Effect of Frying Process on the Flavor Variations of Allium Plants

The Allium plant is widely used in cuisines around the world for its characteristic flavor. The general profile of the plant changes a lot and presents quite different smells after the frying process. In this work, five Allium plants and their fried oils were compared to find out how the frying process impacts the general flavor profile. The results of sensory analysis indicated that the frying process could substantially increase the flavor acceptability of fresh Allium plants. Meanwhile, according to gas chromatography-mass spectrometry (GC-MS) analysis, fewer volatile compounds were detected in fresh Allium plants than in their fried oils. Furthermore, contents of nitrogen-containing compounds (ranging from 0.17 μg/g to 268.97 μg/g), aldehydes (ranging from 71.82 μg/g to 1164.84 μg/g), and lactones (ranging from 0 μg/g to 12.38 μg/g) increased significantly. In addition, more aroma-active substances were identified in the fried Allium oils revealed by gas chromatography-olfactometry (GC-O) analysis. Sulfur-containing compounds were the most abundant in fresh Allium plants, whereas nitrogen-containing compounds dominated in fried oils. The thermal degradation of sugars, amino acids and lipids as well as interactions between carbohydrates, proteins, and fats during the frying process were thought to be the main contributors to these variations. Therefore, this research provides a theoretical basis for the quality control of onion oil flavor and promotes the further development of the onion plant industry. Consequently, the research provided a theoretical basis for the quality control of Allium oils’ flavor and promoted the further development of Allium plant industries.


Introduction
Allium (Allium L.) is distributed in the northern hemisphere, with over 700 species [1]. In many countries, the Allium plant is one of the indispensable food ingredients. The most common Allium plants include garlic (Allium sativum L.), green onion (Allium fistulosum L.), leek (Allium tuberosu mrottl. ex spreng.), onion (Allium cepa L.), chive (Allium schoenoprasum L.), etc. Green onion, which is rich in sugars, proteins, amino acids, as well as some other nutrients [2] is one of the most common seasonings in Chinese cuisine. Garlic is a well-known edible and medicinal plant. Allicin in garlic is regarded to have healthcare functions, such as antiseptic, antibacterial, anti-cancer and anti-aging [3]. With a slightly spicy odor, shallot can be used for flavoring or deodorizing and can also increase appetite. Furthermore, onion, which is widely cultivated throughout the world, is known as the "queen of vegetables" with high nutritional values [4]. Additionally, bunching onion is frequently found in Thai and Cantonese cuisines. What's more, consuming bunching onion can block the growth of cervical and skin cancer cells [5]. Accordingly, the five Allium plants are representative of high dietary and medical values. In general, a frying process can not only weaken the raw green notes of fresh Allium plants but also enhance the salty and roasted aroma characteristics [6,7]. Consequently, it is particularly significant gradually increase. The mixture was continuously stirred during this process to prevent uneven heating. The above process was repeated twice with new oil every time and mixed the oil for subsequent analysis. The different kinds of fresh samples were prepared in the same way. Their fried oils were named green onion-oil, garlic-oil, shallot-oil, onion-oil and bunching onion-oil, respectively.

Sensory Analysis
Sensory analysis was conducted by a panel of ten members (six females and four males, aged 23 to 30). They were trained in quantitative descriptive analysis (QDA) and experienced in the sensory analysis of food samples [13]. The analysis was carried out in a sensory evaluation room at room temperature and relative humidity. The sensory analysis results can be used not only to compare the flavor differences of the samples, but also to assess the consistency of the instrumental analysis.
Samples (10 g) of fresh Allium plants and their fried oils were placed in covered, odorless, transparent polyethylene terephthalate bottles (volume = 30 mL) separately and randomly arranged. Initially, ten sensory panelists were asked to complete a preliminary round of descriptive sensory evaluation. This round of evaluation did not include any descriptors that might influence their decision. Participants were instructed to write down descriptors that may be used to characterize the flavor characteristics they actually perceived in the samples. Following that, all of the descriptors were iteratively debated and re-evaluated to determine whether or not to maintain them. As a result, the organoleptic descriptors for fresh Allium plants adopted by the members were fresh onion-like, fresh garlic-like, green grass, pungent, sweet, and acidic. Additionally, descriptors obtained for the Allium oils included onion-like, garlic-like, salty, fried, burnt, bitter, cooked vegetable, pungent, green grass, and oily. Subsequently, the samples were presented to the evaluators with the letter codes to recognize the aforementioned organoleptic descriptions and rate their scent intensities ranging from 0 (none) to 9 (very strong). The final score for each odor attribute was the average of the ratings of each panelist, and the obtained data were analyzed by principal component analysis (PCA).

Direct Solvent Extraction (DSE) of Fresh Allium Plants
The volatile compounds of fresh Allium plants were extracted by direct solvent extraction at room temperature (20 ± 5 • C). 20 g of fresh Allium samples were chopped and put into an Erlenmeyer flask, 100 mL of distilled water was added into the flask, and then analytical-grade sodium chloride was added to saturate the solution. The above mixture was homogenized at 1500 rpm for 5 min using a SCILO-GEX BlueSpin LED digital hotplate magnetic stirrer (MS-H280-Pro, Berlin, CT, USA). An internal standard solution (1000 µL, 26.0 µg/mL of 1,2-dichlorobenzene) and dichloromethane were added into the mixture for extraction. The mixtures were stirred at 1500 rpm for 30 min at room temperature during extraction. The solvent phase was separated by centrifugation at 3000 rpm and the mixture was centrifuged at 4 • C for 10 min. The procedure was repeated twice and the resulting solvent extract was used for solvent-assisted flavor evaporation (SAFE). The above steps were repeated for each fresh Allium plant.

Solvent-Assisted Flavor Evaporation (SAFE)
50 g of fried oil sample was dissolved in 200 mL of dichloromethane. In addition, 1,2-dichlorobenzene (50 µL, 0.236 µg/µL) was added as an internal standard before the following procedure. The solution was distilled in a high vacuum using SAFE, a method that removes non-volatile materials [14][15][16]. The SAFE distillate was dried over anhydrous Na 2 SO 4 and then filtered, and the solution was concentrated to about 6 mL using a Vigreux column (50 cm × 1 cm). The extracts were concentrated to 0.5 mL using a gentle stream of nitrogen gas. Samples were prepared in triplicate and stored in 2 mL glass vials at −40 • C for instrumental analysis. Each fresh Allium and their fried oil were prepared the same way.

Gas Chromatography-Mass Spectrometry Analysis
The GC-MS analysis was performed using an ISQ LT-TRACE 1310 (Thermo Fisher Scientific, Waltham, MA, USA) coupled with a Thermo Fisher mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA). GCMS-QP2010 (Shimadzu, Japan) was also used. Separations were performed on a TG-WAX column (30 m × 0.25 mm × 0.25 µm, Thermo Fisher Scientific, Waltham, MA, USA). Helium was used as the carrier gas and delivered to the column at a fixed flow rate of 1.0 mL/min. The GC oven temperature was initially 40 • C, followed by a 1 min hold, ramped at 4 • C/min to 220 • C, and held for 8 min. The mass detector conditions were as follows. EI ionization source, ionization energy, 70 eV, ion source temperature, 250 • C, transfer line, 250 • C, mass range, m/z 35-50 amu, injection volume, 1 µL, injector temperature, 250 • C, the sample was injected in a 1:5 split ratio and the mass range was operated with a solvent delay of 6 min for the samples.

Gas Chromatography-Olfactometry Analysis
GC-O analysis was carried out using a trace GC-MS system equipped with an olfactometer detector (ODP3, Gerstel, Mülheim an der Ruhr, Germany). TG-WAX was used as the analytical column, and the other GC parameters were identical to those of GC-MS analysis. Samples were sniffed by three trained and experienced assessors (two females and one male) and retention times (RT) of elution peaks and odor descriptions were recorded during the sniffing process.

Identification and Quantification of the Aroma Compounds
The volatile compounds in fresh Allium plants and fried Allium plants oils were initially identified by comparing their mass spectra with those of the NIST 2014 database. The homologous series of standards of n-alkanes (C6~C30) were analyzed under the same chromatographic conditions. The retention index (RI) of the detected compounds was calculated and compared with the RI in the NIST 2014 database using the same capillary column. The results were used in the equation [17] to calculate the RI of each compound.
The relative concentrations of the aromatic compounds were calculated by correlating the volatiles' peak areas with the internal standard's peak. The amount of each compound can be calculated using the equation [18].

Statistical Analysis
Results are expressed as the mean ± standard deviation (SD) of at least three independent pretreatment experiments (extraction procedure) for different fresh Allium plants and fried oils. Principal component analysis (PCA, carried out by XLSTAT v. 2018) was performed to differentiate the samples. The Venn was conducted by jvenn (jvenn: an interactive Venn diagram viewer (inra.fr)). The circular plot and the heat map were drawn by Origin 2022 software.

Sensory Analysis
PCA analysis was performed based on the sensory analysis results of different kinds of fresh Allium plants and their fried oils. The data were pre-processed before analysis. The mean values of sensory analysis were used for analysis and the outliers were removed.
The PCA biplot of the five fresh Allium plants was shown in Figure 1A. The first principal component (PC1) and second principal component (PC2) accounted for 64.51% and 21.09% of the variance, respectively. These two components together explained 85.60% of the total variance. The variance contribution of PC1 was more outstanding than that of PC2, indicating that the greater the distance of PC1 axis, the greater the sample variation. According to the result, all fresh Allium plant samples were separated on the PCA plot, indicating that the samples differed somewhat in volatile composition. Using the center line as the dividing line, it was obvious that there were significant aromatic differences among the samples. In the plot, onion and green onion were on the right side, and garlic and bunching onion were on the left side. Onion and green onion mainly produced a sweet note. The smell of fresh garlic was mainly fresh garlic-like although the note of bunching onion was primarily strong pungent and acidic.
The PCA biplot of the five fresh Allium plants was shown in Figure 1A. The first principal component (PC1) and second principal component (PC2) accounted for 64.51% and 21.09% of the variance, respectively. These two components together explained 85.60% of the total variance. The variance contribution of PC1 was more outstanding than that of PC2, indicating that the greater the distance of PC1 axis, the greater the sample variation. According to the result, all fresh Allium plant samples were separated on the PCA plot, indicating that the samples differed somewhat in volatile composition. Using the center line as the dividing line, it was obvious that there were significant aromatic differences among the samples. In the plot, onion and green onion were on the right side, and garlic and bunching onion were on the left side. Onion and green onion mainly produced a sweet note. The smell of fresh garlic was mainly fresh garlic-like although the note of bunching onion was primarily strong pungent and acidic. The PCA biplot of the five Allium oils is shown in Figure 1B. The two principal components (PC1 and PC2) accounted for 83.19% of the variance (46.65% and 36.54%, respectively), indicating that the two principal components can reflect the basic information of the original data of the samples. According to the PCA results, sensory attributes, including salty, fried, cooked vegetable, oily, pungent and onion-like, were more related to the characteristics of green onion-oil. Shallot-oil and onion-oil exhibited a relatively high tendency with the characteristics of green grass. Also, bunching onion-oil revealed the characteristics of burnt, bitter, and garlic-like. The flavor profile of garlic-oil showed a more significant difference from the other samples.
Therefore, the thermal treatment process greatly impacted the flavor profile of these fresh Allium plants. This process enhanced the overall profile, which also produced unique odorants. This phenomenon was due to the fact that many chemical reactions occurred during this frying process, such as interactions and thermal degradations between Allium plants and matrix under high-temperature conditions [19]. Furthermore, a large number of aroma substances were generated during this process, leading to the notes of The PCA biplot of the five Allium oils is shown in Figure 1B. The two principal components (PC1 and PC2) accounted for 83.19% of the variance (46.65% and 36.54%, respectively), indicating that the two principal components can reflect the basic information of the original data of the samples. According to the PCA results, sensory attributes, including salty, fried, cooked vegetable, oily, pungent and onion-like, were more related to the characteristics of green onion-oil. Shallot-oil and onion-oil exhibited a relatively high tendency with the characteristics of green grass. Also, bunching onion-oil revealed the characteristics of burnt, bitter, and garlic-like. The flavor profile of garlic-oil showed a more significant difference from the other samples.
Therefore, the thermal treatment process greatly impacted the flavor profile of these fresh Allium plants. This process enhanced the overall profile, which also produced unique odorants. This phenomenon was due to the fact that many chemical reactions occurred during this frying process, such as interactions and thermal degradations between Allium plants and matrix under high-temperature conditions [19]. Furthermore, a large number of aroma substances were generated during this process, leading to the notes of salty, fried, oily, cooked vegetable, and other similar aromas. Thus, the frying process can substantially increase the flavor acceptability of fresh Allium plants.

GC-MS Analysis
A detailed analysis of their aroma substances was carried out to investigate further the factors affecting the flavor profiles of fresh Allium plants. GC-MS analysis was applied to analyze the volatile compounds in the five fresh Allium plant samples, and the results are shown in Table A3. In general, a total of 69 volatile compounds were identified in five fresh Allium plants, including 40 sulfur-containing compounds, 6 alcohols, 7 ketones, 6 aldehydes, 4 acids, 3 esters, 2 heterocyclic compounds and 1 nitrogen-containing compound. Each fresh Allium plant had a similar number of volatile components. Green onion had 29 volatile components, garlic had 28 volatile components, shallot had 27 volatile components, onion had 21 volatile components, and bunching onion had 32 volatile components. The extracts of all five fresh Allium plants exhibited a rich aromatic profile.
In terms of the content of detected components, the highest content of volatile compounds was sulfur-containing compounds (1860.42 µg/g). Besides, contents of ketones (76.98 µg/g), alcohols (47.76 µg/g), aldehydes (71.82 µg/g), and acids (52.48 µg/g) were relatively similar, whereas heterocyclic compounds and nitrogen-containing compounds were the least similar. The concentrations of sulfur-containing compounds in bunching onion, green onion, shallot, garlic and onion ranged from 182.08 µg/g to 599.67 µg/g, indicating that sulfur-containing compounds play a significant role in fresh Allium plants' flavor. Compounds, hexanal and octanoic acid, were detected in all five fresh Allium plants. Their contents did not differ enormously, indicating that they did not play a dominant role in the overall flavor.

GC-O Analysis
The aroma-active substances in fresh Allium plants were characterized by GC-O analysis. It showed that 42 volatile substances were characterized as odor-active compounds, including 2 alcohols, 4 aldehydes, 1 ketone, 1 phenolic, 1 acid, 1 ester, 2 heterocyclics, 26 sulfur-containing compounds, 1 nitrogen-containing compound and 3 unknown compounds (Table A1). As shown in Figure 2A, the Venn figure was plotted based on the identified volatile compounds in different fresh Allium plants. Accordingly, three, eleven, four, two, and five specific volatile compounds were only identified in green onion, garlic, shallot, onion and bunching onion, respectively. It was worth noting that vanillin and furaneol presented only aroma-active qualities in green onion in this work. These two compounds present sweet and creamy aromas, broadly used in the food industry. Additionally, some sulfur-containing compounds, such as diallyl sulfide, methyl allyl disulfide, dimethyl trisulfide and diallyl trisulfide, only presented aroma characteristics in garlic [20]. These compounds were the main elements in the profile of garlic. Odorants were identified in shallot, including propyl sulfide and 2-methoxy-3-(1-methyl-propyl) pyrazine. And nonanoic acid and prenylthiol were only detected in onion. The compounds identified only in bunching onion were dimethyl disulfide, 3-methylthiophene and 2-hexyl-5-methyl-3(2H)-furanone. These results have hardly been researched before, and are rarely reported. In addition, bunching onion and shallot have three compounds in common. 2-Methyl-2pentenal (green), propyl disulfide (alliaceous), 3-mercapto-1-hexanol (sulfurous); this is in line with the results of sensory analysis. That is to say, these compounds are critical elements in the formation of pungent and sour aromas in sensory analysis results.

Sulfur-Containing Compounds
Generally, threshold values of sulfur-containing compounds (44 μg/kg) [21] are the lowest of all types of aroma-active substances. Usually, the contents of sulfur-containing components are pretty low, but they significantly impact the overall profile. Thus, they play a critical role in the overall flavor profile of foods. Sulfur-containing compounds have a pungent aroma. However, when diluted to concentrations of ppb or ppm, they can present aromas of fresh onions and garlic [22]. Combined with the analytical results, it can be observed that sulfur-containing compounds dominated the identified aroma active compounds of fresh Allium plants. A total of 26 sulfur-containing compounds were detected to possess aroma-active characteristics in the fresh Allium plants, which can be broadly classified into monosulfide, disulfide, trisulfide, etc. Among them, the number of trisulfides was relatively higher; they were dimethyl trisulfide, methyl propyl trisulfide, methyl allyl trisulfide, diallyl trisulfide, etc. Similarities can be found between the identified compound species compared to existing researches [23]. Moreover, the differences between different Allium plants may largely stem from the origin and growth environment of the raw materials. As for the concentration, propyl disulfide, which presented an alliaceous note, was the most highly concentrated compound among these fresh Allium plants, at a concentration of 304.01 μg/g. Other compounds with high content were dipropyl trisulfide (155.38 μg/g, garlic) and dimethyl trisulfide (146.58 μg/g, cooked onion). They have been considered key odorant compounds in fresh Allium plants. In addition, thiophene was detected in all fresh samples, which is responsible for the roast flavor. The most abundant thiophene was 3,4-dimethylthiophene at 102.14 μg/g in fresh plants.

GC-MS Analysis
In order to explore how the frying process affected the flavor profile of the fresh Allium plants, the aroma components of their fried oils were carefully conducted. Results of volatile compounds in the five fried Allium oils via GC-MS are shown in Table A4. A total of 150 volatile compounds were identified, including 19 alcohols, 23 aldehydes, 23 ketones, 9 acids, 4 esters, 2 lactones, 28 sulfur-containing compounds, 29 nitrogen-containing compounds, and 13 heterocyclic compounds. Among the five fried Allium oils, sulfur-

Sulfur-Containing Compounds
Generally, threshold values of sulfur-containing compounds (44 µg/kg) [21] are the lowest of all types of aroma-active substances. Usually, the contents of sulfur-containing components are pretty low, but they significantly impact the overall profile. Thus, they play a critical role in the overall flavor profile of foods. Sulfur-containing compounds have a pungent aroma. However, when diluted to concentrations of ppb or ppm, they can present aromas of fresh onions and garlic [22]. Combined with the analytical results, it can be observed that sulfur-containing compounds dominated the identified aroma active compounds of fresh Allium plants. A total of 26 sulfur-containing compounds were detected to possess aroma-active characteristics in the fresh Allium plants, which can be broadly classified into monosulfide, disulfide, trisulfide, etc. Among them, the number of trisulfides was relatively higher; they were dimethyl trisulfide, methyl propyl trisulfide, methyl allyl trisulfide, diallyl trisulfide, etc. Similarities can be found between the identified compound species compared to existing researches [23]. Moreover, the differences between different Allium plants may largely stem from the origin and growth environment of the raw materials. As for the concentration, propyl disulfide, which presented an alliaceous note, was the most highly concentrated compound among these fresh Allium plants, at a concentration of 304.01 µg/g. Other compounds with high content were dipropyl trisulfide (155.38 µg/g, garlic) and dimethyl trisulfide (146.58 µg/g, cooked onion). They have been considered key odorant compounds in fresh Allium plants. In addition, thiophene was detected in all fresh samples, which is responsible for the roast flavor. The most abundant thiophene was 3,4-dimethylthiophene at 102.14 µg/g in fresh plants.

GC-MS Analysis
In order to explore how the frying process affected the flavor profile of the fresh Allium plants, the aroma components of their fried oils were carefully conducted. Results of volatile compounds in the five fried Allium oils via GC-MS are shown in Table A4. A total of 150 volatile compounds were identified, including 19 alcohols, 23 aldehydes, 23 ketones, 9 acids, 4 esters, 2 lactones, 28 sulfur-containing compounds, 29 nitrogen-containing compounds, and 13 heterocyclic compounds. Among the five fried Allium oils, sulfurcontaining compounds and nitrogen-containing compounds were the most abundant, followed by esters and lactones. For the identified volatiles by GC-MS analysis, aldehydes had the highest content, followed by sulfur-containing compounds and heterocyclics. Besides, alcohols, ketones, acids, and nitrogen-containing compounds were moderately abundant in fried Allium oils. Aldehydes were the most abundant; they presented the highest content in onion-oil (433.6 µg/g) and shallot-oil (312.81 µg/g), but less content in garlic-oil (15.58 µg/g). The same phenomenon could be observed in the results of the sensory analysis. The flavor profiles of onion-oil and shallot-oil were closer, although garlicoil differed the most from the other Allium oils. The closer kinship between shallot and onion may be the reason for their similar aroma [24]. The large morphological differences between garlic and other Allium plants may be the reason for the significant distinction in aroma between garlic-oil and other fried Allium oils. Moreover, the difference in the quantity of aroma substances between different fried Allium oils was not apparent. The common compounds of the fried Allium oils were mainly aldehydes and sulfur-containing compounds, such as (E, E)-2,4-heptadienal, (E, E)-2,4-decadienal, dimethyl disulfide, and dimethyl trisulfide.

Comparison of GC-MS Analysis Results
A total of 69 volatile compounds were identified in five fresh Allium plants, and 150 were found in their fried oils. Compared with fresh Allium plants, there was a sharp increase in the amounts of volatile compounds in their fried Allium oils, demonstrating that the frying process significantly affected the flavor profiles. As shown in Figure 3, more sulfur-containing compounds were in the fresh Allium plants than in the fried ones. Conversely, the other types of compounds in fresh Allium plants, such as nitrogen-containing compounds, ketones, aldehydes, alcohols, heterocycles, acids, esters, and lactones, were less than those in the fried oils. In particular, only one nitrogen-containing compound was detected in fresh plants, however, this quantity increased to 29 in their fried oils. In addition, lactones were only presented in the fried Allium oils rather than the fresh ones, which were produced after the frying process and gave creamy and sweet aromas in order to make the overall flavor more layered.
Foods 2023, 12, x FOR PEER REVIEW 9 of 23 detected in fresh plants, however, this quantity increased to 29 in their fried oils. In addition, lactones were only presented in the fried Allium oils rather than the fresh ones, which were produced after the frying process and gave creamy and sweet aromas in order to make the overall flavor more layered.  Table A5.
Meanwhile, it was worth noting that the contents of sulfur-containing compounds dropped significantly after the frying process from 1860.42 μg/g to 468.94 μg/g, whereas nitrogen-containing compounds and aldehydes increased considerably. The content of nitrogen-containing compounds in fresh Allium plants was 0.17 μg/g, which grew to 268.97 μg/g in their fried oils. Simultaneously, aldehydes were detected in fresh Allium plants at a concentration of 71.82 μg/g and in their fried oils at 1164.84 μg/g. The reaction between lipids, sugars and volatile precursors in Allium plants at high temperatures may be a possible reason for the above content changes. Generally, propyl disulfide might play an essential role in the flavor profile of Allium plants due to its low odor threshold value [25]. It was detected in fresh green onion at a maximum concentration of 192.32 μg/g. (E, E)-2,4-decadienal contributed a desirable fried aroma to the fried Allium oils in small amounts, but provided an oily and rancid flavor in high amounts [26]. In this work, (E, E)-2,4-decadienal was detected in all 5 oils with levels not exceeding 83 μg/g. It contributed to the fried note, consequently. In addition, alkyl pyrazines were the primary type of nitrogen-containing compounds, related to the roasted and nutty aroma. They were once reported as a primary contributor to cooked, roasted, and fried foods [26,27]. A total of 13 alkyl pyrazines were identified in this work. Among them, 2,5-dimethylpyrazine was found to have the highest content (64.71 μg/g) and contributed more to the fried flavor.
As an efficient tool for the evaluation of product quality, a heat map can visualize the trend of changes in the content of volatile compounds in tested samples [28]. A total of 9 types of compounds were compared and performed into a heat map, covering the identified components of the five fresh Allium plants and their fried oils (Figure 4). Affected by the frying process, there was a decrease in the number of sulfur-containing compounds,  Table A5.
Meanwhile, it was worth noting that the contents of sulfur-containing compounds dropped significantly after the frying process from 1860.42 µg/g to 468.94 µg/g, whereas nitrogen-containing compounds and aldehydes increased considerably. The content of nitrogen-containing compounds in fresh Allium plants was 0.17 µg/g, which grew to 268.97 µg/g in their fried oils. Simultaneously, aldehydes were detected in fresh Allium plants at a concentration of 71.82 µg/g and in their fried oils at 1164.84 µg/g. The reaction between lipids, sugars and volatile precursors in Allium plants at high temperatures may be a possible reason for the above content changes. Generally, propyl disulfide might play an essential role in the flavor profile of Allium plants due to its low odor threshold value [25]. It was detected in fresh green onion at a maximum concentration of 192.32 µg/g. (E, E)-2,4-decadienal contributed a desirable fried aroma to the fried Allium oils in small amounts, but provided an oily and rancid flavor in high amounts [26]. In this work, (E, E)-2,4-decadienal was detected in all 5 oils with levels not exceeding 83 µg/g. It contributed to the fried note, consequently. In addition, alkyl pyrazines were the primary type of nitrogen-containing compounds, related to the roasted and nutty aroma. They were once reported as a primary contributor to cooked, roasted, and fried foods [26,27]. A total of 13 alkyl pyrazines were identified in this work. Among them, 2,5-dimethylpyrazine was found to have the highest content (64.71 µg/g) and contributed more to the fried flavor.
As an efficient tool for the evaluation of product quality, a heat map can visualize the trend of changes in the content of volatile compounds in tested samples [28]. A total of 9 types of compounds were compared and performed into a heat map, covering the identified components of the five fresh Allium plants and their fried oils (Figure 4). Affected by the frying process, there was a decrease in the number of sulfur-containing compounds, phenols, and an increase in the number of alcohols, aldehydes, ketones, acids, esters, lactones, nitrogen-containing compounds and heterocyclic compounds. Among them, aldehydes, nitrogen-containing compounds and heterocyclic compounds showed the greatest increase, more specifically, unsaturated aldehydes, and furans increased the most in number. Therefore, we can conclude that the most crucial contributors to the flavor of Allium plants after the frying process were unsaturated aldehydes, nitrogen-containing compounds and furans. It means that the frying process significantly changed the main flavor profile of the samples, and the process caused substantial sensory differences. Previous work had reported that the most critical flavor contributors of fried green onion and fried garlic were mainly sulfur-containing compounds, pyrazines, and unsaturated aldehydes [29]. This is consistent with the results of this research. In addition, the sulfurcontaining compounds differed in type and content, such as the more significant flavor contribution of methyl sulfide in green onion-oil [8,9,11,30] and the predominant one in garlic-oil was allyl sulfide [12,31]. This is related to the difference in the sulfur-containing compounds in raw materials of fresh onions and garlic. phenols, and an increase in the number of alcohols, aldehydes, ketones, acids, esters, lactones, nitrogen-containing compounds and heterocyclic compounds. Among them, aldehydes, nitrogen-containing compounds and heterocyclic compounds showed the greatest increase, more specifically, unsaturated aldehydes, and furans increased the most in number. Therefore, we can conclude that the most crucial contributors to the flavor of Allium plants after the frying process were unsaturated aldehydes, nitrogen-containing compounds and furans. It means that the frying process significantly changed the main flavor profile of the samples, and the process caused substantial sensory differences. Previous work had reported that the most critical flavor contributors of fried green onion and fried garlic were mainly sulfur-containing compounds, pyrazines, and unsaturated aldehydes [29]. This is consistent with the results of this research. In addition, the sulfur-containing compounds differed in type and content, such as the more significant flavor contribution of methyl sulfide in green onion-oil [8,9,11,30] and the predominant one in garlic-oil was allyl sulfide [12,31]. This is related to the difference in the sulfur-containing compounds in raw materials of fresh onions and garlic.

Comparison of GC-O Analysis Results
A total of 16 compounds were identified in both fresh Allium plants and their fried oils via GC-O analysis. Specifically, these were five disulfides (dimethyl disulfide, diallyl disulfide, methyl allyl disulfide, methyl propyl disulfide, (E)-methyl-1-propenyl disulfide), five trisulfides (diallyl trisulfide, methyl propyl trisulfide, (E)-1-propenyl propyl trisulfide, methyl allyl trisulfide, dimethyl trisulfide), two unsaturated aldehydes (2-undecenal, 2methyl-2-pentenal), two aldehydes (hexanal, 1-nonanal), one acid (nonanoic acid), and furaneol. It is easy to know that these compounds do not constitute newly generated unique flavors such as roasted aroma after frying. They mainly present the characteristic aroma profile of the Allium plants, such as onion and garlic aromas, as well as other aromas that contribute to the overall aroma profile of the Allium plants.
In Figure 5, the blue series represented fresh Allium plants and the red series represented fried oils. The GC-O information of related compounds was sorted out, and the main volatile components (third circle) were identified in this study. It can be seen that there are similarities and differences in the aroma substances of fresh Allium plants and their fried oils. It could be observed that the fresh Allium plants primarily presented sweet, pungent, and fatty aromas. Meanwhile, the flavor profiles of fried Allium oils were mainly sweet, pungent, oily, and fatty notes. Among them, the same compounds, such as methyl propyl trisulfide, constituted a pungent aroma in fresh Allium plants and their fried oils. Furthermore, different aroma substances contributed the same flavor in the fresh plants and their oils, respectively. For instance, furaneol in fresh Allium plants contributed a sweet aroma profile, and 2-furyl methyl ketone, benzeneacetaldehyde, and (E)-2-pentenal in fried Allium oil embodied the sweet flavor profile. characteristic aroma profile of the Allium plants, such as onion and garlic aromas, as well as other aromas that contribute to the overall aroma profile of the Allium plants.
In Figure 5, the blue series represented fresh Allium plants and the red series represented fried oils. The GC-O information of related compounds was sorted out, and the main volatile components (third circle) were identified in this study. It can be seen that there are similarities and differences in the aroma substances of fresh Allium plants and their fried oils. It could be observed that the fresh Allium plants primarily presented sweet, pungent, and fatty aromas. Meanwhile, the flavor profiles of fried Allium oils were mainly sweet, pungent, oily, and fatty notes. Among them, the same compounds, such as methyl propyl trisulfide, constituted a pungent aroma in fresh Allium plants and their fried oils. Furthermore, different aroma substances contributed the same flavor in the fresh plants and their oils, respectively. For instance, furaneol in fresh Allium plants contributed a sweet aroma profile, and 2-furyl methyl ketone, benzeneacetaldehyde, and (E)-2-pentenal in fried Allium oil embodied the sweet flavor profile.

Analysis of Reasons for Change
Allium plants are rich in proteins and amino acids [32][33][34][35][36], and cooking oil is used as a heat-conducting medium in the frying process. Generally, for the fried Allium oils, 89 aroma-active substances were not present in fresh Allium plants. Those odorants were

Analysis of Reasons for Change
Allium plants are rich in proteins and amino acids [32][33][34][35][36], and cooking oil is used as a heat-conducting medium in the frying process. Generally, for the fried Allium oils, 89 aroma-active substances were not present in fresh Allium plants. Those odorants were considered to be the productions of the interactions and thermal degradations between basic components. The basic components are the fats, proteins and carbohydrates in the system. The interactions between basic components referred to the interaction of carbohydrates, proteins and fats during heating. The Maillard reaction (browning reaction) is the most pivotal process between sugars and amino acids. Strecker aldehydes were often the primary Maillard reaction products during the first stage of the reaction because of the brief heating period and low temperature. These molecules can further combine to create scent compounds that are distinctive, such lactones and furans. Moreover, the Maillard reaction produced substances with baking aromas like pyrazines, pyrroles and pyridines. Combined with the identified compounds, furanol presented in all fried Allium oils. Furaneol is a critical aroma compound with sweet, bakery and caramel aromas [37], which is naturally present in fresh pineapple and beef broth. Nitrogen-containing compounds are also a class of compounds that increase significantly in content and type after the frying process, such as 2-methylpyrazine, 2-ethyl-6-methylpyrazine, and 3-ethyl-2,5-dimethylpyrazine. These compounds typically present a roasted and nutty aroma which also contributes considerably to the flavor profile of the oil. Thermal degradation reactions are divided into thermal degradation of sugars, fats and amino acids. It is known from other studies that the main product of the thermal degradation of sugar is 2-acetylfuran [10]. The sulfur-containing compounds that contributed more to the overall flavor of fried Allium oils were mainly produced by the thermal degradation reactions of amino acids, such as dimethyl trisulfide. Sulfur-containing compounds are characterized by a low threshold but high aroma intensity, and are the main components of meat aroma in Chinese cuisine [38]. Dimethyl trisulfide, which has a similar aroma to fresh onion and garlic, is naturally present in fresh onion with a threshold value of 0.005-0.01 ppb. It is produced through the degradation of sulfur-containing substances such as S-methyl-L-cysteine sulfoxide in onions [39]. Thermal degradation reactions of fats yield unsaturated aldehydes and methyl ketones, lactones and acids. These products are characteristic aromatic compounds of fried foods, which usually have fried and fatty notes. Among them, 2,4-decadienal is the characteristic aromatic compound of fried foods. In summary, the frying process enhances the salty and roasted aroma characteristics of the fresh Allium plants by not only weakening their original green notes.

Conclusions
Due to their characteristic flavor, Allium plants are frequently utilized in cuisines around the world, and the frying process can significantly impact their flavor profiles. This research aimed to explore the flavor changes between fresh Allium plants and their fried oils, as well as the possible causes of the changes. The results of sensory analysis presented significant flavor differences between fresh and fried samples. Fresh Allium plants mainly showed aromas of green, fresh garlic-like and fresh onion-like, whereas fried oils mainly showed fried, salty, fatty, burnt, garlic-like and onion-like notes. GC-MS analysis, GC-O analysis and statistical analysis revealed the flavor composition of fresh Allium plants and fried oils in more detail. Sulfur-containing compounds dominated the aroma-active compounds of the fresh Allium plants. In contrast, unsaturated aldehydes, nitrogen-containing compounds, and furans were the most critical contributors to the flavor of fried Allium oils. Furthermore, it was worth noting that the contents of sulfur-containing compounds dropped significantly after the frying process from 1860.42 µg/g to 468.94 µg/g, whereas nitrogen-containing compounds and aldehydes increased considerably. The reaction between lipids, sugars and volatile precursors in Allium plants at high temperatures may be a possible reason for the above changes. The subsequent study will focus on the contribution of each aroma substance to the overall flavor of the fried Allium oils. This will help to understand further and control the flavor stability of the fried Allium oil. Besides, product improvement and quality control can also be theoretically guided. In addition, the nutritional value, as well as the physicochemical properties of fried Allium oils, were also the focus of future research.

Acknowledgments:
We are greatly thankful to our preceptor N Zhang, H-T Chen and Y-P Liu for their contribution to this work. We also thank the staff of Beijing Key Laboratory of Flavor Chemistry for technical assistance during the course of this work.