 0.10a 15.46 ± 0.06a 16.10 ± 0.03a 14.14 ± 0.04b 13.19 ± 0.05b 12.78 ± 0.03c 10.07 ± 0.7c 11.46 ± 0.03c 8.72 ± 0.06c 9.01 ± 0.03b 7.87 ± 0.05b 8.40 ± 0.05b 6.63 ± 0.05b
Trans-2-heptenal WD frying 20.31 ± 0.72 19.13 ± 1.07 19.29 ± 1.16 ND ND ND ND ND ND ND 14.32 ± 0.92 12.87 ± 0.49 12.97 ± 1.00 12.79 ± 0.82
Trans-2-octenal Oil heating 6.73 ± 0.51c 10.25 ± 0.42b 11.48 ± 0.62b 12.06 ± 0.46a 12.71 ± 0.34a 14.34 ± 0.41a 16.01 ± 0.51a 12.68 ± 0.57a 14.24 ± 0.41a 19.08 ± 0.57a 13.63 ± 0.58a 13.33 ± 0.44a 19.89 ± 0.71a 17.09 ± 0.61a
WD frying 11.05 ± 0.72a 11.63 ± 0.67a 13.10 ± 0.82a 10.27 ± 0.82b 9.12 ± 0.65b 13.12 ± 0.63b 10.03 ± 0.91b 9.38 ± 0.37b 9.71 ± 0.72b 8.42 ± 0.76b 9.54 ± 0.73b 8.57 ± 0.78b 7.83 ± 0.58b 8.65 ± 0.67b
CBM frying 7.60 ± 0.28b 7.47 ± 0.18c 7.74 ± 0.17c 8.50 ± 0.14c 7.78 ± 0.17c 7.07 ± 0.26c 6.97 ± 0.10c 8.30 ± 0.24c 8.11 ± 0.36c 6.17 ± 0.19c 6.05 ± 0.17c 5.91 ± 0.46c 6.23 ± 0.33c 5.37 ± 0.16c
Trans-2-nonenal Oil heating 0.91 ± 0.20a 1.40 ± 0.18a 1.92 ± 0.18a 2.75 ± 0.19a 2.67 ± 0.20a 3.83 ± 0.20a 4.56 ± 0.18a 3.52 ± 0.18a 4.91 ± 0.20a 5.52 ± 0.23a 4.75 ± 0.22a 4.48 ± 0.22a 6.65 ± 0.16a 5.20 ± 0.18a
WD frying 0.76 ± 0.13a 1.24 ± 0.31ab 1.59 ± 0.32a 1.36 ± 0.29b 1.33 ± 0.37c 2.55 ± 0.29b 2.11 ± 0.38b 2.18 ± 0.32c 2.34 ± 0.24c 2.12 ± 0.26b 2.99 ± 0.29b 2.77 ± 0.33b 2.45 ± 0.31c 3.28 ± 0.22b
CBM frying 0.28 ± 0.05b 0.88 ± 0.06b 1.21 ± 0.05b 1.67 ± 0.13b 1.93 ± 0.08b 2.30 ± 0.06b 2.21 ± 0.10b 3.15 ± 0.10b 3.25 ± 0.30b 2.43 ± 0.10b 2.52 ± 0.11c 2.60 ± 0.22b 3.12 ± 0.17b 2.61 ± 0.10c
Trans-2-decenal Oil heating 3.09 ± 0.69a 5.15 ± 0.67a 5.61 ± 0.51a 8.62 ± 0.94a 7.85 ± 0.88a 10.54 ± 1.08a 11.51 ± 1.00a 9.59 ± 1.03a 12.23 ± 1.09a 11.15 ± 0.10a 11.06 ± 1.26a 9.64 ± 1.11a 18.41 ± 1.15a 9.11 ± 0.10a
WD frying 2.66 ± 0.33b 3.27 ± 0.28b 4.21 ± 0.37b 3.12 ± 0.40b 2.38 ± 0.34c 3.73 ± 0.29b 3.69 ± 0.27b 3.54 ± 0.28b 3.41 ± 0.31b 2.66 ± 0.32b 3.90 ± 0.32b 3.61 ± 0.41b 2.44 ± 0.44b 3.30 ± 0.31b
CBM frying 1.10 ± 0.04c 2.73 ± 0.09c 3.02 ± 0.05c 3.13 ± 0.07b 3.29 ± 0.08b 2.67 ± 0.08c 2.47 ± 0.06c 3.2 ± 0.07b ND ND ND ND ND ND
Trans-2-undecenal Oil heating 1.55 ± 0.33a ND ND 5.87 ± 0.73a 4.70 ± 0.44a 7.32 ± 1.07a 8.00 ± 1.01a 6.71 ± 0.88 ND ND 7.78 ± 1.34 6.37 ± 1.01 15.48 ± 2.01 ND
WD frying 1.19 ± 0.04b 2.43 ± 0.05a 3.04 ± 0.07a 2.17 ± 0.08b 1.56 ± 0.08c 2.55 ± 0.05b 2.52 ± 0.06b ND ND ND ND ND ND ND
CBM frying 0.44 ± 0.04c 1.59 ± 0.05b 2.26 ± 0.06b 1.95 ± 0.05c 1.95 ± 0.07b ND ND ND ND ND ND ND ND ND
Trans-2-dodecenal Oil heating ND 2.87 ± 0.41 3.27 ± 0.49 ND ND ND ND ND 8.83 ± 1.01 7.57 ± 1.04 ND ND ND 6.23 ± 0.99
Trans-2-tridecenal WD frying ND ND ND ND ND ND ND 2.49 ± 0.05a 2.41 ± 0.08a 1.64 ± 0.04a 2.84 ± 0.08a 2.37 ± 0.09a 2.48 ± 0.06a 2.05 ± 0.07a
CBM frying ND ND ND ND ND 1.55 ± 0.04 1.34 ± 0.05 1.88 ± 0.07b 2.13 ± 0.06b 1.21 ± 0.05b 1.22 ± 0.06b 1.11 ± 0.09b 1.44 ± 0.10b 0.89 ± 0.07b
Trans-14-hexadecenal Oil heating 0.02 ± 0.01 ND ND ND ND ND ND ND ND ND ND ND 0.25 ± 0.02 ND
WD frying ND ND 0.05 ± 0.02 0.02 ± 0.01 0.27 ± 0.03 ND ND ND ND ND ND ND ND ND
CBM frying ND ND ND ND ND ND ND ND ND ND ND ND ND 0.04 ± 0.01

a Abbreviations: WD wheat dough, CBM chicken breast meat, 12 h oil sample collected at the second hour of the first day, ND not detected

b The relative contents of n-alkenals were express as average value ± confidence interval. The lowercase (a,b,c) behind the content represents the significant difference among the three thermal treatments (p < 0.05, n = 3)

Based on the classical lipid oxidation mechanism (Yin et al. 2011), trans-2-hexenal is not the direct SOP of unsaturated FAs. Due to the features of lipid oxidation, trans-2-hexenal might be formed from cis-3-hexenal by cis-trans isomerization and electronic rearrangement. Cis-3-hexenal can be formed by the keto-enol tautomerism of the combination product between 1,3-hexadienyl radical degraded from Ln-12-hydroperoxide and hydroxyl radical. Therefore, the RAs of trans-2-hexenal formed during the heating and frying processes are lower than those of other major directly oxidative degraded 2-alkenals. The RA of trans-2-hexenal was increased to a maximum value of 1.30 mg/kg in the oil sample of 7–2 h and was then decreased as the processing time increased during the oil heating process (control). However, reduced RAs of trans-2-hexenal were observed during the frying process, especially in the CBM-fried oil samples. It is interesting that the RA of trans-2-hexenal was increased during the WD frying process; however, it was decreased during the CBM frying process as the processing time increased. This might be attributed to the relatively more reaction substrates involved in the CBM frying process.

Trans-2-heptenal, β-homolysis product of L-12-hydroperoxide, might be transformed to cis-2-heptenal by cis-trans isomerization under the adopted experimental conditions (Frankel 2005a). Furthermore, 2-heptenal can also be formed from the decomposition of 2,4-decadienal (Matthews et al. 1971). Cis-2-heptenal was observed during the whole CBM frying process and the interim (from the oil sample of 2–8 h to 5–8 h) of the WD frying process; however, trans-2-heptenal was found during the whole oil heating process (control) and the early and the last stages of WD frying process. This discrepancy might be related to the differences in nature of the reaction substrates involved in the treatments of control, WD-frying, and CBM-frying, and can be also explained by the difference in composition of products formed during the frying process of varying types of food. The RAs of cis-2-heptenal were gradually increased and was then decreased during the oil heating (control) and CBM frying processes. The difference between these two treatments was the occurrence time of maximum value of this product.

The formation of trans-2-octenal can’t be fully explained by the classical oxidation of unsaturated FAs. Therefore, trans-2-octenal was regarded as the oxidative decomposition product (ODP) of 3-nonenal which was formed by the β-homolysis of L-9-hydroperoxide (Frankel 2005b). Oxidation occurred in the methylene between carbonyl and C=C bonds, shortening the carbon chain and forming the 2-octenal. On the other hand, oxidation could also take place in the hydroxyl of cis-2-octen-1-ol, which is the β-homolysis product of L-10-hydroperoxide, producing trans-2-octenal under the conditions of high temperature and occurrence of oxygen. Based on the RAs of trans-2-octenal, the oxidation can intensively take place in the unsaturated SOP of frying oil. Similar trend of change in RA of trans-2-octenal, which was first increased and was then decreased as the processing time increased, was observed during the varying frying treatments. The major difference was the time needed for reaching the maximum value of trans-2-octenal, which also reflect the complexity of the food frying process compared to the control.

3-Nonenal, the combination product of 1,3-nonadienyl radical decomposed from L-9-hydroperoxide and hydroxyl radical, can be converted into trans-2-nonenal by positional isomerization of C=C bonds under the high temperature condition (Frankel et al. 1981). Additionally, trans-2-nonenal was also considered the ODP of 2-undecenal (Warner et al. 2001). An increase and then a decrease in RA of trans-2-nonenal were also observed during the oil heating process (control) and CBM frying process. However, the RA of this aldehyde was increased as the processing time increased during the WD frying process.

Trans-2-decenal is the β-homolysis product of O-9-hydroperoxide. As a direct SOP of O, the RA of trans-2-decenal was increased to a maximum value of 18.41 mg/kg for the oil sample of 7–2 h during the oil heating process (control). Along with the proceeding of heating, trans-2-decenal could contribute to the subsequent reactions, resulting in a decrease in its RA in the frying oil. The RA of trans-2-decenal during the WD frying process was increased to a maximum value of 4.21 mg/kg for the oil sample of 2–2 h and then no significant changes were observed during the rest frying treatment. This might be attributed to the balance between the production and the consumption rate of this aldehyde during the WD frying process. The change profile of trans-2-decenal was different for the CBM frying process. Trans-2-decenal was not observed after the maximum value of 3.32 mg/kg at oil sample of 4–8 h. In addition, the unsaturated SOP could be fully involved in the subsequent reactions during the CBM frying process when the frying treatment was carried out in a certain time.

Similar to trans-2-decenal, trans-2-undecenal is the SOP of O-8-hydroperoxide. Therefore, the RA of trans-2-undecenal was first increased and was then decreased as the processing time increased during the oil heating process (control). Interestingly, trans-2-undecenal was not observed in some oil samples (1–8 h, 2–2 h, 5–2 h, 5–8 h, and 7–8 h) and trans-2-dodecenal was alternatively observed in these samples. First, trans-2-dodecenal might be regarded as the reaction product of trans-2-undecenal under the high temperature condition. Second, due to the same retention time of these two aldehydes, the detected RA might be the total value of these two compounds. An increase and then a decrease (or absence) in RA of trans-2-undecenal was also observed during the frying process. This reflected the major influence resulted from the food components on the unsaturated SOP of frying oil. Thus, trans-2-tridecenal was observed in the food-fried oil samples where no trans-2-undecenal was detected. Furthermore, the RA of trans-2-tridecenal was first increased and was then decreased along with the proceeding of food frying. Comparing with the changes in RA of trans-2-tridecenal and trans-2-undecenal, trans-2-tridecenal might be the ODP resulted from the combination between trans-2-undecenal and alkyl radicals formed under high temperature condition. This can also explain that the addition of food components makes the reactions more intensive and the products more complex during the frying process.

Changes in alkadienals content

Polyunsaturated aldehydes with conjugated diene structure were largely formed during both the oil heating (control) and the food frying processes, which is in agreement with previous studies (Katragadda et al. 2010; Ramírez et al. 2004). Polyunsaturated aldehydes observed in this study included 2,4- and 10,12-alkadienals based on the position of C=C bonds. 2,4-Alkadienals are the most important aldehydes formed during the food frying process from the viewpoint of fried food safety (Dung et al. 2006).

As presented in Table 4, the RA of trans,trans-2,4-heptadienal was first increased to a maximum value of 42.53 mg/kg in oil sample of 5–8 h and was then decreased along with the proceeding of oil heating. However, the maximum RAs of trans,trans-2,4-heptadienal were 38.00 mg/kg in oil sample of 1–2 h and 24.55 mg/kg in oil sample of 1–8 h for the WD and CBM frying processes, respectively. In addition, the time for reaching the maximum value of trans,trans-2,4-heptadienal was significantly short. This can also be attributed to the formation and consumption rates of this aldehyde when the food components were added during the frying process. As a direct β-homolysis product of Ln-12-hydroperoxide, trans,trans-2,4-heptadienal were also quantitatively reported during the heating of extra virgin olive oil at temperature above 180 °C (Katragadda et al. 2010). Therefore, other formation mechanism for the generation of trans,trans-2,4-heptadienal during the heating or frying process is potential.

Table 4.

The changes in amounts of 2,4- and 10,12-alkadienals formed during the soybean oil heating process and food frying process a,b

2,4- and 10,12-Alkadienals Treatments 1–2 ha 1–8 h 2–2 h 2–8 h 3–2 h 3–8 h 4–2 h 4–8 h 5–2 h 5–8 h 6–2 h 6–8 h 7–2 h 7–8 h
Trans,trans-2,4-heptadienal Oil heating 29.74 ± 1.36b 32.25 ± 1.73a 33.18 ± 0.22b 36.27 ± 2.14a 35.42 ± 2.12a 36.80 ± 1.81a 40.02 ± 1.94a 33.12 ± 1.54a 36.80 ± 2.36a 42.53 ± 2.00a 33.52 ± 1.70a 27.81 ± 1.91a 41.92 ± 2.09a 27.16 ± 1.44a
WD frying 38.00 ± 0.98a 34.48 ± 1.26a 36.70 ± 1.04a 27.26 ± 1.00b 21.90 ± 0.87b 33.93 ± 1.23b 26.16 ± 2.19b 23.09 ± 0.98b 23.45 ± 1.52b 18.87 ± 1.67b 21.73 ± 1.55b 18.38 ± 1.56b 16.39 ± 1.29b 15.65 ± 2.17b
CBM frying 19.84 ± 1.98c 24.55 ± 1.31b 23.74 ± 2.10c 23.89 ± 1.22c 22.36 ± 1.45b 19.53 ± 1.82c 18.88 ± 0.89c 16.23 ± 1.76c 18.33 ± 1.42c 10.56 ± 0.88c 12.78 ± 1.02c 11.71 ± 1.01c 12.39 ± 0.88c 9.83 ± 0.98c
Trans,trans-2,4-nonadienal Oil heating 0.31 ± 0.11a 0.48 ± 0.15b 0.75 ± 0.10b 1.48 ± 0.35a 1.33 ± 0.37a 2.49 ± 0.32a 3.06 ± 0.42a 2.13 ± 0.28a 3.11 ± 0.29a 3.61 ± 0.10a 3.10 ± 0.33a 2.72 ± 0.18a 6.40 ± 0.25a 2.74 ± 0.32a
WD frying 0.26 ± 0.10a 0.77 ± 0.05a 1.35 ± 0.10a 1.00 ± 0.21a 0.86 ± 0.22ab 1.00 ± 0.15b 1.26 ± 0.33b 1.30 ± 0.16b 1.40 ± 0.25b 1.02 ± 0.09b 1.58 ± 0.22b 1.40 ± 0.20b 0.96 ± 0.26b 1.24 ± 0.15b
CBM frying ND ND 0.61 ± 0.06b 0.73 ± 0.05b 0.79 ± 0.08b 0.59 ± 0.05c 0.58 ± 0.02c 0.76 ± 0.06c 0.94 ± 0.05c 0.54 ± 0.02c 0.55 ± 0.02c 0.55 ± 0.03c 0.56 ± 0.04c 0.41 ± 0.06c
Trans,trans-2,4-decadienal Oil heating 46.13 ± 4.15a 78.77 ± 2.32a 85.87 ± 4.76a 127.46 ± 2.08a 104.60 ± 3.23a 133.13 ± 3.81a 139.70 ± 4.22a 109.29 ± 4.08a 140.03 ± 3.34a 109.67 ± 4.00a 119.60 ± 2.77a 89.94 ± 3.72a 185.07 ± 1.29a 80.39 ± 3.12a
WD frying 35.66 ± 3.11b 46.27 ± 1.66b 58.98 ± 3.62b 40.93 ± 1.34b 30.09 ± 3.12c 48.82 ± 2.94b 49.67 ± 1.49b 44.05 ± 2.19b 42.46 ± 3.61b 28.17 ± 2.44c 44.35 ± 3.20b 38.73 ± 2.16b 23.44 ± 3.47b 31.08 ± 2.20b
CBM frying 15.76 ± 1.98c 36.46 ± 1.56c 40.02 ± 2.03c 37.83 ± 1.88c 39.90 ± 3.00b 35.29 ± 1.35c 31.80 ± 3.00c 43.26 ± 1.85b 44.12 ± 2.11b 22.21 ± 1.87c 22.71 ± 1.02c 20.03 ± 2.00c 25.20 ± 1.25c 16.00 ± 3.11c
Trans,trans-2,4-undecadienal Oil heating ND ND ND 0.47 ± 0.09 0.14 ± 0.07 0.43 ± 0.10 ND 0.36 ± 0.08 ND ND 0.67 ± 0.11 0.35 ± 0.09 1.10 ± 0.10 0.65 ± 0.08
Oil heating ND ND ND ND ND ND 0.41 ± 0.04 ND 0.74 ± 0.05 0.34 ± 0.05 ND ND ND ND
Trans,trans-2,4-dodecadienal WD frying ND ND ND 0.06 ± 0.02 0.27 ± 0.05a ND ND ND ND ND ND ND ND ND
CBM frying 0.01 ± 0 ND ND ND 0.03 ± 0.01b 0.05 ± 0 ND ND ND ND ND ND ND ND
Cis,cis-10,12-hexadecadienal Oil heating ND ND ND 0.10 ± 0.03 0.12 ± 0.04 0.09 ± 0.03 0.15 ± 0.04 0.13 ± 0.04a 0.29 ± 0.05a ND 0.16 ± 0.03 ND 0.71 ± 0.04 ND
WD frying ND ND ND ND ND ND ND ND ND 0.04 ± 0.01 ND ND ND ND
CBM frying ND ND ND ND ND ND ND 0.06 ± 0.01b 0.06 ± 0.02b ND ND ND ND 0.05 ± 0.02

a Abbreviations: WD wheat dough, CBM chicken breast meat, 12 h oil sample collected at the second hour of the first day, ND not detected

b The relative contents of 2,4- and 10,12-alkadienals were express as average value ± confidence interval. The lowercase (a,b,c) behind the content represents the significant difference among the three thermal treatments (p < 0.05, n = 3)

The changes in RA of trans,trans-2,4-nonadienal were similar during both the oil heating (control) and food frying processes except for the occurrence time of the maximum RA of this aldehyde. Trans,trans-2,4-nonadienal is not the direct reaction product of the unsaturated FAs based on the classical lipid oxidation mechanism, resulting in relatively lower RAs of this type of conjugated aldehyde than those of other SOP during the oil heating (control) and food frying processes. It was proposed that trans,trans-2,4-nonadienal was the peroxidation product of 3-nonenal generated from the decomposition of L-9-hydroperoxide (Warner et al. 2001). Hydroperoxidation occurs at the C5 position of 3-nonenal and intramolecular rearrangement of electronic then takes place to form 2-nonenal-4-hydroperoxide. Loss of a hydrogen peroxide leads to produce trans,trans-2,4-nonadienal. Furthermore, the aforementioned trans-2-nonenal can be directly formed from 3-nonenal. Comparing the changes in RAs of trans,trans-2,4-nonadienal and trans-2-nonenal, the same trend of change was observed for these two aldehydes. Therefore, trans,trans-2,4-nonadienal could be directly formed by the hydroperoxidation of trans-2-nonenal under the conditions of high temperature and oxygen. Briefly, hydroperoxidation occurs at the C4 position of trans-2-nonenal to form the trans-2-nonenal-4-hydroperoxide and the loss of hydrogen peroxide then occurs to directly produce the trans,trans-2,4-nonadienal. According to this proposed mechanism, the formation of trans,trans-2,4-heptadienal can be considered the peroxidation product of trans-2-heptenal during the processes of oil heating and food frying.

In general, because the trans,trans-2,4-decadienal is the decomposition product of L-9-hydroperoxide (Porter et al. 1995), its RA was the highest among the detected VAs during the theating of SBO. In addition, according to the above-mentioned formation mechanism of 2,4-dienal from 2-enal, trans,trans-2,4-decadienal can also be formed by the peroxidation of trans-2-decenal. It's worth noting that hydroperoxidation and decomposition reaction can also take place in 2,4-decadienal under the conditions of high temperature and occurrence of oxygen to form some lower-molecular-weight decomposition products, such as pentanal, 2-octanal, and 2-nonanal (Yu et al. 1994). This might be the reason for the changes in RAs of trans,trans-2,4-decadienal during the oil heating (control) and the food frying processes (Shown in Table 4). Oil samples of 7–2 h, 2–2 h, and 5–2 h showed the highest RAs of trans,trans-2,4-decadienal during the oil heating (control), WD frying, and CBM frying processes, respectively. However, the changes in RA of trans,trans-2,4-decadienal could be described as the alternation of the increase and the decrease during all the thermal treatments. The food components intensified the frying-induced chemical reactions to make trans,trans-2,4-decadienal contribute to the consequent reactions. The occurrence time of the maximum RA of trans,trans-2,4-decadienal for the CBM frying process was later compared with that for the WD frying process, which might be attributed to the relatively high content of in CBM lipids. Furthermore, the relatively lower RA of trans,trans-2,4-decadienal detected during the CBM frying process may be attributed to the relatively high content of CBM proteins due to the possible Maillard reaction taken place between this 2,4-dienal and the amino-containing compounds. Moreover, the absorption of trans,trans-2,4-decadienal by the fried food might lead to a decrease in its RAs in the frying oil (Boskou et al. 2006).

It is interesting that trans,trans-2,4-undecadienal and trans,trans-2,4-dodecadienal showed the same retention time and were alternatively presented during the oil heating process (control). However, trans, trans-2,4-dodecadienal was only observed for some oil samples during the frying process. A maximum RA of 1.10 mg/kg was found for trans,trans-2,4-undecadienal in the oil sample of 7–2 h. The RAs of trans,trans-2,4-dodecadienal were relatively low. Based on the aforementioned mechanism of 2,4-dienal from the 2-enal, trans,trans-2,4-undecadienal and trans,trans-2,4-dodecadienal could be the desaturation products of trans-2-undecenal and trans-2-dodecenal, respectively. Along with the addition of food components, such as water, protein, carbohydrate, and lipid, the oxidation intensity of unsaturated FAs was increased and the SOP were more prone to be involved in further oxidation or been reacted with each other. Therefore, the precursors of these two 2,4-dienals might be involved in the formation of other products. The formation of trans-2-tridecenal is the case for the other reaction occurred in trans-2-undecenal during the food frying process.

Cis,cis-10,12-hexadecadienal is the sole polyunsaturated aldehyde with two cis configurations. As presented in Table 4, cis,cis-10,12-hexadecadienal was mainly observed during the oil heating process (control) and reached a maximum RA of 0.71 mg/kg in the oil sample of 7–2 h. However, cis,cis-10,12-hexadecadienal was only presented in the oil sample of 5–8 h for the WD frying process and in the oil samples of 4–8 h, 5–2 h, and 7–8 h for the CBM frying process. From the molecular structure of this aldehyde, cis,cis-10,12-hexadecadienal is not the ODP of any unsaturated FAs contained in SBO. Nevertheless, based on the structures of the reaction substrates and the main reaction type (i.e., free radical reaction) involved in the heating and frying processes, cis,cis-10,12-hexadecadienal could be the combined product of the unsaturated radicals. Cis,cis-10,12-hexadecadienal might be unstable under the conditions of high temperature and oxygen; therefore, it is rarely to be observed during the frying process.

Changes in substituted alkanals content

Except for the linear chain aldehydes, several aldehydes with substituent were observed during the thermal treatments of SBO. As shown in Table 5, four substituted aldehydes belong to the saturated aldehydes were detected. Comparing to the above-mentioned aldehydes, the substituted aldehydes are not the direct products of the unsaturated FAs oxidation but are related to the free radical reaction.

Table 5.

The changes in amounts of substituted alkanals formed during the soybean oil heating process and food frying process a,b

Substituted alkanals Treatments 1–2 h 1–8 h 2–2 h 2–8 h 3–2 h 3–8 h 4–2 h 4–8 h 5–2 h 5–8 h 6–2 h 6–8 h 7–2 h 7–8 h
Benzaldehyde Oil heating 0.63 ± 0.01b 1.16 ± 0.01 1.34 ± 0.02a 1.45 ± 0.02a 0.65 ± 0.01 1.55 ± 0.02 1.83 ± 0.03 1.83 ± 0.02a 1.91 ± 0.01 ac 1.97 ± 0.02a 1.86 ± 0.01a 1.89 ± 0.04a 2.04 ± 0.08a 1.60 ± 0.01a
WD frying ND ND ND 0.22 ± 0.03b ND ND ND ND 0.72 ± 0.02b ND 0.09 ± 0.04c 0.65 ± 0.06b 0.08 ± 0.08c ND
CBM frying 0.79 ± 0.03a ND 0.96 ± 0.04b ND ND ND ND 0.83 ± 0.05b 0.63 ± 0.01c 0.67 ± 0.04b 0.70 ± 0.07b 0.73 ± 0.06b 0.76 ± 0.08b 0.70 ± 0.08b
Cyclohexanecarboxaldehyde, 4-(hydroxymethyl)- Oil heating ND ND ND ND ND ND ND ND ND ND ND ND 0.80 ± 0.04a 0.25 ± 0.04b
3-Methylbutanal Oil heating ND ND ND ND ND 0.17 ± 0.04 0.05 ± 0.02 ND 0.28 ± 0.03 0.43 ± 0.03 0.49 ± 0.03 ND 0.02 ± 0.01c ND
WD frying ND ND ND ND ND ND ND ND ND ND ND 0.34 ± 0.07a 0.43 ± 0.07a 0.39 ± 0.07a
CBM frying ND ND ND ND ND ND ND ND ND ND ND 0.03 ± 0.01b 0.07 ± 0.02b 0.12 ± 0.03b
4-Oxononanal Oil heating 0.95 ± 0.32ab 1.46 ± 0.38b 2.01 ± 0.49b 4.52 ± 0.71a 2.92 ± 0.48b 5.17 ± 1.17a 5.53 ± 1.12a 5.52 ± 0.88a 6.99 ± 1.01a 4.82 ± 0.81a 7.34 ± 1.15a 5.55 ± 0.72a 9.00 ± 1.35a 4.52 ± 0.92a
WD frying 1.38 ± 0.15a 2.41 ± 0.38a 3.26 ± 0.29a 2.19 ± 0.17c 1.73 ± 0.38c 2.76 ± 0.45b 2.76 ± 0.17b 2.49 ± 0.10c 2.42 ± 0.46b 1.74 ± 0.56b 3.19 ± 0.62b 2.40 ± 0.37b 1.73 ± 0.35b 1.93 ± 0.62b
CBM frying 0.61 ± 0.05b 2.38 ± 0.06a 3.30 ± 0.06b 3.65 ± 0.03b 4.12 ± 0.05a 2.95 ± 0.04b 2.67 ± 0.09b 2.89 ± 0.07b 2.86 ± 0.07b 1.37 ± 0.05b 1.64 ± 0.10c 0.86 ± 0.12c 1.21 ± 0.13c 0.48 ± 0.02c

a Abbreviations: WD wheat dough, CBM chicken breast meat, 12 h oil sample collected at the second hour of the first day, ND not detected

b The relative contents of substituted alkanals were express as average value ± confidence interval. The lowercase (a,b,c) behind the content represents the significant difference among the three thermal treatments (p < 0.05, n = 3)

Benzaldehyde was formed by a series of oxidation and dehydration reactions of L under the conditions of heating and occurrence of oxygen in previous study (Kawada et al. 1967). The RA of benzaldehyde in the volatiles released from the heating of trilinlein was higher than that in the volatiles released from the heating of triolein, corn oil, and hydrogenated cottonseed oil (Matthews et al. 1971). In addition, benzaldehyde was also detected during the autoxidation treatment of 2,4-decadienal. Therefore, both of these two formation mechanisms can be used to explain the formation and detection of benzaldehyde in this study due to the high amounts of L and Ln in SBO. The RA of benzaldehyde was first increased to a maximum value of 2.04 mg/kg in the oil sample of 7–2 h and was then decreased as the processing time increased during the heating process. However, benzaldehyde was not observed or detected in lower RA in the food-fried oil samples, especially for the WD-fried oil samples. This is mainly attributed to the instability of benzaldehyde when high temperature, oxygen, and food components are present. Relatively higher RAs and more detection frequency of benzaldehyde were found during the CBM frying process than those during the WD frying process, which could be related to the higher content of CBM lipids.

4-(Hydroxymethyl)cyclohexanecarbaldehyde was only observed at the last day of the heating process. This hydroxyl-containing aldehyde has the property of instability due to its molecular structure. That’s why this compound was detected at the last stage of the oil heating process (control) and was not observed during the food frying process. Based on the possible free radical reaction and the detected cyclic alkanes (such as ethylidenecyclohexane and 3-propyl-cyclohexene, data are not shwon) during the long-time thermal treatments, 4-(hydroxymethyl)cyclohexanecarbaldehyde could be generated by the oxidation of the structure-similar alkanes, alkenes, and other compounds.

3-Methylbutanal was regarded as the degradation product of leucine by a nonenzymatic strecker reaction in previous study (Meynier et al. 1999). However, 3-Methylbutanal was obviously observed in the middle and later stages of the heating process and in the late stage of the frying processes. Therefore, 3-methylbutanal could be produced from the reactions occurred in lipid components. According to the properties of SBO constituents, 3-methylbutanal might be formed through the oxidation decomposition reaction of Ln under the high temperature condition. The contribution of the food components affected the production and the consumption rate of 3-methylbutanal during the food frying process.

Similar to 4-oxo-2-alkenal (C7-C14), 4-hydroxy-2-alkenal (C6-C9), and 4,5-epoxy-2-alkenal (C7 or C10), 4-oxononanal which belongs to 4-oxo-alkanal (C6-C10) is also considered a type of toxic aldehyde formed during the deep-fat frying process (Guillén and Uriarte 2012). Therefore, the formation of 4-oxononanal are very important in terms of fried food and frying oil safety. As presented in Table 5, the RA of 4-oxononanal was first increased to a maximum value of 9.00 mg/kg in the oil sample of 7–2 h and was then decreased during the oil heating process (control) as the processing time increased. However, a decrease in RAs of 4-oxononanal was found during the food frying process and advancing to earlier stages in time needed for reaching the maximum RAs was observed. The food components might not only hinder the formation of 4-oxononanal but also accelerate the consumption of 4-oxononanal. 4-Oxononanal, for instance, can be reacted with amino acids to form the imine linkage-containing compounds (Zamora and Hidalgo 2005). The formation mechanism of 4-oxononanal has been proposed as the consecutive oxidation reactions of L and its SOPs (Neff et al. 1983). Briefly, 9-hydroperoxy-10,13-epidioxy-11-octadecenoic acid is formed by photooxidation of L at the presence of photosensitizer. This 6-membered hydroperoxy epidioxides can decompose to 4-oxo-2-nonenal by β-homolysis and breakage of oxygen-oxygen single bond under the heating condition. 4-Oxononanal is generated from the further hydroperoxidation of 4-oxo-2-nonenal. Due to the diversity of the reaction sites in the unsaturated aliphatic chain, the types of the oxidation products could be diverse. Moreover, interaction could occur among the formation mechanisms of these oxidation products, which could be reflected by the decrease in RA and the time for reaching the maximum RA of 4-oxononanal during the food frying process.

Based on the molecular structure, all the detected VAs are generated from O, L, and Ln through a series of oxidation or free radical reactions when SBO was thermally treated without or with food. During the heating process, most of the RAs of the detected aldehydes were gradually increased to the maximum in the oil sample of 7–2 h and were then decreased as the processing time increased. However, similar trend of change in RA of these aldehydes was observed during the frying processes, but relatively lower RA and shorter time for reaching the maximum RA values of these aldehydes were found when SBO was fried with WD and CBM compared with the control. Therefore, the fried food has a significant influence on the formation of the detected aldehydes. Firstly, the intensity and complexity of the occurred reactions were increased by the addition of food components, which resulted in an increase in production of the aldehydes. However, the formation of aldehydes might be influenced or inhibited to a certain degree and most of the aldehydes easily volatilized away under the boiling condition. Secondly, under the conditions of high temperature and occurrence of oxygen, further hydroperoxidation could take place in these unsaturated aldehydes (such as trans,trans-2,4-decadienal), which reflected by the reduction in RAs of the detected aldehydes. Finally, the detected aldehydes might react with the food component (such as the Maillard reaction) to expedite the reaction consumption of these unstable aldehydes. According to the RAs of the detected aldehydes and their change profiles, components of CBM showed the most significant influence on the formation of aldehydes during the frying processes.

On the other hand, many of the aforementioned aldehydes have been considered the mutagenic compounds associated with the genotoxicity, carcinogenicity, and oxidative stress (Dung et al. 2006). According to the data in this study, the potential harmful aldehydes, such as nonanal, trans-2-octenal, trans,trans-2,4-decadienal, and 4-oxononanal, were detected in significant amounts during both the heating process and the frying process. Because of the effect of boiling and the intensified chemical reactions and the incomplete extraction of the adopted SMPE method, the practical amount of these aldehydes formed during the deep-fat frying process might be higher than those measured values. In other words, a large amount of harmful aldehydes could release into the environment during the deep-fat frying process, especially the CBM frying process. Therefore, more attention should be paid to the safety measures of the fried food.

Conclusion

VAs were significantly formed from the unsaturated FAs when SBO was used as the frying medium of WD and CBM. Free radical reaction, particularly the cis-trans isomerization and positional isomerization of C=C bonds, might be the main reaction occurred during the deep-fat frying process. According to these reaction mechanisms, alkanals, n-alkenals, alkadienals, and substituted alkanals were generated. Furthermore, based on the molecular structures of the detected VAs, some of them were the further oxidation products of the SOPs of O, L, and Ln under the conditions of high temperature and oxygen.

The RAs of VAs were firstly increased to a maximum value in oil sample of 7–2 h and was then decreased as the processing time increased during the oil heating process (control). However, RAs of the typical aldehydes were mostly reduced and the time for reaching the maximum RAs was advanced to the earlier or medium stages of the treatments during the frying process. Therefore, the food component, such as water, protein, lipid, and carbohydrate, intensified the chemical reactions taken place in the unsaturated FAs. Significant changes were observed during the CBM frying process, which might be related to the relatively high contents of water, protein, and lipids. In view of the health risks of the studied VAs (such as trans,trans-2,4-decadienal, trans-2-decenal, and 4-oxononanal) and their relatively high content during the frying process, improvement of measures for determining of fried food quality and safety is very important. Further studies should be carried out to investigate the amounts of these aldehydes in the fried food to help in evaluation of the safety of the whole deep-fat frying process.

Acknowledgments

The authors would like to thank Prof. Wu Jihong who works in College of Food Science and Nutritional Engineering, China Agricultural University for her kind help during the operating of SPME-GC-MS.

References

  1. Bianchi F, Careri M, Mangia A, Musci M. Retention indices in the analysis of food aroma volatile compounds in temperature-programmed gas chromatography: database creation and evaluation of precision and robustness. J Sep Sci. 2007;30:563–572. doi: 10.1002/jssc.200600393. [DOI] [PubMed] [Google Scholar]
  2. Boskou G, Salta FN, Chiou A. Content of trans, trans-2,4-decadienal in deep-fried and pan-fried potatoes. Eur J Lipid Sci Technol. 2006;108:109–115. doi: 10.1002/ejlt.200500236. [DOI] [Google Scholar]
  3. Cao J, Ding R, Wang Y, Chen DJ, Guo DM, Liang CM, Feng ZW, Che Z. Toxic effect of cooking oil fumes in primary fetal pulmonary type II-like epithelial cells. Environ Toxicol Pharmacol. 2013;36:320–331. doi: 10.1016/j.etap.2013.04.011. [DOI] [PubMed] [Google Scholar]
  4. Chung J, Lee J, Choe E. Oxidative stability of soybean and sesame oil mixture during frying of flour dough. J Food Sci. 2004;69:574–578. doi: 10.1111/j.1365-2621.2004.tb13652.x. [DOI] [Google Scholar]
  5. Dung C, Wu S, Yen G. Genotoxicity and oxidative stress of the mutagenic compounds formed in fumes of heated soybean oil, sunflower oil and lard. Toxicol In Vitro. 2006;20:439–447. doi: 10.1016/j.tiv.2005.08.019. [DOI] [PubMed] [Google Scholar]
  6. Frankel EN. Frying fats. In: Frankel EN, editor. Lipid oxidation. Dundee: The Oily Press Ltd.; 2005. pp. 227–248. [Google Scholar]
  7. Frankel EN. Hydroperoxide decomposition. In: Frankel EN, editor. Lipid oxidation. Dundee: The Oily Press Ltd.; 2005. pp. 55–77. [Google Scholar]
  8. Frankel EN, Neff WE, Selke E. Analysis of autoxidized fats by gas chromatography–mass spectrometry: VII. Volatile thermal decomposition products of pure hydroperoxides from autoxidized and photosensitized oxidized methyl oleate, linoleate and linolenate. Lipids. 1981;16:279–285. doi: 10.1007/BF02534950. [DOI] [Google Scholar]
  9. Guillén MD, Uriarte PS. Aldehydes contained in edible oils of a very different nature after prolonged heating at frying temperature: Presence of toxic oxygenated α, β unsaturated aldehydes. Food Chem. 2012;131:915–926. doi: 10.1016/j.foodchem.2011.09.079. [DOI] [Google Scholar]
  10. Jerković I, Mastelić J, Tartaglia S. A study of volatile flavour substances in Dalmatian traditional smoked ham: impact of dry-curing and frying. Food Chem. 2007;104:1030–1039. doi: 10.1016/j.foodchem.2007.01.013. [DOI] [Google Scholar]
  11. Katsuta I, Shimizu M, Yamaguchi T, Nakajima Y. Emission of volatile aldehydes from DAG-rich and TAG-rich oils with different degrees of unsaturation during deep-frying. J Am Oil Chem Soc. 2008;85:513–519. doi: 10.1007/s11746-008-1236-8. [DOI] [Google Scholar]
  12. Katragadda HR, Fullana A, Sidhu S, Carbonell-Barrachina ÁA. Emissions of volatile aldehydes from heated cooking oils. Food Chem. 2010;120:59–65. doi: 10.1016/j.foodchem.2009.09.070. [DOI] [Google Scholar]
  13. Kawada T, Krishnamurthy R, Mookherjee B, Chang S. Chemical reactions involved in the deep fat frying of foods. II. Identification of acidic volatile decomposition products of corn oil. J Am Oil Chem Soc. 1967;44:131–135. doi: 10.1007/BF02558172. [DOI] [Google Scholar]
  14. Lee T, Gany F. Cooking oil fumes and lung cancer: a review of the literature in the context of the U.S. population. J Immigr Minor Health. 2013;15:646–652. doi: 10.1007/s10903-012-9651-1. [DOI] [PubMed] [Google Scholar]
  15. Matthews RF, Scanlan RA, Libbey LM. Autoxidation products of 2,4-decadienal. J Am Oil Chem Soc. 1971;48:745–747. doi: 10.1007/BF02638534. [DOI] [Google Scholar]
  16. Meynier A, Novelli E, Chizzolini R, Zanardi E, Gandemer G. Volatile compounds of commercial Milano salami. Meat Sci. 1999;51:175–183. doi: 10.1016/S0309-1740(98)00122-3. [DOI] [PubMed] [Google Scholar]
  17. Neff WE, Frankel EN, Selke E, Weisleder D. Photosensitized oxidation of methyl linoleate monohydroperoxides: hydroperoxy cyclic peroxides, dihydroperoxides, keto esters and volatile thermal decomposition products. Lipids. 1983;18:868–876. doi: 10.1007/BF02534564. [DOI] [PubMed] [Google Scholar]
  18. Porter NA, Caldwell SE, Mills KA. Mechanisms of free radical oxidation of unsaturated lipids. Lipids. 1995;30:277–290. doi: 10.1007/BF02536034. [DOI] [PubMed] [Google Scholar]
  19. Ramírez MR, Estévez M, Morcuende D, Cava R. Effect of the type of frying culinary fat on volatile compounds isolated in fried pork loin chops by using SPME-GC-MS. J Agric Food Chem. 2004;52:7637–7643. doi: 10.1021/jf049207s. [DOI] [PubMed] [Google Scholar]
  20. Selke E, Rohwedder WK. Volatile components from trilinolenin heated in air. J Am Oil Chem Soc. 1983;60:1853–1858. doi: 10.1007/BF02901537. [DOI] [Google Scholar]
  21. Selke E, Rohwedder WK, Dutton HJ. Volatile components from triolein heated in air. J Am Oil Chem Soc. 1977;54:62–67. doi: 10.1007/BF02912391. [DOI] [Google Scholar]
  22. Selke E, Rohwedder WK, Dutton HJ. Volatile components from trilinolein heated in air. J Am Oil Chem Soc. 1980;57:25–30. doi: 10.1007/BF02675520. [DOI] [Google Scholar]
  23. Soheili KC, Artz WE, Tippayawat P. Pan-heating of low-linolenic acid and partially hydrogenated soybean oils. J Am Oil Chem Soc. 2002;79:287–290. doi: 10.1007/s11746-002-0475-9. [DOI] [Google Scholar]
  24. Warner K, Neff WE, Byrdwell WC. Effect of oleic and linoleic acids on the production of deep-fried odor in heated triolein and trilinolein. J Agric Food Chem. 2001;49:899–905. doi: 10.1021/jf000822f. [DOI] [PubMed] [Google Scholar]
  25. Yin HY, Xu LB, Porter NA. Free radical lipid peroxidation: mechanisms and analysis. Chem Rev. 2011;111:5944–5972. doi: 10.1021/cr200084z. [DOI] [PubMed] [Google Scholar]
  26. Yu TH, Lee MH, Wu CM, Ho CT (1994) Volatile compounds generated from thermal interaction of 2,4-decadienal and the flavor precursors of garlic. In: Ho CT, Hartman TG (eds.) Lipids in food flavors American Chemical Society, Washington DC, pp. 61–76
  27. Zamora R, Hidalgo FJ. Coordinate contribution of lipid oxidation and Maillard reaction to the nonenzymatic food browning. Crit Rev Food Sci Nutr. 2005;45:49–59. doi: 10.1080/10408690590900117. [DOI] [PubMed] [Google Scholar]
  28. Zhang Q, Saleh ASM, Chen J, Shen Q. Chemical alterations taken place during deep-fat frying based on certain reaction products: a review. Chem Phys Lipids. 2012;165:662–681. doi: 10.1016/j.chemphyslip.2012.07.002. [DOI] [PubMed] [Google Scholar]

Articles from Journal of Food Science and Technology are provided here courtesy of Springer

RESOURCES