Revealing the aroma profile, aroma compounds, and nutritional quality changes of sesame oil under different pre-treatments

Abstract

Herein, the study aimed to investigate pre-treatments processing on aroma profiles and nutritional quality of cold-pressing sesame oil (CPSO), roasting sesame oil (RSO), microwave sesame oil (MSO), steam explosion sesame oil at 0.8 MPa and 1.2 MPa (SESO₁, SESO₂). The GC × GC-TOFMS-O identified 82 aroma compounds in sesame oils. CPSO was characterised by aldehydes and and phenols, whereas RSO, MSO, SESO₁ and SESO₂ markedly augmented pyrazines and thiazoles concentration, conferring roasted and nutty aromas. From a nutritional standpoint, MSO, SESO₁ and SESO₂ exhibited a significant elevation in total phenolic, sesamol, and antioxidant capacity, notwithstanding partial degradation of sesamin and sesamolin. Steam explosion (1.5–2.0 min) achieved comparable or even superior results to microwave (8 min) and roasting (20 min) in terms of aroma complexity. These findings not only provide a scientific basis for the precise regulation of sesame oil flavor but also offer a high-efficiency alternative for the sesame oil processing.

Graphical abstract

Highlights

• •Microwave and steam explosion increasing total phenols and sesamol in sesame oil.
• •Pyrazines, thiazoles, furans, pyrroles were increased after different pre-treatments.
• •GC × GC-TOFMS-O and GC-IMS can distinguish sesame oil from different pre-treatments.
• •SESO and MSO exhibits excellent antioxidant activity compared to CPSO.

1. Introduction

Sesame (Sesamum indicum L.), a traditional oilseed crop renowned for its abundant oil content, high-quality protein, and bioactive lignans (sesamin and sesamolin), has secured widespread global utilization due to its exceptional nutritional profile and versatile applications (Chau et al., 2021; Chen et al., 2023; Wei et al., 2022; Yang et al., 2025). In this context, flavor is a key quality determinant of sesame oil, primarily determined by its aroma compounds, which are significantly influenced by the pre-treatment methodologies employed (Yin et al., 2021). Concurrently, the preservation of nutritionally salient components such as lignans, polyunsaturated fatty acids, and tocopherols are closely associated with pre-treatments parameters, thereby exerting a direct impact on the health-promoting potential of sesame oil (Huang et al., 2023; Lee et al., 2010). In recent years, growing consumer demand for natural, high-quality edible oils has catalysed research endeavours aimed at optimizing pre-treatments to achieve a synergistic enhancement of both flavor complexity and nutritional attributes.

Commercial production of sesame oil predominantly employs thermal pre-treatment technologies, including roasting and microwave (Lee et al., 2010; Yin et al., 2021; Yin et al., 2023; Zhang, Wang, et al., 2024), in conjunction with diverse pressing methodologies such as hot pressing, cold pressing, and aqueous extraction (Chen et al., 2022; Feng et al., 2025; Huang et al., 2024; Yu & Jia, 2025; Zhang, Zhang, et al., 2024). These processing modalities substantially modulate the aroma profile of sesame oil by altering the concentration of aroma compounds. For instance, Huang et al. elucidate the dynamic evolution of aroma compounds throughout the aqueous extraction of small mill sesame oil (namely Xiaomo sesame oil), identifying roasting and hot water extraction as critical stages in flavor formation (Huang et al., 2024). Moreover, Yu et al. demonstrate that this process may engender undesirable sweaty off-flavors for small mill sesame oil, primarily caused by butanoic acid, 2-methylbutanoic acid, and 2-methylpropanoic acid (Yu et al., 2025). Although traditional roasting imparts the characteristic sesame oil aroma, it is concomitant with high energy consumption, degradation of nutritional constituents, and the generation of off-flavors. Consequently, the development of innovative processing technologies capable of mitigating these drawbacks emerged as research imperative. For example, microwave treatment has been shown to expedite the formation of heterocyclics and phenolic compounds relative to roasting, thereby augmenting aroma profile (Zhang, Wang, et al., 2024). Steam explosion, a novel processing technique, has been reported to markedly enhance lignan and vitamin E content in sesame oil (Huang et al., 2023; Yi et al., 2019). Nevertheless, extant literature focus solely on optimizing roasting parameters, with a paucity of systematic comparative analyses encompassing cold pressing, roasting, microwave, and steam explosion—particularly with respect to elucidating the mechanistic underpinnings of flavor formation during steam explosion.

Quantitative descriptive analysis (QDA) currently provides valuable insights into aroma profile; however, it often lacks molecular-level information (Yang et al., 2024). In contrast, solvent-assisted flavor evaporation (SAFE) coupled with gas chromatography–mass spectrometry-olfactometry (GC–MS-O) facilitates precise identification of aroma compounds, thereby enabling comprehensive aroma profile in sesame hulls at various roasting temperatures (Wang et al., 2024). Nonetheless, the accuracy of GC–MS-O is contingent upon extraction efficiency, assessor olfactory acuity, and instrumental detection thresholds. Notably, comprehensive two-dimensional gas chromatography time-of-flight mass spectrometry-olfactometry (GC × GC-TOFMS-O) demonstrates superior sensitivity and resolution relative to GC–MS-O in the aroma compounds screening (Jia et al., 2025; Zhao et al., 2022). Additionally, gas chromatography-ion mobility spectrometry (GC-IMS) fingerprinting was extensively utilised to differentiate aroma compounds among sesame oil samples derived from 24 distinct sources (Dou et al., 2022), as well as to monitor aroma compounds changes across the sequential processing stages of black sesame seeds (Zhang et al., 2020).

In light of the foregoing, the present study aims to: (i) screening and revealing the aroma compounds changes of sesame oils produced via cold pressing, roasting, microwave, and steam explosion, utilizing a complementary approach with GC × GC-TOFMS-O and GC-IMS; (ii) comparatively assess the nutritional composition and antioxidant activity of sesame oils derived from distinct pre-treatments; and (iii) elucidate the interrelationships among thermal pre-treatments, aroma profiles, aroma-active compounds, and nutritional quality through advanced stoichiometry. The outcomes of this investigation are anticipated to offer a robust scientific foundation for the innovation and optimisation of sesame oil processing technologies, thereby informing the development of sesame oil products with distinctive aroma profiles and increased health-improving effects.

2. Materials and methods

2.1. Chemicals

The labeled standards (²H₆-dimethyl trisulfide) was obtained from FishReag (Nanjing, China). Sesame lignans (sesamin, sesamolin, and sesamol; purity ≥97%), tocopherols (α, β, γ, and δ- isomers; purity ≥96%), phytosterols (brassicasterol, stigmasterol, and β-sitosterol; purity ≥95%), fatty acid methyl ester (FAME) qualitative mixtures, Folin–Ciocalteu reagent, n-ketones (C₄–C₉), n-alkanes (C₇–C₃₀), and 2-methyl-3-heptanone were all sourced from Sigma-Aldrich Co. (St. Louis, MO, USA). Dichloromethane and ethanol (both HPLC grade) were supplied by Merck (Darmstadt, Germany). Standards used for aroma compounds obtained from various reagent suppliers are as detailed in Table S1.

2.2. Preparation of sesame oils

Sesame seeds (cultivar Yuzhi No. 2) were sourced from Zhumadian, Henan Province, China. Five distinct processing methods were employed as follows: (i) untreated (no thermal pre-treatment); (ii) roasting pre-treatment: raw sesame seeds (500 g) were roasted for 20 min at 180 °C (roasting machine GSCH, Henan Ruiguang Machinery Co., Ltd.,China) with continuous agitation; (iii) microwave pre-treatment: sesame seeds (50 g) were placed in each 9 cm diameter Pyrex Petri dish, and eight dishes were simultaneously microwaved in a MARS-X system (CEM Corporation, USA) at 2450 Hz and 1200 W for 8 min; (iv) steam explosion: raw sesame seeds (500 g) were loaded into the reactor of a puffing machine (XSS-QPD, KINHE Food Machinery Co., Ltd., Wuhan, China), and heated until the system pressure reached the preset target pressures (0.8 MPa and 1.2 MPa, requiring approximately 1.5 min and 2 min, respectively). Subsequently, the pressure was rapidly released within approximately 0.00857 s without any holding period (Huang et al., 2023). The moisture content of the pretreated sesame seeds was adjusted to 6%, then stored overnight at 4 °C and were pressed using an oil expeller. Then crude sesame oils were centrifuged for 20 min at 8000 rpm, upper oil phase was taken for further assessment. The sesame oils obtained from untreated, roasted, microwaved, medium-pressure steam explosion (0.8 MPa), and high-pressure steam explosion (1.2 MPa) were designated CPSO, RSO, MSO, SESO₁, and SESO₂, respectively.

2.3. Electronic nose (E-nose) analysis

Aroma profiles of the different sesame oil samples were analysed by using an E-nose system (PEN3, AIRSENSE, Germany). The core hardware comprises a sensor array of ten different metal oxide semiconductor sensors. Samples of the sesame oils (1 g each) were placed in 10 mL headspace vial, from which headspace gas was injected into sensor chamber. Data acquisition time was set to 180 s, with a cleaning time of 90 s. All the samples were evaluated in triplicate.

2.4. Identification of aroma compounds

2.4.1. Isolation aroma compounds isolation using SAFE

Aroma compounds isolation was performed following the method of Engel et al. (1999). Briefly, sesame oil (50 g) was added in a 250 mL flask with 150 mL dichloromethane and exposed to ultrasonic extraction for 60 min at 4 °C and transferred to a SAFE apparatus, operating under a vacuum below 2 × 10⁻³ kPa, while the collection flask was immersed in liquid nitrogen. The collected extract was dried over anhydrous sodium sulphate, concentrated to approximately 2 mL using a Vigreux column, and further reduced to 200 μL under a stream of nitrogen (99.999%). Each sample was extracted in triplicate. 1.0 μL of SAFE extract was checked by GC × GC-TOFMS-O and GC–MS/MS in splitless mode at 250 °C, with a 5 min run time.

2.4.2. GC × GC–TOFMS–O analysis

GC × GC–TOFMS–O analysis was done as explained by Zhao et al. (2022), using an Agilent GGT 0620 system (Guangzhou Hexin Instrument Co., Ltd., China) combined to olfactometer (ODP3, Gerstel, Germany). The first-dimension column was a DB-WAX UI capillary column (60 m × 0.25 mm × 0.25 μm, Agilent, USA), and second-dimension column was a DB-17 capillary column (0.65 m × 0.18 mm × 0.18 μm, Agilent, USA). The oven temperature was programmed as: 40 °C (1 min), ramped at 4 °C/min to 220 °C, then at 5 °C/min to 250 °C (kept for 2 min), for a 54 min total run time. The HV modulation column (1.2 m × 0.25 mm, Shanghai Snowking Technology, China) operated with a 5 s cycle and a 1 s analysis time. Helium (99.999%) was used as a carrier gas at a constant 1.5 mL/min (flow rate).

TOFMS–O conditions: EI source voltage was 70 eV, scan range was 35–500 m/z, acquisition rate was 100 spectra/s, and solvent delay was 3 min. The transfer line, and ion source, olfactometer maintained at 250 °C, 230 °C and 100 °C. Calibration with (−)-limonene, citral (cis/trans), and methyl dihydrojasmonate (cis/trans) was performed before odor evaluation. Panellists used a manual controller to record odor intensity (rated from 1 to 4) and sniff time. Each sample was assessed three times by each of three panellists (aged 23–32, 2 females and 1 males), and only results that are consistent two or more times are used for statistical analysis. Results were processed using Canvas software (version 1.0.0.25117, J&X Technologies, China) with the NIST 20 mass spectral library. Commercial standards for n-alkanes (C₇–C₃₀) and aroma compounds were tested under similar conditions. The retention index (RI) was determined by comparing the peak time of the target aroma compound with a series of n-alkanes (C₇-C₃₀) as outlined in previous research (Zhao et al., 2022).

2.4.3. GC–MS/MS analysis

Qualitative analysis of aroma compounds was carried out using Agilent 8890-7000D GC–MS/MS system (Agilent Technologies, USA) as before mentioned. Each sample was separated on a J&W HP-5 capillary column (30 m × 0.25 mm × 0.25 μm, Agilent, USA) with following oven programme: 40 °C (held for 2 min), ramped at 4 °C/min to 200 °C, then at 5 °C/min to 250 °C (held for 2 min); total run time was 54 min. Injections were in splitless mode at 250 °C, with helium (99.999%) as the carrier gas at 1.5 mL/min. The EI source was set at 70 eV; ion source, quadrupole, and transfer line temperatures were kept at 250 °C, 150 °C, and 280 °C, respectively, with a solvent delay of 3 min. Detection was done over range of 35–500 m/z. Data were acquired and processed using an Agilent MassHunter Workstation.

2.5. Aroma extract dilution analysis (AEDA)

AEDA was performed to determine flavor dilution (FD) factors of aroma compounds (Yu & Jia, 2025). The SAFE extract was serially diluted with dichloromethane at ratios of 1:3, 1:9, 1:27, 1:81, 1:243, 1:729, 1:2187 and 1:6561. Each dilution, along with the original extract, was analysed by GC × GC–TOFMS–O for olfactometric evaluation. The FD factor of an odor compound was defined as the highest dilution at which its odor was still detectable.

2.6. Quantification of aroma compounds in sesame oil

Aroma compounds were quantified using external standard curve method (Jia et al., 2024; Jia et al., 2025). Briefly, 50.0 g of sesame oil and 150 mL of dichloromethane added to a 250 mL Erlenmeyer flask, after which 100 μL of 2-methyl-3-heptanone (0.816 mg/mL) was added as the internal standard. Aroma compounds were collected using SAFE and examined by GC–MS/MS in selected ion monitoring (SIM) mode. Subsequently, standards of known concentrations were mixed in dichloromethane, with 100 μL of 0.816 mg/mL 2-methyl-3-heptanone, and diluted to final volume of 200 μL. Standards were analysed by GC–MS/MS in SIM mode. Calibration curves were constructed by plotting ratio of concentration to internal standard concentration versus ratio of their peak areas. The aroma compound (aldehydes, ketones, alcohols, acids, esters, heterocyclics) concentration in sesame oil was calculated using the previously described method (Jia et al., 2024). Employing the same procedure, establish calibration curves for sulfides against ²H₆-dimethyl trisulfide, and terpenes and phenols against cyclohexanone (Jia et al., 2025).

2.7. Odor threshold and Odor activity value (OAV) analysis

OAV of each aroma compound was measured as the ratio of its concentration in sesame oil sample to its odor threshold (OT) in oil. OTs were obtained from published literature (Jia et al., 2019; Jia et al., 2024) or determined using a three-alternative forced-choice (3-AFC) method. In this sensory test, standard solutions of each aroma compound (prepared in ethanol) were added to a refined sunflower seed oil. Twelve trained panellists (aged 23–32, 7 females and 5 males) compared the odor of spiked and unspiked standards. The calculation method for OTs followed previously published procedures (Jia et al., 2024; Matheis & Granvogl, 2019).

2.8. Aroma omission and reconstitution experiments

Aroma profiling analysis (APA) was performed by a panel of 18 trained individuals (6 men and 12 women, 22–32 of age) in a sensory evaluation room (Neugebauer et al., 2020). Aroma profiles were rated on a 0–5 scale (at intervals of 0.5), where 0 indicated “not perceptible” and 5 indicated “intense”. All assessors underwent specific odor training prior to evaluation, based on descriptors and reference standards listed in Table S2. Every sample (5 g) was put in a glass bottle (brown) for assessment according to the APA protocol. For aroma recombination, aroma compounds with an OAV ≥ 1 were added to refined sunflower seed oil at their measured concentrations, thus creating a reconstituted model oil. Native and recombined oils were evaluated by APA to generate aroma sensory radar plots. In the omission experiments, specific classes of aroma compounds were omitted from the model. Each set, comprising two complete reconstituted samples and one omission model, was randomly coded and evaluated by panellists using triangle test protocols to detect significant differences.

2.9. GC–IMS analysis

The GC–IMS (FlavorSpec®, G.A.S., Dortmund, Germany) was employed to identify aroma compounds in sesame oil samples (Zhang et al., 2020). Oil sample (2.0 g) was placed into 20 mL headspace vial and incubated for 20 min at 60 °C. Afterward, 500 μL of the headspace vapour was injected (splitless mode) into GC inlet held at 85 °C. The DB-WAX column (30 m × 0.53 mm i. d., film thickness 1.00 μm, USA) and IMS detector temperatures were set at 60 °C and 45 °C, respectively. Ultra-high-purity nitrogen (≥99.999%) was used as a carrier gas with flow rate programmed as follows: initial 2 mL/min for 2 min, ramped to 10 mL/min over 10 min, then to 100 mL/min over 20 min, and held at 100 mL/min for 30 min. Data were processed and visualised with VOCal 0.4.03 software (G.A.S., Dortmund, Germany), and sample differences were investigated using Reporter and Gallery Plot plug-ins.

2.10. Analysis of nutritional components

2.10.1. Determination of Total phenolics

Total phenolics (TP) content was assessed determined using method of Huang et al. (2023). Briefly, 1.25 g of sesame oil was mixed in n-hexane and washed 3 times with 1.5 mL of methanol/deionised water (80/20, v/v). The aqueous phase was collected for analysis, and TP was measured from UV–Vis spectrophotometer (DU800, Beckman Coulter, Brea, CA, USA) based on Folin–Ciocalteu colorimetric method. The results were presented as milligrams of gallic acid equivalents per 100 g of oil sample (mg GAE/100 g).

2.10.2. Determination of tocopherols

Tocopherol content was analysed as presented by Zhang et al. (2019). 0.8 g of sesame oil samples were added in n-hexane (10 mL) and passed through a 0.22 μm organic membrane filter. The HPLC analysis was performed with an LC-6 CE (Shimadzu, Tokyo, Japan) equipped with a diode array detector (SPDM20 A, Shimadzu, Tokyo, Japan). Identification and quantification of tocopherols were done by contrast with calibration standard curves.

2.10.3. Determination of phytosterol

The analytical method for phytosterols was adapted from Huang et al. (2023), employing a DB-5HT column (30 m × 0.22 mm × 0.1 μm, Agilent, Santa Clara, CA, USA) with helium (99.999%) as the carrier gas at a flow rate of 1.5 mL/min. Identification relied on comparison with retention times of standard, while quantification relied on the peak area relative to the internal standard.

2.10.4. Determination of sesame lignans

The contents of sesame lignans (sesamol, sesamin, and sesamolin) were determined by HPLC utilizing a previously described method (Yin et al., 2023). A 10 μL aliquot was injected into an LC-20 A HPLC system (Shimadzu, Tokyo, Japan) equipped with an Inertsil OSD-EP column (4.6 mm × 250 mm, 5 μm). The column temperature was set at 30 °C, and mobile phase was water/methanol (70/30, v/v), eluted at a constant flow rate of 0.8 mL/min. Sesame lignans were quantified by comparison with standards at 290 nm.

2.10.5. Antioxidant capacity

The DPPH, FRAP, and oxidative stability index (OSI) were determined following previously used method (Huang et al., 2023).

2.11. Statistical analysis

All data are expressed as mean ± standard deviation (three parallel trials). Duncan's test (P < 0.05) was used to calculate the statistical comparisons using SPSS 22.0 software. Graphs were generated using Origin 2021, ChemDraw 19.0, Simca 14.0, TBtools-II, and Cytoscape v3.10.0.

3. Results and discussion

3.1. E-nose analysis of sesame oils with different pre-treatments

E-nose comprising multiple sensor arrays, serves as substitute to human olfactory system and is commonly employed to discriminate vegetable oils based on their aroma profiles. As shown in Fig. 1A, the E-nose sensor analysis revealed distinct differences in the sesame oils subjected to various different pre-treatment. CPSO, RSO, MSO, SESO₁, and SESO₂ exhibited varying degrees of response on three characteristic sensors: W1W (primarily sensitive to sulfides and terpenes), W2W (responsive to sulfides and aroma compounds), and W5S (closely related to nitrogen oxides) (Chen et al., 2022). CPSO exhibited only weak responses on the W5S, W1W, and W2W sensors, resulting in a lighter overall aroma. In contrast, sesame oils produced via roasting, microwave, and steam explosion exhibited strong sensor responses due to high-temperature-induced Maillard reactions and unsaturated fatty acids degradation, which leading to more complex and intense aromas (Yin et al., 2021; Yu & Jia, 2025; Zhang et al., 2021). Previous studies have also demonstrated that E-nose sensors can effectively discriminate between sesame oils that were processed under different microwave time (Jia et al., 2019) and flaxseed oils that were subjected to varying thermal pre-treatments (Yu & Jia, 2025).

Fig. 1.

E-nose sensor responses (A) and PCA score distribution diagram (B); GC × GC–TOFMS-O 3D chromatograms (C); partial least squares regression (PLSR) biplot of aroma compounds based on GC × GC–TOFMS-O data (peak area)(D) in CPSO, RSO, MSO, SESO₁ and SESO₂.

Fig. 1B illustrating the clustering and distribution of sesame oil samples along first principal component (PC₁, 94.56%) and second principal component (PC₂, 5.48%). Notably, the samples exhibit a left-to-right shift along the PC1 axis, with CPSO distinctly separated from RSO, MSO, SESO₁, and SESO₂, indicating that these pre-treatments markedly alter the aroma profiles of sesame oil. SESO₂ is notably differentiated from the other sesame oil samples, which may be attributed to high-pressure-induced changes in the sesame seed structure (Huang et al., 2023; Yi et al., 2019).

3.2. Screening of aroma compounds in sesame oils with different pre-treatments

Aroma profiles of sesame oil results from the interactions among numerous aroma compounds. As summarised in Table 1, 82 aroma compounds were determined across CPSO, RSO, MSO, SESO₁, and SESO₂ using GC × GC–TOFMS–O, including aldehydes (nos. 1–13), ketones (nos. 14–19), alcohols (nos. 20–23), acids (nos. 24–28), esters (nos. 29–31), pyrazines (nos. 32–42), thiazoles (nos. 43–49), furans (nos. 50–55), pyridines (nos. 56–58), pyrroles (nos. 59–61), thiophenes (nos. 62–66), sulfides (nos. 67–70), and others (nos. 71–82). As illustrated in Fig. 1C, the number of aroma compound peaks was markedly higher in RSO, MSO, SESO₁, and SESO₂ than in CPSO. Furthermore, partial least squares regression (PLSR) effectively distinguished sesame oils prepared via different pre-treatments, with most aroma compounds being highly concentrated in SESO₂ (Fig. 1D).

Table 1.

Information on aroma compounds identified in sesame oils from different pre-treatments based on GC × GC–TOFMS-O and GC–MS/MS analysis.

No.	Aroma compounds	RI		Odor notes	FD factora					Identificationb
		DB-WAX	HP-5MS		CPSO	RSO	MSO	SESO1	SESO2
Aldehydes
1	Pentanal	901	694	green, grassy, pungency	NS	1	NS	1	1	MS, RI, S, O
2	2-Methylbutanal	949	647	cocoa, almond, woody	3	9	9	9	81	MS, RI, S, O
3	Hexanal	1064	786	grass, tallow, fatty	3	3	3	9	9	MS, RI, S, O
4	Heptanal	1172	892	fatty, green	NS	1	1	1	1	MS, RI, S, O
5	(E)-2-Hexenal	1211	846	green	NS	NS	NS	3	1	MS, RI, S, O
6	Octanal	1304	1013	fatty, green	NS	9	3	9	9	MS, RI, S, O
7	(E)-2-Octenal	1442	1061	green, nutty, fatty	NS	NS	NS	NS	NS	MS, RI, S
8	Decanal	1489	1189	soap, orange peel, tallow	NS	NS	NS	NS	NS	MS, RI, S
9	Benzaldehyde	1502	955	almond, burnt sugar	NS	NS	NS	1	1	MS, RI, S, O
10	(E)-2-Nonenal	1558	1176	cucumber, fatty, green	1	1	1	1	3	MS, RI, S, O
11	(E)-2-Decenal	1650	1202	fatty	NS	NS	NS	3	1	MS, RI, S, O
12	Phenylacetaldehyde	1661	1062	honey, sweet	9	27	27	81	243	MS, RI, S, O
13	(E,E)-2,4-Decadienal	1796	1309	fried, wax, nutty	1	9	9	9	27	MS, RI, S, O
Ketones
14	2,3-Pentanedione	1051	695	cream, butter	3	27	9	9	9	MS, RI, S, O
15	3-Hydroxy-2-butanone	1287	722	butter, cream	1	3	1	9	9	MS, RI, S, O
16	2-Octanone	1300	1005	earthy, woody, herbal	1	3	3	1	1	MS, RI, S, O
17	1-Octen-3-one	1331	993	mushroom, metal	NS	3	9	NS	NS	MS, RI, S, O
18	6-Methyl-5-hepten-2-one	1319	987	pepper, mushroom, rubber	NS	NS	NS	NS	NS	MS, RI, S
19	3-Methyl-1,2-cyclopentandione	1777	ND	caramel, coffee	NS	NS	NS	1	1	MS, RI, S, O
Alcohols
20	3-Methyl-1-butanol	1185	734	whiskey, malt	9	9	27	27	3	MS, RI, S, O
21	1-Pentanol	1252	768	balsamic	NS	NS	NS	NS	NS	MS, RI, S
22	1-Hexanol	1360	867	resin, flower, green	1	1	1	3	3	MS, RI, S, O
23	1-Octen-3-ol	1430	985	mushroom	3	27	27	9	27	MS, RI, S, O
Acids
24	Acetic acid	1431	672	vinegar	1	3	1	3	9	MS, RI, S, O
25	Butanoic acid	1594	793	cheese, butter	NS	3	1	3	9	MS, RI, S, O
26	Hexanoic acid	1843	975	sweat	NS	1	NS	1	1	MS, RI, S, O
27	Heptanoic acid	1956	1076	rancid, sour	NS	NS	NS	NS	1	MS, RI, S, O
28	Octanoic acid	2041	1182	sweat, cheese	1	3	1	3	9	MS, RI, S, O
Esters
29	γ-Caproalctone	1694	1056	coconut, sweet	3	9	9	9	3	MS, RI, S, O
30	γ-Octalactone	1910	1260	coconut	NS	NS	NS	NS	NS	MS, RI, S
31	γ-Nonalactone	2023	1362	coconut, peach	NS	NS	NS	NS	NS	MS, RI, S
Pyrazines
32	2,5-Dimethylpyrazine	1323	927	nutty, roasted sesame	NS	2187	729	729	2187	MS, RI, S, O
33	2,6-Dimethylpyrazine	1328	909	roasted nut	NS	9	9	9	81	MS, RI, S, O
34	2-Ethylpyrazine	1333	929	peanut butter	NS	81	27	27	243	MS, RI, S, O
35	2,3-Dimethylpyrazine	1346	917	nutty, peanut butter, cocoa	NS	1	1	1	3	MS, RI, S, O
36	2-Ethyl-6-methylpyrazine	1385	997	nutty	NS	27	9	9	81	MS, RI, S, O
37	2-Ethyl-5-methylpyrazine	1392	1002	nutty	NS	1	1	1	3	MS, RI, S, O
38	2,3,5-Trimethylpyrazine	1406	1000	roast, potato	NS	81	27	27	81	MS, RI, S, O
39	3-Ethyl-2,5-dimethylpyrazine	1447	1071	potato, roast	NS	81	81	81	243	MS, RI, S, O
40	2-Ethenyl-6-methylpyrazine	1488	1029	earthy, coffee	NS	27	9	9	81	MS, RI, S, O
41	2-Isobutyl-3-methoxypyrazine	1520	1199	bell pepper-like	NS	3	3	3	9	MS, RI, S, O
42	2-Acetylpyrazine	1622	ND	popcorn, nutty	NS	27	9	27	243	MS, RI, S, O
Thiazoles
43	Thiazole	1248	ND	meaty, nutty	NS	9	3	9	9	MS, RI, S, O
44	2,4-dimethylthiazole	1263	ND	roasted, coffee	NS	3	1	3	9	MS, RI, S, O
45	4-Methylthiazole	1301	ND	roasted meat	NS	9	9	3	27	MS, RI, S, O
46	2,4,5-Trimethylthiazole	1390	980	earthy	1	9	1	9	27	MS, RI, S, O
47	2-Acetyl-2-thiazoline	1712	ND	roast, popcorn	NS	243	81	243	279	MS, RI, S, O
48	Benzothiazole	1897	1226	gasoline, rubber	NS	NS	NS	NS	NS	MS, RI
49	4-Methyl-5-thiazoleethanol	2311	1280	sulfur	NS	NS	NS	NS	NS	MS, RI, S
Furans
50	2-Pentylfuran	1230	989	green bean, butter	9	9	9	27	27	MS, RI, S, O
51	Furfural	1460	820	bread, almond, sweet	NS	3	1	1	3	MS, RI, S, O
52	2-Acetylfuran	1501	901	sweet, popcorn	NS	3	3	3	9	MS, RI, S, O
53	5-Methyl-2-furanaldehyde	1570	977	almond, caramel, burnt	NS	9	9	9	27	MS, RI, S, O
54	2(5H)-Furanone	1767	916	roasted	NS	27	9	27	81	MS, RI, S, O
55	5-Hydroxymethylfurfural	2526	1247	caramel	NS	3	3	9	9	MS, RI, S, O
Pyridines
56	Pyridine	1186	740	brunt	NS	1	NS	NS	NS	MS, RI, S, O
57	2-Methylpyridine	1204	825	sweat	NS	1	NS	NS	NS	MS, RI, S, O
58	1-(2-Pyridinyl)ethanone	1602	1045	popcorn	NS	NS	NS	NS	3	MS, RI, S, O
Pyrroles
59	1-Methyl-1H-pyrrole-2-carboxaldehyde	1626	1000	popcorn, roasted meat	NS	27	27	27	243	MS, RI, S, O
60	2-Acetylpyrrole	1971	1063	walnut, bread	NS	1	NS	1	3	MS, RI, S, O
61	1H-Pyrrole-2-carboxaldehyde	2028	1015	beefy, coffee	NS	9	9	9	27	MS, RI, S, O
Thiophenes
62	Thiophene	1025	670	roasted garlic	NS	1	1	1	1	MS, RI, S, O
63	3-Methylthiophene	1105	780	roasted, nutty	NS	1	1	1	1	MS, RI, S, O
64	Dihydro-3-(2H)-thiophenone	1526	ND	mushroom	NS	27	3	9	81	MS, RI, S, O
65	5-Methyl-2-thiophenecarboxaldehyde	1767	ND	meaty, nutty	NS	NS	NS	NS	1	MS, RI, S, O
66	2-Acetylthiophene	1785	ND	sulfur, nutty	NS	NS	NS	NS	NS	MS, RI, S
Sulfides
67	Dimethyl disulfide	1065	746	sulfurous, cabbage, onion	81	729	243	729	729	MS, RI, S, O
68	Dimethyl trisulfide	1383	980	fresh onion, mint, spicy	81	729	729	729	2187	MS, RI, S, O
69	Methional	1480	988	onion, potato	81	243	81	243	243	MS, RI, S, O
70	Dimethyl Sulfoxide	1553	ND	garlic	NS	NS	NS	NS	NS	MS, RI
Others
71	3-Carene	1142	899	lemon, resin	3	3	3	3	1	MS, RI, S, O
72	Styrene	1238	1412	balsamic, gasoline	NS	NS	NS	NS	NS	MS, RI, S
73	α-Ionone	1854	1078	woody	1	3	1	1	1	MS, RI, S, O
74	2-Methoxyphenol	1862	1192	gammon-like, smoky	9	27	27	27	81	MS, RI, S, O
75	Creosol	1959	1098	woody	1	1	1	NS	NS	MS, RI, S, O
76	Maltol	1965	985	caramel	NS	NS	1	NS	NS	MS, RI, S, O
77	Phenol	2008	1102	phenol	1	1	1	3	9	MS, RI, S, O
78	4-Methylphenol	2076	1303	smoky	NS	NS	NS	NS	3	MS, RI, S, O
79	2-Methoxy-4-vinylphenol	2156	1362	smoked, burnt, clove-like	NS	9	3	27	27	MS, RI, S, O
80	Eugenol	2167	1381	clove-like, honey	NS	NS	NS	NS	NS	MS, RI, S
81	2,6-Dimethoxyphenol	2269	1402	smoky	NS	NS	NS	NS	1	MS, RI, S, O
82	Vanillin	2566	899	vanilla-like	NS	NS	NS	NS	NS	MS, RI

^aCPSO: cold-pressed sesame oil, RSO: roasted sesame oil (180 °C, 20 min), MSO: microwaved sesame oil (1200 W, 8 min), SESO₁: steam-exploded sesame oil (0.8 MPa), SESO₂: steam-exploded sesame oil (1.2 MPa). ^d Identification method. MS: NIST 20 databases; S: standard identified; O: sniff identified by GC × GC–TOFMS-O; RI: aroma compounds were identified by theoretical RI (http://webbook.nist. gov/chemistry/); ‘ND’: cannot be identified. ‘NS’: cannot be sniffed.

3.2.1. Aldehydes, ketones, alcohols, acids, and esters

The FD factor indicates the aroma intensity from aroma compounds. By AEDA and GC × GC–TOFMS–O analysis, 24, 58, 55, 58, and 63 characteristic aroma compounds (FD factor ≥ 1) were identified in CPSO, RSO, MSO, SESO₁, and SESO₂, respectively. As shown in Table 1, FD factors of aroma compounds in CPSO were generally low, with aroma compounds including: 2-methylbutanal (FD = 3, cocoa/almond), hexanal (FD = 3, grassy, fatty), phenylacetaldehyde (FD = 9, honey, sweet), (E)-2-nonenal (FD = 1, cucumber, fatty, green), (E, E)-2,4-decadienal (FD = 1, fried, waxy, nutty), 2,3-pentanedione (FD = 3, cream, butter), 2-octanone (FD = 1, earthy, woody, herbal), 3-methyl-1-butanol (FD = 9, whisky, malty), 1-hexanol (FD = 1, resin, floral, green), and 1-octen-3-ol (FD = 3, mushroom). For most of these were increased in RSO, MSO, SESO₁, and SESO₂. These aldehydes, ketones, and alcohols conferring green, grassy, fatty, caramel, and nutty notes to sesame oil. Feng et al. reported that hexanal, (E)-2-nonenal, 2-methoxyphenol, and phenol (FD > 1) were key aroma-active compounds in CPSO (Feng et al., 2025). Yin et al. (2021) similarly identified 1-octen-3-ol (FD = 32), hexanal (FD = 16), (E, E)-2,4-decadienal (FD = 8), and 1-hexanol (FD = 8) as major aroma contributors in CPSO. Notably, (E)-2-hexenal (green), benzaldehyde (almond, burnt sugar), (E)-2-decenal, and 3-methyl-1,2-cyclopentanedione (caramel, coffee) were only sniffed in SESO₁ and SESO₂. Acids were primarily formed via hydrolysis of fatty acids, which may intensified unpleasant notes in SESO₂. Yu et al. (2025) confirmed that 2-methylbutanoic acid (FD > 64) contributed off-flavor (specifically, sweaty odor) to small mill sesame oil.

3.2.2. Pyrazines, thiazoles, furans, pyridines, pyrroles, and thiophenes

As shown in Table 1, with the exception of 2-pentylfuran (FD = 9, green bean, buttery) and 2,4,5-trimethylthiazole (FD = 1, earthy), no pyrazines, thiazoles, furans, pyridines, pyrroles, or thiophenes were sniffed in CPSO. Remarkably, SESO₂ contained several heterocyclics with high FD factors (≥243), including 2,5-dimethylpyrazine, 2-ethylpyrazine, 3-ethyl-2,5-dimethylpyrazine, 2-ethenyl-6-methylpyrazine, 2-acetylpyrazine, 2-acetyl-2-thiazoline, and 1-methyl-1H-pyrrole-2-carboxaldehyde. Most detectable compounds were consistent with previous studies. For instance, Yin et al. reported 2,5-dimethylpyrazine, 3-ethyl-2,5-dimethylpyrazine, 2-ethylpyrazine, 2,3,5-trimethylpyrazine, and 2-pentylfuran (FD ≥ 16) as key aroma compounds in RSO (Yin et al., 2021). Notably, compounds with FD factors ≥27, such as 4-methylthiazole (roasted meat), 5-methyl-2-furancarboxaldehyde (burnt, almond, and caramel-like), 2(5H)-furanone (roasted), and 1H-pyrrole-2-carboxaldehyde (beefy, coffee), each contributed significantly to the richness and complexity of the SESO₂ aroma. Heterocyclics were the most abundant aroma types in roasted and microwaved sesame samples (Hu et al., 2025; Yang et al., 2024; Yang et al., 2025; Yin et al., 2021; Yin et al., 2023; Yu et al., 2025). Additionally, several newly identified heterocyclics by GC × GC-TOFMS-O included 2-isobutyl-3-methoxypyrazine (no. 41), 2,4,5-trimethylthiazole (no. 46), 2-acetyl-2-thiazoline (no. 47), 5-hydroxymethylfurfural (no. 55), and 5-methyl-2-thiophenecarboxaldehyde (no. 65). These newly discovered heterocyclics are crucial contributors to the complex and rich aroma profiles of sesame oil.

3.2.3. Sulfides and others

Sulfides also play an important role in sesame oil aroma. Dimethyl disulfide (FD = 81–729), dimethyl trisulfide (FD = 81–2187), and methional (FD = 27–243) were sniffed with high FD factors, conferring distinctive sulfurous, spicy, cabbage, and onion notes. The presence of sulfides notably enhanced the complexity of RSO, MSO, and small mill sesame oil (Yang et al., 2025). As weakly acidic compounds, phenolic compounds derive from thermal degradation of natural phenolics during sesame roasting, typically imparting smoky, sweet, and woody notes to sesame oil (Yang et al., 2025). The 2-methoxyphenol, creosol, maltol, phenol, 4-methylphenol, 2-methoxy-4-vinylphenol, and 2,6-dimethoxyphenol were sniffed in sesame oil (Table 1). Among these, 4-methylphenol (FD = 1) and 2,6-dimethoxyphenol (FD = 1) were only detected in SESO₂, while maltol (FD = 1) was unique to MSO. Previous studies reported that 2-methoxy-4-vinylphenol (FD = 1024) and 2-methoxyphenol (FD = 64) were major aroma compounds in small mill sesame oil (Yu et al., 2025). Furthermore, 4-methoxyphenol, 2-methoxy-4-vinylphenol, and 3-carene were sniffed in sesame oil during hot water extraction (Huang et al., 2024).

3.3. Effect of different pre-treatments on the concentration and Odor activity value (OAV) of aroma compounds in sesame oil

The calibration curves for quantifying aroma compounds in sesame oil were established (Table S3). Fig. 2A illustrates that both the composition and concentration of aroma compounds differ significantly within sesame oils subjected to different pre-treatments. Following roasting, microwave, and steam explosion, the concentrations of pyrazines (10.41–32.43 mg/kg), aldehydes (4.83–10.97 mg/kg), “others” (5.20–9.69 mg/kg), pyrroles (1.33–9.07 mg/kg), furans (3.50–7.35 mg/kg), thiazoles (0.78–3.33 mg/kg), pyridines (0.34–1.23 mg/kg), and thiophenes (0.17–0.99 mg/kg) were significantly elevated compared to CPSO CPSO (Fig. 2B-F). Previous studies have demonstrated that pyrazines, pyrroles, and pyridines, primarily generated via Maillard reactions, are the principal contributors to roasted, nutty, meaty, and caramel-like aromas in fragrant vegetable oils (Chen et al., 2022; Hou et al., 2021; Jia et al., 2024; Yu & Jia, 2025). Furans, formed via lipid oxidation or caramelisation, predominantly contribute popcorn-like, roasted, and caramel-like aromas. Sulfur-containing amino acids act as precursors in Maillard reactions or Strecker degradation pathways, leading to formation of thiophenes and thiazoles (Wang et al., 2024). Roasting also significantly increased the levels of 2,5-dimethylpyrazine, 2-methylpyrazine, 2,3,5-trimethylpyrazine, and 2-ethylpyrazine compared to CPSO (Yin et al., 2021). Liu et al. similarly identified 2,5-dimethylpyrazine, 2-methylpyrazine, and 3-ethyl-2,5-dimethylpyrazine as major aroma contributors in RSO (Liu et al., 2025). Zhang et al. (2025) reported that furan concentrations (e.g., furfural, 2-pentylfuran) were higher in walnut oil subjected to stir-frying and microwave than in cold-pressed oil. Yin et al. found that, compared to CPSO, microwave treatment led to the highest contents of pyrazines (77.34 mg/kg), furans (16.27 mg/kg), thiazoles (8.36 mg/kg), and pyridines (4.05 mg/kg), while for 10 min resulted in the highest levels of pyrroles (19.94 mg/kg) and thiophenes (0.74 mg/kg) in MSO (Yin et al., 2023). Notably, the alcohols was highest in MSO (2.69 mg/kg), followed by SESO₁ (1.40 mg/kg), SESO₂ (1.24 mg/kg), RSO (1.21 mg/kg), and CPSO (0.79 mg/kg). Alcohols are generally associated with fruity and floral aromas, and their relatively high content in MSO may be due to milder conditions that suppress further oxidation to aldehydes or acids. Ester was highest in SESO₁ (0.60 mg/kg), typically imparting fruity aroma. Apart from alcohols and esters, SESO₂ exhibited the highest concentrations of pyrazines, aldehydes, pyrroles, furans, thiazoles, and pyridines. These new findings indicate that steam explosion has a greater effect on the formation of heterocyclics than either roasting or microwave, with higher puffing pressures yielding even greater levels of aroma compounds.

Fig. 2.

The concentration and OAV of aroma compound in the five sesame oils: Total concentration of aroma compounds in different categaries (A); A pie chart of the aroma compounds concentration percentage in the CPSO (B), RSO (C), MSO (D), SESO₁ (E), SESO₂(F); A heatmap of aroma-active compounds (OAV ≥ 1) in five sesame oils (G)_. Abbreviations: CPSO: cold-pressed sesame oil, RSO: roasted sesame oil (180 °C, 20 min), MSO: microwaved sesame oil (1200 W, 8 min), SESO₁: steam-exploded sesame oil (0.8 MPa), SESO₂: steam-exploded sesame oil (1.2 MPa).

OAV is a key parameter for identifying aroma-active compounds. Key contributors (OAV ≥ 1) included methional (OAV = 256.7–658.4), 2-acetyl-2-thiazoline (OAV = 0–233.3), 2-acetylpyrazine (OAV = 0–294.2), 1-methyl-1H-pyrrole-2-carboxaldehyde (OAV = 0–149.0), 2-methoxyphenol (OAV = 24.0–144.0), phenylacetaldehyde (OAV = 16.4–117.3), 2-isobutyl-3-methoxypyrazine (OAV = 0–100.0), dimethyl trisulfide (OAV = 21.8–90.3), dimethyl disulfide (OAV = 12.9–77.4), 2-ethenyl-6-methylpyrazine (OAV ≤1–75.6), dihydro-3-(2H)-thiophenone (OAV = 0–52.6) (Table 2). Similar key odorants in RSO were 2-methoxyphenol (OAV = 114) and phenylacetaldehyde (OAV = 49) (Yin et al., 2021). Wang et al. also identified 2-methoxyphenol (OAV = 84), 2-acetylpyrazine (OAV = 7), dihydro-3-(2H)-thiophenone (OAV = 1), dimethyl disulfide (OAV = 29), and dimethyl trisulfide (OAV = 6) as principal aroma markers in roasted sesame kernels (Wang et al., 2024). A heatmap of OAV ≥ 1 revealed that RSO, MSO, SESO₁, and SESO₂ featured a wide variety of aroma-active compound classes, notably aldehydes (nos. 1–3, 6, 9, 12–13, gray triangle), pyrazines (nos. 32–42, red dot), thiazoles (nos. 43–47, Pink dot), furans (nos. 50–51, 53–55, yellow dot), pyrroles (nos. 59–61，blue dot), sulphides (nos. 67–69, purple square), and miscellaneous categories (nos. 73–74, 77–79, 81, pink square) (Fig. 2G). Specifically, furfural, α-ionone, and 2,3-pentanedione exhibited the highest OAV in RSO; 1-octen-3-one and 3-methyl-1-butanol were aroma markers in MSO; γ-caproalactone and 2-pentylfuran were the aroma markers in SESO₁; 4-methylphenol, 2,6-dimethoxyphenol, 2-acetylpyrazine, 3-ethyl-2,5-dimethylpyrazine; 2-ethenyl-6-methylpyrazine were the aroma markers in SESO₂. Notably, previous research found that steam explosion promoted Maillard reaction (Yu & Jia, 2025), resulting in a marked increase in the OAV of pyrazines in flaxseed oil, demonstrating a consistent pattern with our results in a sesame oil.

Table 2.

The odor activity values (OAVs) of key aroma compounds in sesame oils from different pre-treatments.

No.	Aroma compounds	OToil(mg/kg)	CPSO	RSO	MSO	SESO1	SESO2
1	Pentanal	0.24a	ND	1.2	<1	3.2	4.2
2	2-Methylbutanal	0.11b	2.8	17.8	14.5	18.8	27.4
3	Hexanal	0.3b	1.1	1.4	1.3	2.6	2.7
6	Octanal	0.14b	<1	4.4	3.7	4.2	5.1
9	Benzaldehyde	0.69b	<1	<1	<1	1.5	2.7
12	Phenylacetaldehyde	0.022c	16.4	69.5	56	85.5	117.3
13	(E,E)-2,4-Decadienal	0.066b	<1	3.5	2.2	4.6	12.4
14	2,3-Pentanedione	0.02f	1.5	15.5	4.8	5.0	6.0
15	3-Hydroxy-2-butanone	0.014b	1.1	5.9	2.6	13.9	15.6
17	1-Octen-3-one	0.003f	ND	10.0	28.1	ND	ND
20	3-Methyl-1-butanol	0.112e	3.9	3.7	4.4	4.2	2.5
23	1-Octen-3-ol	0.038e	2.6	8.7	9.2	7.4	10.8
24	Acetic acid	0.35c	1.2	2.6	1.4	3.8	6.8
25	Butanoic acid	0.034b	<1	1.8	1.2	2.7	3.2
26	Hexanoic acid	0.46a	<1	1.3	<1	1.6	1.8
28	Octanoic acid	0.0724e	2.5	6.5	4.6	9.4	13.3
29	γ-Caproalctone	0.018c	6.7	19.4	16.7	20.6	11.7
32	2,5-Dimethylpyrazine	2b	<1	1.5	1.2	1.1	1.6
33	2,6-Dimethylpyrazine	0.951c	ND	1.6	1.9	1.9	7.5
34	2-Ethylpyrazine	0.095b	ND	9.2	5.7	6.5	16.4
35	2,3-Dimethylpyrazine	0.1*	ND	10.0	4.6	7.5	20.7
36	2-Ethyl-6-methylpyrazine	0.051d	<1	11.6	6.3	7.1	28.6
37	2-Ethyl-5-methylpyrazine	1d	<1	1.5	1.3	1.3	4.4
38	2,3,5-Trimethylpyrazine	0.29e	<1	9.6	6.4	7.8	14.1
39	3-Ethyl-2,5-dimethylpyrazine	0.079d	ND	15.3	19.9	18.2	37.7
40	2-Ethenyl-6-methylpyrazine	0.027c	ND	21.9	4.1	7.0	75.6
41	2-Isobutyl-3-methoxypyrazine	0.0004*	ND	50.2	60.4	50.3	100.1
42	2-Acetylpyrazine	0.012*	ND	37.5	10.2	52.5	294.2
43	Thiazole	0.038f	ND	12.6	8.7	13.4	16.3
44	2,4-dimethylthiazole	0.19f	ND	1.3	<1	1.4	6.5
45	4-Methylthiazole	0.01*	ND	11.3	9.2	5.0	41.2
46	2,4,5-Trimethylthiazole	0.05*	<1	2.8	<1	2.4	6.2
47	2-Acetyl-2-thiazoline	0.002*	ND	138.9	66.7	116.7	233.3
50	2-Pentylfuran	0.116e	2.2	3.1	2.6	7.2	6.8
51	Furfural	0.7d	<1	3.7	2.4	1.7	3.2
53	5-Methyl-2-furanaldehyde	0.26c	ND	5.8	5.3	4.1	9.3
54	2(5H)-Furanone	0.02e	ND	11.0	7.4	11.5	42.0
55	5-Hydroxymethylfurfural	0.05d	ND	5.8	7.6	11.2	12.8
58	1-(2-Pyridinyl)ethanone	0.5*	ND	<1	ND	<1	1.4
59	1-Methyl-1H-pyrrole-2-carboxaldehyde	0.02e	ND	35.0	23.4	25.5	149.0
60	2-Acetylpyrrole	0.3536c	ND	2.0	<1	1.5	5.8
61	1H-Pyrrole-2-carboxaldehyde	0.156b	<1	6.1	4.1	7.8	26
62	Thiophene	0.08*	ND	1.2	1.1	1.0	1.3
64	Dihydro-3-(2H)-thiophenone	0.01*	ND	33.4	5.0	12.1	52.6
67	Dimethyl disulfide	0.0008*	12.9	51.5	39.9	69.9	77.4
68	Dimethyl trisulfide	0.004*	21.8	64.6	77.8	71.2	90.3
69	Methional	0.00005a	256.7	395.6	198.1	524.4	658.4
73	α-Ionone	0.15*	1.2	3.1	1.5	1.4	1.1
74	2-Methoxyphenol	0.015c	24.0	80.7	62.7	92.7	144.0
77	Phenol	0.1*	<1	<1	<1	1.5	3.7
78	4-Methylphenol	0.07*	ND	ND	ND	ND	1.8
79	2-Methoxy-4-vinylphenol	0.2d	<1	10.3	7.5	16.4	24.6
81	2,6-Dimethoxyphenol	0.26*	ND	<1	ND	<1	1.4

a: Jia et al., 2019; b: Jia et al., 2024; c: Jia et al. 2021; d: Zhou et al., 2019; e: Jia et al., 2020; *:Odor thresholds were newly determined in refined oil.

3.4. GC-IMS analysis of sesame oils with different pre-treatments

As listed in Table S4, a total of 102 signal peaks (involving monomers and dimers) were observed across sesame oil samples. Among these, 80 distinct aroma compounds were found, such as 16 aldehydes, 9 ketones, 15 alcohols, 5 acids, 8 esters, 8 pyrazines, 1 thiazole, 5 furans, 1 pyridine, 5 sulfides, and 7 “others”. Compared with the GC × GC–TOFMS–O analysis, GC–IMS did not detect pyrroles and thiophenes. GC–IMS provides complementary insight into impact of pre-treatments on aroma profiles of sesame oil from different analytical perspectives. Most aroma compounds appeared within the 0–1000 s retention time interval (Fig. 3A). Each signal spot corresponds to a unique aroma compound, with larger areas indicating higher concentrations. Compared with CPSO, the signal intensities for most volatile compounds in RSO, MSO, SESO₁, and SESO₂ were substantially elevated. Taking CPSO as a reference, progressive colour intensity (e.g., deepening red) indicates an increase in volatile compound concentrations following processing pre-treatments based on difference plot result (Fig. 3B). Fingerprint mapping (Fig. 3C) summarises the characteristic peaks of all aroma compounds in bar graphs, enabling the identification of “aroma markers” sensitive to specific pre-treatments (Li et al., 2024; Zhang et al., 2025). In CPSO, the primary aroma compounds are concentrated in region I, including 2-butanol, 2-pentanol, dimethyl sulphide, and ethyl pentanoate. In RSO, major aroma compounds include furfural, (E)-2-pentenal, ethyl 2-methylbutanoate, 2,3-pentanedione, and methyl acetate (region II). MSO is characterised by 1-hexanol and 3-pentanol (region III). SESO₁ contains 1-butanol, (E)-3-hexen-1-ol, 2-pentylfuran, (E)-2-hexenal, and (E)-2-heptenal (region IV). SESO₂ features 2,3,5-trimethylpyrazine, 2-furfurylthiol, 5-ethyl-2-furanaldehyde, 2-methylpyrazine, 2,4-hexadienal, (E,E)-2,4-heptadienal, and γ-butyrolactone (region V). In summary, GC-IMS-fingerprint effectively distinguish sesame oils subjected to different pre-treatments. The variation patterns of most aroma compounds reflected by GC-IMS are consistent with those observed in Section 3.3.

Fig. 3.

Top-view plot (A), difference plot (B), gallery plot (C) of aroma compounds in CPSO, RSO, MSO, SESO₁ and SESO₂ based on the signal intensity via HS-GC–IMS.

3.5. Correlation analysis of aroma-active compounds and aroma profiles in sesame oils with different pre-treatments

Fig. 4A shows the sensory scores of five sesame oils (CPSO, RSO, MSO, SESO₁, and SESO₂) across 12 odor attributes, demonstrating the significant impact of different pre-treatments on aroma profiles. CPSO exhibited a raw sesame odor primarily described as “grass”, “earthy”, and “woody”, it lacked desirable process-induced notes (roasted, nutty). Both RSO and MSO were characterised by roasted, nutty, and popcorn-like aromas. Peng et al. (2025) reported that a sweet aroma in roasted peanut oil (140 °C) was attributed to elevated aldehyde, nevertheless, pyrazines, pyrroles, and furans rapidly accumulate in roasted peanut oil (170 °C), resulted in progressively dominant roasted, nutty, and even burnt notes (Peng et al., 2025). For SESOs, the perceived intensities of roasted, popcorn-like, burned, and caramel-like notes increased with elevated explosion pressure (0.8 Mpa and 1.2 MPa), although high pressures risked producing excessive burnt or bitter notes (e.g., the “burnt” score for SESO₂ was 1.0). This phenomenon is attributed to the rupture of sesame seeds' cells caused by high temperatures and pressures during steam explosion, producing more aroma compounds (Yu et al., 2020).

Fig. 4.

Spider diagram of sensory scores (A), spider diagram of sensory scores between sesame oils and reconstituted samples (B-F), network diagram (G-H) between aroma-active compounds (OAV ≥ 1) and aroma profiles in CPSO, RSO, MSO, SESO₁ and SESO₂.

Based on the results of the reconstitution experiments, it was found that the screened key aroma compounds could reproduce the characteristic aroma of sesame oil (Fig. 4B-F). Table 3 summarises the results of omission testing for different classes of aroma compounds. For CPSO, omission of aldehydes significantly affected its characteristic aroma (P < 0.01), while in RSO, MSO, SESO₁, and SESO₂, omission of aldehydes showed an even stronger correlation (P < 0.001), showing that aldehydes are core contributors to overall aroma of sesame oil. Thiazoles, furans, and pyrroles showed marked influence on SESO₂ (P < 0.01) than on the other oils. Notably, omission of pyrazines and thiophenes resulted in a significant decrease in characteristic roasted, nutty, burned aroma for SESO₂ (P < 0.001). Hou et al. reported that both pyrazines and aldehydes significantly increased after roasting and microwave, imparting a pleasant roasted aromas (Hou et al., 2021). For instance, pyrazines (e.g., 3-ethyl-2,5-dimethylpyrazine, 2,3,5-trimethylpyrazine, 2-ethyl-6-methylpyrazine) as key contributors to characteristic roasted and “caramel-like” notes (Xu et al., 2025). Hu et al. demonstrated that thiophenes were positively correlated with roasted, burnt, nutty, and baking notes, serving as important contributors to typical aroma profile of sesame oil (Hu et al., 2025). Additionally, esters and pyridines had negligible effects on aroma, indicating a minor contribution. The network diagram in Fig. 4G illustrates the associations between aroma-active compounds and odor attributes in a sesame oil under pre-treatments. The largest area in inner circle indicates greatest figure of aroma-active compounds (OAV ≥ 1). According to principle that higher OAV values indicate stronger sensory contributions, the overall odor intensity followed the order: SESO₂ > SESO₁ > RSO > MSO > CPSO. Notably, steam explosion treatments markedly promoted the formation of roasted, nutty, and popcorn-like notes, which became the dominant aroma attributes (Fig. 4H).

Table 3.

Omission model analysis of aroma-active compounds in sesame oils from different pre-treatments.

Omitted odorants	Significance
	CPSO	RSO	MSO	SESO1	SESO2
all aldehydes	**	***	***	***	***
all ketones	*	**	**	*	*
all alcohols	*	*	*	*	*
all acids	*	*	*	*	*
all esters	/	/	/	/	/
all pyrazines	/	***	***	***	***
all thiazoles	/	*	*	*	**
all furans	*	*	*	*	**
all pyridines	/	/	/	/	/
all pyrroles	/	*	*	*	**
all thiophenes	/	**	*	*	***
all sulfides	*	*	*	*	*
all others	/	*	*	**	**

‘*’ Indicates significance at P < 0.05; ‘**’ indicates significance at P < 0.01; ‘***’ indicates significance at P < 0.001.

Further analysis of sesame seeds revealed significant variations in amino acid, reducing sugar, and fatty acid levels across different pre-treatment groups (Fig. S1). In Fig. 5A,the size of the nodes represents the correlation between precursors and heterocyclic compounds (the larger the circle, the higher the correlation), while the edges indicate statistically significant correlations between them (|r| > 0.8, P < 0.05). The results indicate that the pre-treatments significantly altered the 3-ethyl-2,5-dimethylpyrazine, 2-isobutyl-3-methoxypyrazine, 2,5-dimethylpyrazine, 5-methyl-2-furanaldehyde, 5-hydroxymethylfurfural of sesame oil by modulating Maillard reaction, with precursors including reducing sugars such as glucose, ribose, and fructose, as well as amino acids such as serine, cysteine, arginine, and lysine. Similarly, hexanal, (E,E)-2,4-decadienal, pentanal, octanal in sesame oil are closely associated with lipid oxidation/degradation, and their main precursors include linoleic acid, oleic acid, and linolenic acids (Fig. 5B). The formation pathways of key aroma compounds are illustrated in Fig. 5 (C—D), the amino groups of amino acids condense with the carbonyl groups of reducing sugars to form unstable Schiff bases, which then cyclize to form Amadori rearrangement products, subsequently generated 5-hydroxymethylfurfural and 2,5-dimethylpyrazine. In addition, unsaturated fatty acids produce hexanal, (E,E)-2,4-decadienal, octanal through free radical chain reactions under the action of lipoxygenase (LOX) or thermal processing. This results clearly indicate that diverse pre-treatments exert distinct effects on sesame oil aroma by selectively modulating the Maillard reaction and lipid oxidation/degradation. This conclusion aligns with Hu et al. (2025), who showed that combining lipoxygenase treatment and Maillard reaction modeling mimicked roasted sesame oil flavor.

Fig. 5.

Network diagram between heterocyclic compounds (OAV ≥ 1) and amino acids as well as reducing sugars (A); network diagram between aroma compounds (OAV ≥ 1) and fatty acids (B). (The larger the circle, the higher the correlation). The possible formation pathway of key aroma compounds: Maillard reaction (C); Lipid degradation (D). LOX: Lipoxygenase, HPL: hydroperoxide lyase; △: Thermal processing.

3.6. Effects of different pre-treatments on nutritional quality of sesame oil

Fig. 6 (A–H) depict the influence of pre-treatments on TP content, γ-tocopherol, phytosterols, and sesame lignans in sesame oil. The TP increased in the order: CPSO (40.4 GAE/100 g) < RSO (43.5 GAE/100 g) < SESO₁ (53.4 GAE/100 g) < MSO (59.6 GAE/100 g) < SESO₂ (59.8 GAE/100 g) (Table S5), indicating that both steam explosion and microwave significantly enhance TP content. These findings are consistent with recent studies on the effects of microwave on TPC in flaxseed oil (Suri et al., 2020), where the level of polyphenols in roasted flaxseed oil (170.7–235.2 mg/100 g) were higher than cold pressing flaxseed oil (158.1 mg/100 g). Similarly, Sun et al. reported increased TP content of corn oil by elevated roasting temperature (Sun et al., 2025). Steam explosion also increased TP content, as puffing pressure increased from 0 to 1.6 MPa; however, excessive pressure (1.8–2.3 MPa) resulted in a decrease in TP content (Zhang et al., 2019). Compared to CPSO (575.9 mg/kg), γ-tocopherol content decreased slightly after roasting and microwave (by 2.3–4.3%). In contrast, steam explosion increased γ-tocopherol content, reaching 650.1 mg/kg in SESO₂, likely due to reduced oxidative loss under high-pressure conditions. SESO₂ contained the highest brassicasterol (903.8 mg/kg) and β-sitosterol (2787.0 mg/kg) levels, whereas CPSO exhibited the highest stigmasterol content (449.7 mg/kg). CPSO contained the highest levels of sesamin (6569.1 mg/kg) and sesamolin (3320.8 mg/kg), both of which decreased following roasting, microwave, and steam explosion. Notably, sesamol was increased significantly after roasting, microwave, and steam explosion, with the highest level observed in SESO₂ (58.4 mg/kg). This increase likely results from the thermal degradation and conversion of sesamolin to sesamol during pre-treatment (Yang et al., 2025). As shown in Fig. S2, DPPH, FRAP, and OSI of SESO₂, SESO₁, MSO, and RSO were significantly higher than CPSO. This enhancement is likely related to SESO₂'s higher content of antioxidant components such as TP and sesamol. Previous studies confirmed that thermal treatments—including wet heat, microwave and roasting—can improve oxidative stability and antioxidant activity of a sesame oil (Xiang et al., 2025).

Fig. 6.

Contents of total phenolic (A), γ-tocopherol (B), brassicasterol (C), stigmasterol (D), β-sitosterol (E), sesamol (F), sesamin (G), and sesamolin (H) in sesame oils produced through different processes.

4. Conclusions

This study utilised integrated multi-omics analyses to systematically elucidate the mechanisms by which heating pre-treatments impact the aroma profile and nutritional quality of sesame oil. For the first time, it systematically compares the effects of pre-treatment processes (cold pressing, roasting, microwave, and steam explosion) on sesame oil aroma, filling the gap in comparative studies between steam explosion and other processes. Distinct differences in aroma compounds among CPSO, RSO, MSO, SEFO₁, and SEFO₂ were characterised using GC × GC–TOFMS-O and GC–IMS, and found characteristic aroma markers for different pre-treatment processes. Newly discovered 2-acetyl-2-thiazoline, 5-hydroxymethylfurfural, and 5-methyl-2-thiophenecarboxaldehyde were crucial contributors to the complex and rich aroma profiles of sesame oil.

Cold-pressed excels at preserving heat-sensitive nutrients such as sesamin and sesamolin, while CPSO exhibited the simplest aroma compounds and aroma complexity and intensity of CPSO are relatively low. Roasting significantly accelerates the formation of aroma compounds such as 2,3-pentanedione, γ-caproalctone, 2,5-dimethylpyrazine while enriching flavor diversity of RSO. However, the roasting process is time-consuming, and over-roasting can produce undesirable flavors while causing substantial loss of heat-sensitive nutrients. Microwave processing enables rapid heating and imparts sesame oil with distinctive aromatic profiles (nutty and roasted) as well as antioxidant activities. However, it may encounter localized overheating issues during large-scale production, and the equipment costs are high. Steam explosion uniquely enhanced the formation of pyrazines, pyrroles, and thiazoles, resulting in the highest overall odor intensity (SESO₂ > SESO₁ > RSO > MSO > CPSO). This process effectively balances aroma intensity, nutrient retention, and antioxidant activity, highlighting its industrial application potential. Compared to traditional roasting process, it drastically reduces processing time but requires strict adherence to high-pressure equipment safety protocols and professional operation, with overprocessing potentially causing flavor degradation. Overall, these findings advance our understanding of sesame oil aroma and provide a theoretical basis for developing high-value, aroma-enhanced oil products through targeted pre-treatment strategies. Future studies will employ molecular sensory science, metabolomics, and stable isotope tracing to comprehensively elucidate how the process alters metabolic pathways of non-volatile flavor precursors in sesame seeds and identify key regulatory sites.

CRediT authorship contribution statement

Gaiwen Yu: Writing – original draft, Formal analysis, Methodology. Xiao Jia: Writing – review & editing, Project administration, Investigation, Funding acquisition.

Ethical statement

When conducting experiments involving human sensory perception, verbal informed consent was obtained from all participants prior to their involvement. Sesame oil is safe for consumption and poses no health risks. In addition to verbal informed consent, all participants were provided with a detailed written explanation of the study's purpose, procedures, and voluntary participation rights. The study protocol adhering strictly to the Declaration of Helsinki and national ethical guidelines for human research.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary material

: Details of chemicals and standards (Table S1). The nine odor attributes and referenced standards for sensory evaluation of sesame oils (Table S2). The standard calibration curves for aroma compounds (Table S3). Aroma compounds identified in sesame oils via GC–IMS (Table S4). The impact of different pre-treatments on nutritional components (Table S5). Variations in amino acids, reducing sugars (in sesame seeds) and fatty acids (in sesame oils) across different pre-treatments (Fig. S1). Effects of different pre-treatments on the antioxidant capacity of sesame oil (Fig. S2).

Data availability

Data will be made available on request.