Effect of Spices on the Formation of VOCs in Roasted Mutton Based on GC-MS and Principal Component Analysis
Peppertree prickly ash, Amomum tsao-ko, cumin, and ginger have long been used in Asian countries to modify the flavor and to partially neutralize any unpleasant odors present in roast lamb. The purpose of this study was to evaluate the change in the amount of volatile components present in roast lamb compared to meat added with peppertree prickly ash, Amomum tsao-ko, cumin, and ginger. Principal component analysis was carried out on the 27 initially selected from 88 volatile substances, and 15 substances with a projection of more than 0.25 in the load matrix were used as indicators to study the different contents in roasted mutton and lamb prepared by adding peppertree prickly ash, Amomum tsao-ko, cumin, and ginger. The types of VOCs (volatile organic compounds) detected in roast meat without adding spices were the least. Roast meat with the addition of cumin leads to the strongest content of aldehydes, followed by the addition of Amomum tsao-ko. Additionally, roast meat with the addition of Chinese prickly ash leads to the strongest content of terpenes, followed by the addition of ginger. Moreover, with the addition of spices, the content of volatiles responsible for the presence of a mutton odor (such as hexanal, heptanal, pentanal, (z)-4-decenal, benzaldehyde, p-propyl-anisole, and dimethyl ether) was not significantly decreased, and in fact some volatiles increased in amount such as pentanal, hexanal, octanal, and (z)-4-decenal. In conclusion, the effect of addition of spices on the volatile profile of roasted mutton and lamb can be attributed to the generation of flavor volatiles mainly derived from raw spices’ hot action, with few additional volatiles formed during boiling.
Roast lamb or mutton meat is spreading in many countries as the most popular traditional foods [1, 2]. Consumer acceptance of mutton from farm to fork is strongly influenced by the eating quality including flavor development . Aromas of cooked meat are predestined from volatile compounds . When meat is cooked, various volatile compounds are released by heat-induced reactions, mainly Strecker and Maillard reaction, lipid oxidation and degradation, thiamin degradation, and interaction of lipid-oxidized products with Strecker and Maillard products . However, consumers may dislike the strong flavor of mutton [5, 6]. Several studies concerning the factors affecting components of the goaty flavor in mutton such as fat content, fatty acid composition, and amino acid content and composition, and reducing sugar content, and composition of meat , which are affected by breed, sex, and maturity of animals and aging time with proteolytic and lipolytic enzyme activities of meat [8–10], have been conducted, and some methods have been gradually developed to reduce or modify the species flavor, for instance, washing, hot processing, extrusion with nonmeat ingredients, and adding spices .
In meat processing, spices are used to improve shelf life, appearance, or even flavor of meat products . The basic effects of spices used in cooking are an enhancement or modification in the flavor of the products , partially removing off-flavor and plumping the flavoring properties . The masking function of the spice odor was not through physical absorption or chemical conversion of the compounds responsible for the undesirable odors but through the release of strongly flavored compounds covering the unpleasant smell or mixed with the odor compounds . For example, the addition of black pepper and cumin extracts improved the quality of Bulgarian-type dry-cured sausages “Sudjuk” . Jung  found that the addition of spices to the beef patty changes the volatile compounds released and interact with meat aroma significantly. In addition, most spices of an Indian spice blend (garam masala) imparted distinctive flavors to cooked ground beef , as well the addition of garlic or onion greatly increased the amounts of sulfur volatiles and masked or changed off-odor and off-flavor from cooked ground beef . Proverbially, lemon is often used to remove smells of meat in Western cooking, while Chinese prickly ash, cumin, and ginger serve a similar function in Chinese cooking .
Previous studies reported that various volatile compounds in meat product originated from spices [21, 22]. Roasted mutton is a widely acclaimed traditional meat product in China. However, the change in the amount and composition of volatile compounds from meat product by adding each spice has not been reported. Therefore, this study was carried out to investigate the change in the amount and composition of the volatile compounds emanating from roasted lamb/mutton with added spices such as peppertree prickly ash, Amomum tsao-ko, cumin, or ginger, which are Chinese familiar spices used as habitual and classical additives in meat products [23–26].
2. Materials and Methods
2.1. Materials and Sample Preparation
An internal standard 1,2-dichlorobenzene purchased from Dr. Ehrenstorfer was used for the determination of the content of volatile matters. All other chemicals and solvents were of analytical grade. Chinese prickly ash (Zanthoxylum bungeanum Maxim.), ginger (Zingiber officinale Roscoe), cumin (Cuminum cyminum L.), and Amomum tsao-ko were purchased from a local market in Shihezi (HaoJiaxiang Department Store, China). The hind legs and tail oil of Xinjiang sheep meeting the unified sales standards were provided by Xinjiang Beibei Ehe Grassland Food Co., Ltd.
2.2. Sample Preparation and Headspace Solid-Phase Microextraction Conditions
For the roasted mutton samples, the hind legs’ meat and tail oil were cut into 2 × 2 × 2 cm3 blocks. Each container was added with a spice to test roasted mutton with one of four spices, and the mixture was pickled for 120 min. The mince of the fifth batch was salted without adding spice and was used as a control. The kinds of spices used in experiments were encoded as follows: sample A (1.0% salt), sample B (0.3% Chinese prickly ash), sample C (1.3% A. tsao-ko), sample D (1.5% cumin), and sample E (1.8% ginger). The meat and suet were put in a preheated oven and baked for 15 min at 210°C. The roasted mutton was mashed until the size of a grain of rice using a DFT-100 Portable high-speed Chinese medicine pulverizer (Longtuo Equipment Co., Ltd., Shanghai, China). Aliquots (5 g) of comminuted rusted mutton were transferred into 15 mL gas-tight glass vessels (Supelco) and saturated with 1 mL NaCl solution. Prior to analysis, 5 μL of 1,2-dichlorobenzene (0.624 g/mL in methanol), as an internal standard, was added and mixed. The flask was shaken while the mixture was added at a uniform rate. After being equilibrated at 50°C for 25 min until the sealed bottle had small water droplets attached to the inner wall, the sample was extracted with 75 μm CAR/PDMS fibre [27, 28] by continuous heating and agitation for 25 min.
After extraction, the fibre was inserted into the GC injector at 250°C for 7 min to desorb analytes. Prior to extraction, the vial of the meat sample was preheated and shaken at intervals to facilitate the volatiles into headspace and to simulate a reheating treatment prior to consumption.
2.3. Gas Chromatography-Mass Spectrometry Analysis
Chromatographic separation of the analytes was achieved with an Agilent 5977A gas chromatograph with 7890B mass selective detector (Agilent Technologies, USA) with the injector temperature of 280°C, equipped with a HP-INNOWAX column (30 × 0.25 mm i.d. fused-silica capillary column chemically bonded with a 0.25-μm cross-bond, 5% diphenyl/95% dimethyl polysiloxane), using highly pure helium as the carrier gas (99.99%) with a constant flow rate of 1.3 mL/min. The GC oven temperature was initially held at 40°C for 8 min, increased at a rate of 5°C/min to 120°C, and then increased to 240°C at 20°C/min, where it was held for an additional 15 min. Helium was used as the carrier gas at a flow rate of 1.3 mL/min in split mode (1 : 20 split ratio). The hydrogen flow rate was 30 mL/min. The MS was acquired in full-scan mode, and the eluate was introduced through the transfer line into the mass spectrometer where the molecules were ionized with a current beam of 70 eV. The temperature of the source and the detector were 150°C and 230°C, respectively, while the MS transfer line was kept at 280°C with a solvent delay of 1 min. A mixture of n-alkanes (C7–C40, Sinopharm Chemical, Shanghai, China) was analyzed under the same conditions to calculate the retention indices of the volatiles.
The data whose matching degrees were more than 80 could be used for qualitative analysis and quantitative analysis. For quantitative analysis, the roasted lamb sample was analyzed using 1,2-dichlorobenzene (0.624 g/mL in methanol) as the internal standard. For qualitative analysis, the identification of the components was performed by matching their mass spectra with the Wiley and NIST 14 mass spectral database (NIST, National Institute of Standards and Technology), relative retention index library, and authentic reference samples. The LRI (Linear Retention Index) according to the method of HVDD  and the RI (Retention Index) were evaluated by using the following equation :where tRi is the retention time of an unknown compound i, z represents the number of carbon atoms of the n-alkane eluted before the unknown i, tRz denotes the number of carbon atoms of the n-alkane eluted after the unknown compound z, and tRz and tRz+1 are the retention times of the n-alkanes with carbon numbers of z and z + 1.
2.4. Statistical Analysis
All analyses were conducted in triplicate. Each roasted lamb sample (object) was considered as an assembly of 88 variables represented by the chemical data. The results reported were the average of these three replicates, and standard deviations were performed with SPSS23.0 software (SPSS, IBM Corp., Armonk, NY, USA), and the Multivariate Statistical Package (MVSP) (Kovach Computing Services, Wales, UK) was used for PCA analysis.
3. Results and Discussion
3.1. Multiple Comparisons of Concentrations of Volatile Substances
Unambiguous identification of 53 of the 88 relevant VOCs was performed by using NIST 14 and correlation with linear retention indices (LRIs) reported in the literature, while the remaining 35 were tentatively identified by using mass spectral databases only (Table 1). Free aroma compounds in VOCs, listed in Table 1, largely belonged to terpenoids or their derivatives. They were classified as aldehydes (19), ketones (7), alcohols (18), ethers (5) (e.g., estragole), phenols (5) (e.g., eugenol), acids and esters (8), terpenes (14), and others (14) (e.g., furans). Among them, 14 substances were found in five samples at the same time, i.e., (z)-4-decenal, nonanal, benzaldehyde, 1-hexanol, 1-octen-3-ol, 1-heptanol, 2-ethyl-1-hexanol, 2-methyl-3-furanthiol, dimethyl ether, butylated hydroxytoluene, 4-methylphenol, eugenol, 4-methylnonanoic acid, and 2-pinene.
The volatile substances of roasted mutton were more abundant in spices-treated samples. From Figure 1, it can be stated that the types of VOCs detected in sample A without adding spices were the least with a higher content in aldehydes and a lower content in terpenes and other compounds. All the other samples (i.e., those with added spices) were similar with small differences. In particular, the cumin-treated sample (sample D) was the strongest in the content of aldehydes, followed by the Amomum tsao-ko-treated sample (sample C). Additionally, the Chinese prickly ash-treated sample (sample B) was the strongest in content of terpenes, followed by ginger-treated sample (sample E). Moreover, Amomum tsao-ko-treated sample (sample C) had the most obvious growth, followed by Chinese prickly ash-treated sample (sample B).
The main sources of aldehydes are lipid oxidation and degradation, as well as Strecker amino acid reactions. Most of aldehydes detected in roast meat are hexanal, heptanal, pentanal, (z)-4-decenal, benzaldehyde, and 3-phenyl-2-propenal. Zhao et al.  proposed that nonanal, hexanal, and octanal in stewed mutton were the key aroma substances. Xu et al.  found that (E)-2-enealdehydes can significantly affect the formation of thiols and heterocyclic thiols and compounds in Maillard reaction systems. Su et al.  showed that the flavor compounds of reheated refrigerated roast whole-sheep were mainly aldehydes. It was worth mentioning that a large amount of hexanal is generated during lipid oxidation at low temperatures, while (E,E)-2,4-decadienal is more easily oxidized under high temperatures over a long time.
The olfactory and perceived thresholds of ketone are much higher than that of their isomeric aldehyde; therefore, their contributions to mutton flavor are relatively small. Nevertheless, ketone should still be the focus of attention, such as DL-2,3-butanediol and 2-methyl-ketene (meat and strong fat smell) .
Higher alcohols have usually a strong, pungent smell; nevertheless, the aliphatic aldehydes have a different odor, from the ripe apple (acetaldehyde) to a pleasant fresh odor as the cucumber (C8–C18 aldehydes), whereas the aromatic aldehydes (benzaldehyde) have odor of bitter almonds. Alcohols are mainly derived from lipid oxidation. The main contents of alcohols are saturated alcohols in our experiments and tests, such as 1-pentanol, 2-heptanol, 1-heptanol and so on, which have higher olfactory thresholds and contribute less to the overall flavor of the roasted mutton compared with unsaturated alcohols, such as 1-octen-3-ol. 1-Octen-3-ol has a pleasant mushroom flavor and a grassy scent and plays an important role in the aroma of roast meat .
Fatty acids are substances produced by the oxidation of aldehydes and ketones or the degradation of fatty acids, which can effectively modify flavor. There is a high content of 4-methylnonanoic acid and acetic acid in roasted mutton, which, at low concentrations, produce the fat flavor, while at high concentrations may be part of the smell of mutton.
3.2. Principal Component Analysis
The discriminating power of all the 88 compounds listed in Table 1 was tested by means of the analysis of variance and the number of homogeneous groups (HG) formed at statistical level, according to the Fisher’s least significant difference method applied to their respective relative areas. Among these compounds, 27 VOCs (z1, z3, z4, z5, z6, z8, z10, z11, z16, z23, z24, z43, z46, z47, z51, z54, z57, z59, z60, z62, z63, z65, z70, z72, z77, z86, and z87) were selected by significant comparison and established three HG to apply other statistical analysis such as principal component analysis (PCA), and the data were square-root transformed  and standardized, in accordance with Vararu et al. . Taking z1 and z2 as an example, z1 was selected since there were three HG of a, b, and c, while z2 was not selected since there were two HGs of a and b.
The loading plot to the projection was based on 88 VOCs and generated from the data matrix (Figure 2). As shown in Figure 2(a), by principal component analysis, 5 samples were classified and the repeated results of the tests were appropriate. Five samples were clearly distinguished, with three principal components representing a total of 88.31% of the data information—PC1 explained 49.13% of the data change, PC2 explained 20.97% of the data change, and PC3 explained 18.31% of the data change.
It can be seen from Figure 2(b) that after 27-dimensional analysis of the volatile substances in three significant grades, the two principal components represent 70.96% of the data information—PC1 (42, 40% of the variability) and PC2 (28.56% of the variability). The projection of 15 volatile substances in the load matrix exceeds 0.25.
The projection coefficient of load matrix can comprehensively reflect the influence of each volatile flavor substance in roast lamb on each principal component. The larger the projection coefficient of load matrix, the stronger the representativeness of the principal component to the variable. Therefore, 15 kinds of volatile substances with a projection of more than 0.25 in the load matrix (Table 2) were used as indicators to explore the effect of adding spices on VOCs of roasted mutton . As can be seen from Table 2, a total of 5 substances out of 15 substances were detected simultaneously in 5 samples: (z)-4-decenal, benzaldehyde, (+)-α-phellandrene, 2-methyl-3-furanthiol, and dimethyl ether are detected simultaneously in five samples. Among them, hexanal is a degradation product of linoleic acid, probably because of the higher content of mutton fat, showing apple aroma and raw oil flavor. Benzaldehyde probably comes from the degradation of phenylalanine  and exhibits a fruity, bitter almond flavor that gives a strong fatty smell to roast meat. Azelaic acid (fruity aroma) is a product of oleic acid oxidation. Hexanal, octanal (green peel flavor and fat flavor), enanthaldehyde (nutty, fruity, and fat flavor), n-valeraldehyde (pungent), and nonanal (rose, citrus, and strong fat flavor) are significant aroma components of roasted mutton [27, 40].
Among them, the contents of z1, z8, z28, z16, z40, and z45 in sample A are relatively high. However, almost no z1 is detected in sample B and sample C probably due to the addition of Chinese prickly ash and Amomum tsao-ko, 4-terpineol and α-terpineol are important volatile substances in pepper, while α-terpinol and geraniol are important materials in A. tsao-ko . Similarly, it can be speculated that the addition of Chinese prickly ash, Amomum tsao-ko, and ginger may result in a decrease in z8 and z28, while the addition of Chinese prickly ash and ginger may result in a significant decrease in z16.
Obviously, the content of z3, z5, z49, z11, and z10 in sample B is higher than that of other samples; trans-geranyl acetate, 2-pinene in Chinese prickly ash, and 2-pinene and gamma-terpinene were reported by Iseli et al. . p-Propyl-anisole, dimethyl ether, p-propyl-anisole, and trans-geranyl acetate were provided by pepper, thus contributing to the aroma structure of roasted mutton.
The content of z40 in sample C was higher than that in the others. A. tsao-ko essential oils mainly consisted of monoterpene hydrocarbons and oxygenated monoterpenes such as gamma-terpinene, terpineol, geraniol, and geranial . It may be because the terpenoids contribute to the production of 2-methyl-3-furanthiol in meat products at high temperatures.
The other 13 substances in sample D except z10 were higher than in the other samples. Aromatic compounds mainly come from Maillard reactions and thermal degradation of amino acids and thiamine , such as ρ-cymene, probably coming from a high-temperature change in cumin.
The content of (z)-4-decenal in sample E is higher than that in the other samples. It is noteworthy at this point that α-terpinol and decanal maybe emanating from ginger mostly. Pang et al.  consider α-pinene and eucalyptol were important substances in ginger.
Overall, with the addition of only salt (sample A), Chinese prickly ash (sample B), Amomum tsao-ko (sample C), cumin (sample D), and ginger (sample E), the content of volatiles responsible for the presence of a mutton odor (such as hexanal, heptanal, pentanal, (z)-4-decenal, benzaldehyde, p-propyl-anisole, and dimethyl ether) was not significantly decreased, and in fact some volatiles increased in amount.
The roasted lamb after adding spices shows a unique flavor due to the increased content of (z)-4-nonenal, benzaldehyde, decanal, and benzaldehyde.
Some VOCs of food product are not detected by consumers because of their thresholds and interactions with other compounds, but they are also important for food flavor. The tendency of ever-growing and changing consumer preferences for meat products are particularly worthy of academic investigation, including the effect of adding different spices on the quality of roast meat. In order to grasp the roasted mutton flesh flavor, SPME/GC-MS and PCA were used. The study focussed on 15 significant volatile compounds. Among these, several aldehyde- and alcohol-containing compounds, as well as terpenoids, were identified and considered essential for the flavor profile. It can be postulated that aroma compounds including aldehydes, alcohols, ether, esters, and ketone resulted in subtle differences between the samples with different spices. The masking function of Chinese prickly ash, Amomum tsao-ko, cumin, and ginger of the astringency odor was ascribed to “sensational deodorizing.”
The data used to support the findings of this study are included within the article and the supplementary information file.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
This study was funded by the National Key Research and Development Program (Grant no. 2016YFD0400705), China Postdoctoral Science Foundation of China (Grant no. 2016M591029), and Key Research and Development Projects of Shaanxi Province, China (Grant no. 2019NY-147).
The supplementary material contains the data of Figure 1 and data of Figure 2. (Supplementary Materials)
J. Song, Y. Yuan, Z. Wang et al., “Suitability of different cuts of lamb for roasting,” Food Science, vol. 38, no. 15, pp. 108–114, 2017.View at: Google Scholar
J. J. Jamora and K. S. Rhee, “Flavor of lamb and mutton,” in Quality Attributes of Muscle Foods, Springer, Berlin, Germany, 1999.View at: Google Scholar
P. M. Brown, “Spices, seasonings, and flavors,” in Ingredients in Meat Products, T. Rodrigo, Ed., pp. 199–210, Springer, New York, NY, USA, 2009.View at: Google Scholar
K. Hirasa and M. Takemasa, Spice Science and Technology, Marcel Dekker, New York, NY, USA, 1998.
D. Balev, N. Nenov, S. Dragoev et al., “Comparison of the effect of new spice freon extracts towards ground spices and antioxidants for improving the quality of Bulgarian-type dry-cured sausage,” Polish Journal of Food and Nutrition Sciences, vol. 67, no. 1, pp. 59–66, 2017.View at: Publisher Site | Google Scholar
E. Demirok, M. Kiralan, and A. Ángel, “Determination and classification of volatile compounds of pastirma using solid phase microextraction/gas chromatography/mass spectrometry,” Journal of Microbiology Biotechnology & Food Sciences, vol. 3, no. 2, pp. 105–109, 2013.View at: Google Scholar
G. Zhang, J. Cheng, X. Yue, M. Yu, and J. Ding, “Chinese prickly ash affects the microflora of harbin air-dried sausage and therefore its flavor,” Food Science, vol. 38, no. 8, pp. 184–190, 2017.View at: Google Scholar
S. Wu, G. Zhao, Y. Liu et al., “Effect of Amomum tsao-ko on the formation of volatile flavor compounds of stewed chicken,” Food Science & Technology, vol. 39, no. 07, pp. 112–117, 2014.View at: Google Scholar
G. Amin, “Cumin,” Handbook of Herbs and Spices, vol. 1, Woodhead Publishing, Cambridge, UK, 2012.View at: Google Scholar
Y. Shi, H. Jin, Y. Yang et al., “Herbal verification of medicinal Amomum tsao-ko,” Medicinal Plant, vol. 5, no. 9, pp. 16–19, 2014.View at: Google Scholar
H. van Den Dool and P. D. Kratz, “A generalization of the retention index system including linear temperature programmed gas-liquid partition chromatography,” Journal of Chromatography A, vol. 11, pp. 463–471, 1963.View at: Google Scholar
S. Song, X. Zhang, K. Hayat, Z. Xiao, Y. Niu, and K. Eric, “Coordinating fingerprint determination of solid-phase microextraction/gas chromatography-mass spectrometry and chemometric methods for quality control of oxidized tallow,” Journal of Chromatography A, vol. 1278, pp. 145–152, 2013.View at: Publisher Site | Google Scholar
M. Zhao, J. Zhao, and L. Hou, “Characterization of the aroma compounds from stewing lamb,” Science and Technology of Food Industry, vol. 37, no. 19, pp. 284–293, 2016.View at: Google Scholar
Y. Xu, J. Li, H. Zhao et al., “Effect of (E)-2-Octenal on volatile products derived from maillard model reaction systems,” Food Science, vol. 35, no. 3, pp. 50–54, 2014.View at: Google Scholar
G. Su, F. Wang, X. Liu et al., “Flavor compounds of refrigerated roasted whole sheep reheated in an electric infrared oven,” Meat Research, vol. 31, no. 1, pp. 32–36, 2017.View at: Google Scholar
J. Xie, B. Sun, F. Zheng et al., “Study on thermal reaction Technology with oxidized lamb fat appended to develop characteristic mutton flavor,” Food & Fermentation Industries, vol. 33, no. 4, pp. 1–5, 2007.View at: Google Scholar
J. Xi, P. Zhan, H. Tian et al., “Comparison of different extraction fibers for analysis of characteristic aroma compounds in Xinjiang roasted lamb by SPME-GC-MS and PCA,” Food Science, vol. 39, no. 10, pp. 234–241, 2018.View at: Google Scholar
V. Iseli, O. Potterat, L. Hagmann et al., “Characterization of the pungent principles and the essential oil of zanthoxylum schinifolium pericarp,” Pharmazie Die, vol. 62, no. 5, pp. 396–400, 2007.View at: Google Scholar