Effects of drying methods on the functional components and volatile organic compounds of Atractylodes chinensis Authors: Yichao Liu, Mingjiu Wang, Zhijun Wang, Guanzhi Liu, Yalaqiqige Su, Junfeng Hao, Dong Luo, Rui Yang, Gentu Ge, Yushan Jia Published: 2026 ABSTRACT Different drying methods affect the functional components and the odor composition of medicinal plants. This study investigated the effects of sun drying (S), shade drying (Y), hot air drying (H), microwave drying (W), and freeze drying (D) on the functional components, odor response intensity, and volatile organic compound composition (VOCs) of Atractylodes chinensis using an electronic nose and GC-MS. Results showed that the content of functional components was higher in groups D and Y, and lower in group H. The electronic nose had the highest response values in the W1W (terpenes), W2W (organic sulfides), and W5S (nitrogen oxides) channels. The odor response intensity of groups D and Y was higher, while group H was lower, showing a strong positive correlation with VOCs. The VOCs of A. chinensis were mainly terpenoids, esters, phenols and ketones, and the contents of differential VOCs in group D were mainly atractylodin, atractylone, bornyl acetate, and other substances with woody, camphor, and bitter and hot odors. The biosynthesis of secondary metabolites and the biosynthesis of terpenoids and steroids were the main metabolic pathways. In conclusion, drying methods affect the functional ingredients and odor of A. chinensis. Hot air drying lost more of these functional components and odors, while more were retained through freeze drying. Introduction Atractylodes chinensis is a perennial herb in the chrysanthemum family with irregularly cylindrical or conical roots. Its skin is brown or dark brown, and the interior is white or yellowish white, with a strong and fragrant smell [1]. A. chinensis contains rich volatile oils, most of which are present in the roots, followed by flowers, leaves, and stems, with the main components including atractylodin, atractylone and β-eudesmol [2]. These ingredients give A. chinensis a special odor and have antibacterial, anti-inflammatory, and dehumidification effects. Xu et al. [3] found that atractylodin had powerful antioxidant and anti-inflammatory effects on listeria monocytogenes infection. A study by He et al. [4] showed that the mechanism of atractylone antibacterial activity is to destroy cell membranes, and concludrd that it can be used as a natural antioxidant and antibacterial agent. Chen et al. [5] found that the sesquiterpenoids and polyacetylenes biosynthetic pathways were related to changes in the contents of these ingredients. This suggested that understanding the association between A. chinensis processing methods and its functional components could point to ways to extend products’ storage time and improve the efficiency of the industry. The odor of plants comes from volatile organic compounds (VOCs). Gas chromatography-mass spectrometry (GC-MS) analysis can identify and quantitatively analyze volatile and semi-volatile components in plants [6]. It can evaluate and distinguish the quality of plants from different producing areas, harvesting periods, or processing methods, so analysis using this method can help identify ways to ensure the stability and effectiveness of plant components. A study by Zhu et al. [7] found that the flower sources of honey were significantly correlated with volatiles, and that the source of honey could be identified by analyzing the VOCs. Yu et al. [8] used GC-MS and electronic nose (e-nose) to successfully distinguish the varieties and origin of dried ginger from seven major producing areas in China. Turner et al. [9] found that the harvest stage had a greater impact on the aroma quality of celery. As celery matures, monoterpenes, sesquiterpenes, and phthalates began to decrease, while alcohols increased. Hence, we posited that the correlation between functional components and volatile organic compounds of A. chinensis could provide a reference for the classification and quality characterization of this plant. After harvesting, fresh A. chinensis has higher water content, is metabolically active, and easily degrades during storage, resulting in the loss of functional components and a decline in product quality decline [10]. Jin et al. [11] studied the differences in the volatile distribution of Perilla leaves after drying at different temperatures and found that most of the volatile flavor substances could be better retained using lower drying temperature and shorter pretreatment duration. In the hawthorn roasting process, Fei et al. [12] identified a total of 54 volatile flavor substances, among which aldehydes, ketones, esters, and furans underwent the most significant changes. Wang et al. [13] showed that black tea would produce different types of aroma types after being dried for different durations, and identified 45 min as the key time point for the formation of tea aroma quality, after which aroma quality would decline. Drying methods are very important in the processing, storage, and transportation of agricultural products due to their effects on product functional characteristics and quality, so identifying the appropriate drying method is of great significance for the improvement of product quality and the retention of plant functional components. To date, there have been few reports on the comprehensively analyzing of the quality and functional components of A. chinensis under different drying methods. In order to fill this research gap, we used five different drying methods: sun drying (using solar radiation’s heat energy, heating the surface of the materials is heated by infrared rays, and the moisture evaporates through natural convection), shade drying (in a shaded and ventilated environment, relying on air flow to remove moisture, avoiding direct exposure to ultraviolet rays and high temperatures), hot air drying (forced convection hot air comes into contact with the surface of the materials, through conduction and convective heat transfer, the moisture diffuses from the interior to the surface to the interior and then evaporates), microwave drying (using microwaves to excite the polar molecules of the materials to rotate at high speed, achieving volume heating through molecular friction, and the heat is generated from the inside out), and freeze drying (based on the sublimation principle, moisture is first frozen at low temperatures, and then in a vacuum environment, the ice crystals directly sublime from the solid state to the gaseous state) [14]. The aim was to compare the effects of different drying methods on the functional components and volatile organic compounds of A. chinensis, so as to provide a scientific basis for selecting the best drying method and improving the storage and transportation quality of A. chinensis. Materials and methods Sample preparation Fresh A. chinensis samples were purchased from the planting base in Sanshan Township, Balin Left Banner, Chifeng City, Inner Mongolia Autonomous Region, China. Atractylodes with uniform growth and no pests or diseases were randomly selected as test samples. The roots and soil were removed, and samples were sliced (2–4 mm) before applying different drying treatments: sun drying (S), shade drying (Y), hot air drying (H), microwave drying (W), and freeze drying (D), with 6 replicates of each treatment. Plant samples were dried and weighed until samples approached a constant weight. After cooling for 30 min, samples were weighed and re-weighted until the difference between the two weight measurements was less than ± 0.5%. Specific treatments were implemented as follows. In the sun drying treatment, plant slices were placed in a sunny and ventilated area for drying, and the drying process concluded when there was no variation in weight. For shade drying, slices were placed in the shade and the drying process ended when a constant weight was achieved. For the hot air drying treatment, slices were placed in an electric blast drying oven (Memmert, UN260), the oven temperature set to 65℃, and the drying process was completed when a constant weight was achieved. In microwave drying, slices were placed in a microwave device (Meidi, EM7KCGW3), with the current set at 250 mA and the temperature at 65℃, and samples were dried to a constant weight. For freeze drying, slices were placed in a freeze dryer (Christ, Alphal-4LD) with temperature set to -80℃, and the drying process finished when constant weight was observed [15]. Functional component analysis Standard reference substances for atractylodin, atractylone, hinesol, and β-eudesmol were weighed (Shanghai Yuanye Biotechnology Co., Ltd), and methanol was added to prepare a mixed reference solution. A. chinensis powder (0.5 g) was weighed and placed in a conical bottle with a stopper. 20 mL of methanol was added quickly, and the conical bottle mouth was sealed with a sealing film. It was then wrapped tightly with tin paper and placed in an ultrasonic cleaner at constant water temperature. After ultrasonic extraction, it was cooled to room temperature, and weighed again. The weight loss was offset with methanol, and the filtrate was filtered with a filter membrane to obtain a test solution. The mixed reference solution and the test solution were taken and determined by HPLC (Waters, e2695). The chromatographic column settings were as follows: C18 (50 mm × 2.1 mm, 1.7 μm); Mobile phase: acetonitrile - aqueous solution; Gradient elution: 0–2 min with 50% acetonitrile, 2–4 min with 50% − 60% acetonitrile, 4–6 min with 60% − 70% acetonitrile, 6–7 min with 70% − 95% acetonitrile, 7–9 min with 95% acetonitrile, 9–10 min with 95% − 50% acetonitrile; flow rate: 0.5 mL/min; detection wavelength: 203 nm; column temperature: 35 ℃; injection volume: 2 µL. The ultrasonic conditions were 240 W, 40 kHz, with an ultrasonic duration of 30 min and a temperature of 25℃ [16]. Response intensity analysis The odor response intensity of the samples was measured using a PEN3 electronic nose from Ensoul Technology Ltd., Germany. The performance characteristics of each sensor in the electronic nose can be found in Table 1. A 2 g sample was placed into a 20 ml sample bottle, and the injection needle was cleaned for 95 s before sampling. The zero point was adjusted for 5 s, and the preparation time was 5 s. The initial injection flow rate and the sample chamber flow rate were both set at 300 mL/min. The measurement period duration was 100 s, with data recorded every 2 s with an acquisition interval of 1 s. Prior to subsequent injections, the system required a 95-second period to cleanse the injection needle in order to establish a stable baseline. The samples were evaluated in triplicate, and the relative odor response intensity was determined by calculating the ratio of the resistance of the volatile gases in the sample (G) to the resistance of the corresponding carrier gas (G0) over time [17]. Table 1Performance description of the sensorSerial numberSensor namePerformance description1W1CSensitive to aromatic components2W5SSensitive to nitrogen oxides3W3CSensitive to aromatic ammonia4W6SSensitive to hydrides5W5CSensitive to aromatic alkanes6W1SSensitive to methyl compound7W1WSensitive to terpenes and inorganic sulfides8W2SSensitive to alcohols, aldehydes, and ketones9W2WSensitive to organic sulfides10W3SSensitive to alkanes Table 1Performance description of the sensorSerial numberSensor namePerformance description1W1CSensitive to aromatic components2W5SSensitive to nitrogen oxides3W3CSensitive to aromatic ammonia4W6SSensitive to hydrides5W5CSensitive to aromatic alkanes6W1SSensitive to methyl compound7W1WSensitive to terpenes and inorganic sulfides8W2SSensitive to alcohols, aldehydes, and ketones9W2WSensitive to organic sulfides10W3SSensitive to alkanes Performance description of the sensor Volatile organic compounds analysis The sample was squeezed with liquid nitrogen, and 500 mg of powder was mixed into 20 mL vials. Saturated NaCl solution was added to prevent enzyme reactions, and the vial was sealed with agile overhead silicone. Each vial was heated at 60 °C for 5 min before solid-phase microextraction. An Agilent fiber optic was then inserted into the headspace vials and placed at 60 °C for 15 min to capture VOCs in the sample. VOCs were identified and quantified using an Agilent 8890 7000D mass spectrometer. A helium-loaded DB-5MS capillary column (5% phenyl-polymethylsiloxane) was used to inhale at a rate of 1.2 mL/min. The injector temperature was maintained at 250 °C, and the detector temperature was set at 280 °C. The oven temperature was set to start at 40 °C for 3.5 min and then increase at 10 °C per minute until 100 degrees Celsius was reached. This was followed by a 7 °C rise per minute to 180 °C, then a 25 °C rise per minute to 280 °C, and finally, the temperature was held constant for 5 min. Mass spectra were acquired using the electron impact (EI) ionization technique with an energy of 70 electron volts (eV). The temperatures of the quadrupole mass detector, ion source, and transmission lines were 150 °C, 230 °C and 280 °C, respectively. The selective ion monitoring (SIM) technique was used to identify and quantify the substances. Integration and calibration correction procedures were performed using the MassHunter quantitative software with 3-hexanone (CAS 24588-54-3) as the internal standard. The relative content of each substance was calculated using the peak area normalization method. The annotation results of all VOCs metabolites in the KEGG database were classified according to the types of pathways [18]. Statistical analysis SAS 9.2 was employed for analysis of variance (ANOVA) with six replicates in each group. Statistical significance was determined by considering p-values less than 0.05. The reliability of the sample means was assessed using the mean ± standard error of the mean. Microsoft Excel 2010 was used for table creation, while GraphPad Prism 8.0.2 and R 4.1.2 were employed for graph generation. The identified metabolites were annotated using the Kyoto Encyclopedia of Genes and Genomes (KEGG) compound database, and then the annotated metabolites were mapped to the KEGG pathway database. Results Functional components of Atractylodes chinensis The contents of functional components in each treatment group are shown in Table 2. The table indicates that different drying methods have significant effects on each component (P < 0.05). The contents of atractylodin and atractylone in group D were significantly higher than in other groups (P < 0.05), while the contents of hinesol and β-eudesmol were not significantly different from those in group Y (P > 0.05). The contents of atractylodin, atractylone, hinesol, and β-eudesmol in groups H and W were significantly lower than those in other groups (P < 0.05), and the contents of atractylodin, hinesol, and β-eudesmol in group W were higher than those in group H. The contents of atractylodin, atractylone, hinesol, and β-eudesmol in group Y were significantly higher than those in group S (P < 0.05). Table 2Functional components of Atractylodes chinensisItemsSYHWDP-valueAtractylodin (mg/g)3.98 ± 0.09c4.51 ± 0.07b3.28 ± 0.07d3.44 ± 0.08d5.61 ± 0.18a< 0.01Atractylone (mg/g)5.23 ± 0.11c7.25 ± 0.13b4.32 ± 0.23d4.19 ± 0.15d8.45 ± 0.31a< 0.01Hinesol (mg/g)3.13 ± 0.10b4.27 ± 0.12a2.75 ± 0.10c3.13 ± 0.11c4.45 ± 0.16a< 0.01β-Eudesmol (mg/g)3.14 ± 0.07b4.31 ± 0.12a2.58 ± 0.06c3.04 ± 0.09 b4.53 ± 0.13a< 0.01Sun drying (S), shade drying (Y), hot air drying (H) microwave drying (W) and freeze drying (D)Different letters indicate significant differences among different treatments (P < 0.05) Table 2Functional components of Atractylodes chinensisItemsSYHWDP-valueAtractylodin (mg/g)3.98 ± 0.09c4.51 ± 0.07b3.28 ± 0.07d3.44 ± 0.08d5.61 ± 0.18a< 0.01Atractylone (mg/g)5.23 ± 0.11c7.25 ± 0.13b4.32 ± 0.23d4.19 ± 0.15d8.45 ± 0.31a< 0.01Hinesol (mg/g)3.13 ± 0.10b4.27 ± 0.12a2.75 ± 0.10c3.13 ± 0.11c4.45 ± 0.16a< 0.01β-Eudesmol (mg/g)3.14 ± 0.07b4.31 ± 0.12a2.58 ± 0.06c3.04 ± 0.09 b4.53 ± 0.13a< 0.01Sun drying (S), shade drying (Y), hot air drying (H) microwave drying (W) and freeze drying (D)Different letters indicate significant differences among different treatments (P < 0.05) Functional components of Atractylodes chinensis Sun drying (S), shade drying (Y), hot air drying (H) microwave drying (W) and freeze drying (D) Different letters indicate significant differences among different treatments (P < 0.05) Odor response intensity of Atractylodes chinensis The electronic nose can distinguish the differences in odor of sample through the response intensity of the sensor. The odor characteristics of A. chinensis under different drying methods are shown in Fig. 1A. The sensors with high odor response intensity were W1W, W2W, W5S, and W1S, corresponding to terpenes, inorganic sulfides, organic sulfides, nitrogen oxides, and methyl compound, respectively. The most obvious difference among samples was W1W, followed by W2W and W5S, indicating that drying method had a greater effect on terpenes, inorganic sulfides, organic sulfides, and nitrogen oxides. Group Y and group D had the largest response intensities to these three sensors, followed by group S and group W, while H group had the smallest response intensities. Fig. 1Radar plots of response intensity for each group (A). OPLS-DA scores plot of response intensity (B). Sun drying (S), shade drying (Y), hot air drying (H) microwave drying (W) and freeze drying (D). OPLS-DA model evaluation parameters: R2X = 0.869, R2Y = 0.906, Q2 = 0.865 Radar plots of response intensity for each group (A). OPLS-DA scores plot of response intensity (B). Sun drying (S), shade drying (Y), hot air drying (H) microwave drying (W) and freeze drying (D). OPLS-DA model evaluation parameters: R2X = 0.869, R2Y = 0.906, Q2 = 0.865 Orthogonal partial least squares discriminant analysis (OPLS-DA) was used to assess the differences between and within the sample groups. In Fig. 1, the horizontal direction represents differences between groups, and the vertical direction represents differences within groups, and the percentage represents the interpretation rate. Figure 1B shows that A. chinensis samples treated with different drying methods were significantly separated between groups but showed obvious aggregation within groups, indicating significant differences in odor between groups and small differences within groups. Among them, groups Y, H, and D were distributed far from each other, indicating large differences in odor characteristics, while groups S and W were distributed close to each other, indicating smaller differences in odor characteristics. VOCs composition of Atractylodes chinensis The results of GC-MS showed that there were 41 volatile substances in all A. chinensis treatment groups, among which terpenoids were the main compounds, followed by esters, and a small number of ketones and phenols (Fig. 2A). The composition, retention time, and relative contents of each group are shown in Table 3. Atractylone, atractylodin, hinesol, β-eudesmol, caryophyllene, and γ-elemene were the most abundant substances. Among them, atractylone, atractylodin, hinesol, and β-eudesmol contents in group D were higher, while those in group H were lower, which was the same as the previous determination results. The results of OPLS-DA showed that there were significant differences among different sample groups, and small differences within sample groups, indicating that different drying methods had significant effects on the VOCs of A. chinensis. Group D was significantly separated from the other groups, indicating a large difference from the other groups. The distribution of Y, H, and D groups were far from each other, indicating large differences among these three groups. The distribution of groups S and W was close, so the difference between these two groups was small (Fig. 2B). Table 3VOCs composition of Atractylodes chinensisNORTItemsCASRelative content (%)SYHWD16.29α-Pinene7785-20-40.25 ± 0.04b0.31 ± 0.03a0.17 ± 0.01c0.28 ± 0.04b0.08 ± 0.10d27.24Geranyl acetate105-87-30.42 ± 0.05c0.72 ± 0.03a0.34 ± 0.03c0.51 ± 0.01b0.07 ± 0.03d37.73β-Pinene127-91-30.63 ± 0.03b0.65 ± 0.02b0.82 ± 0.06a0.74 ± 0.02b0.15 ± 0.04d48.12δ-elemene20307-84-00.50 ± 0.11b0.75 ± 0.04a0.60 ± 0.02b0.33 ± 0.12c0.11 ± 0.08d58.48Isoborneol124-76-50.71 ± 0.08a0.70 ± 0.02a0.36 ± 0.04b0.38 ± 0.11b0.24 ± 0.04c69.89Caryophyllene13877-93-54.70 ± 0.23b4.71 ± 0.37b6.62 ± 0.38a6.04 ± 0.30a2.13 ± 0.20c710.22Myrcene123-35-30.89 ± 0.05a0.86 ± 0.07a0.87 ± 0.29a0.41 ± 0.03b0.14 ± 0.03c810.514-Carene29050-33-70.68 ± 0.04a0.69 ± 0.05a0.57 ± 0.03b0.59 ± 0.03b0.13 ± 0.07c911.38Carvone99-49-00.90 ± 0.07a1.02 ± 0.07a0.71 ± 0.08b0.88 ± 0.04a0.22 ± 0.321013.35Aromadendrene489-39-42.45 ± 0.24a1.48 ± 0.28b2.42 ± 0.26a2.24 ± 0.05a0.27 ± 0.03c1114.37Bornyl acetate76-49-30.71 ± 0.15b0.68 ± 0.07b0.78 ± 0.06b1.03 ± 0.03a0.25 ± 0.05c1215.48γ-Elemene29873-99-22.44 ± 0.31b1.50 ± 0.21c3.04 ± 0.46a2.54 ± 0.73b2.14 ± 0.49b1318.41Limonene138-86-30.73 ± 0.26b0.89 ± 0.09a0.51 ± 0.05bc0.62 ± 0.03b0.11 ± 0.01c1419.05Citronellyl acetate67650-82-20.91 ± 0.30a0.91 ± 0.11a0.11 ± 0.01b0.85 ± 0.02a0.23 ± 0.02b1521.19Cadinene29350-73-00.07 ± 0.01c0.09 ± 0.01c0.52 ± 0.04a0.07 ± 0.01c0.22 ± 0.05b1622.86α-Guaiene3691-12-10.12 ± 0.01c0.02 ± 0.01c0.42 ± 0.02a0.09 ± 0.02c0.25 ± 0.03b1723.55β-Elemene515-13-90.32 ± 0.02c0.53 ± 0.04b1.37 ± 0.28a0.02 ± 0.01d0.02 ± 0.01d1824.73Atractylodin55290-63-615.67 ± 0.46b16.41 ± 1.21b16.26 ± 0.77b16.37 ± 0.81b21.61 ± 1.03a1925.38α-Humulene6753-98-60.83 ± 0.02a0.78 ± 0.09a0.42 ± 0.01b0.83 ± 0.03a0.13 ± 0.02c2026.62α-Selinene473-13-20.78 ± 0.02a0.87 ± 0.11a0.78 ± 0.03a0.78 ± 0.07a0.34 ± 0.03b2128.57β-Maaliene489-29-21.80 ± 0.19c1.02 ± 0.03d2.39 ± 0.19b2.91 ± 0.47a2.56 ± 0.21b2229.25Elixene3242-08-81.04 ± 0.03a0.71 ± 0.01b0.70 ± 0.02b1.01 ± 0.50a0.50 ± 0.03c2332.44Elemol639-99-61.10 ± 0.16b0.50 ± 0.05c1.83 ± 0.09a0.73 ± 0.03c1.56 ± 0.38 a2435.81Eremanthin37936-58-60.46 ± 0.10b0.76 ± 0.02a0.26 ± 0.06c0.46 ± 0.11b0.06 ± 0.01d2536.05Atractylone6989-21-524.28 ± 1.09b24.79 ± 0.84b20.26 ± 0.81c20.21 ± 0.78c27.65 ± 0.93a2636.42Hinesol23811-08-715.02 ± 0.76bc17.68 ± 0.74b14.11 ± 0.52c16.33 ± 0.57bc19.43 ± 0.44a2737.09Diethyl Phthalate84-66-20.76 ± 0.10a0.31 ± 0.06b0.38 ± 0.04b0.75 ± 0.03a0.71 ± 0.08a2837.54β-Eudesmol473-15-413.04 ± 0.59a12.86 ± 0.48a13.02 ± 0.66a14.42 ± 1.09 a14.88 ± 0.62a2937.98α-Eudesmol473-16-50.10 ± 0.03c0.60 ± 0.10b0.90 ± 0.03a0.20 ± 0.02c0.12 ± 0.03c3038.46Azulene53863-54-00.64 ± 0.07b0.14 ± 0.04c0.77 ± 0.02a0.64 ± 0.27b0.39 ± 0.08c3139.78Spathulenol6750-60-30.14 ± 0.02b0.94 ± 0.11a0.89 ± 0.04a0.14 ± 0.01b0.13 ± 0.02b3241.31Dehydrofukinone19598-45-91.89 ± 0.31a0.53 ± 0.02c2.03 ± 0.07a1.84 ± 0.16a1.33 ± 0.27b3341.38Butylidenephthalide551-08-60.52 ± 0.04a0.42 ± 0.01a0.45 ± 0.03a0.77 ± 0.06a0.10 ± 0.02b3441.792-Ethylpropiophenone16819-79-70.41 ± 0.02b0.29 ± 0.01b1.02 ± 0.09a0.40 ± 0.02b0.33 ± 0.03b3543.26Biphenylcarbowaldehyde50670-58-10.73 ± 0.08a0.81 ± 0.04a0.81 ± 0.03a0.34 ± 0.01b0.40 ± 0.07b3645.25Menthofurane494-90-60.75 ± 0.05a0.50 ± 0.02ab0.63 ± 0.06ab0.79 ± 0.03a0.27 ± 0.02b3745.36Cubenol21284-22-00.51 ± 0.05a0.57 ± 0.01a0.50 ± 0.02a0.62 ± 0.05a0.11 ± 0.02b3845.754-Heptylphenol1987-50-40.64 ± 0.08ab0.82 ± 0.03a0.47 ± 0.03ab0.81 ± 0.09a0.10 ± 0.01b3946.19Costunlide553-21-90.50 ± 0.07a0.64 ± 0.03a0.40 ± 0.02ab0.53 ± 0.03a0.31 ± 0.03b4048.95Velleral50656-61-61.30 ± 0.05a0.38 ± 0.01c0.79 ± 0.07b0.69 ± 0.07b0.21 ± 0.10c4156.03Linderene26146-27-00.01 ± 0.01c0.07 ± 0.01c0.78 ± 0.02a0.47 ± 0.04b0.02 ± 0.01cSun drying (S), shade drying (Y), hot air drying (H) microwave drying (W) and freeze drying (D), retention time (RT)Different letters indicate significant differences among different treatments (P < 0.05) Table 3VOCs composition of Atractylodes chinensisNORTItemsCASRelative content (%)SYHWD16.29α-Pinene7785-20-40.25 ± 0.04b0.31 ± 0.03a0.17 ± 0.01c0.28 ± 0.04b0.08 ± 0.10d27.24Geranyl acetate105-87-30.42 ± 0.05c0.72 ± 0.03a0.34 ± 0.03c0.51 ± 0.01b0.07 ± 0.03d37.73β-Pinene127-91-30.63 ± 0.03b0.65 ± 0.02b0.82 ± 0.06a0.74 ± 0.02b0.15 ± 0.04d48.12δ-elemene20307-84-00.50 ± 0.11b0.75 ± 0.04a0.60 ± 0.02b0.33 ± 0.12c0.11 ± 0.08d58.48Isoborneol124-76-50.71 ± 0.08a0.70 ± 0.02a0.36 ± 0.04b0.38 ± 0.11b0.24 ± 0.04c69.89Caryophyllene13877-93-54.70 ± 0.23b4.71 ± 0.37b6.62 ± 0.38a6.04 ± 0.30a2.13 ± 0.20c710.22Myrcene123-35-30.89 ± 0.05a0.86 ± 0.07a0.87 ± 0.29a0.41 ± 0.03b0.14 ± 0.03c810.514-Carene29050-33-70.68 ± 0.04a0.69 ± 0.05a0.57 ± 0.03b0.59 ± 0.03b0.13 ± 0.07c911.38Carvone99-49-00.90 ± 0.07a1.02 ± 0.07a0.71 ± 0.08b0.88 ± 0.04a0.22 ± 0.321013.35Aromadendrene489-39-42.45 ± 0.24a1.48 ± 0.28b2.42 ± 0.26a2.24 ± 0.05a0.27 ± 0.03c1114.37Bornyl acetate76-49-30.71 ± 0.15b0.68 ± 0.07b0.78 ± 0.06b1.03 ± 0.03a0.25 ± 0.05c1215.48γ-Elemene29873-99-22.44 ± 0.31b1.50 ± 0.21c3.04 ± 0.46a2.54 ± 0.73b2.14 ± 0.49b1318.41Limonene138-86-30.73 ± 0.26b0.89 ± 0.09a0.51 ± 0.05bc0.62 ± 0.03b0.11 ± 0.01c1419.05Citronellyl acetate67650-82-20.91 ± 0.30a0.91 ± 0.11a0.11 ± 0.01b0.85 ± 0.02a0.23 ± 0.02b1521.19Cadinene29350-73-00.07 ± 0.01c0.09 ± 0.01c0.52 ± 0.04a0.07 ± 0.01c0.22 ± 0.05b1622.86α-Guaiene3691-12-10.12 ± 0.01c0.02 ± 0.01c0.42 ± 0.02a0.09 ± 0.02c0.25 ± 0.03b1723.55β-Elemene515-13-90.32 ± 0.02c0.53 ± 0.04b1.37 ± 0.28a0.02 ± 0.01d0.02 ± 0.01d1824.73Atractylodin55290-63-615.67 ± 0.46b16.41 ± 1.21b16.26 ± 0.77b16.37 ± 0.81b21.61 ± 1.03a1925.38α-Humulene6753-98-60.83 ± 0.02a0.78 ± 0.09a0.42 ± 0.01b0.83 ± 0.03a0.13 ± 0.02c2026.62α-Selinene473-13-20.78 ± 0.02a0.87 ± 0.11a0.78 ± 0.03a0.78 ± 0.07a0.34 ± 0.03b2128.57β-Maaliene489-29-21.80 ± 0.19c1.02 ± 0.03d2.39 ± 0.19b2.91 ± 0.47a2.56 ± 0.21b2229.25Elixene3242-08-81.04 ± 0.03a0.71 ± 0.01b0.70 ± 0.02b1.01 ± 0.50a0.50 ± 0.03c2332.44Elemol639-99-61.10 ± 0.16b0.50 ± 0.05c1.83 ± 0.09a0.73 ± 0.03c1.56 ± 0.38 a2435.81Eremanthin37936-58-60.46 ± 0.10b0.76 ± 0.02a0.26 ± 0.06c0.46 ± 0.11b0.06 ± 0.01d2536.05Atractylone6989-21-524.28 ± 1.09b24.79 ± 0.84b20.26 ± 0.81c20.21 ± 0.78c27.65 ± 0.93a2636.42Hinesol23811-08-715.02 ± 0.76bc17.68 ± 0.74b14.11 ± 0.52c16.33 ± 0.57bc19.43 ± 0.44a2737.09Diethyl Phthalate84-66-20.76 ± 0.10a0.31 ± 0.06b0.38 ± 0.04b0.75 ± 0.03a0.71 ± 0.08a2837.54β-Eudesmol473-15-413.04 ± 0.59a12.86 ± 0.48a13.02 ± 0.66a14.42 ± 1.09 a14.88 ± 0.62a2937.98α-Eudesmol473-16-50.10 ± 0.03c0.60 ± 0.10b0.90 ± 0.03a0.20 ± 0.02c0.12 ± 0.03c3038.46Azulene53863-54-00.64 ± 0.07b0.14 ± 0.04c0.77 ± 0.02a0.64 ± 0.27b0.39 ± 0.08c3139.78Spathulenol6750-60-30.14 ± 0.02b0.94 ± 0.11a0.89 ± 0.04a0.14 ± 0.01b0.13 ± 0.02b3241.31Dehydrofukinone19598-45-91.89 ± 0.31a0.53 ± 0.02c2.03 ± 0.07a1.84 ± 0.16a1.33 ± 0.27b3341.38Butylidenephthalide551-08-60.52 ± 0.04a0.42 ± 0.01a0.45 ± 0.03a0.77 ± 0.06a0.10 ± 0.02b3441.792-Ethylpropiophenone16819-79-70.41 ± 0.02b0.29 ± 0.01b1.02 ± 0.09a0.40 ± 0.02b0.33 ± 0.03b3543.26Biphenylcarbowaldehyde50670-58-10.73 ± 0.08a0.81 ± 0.04a0.81 ± 0.03a0.34 ± 0.01b0.40 ± 0.07b3645.25Menthofurane494-90-60.75 ± 0.05a0.50 ± 0.02ab0.63 ± 0.06ab0.79 ± 0.03a0.27 ± 0.02b3745.36Cubenol21284-22-00.51 ± 0.05a0.57 ± 0.01a0.50 ± 0.02a0.62 ± 0.05a0.11 ± 0.02b3845.754-Heptylphenol1987-50-40.64 ± 0.08ab0.82 ± 0.03a0.47 ± 0.03ab0.81 ± 0.09a0.10 ± 0.01b3946.19Costunlide553-21-90.50 ± 0.07a0.64 ± 0.03a0.40 ± 0.02ab0.53 ± 0.03a0.31 ± 0.03b4048.95Velleral50656-61-61.30 ± 0.05a0.38 ± 0.01c0.79 ± 0.07b0.69 ± 0.07b0.21 ± 0.10c4156.03Linderene26146-27-00.01 ± 0.01c0.07 ± 0.01c0.78 ± 0.02a0.47 ± 0.04b0.02 ± 0.01cSun drying (S), shade drying (Y), hot air drying (H) microwave drying (W) and freeze drying (D), retention time (RT)Different letters indicate significant differences among different treatments (P < 0.05) VOCs composition of Atractylodes chinensis Sun drying (S), shade drying (Y), hot air drying (H) microwave drying (W) and freeze drying (D), retention time (RT) Different letters indicate significant differences among different treatments (P < 0.05) Fig. 2Ring graph of VOCs classification (A). OPLS-DA score plot of VOCs (B). Sun drying (S), shade drying (Y), hot air drying (H) microwave drying (W) and freeze drying (D). OPLS-DA model evaluation parameters: R2X = 0.507, R2Y = 0.970, Q2 = 0.887 Ring graph of VOCs classification (A). OPLS-DA score plot of VOCs (B). Sun drying (S), shade drying (Y), hot air drying (H) microwave drying (W) and freeze drying (D). OPLS-DA model evaluation parameters: R2X = 0.507, R2Y = 0.970, Q2 = 0.887 Differential VOCs of Atractylodes chinensis The composition of VOCs in A. chinensis treated with different drying methods was significantly different. K-means analysis was conducted for substances in each group to reveal their differences (Fig. 3). The 41 substances in each group were divided into 6 different profiles, among which the content of substances in profile 1 (mainly caryophyllene, β-elemene, and spathulenol) was higher in group Y. The substance content in profile 2 (mainly α-pinene, carvone, and bornyl acetate) was higher in group W. The substance content in profile 3 (mainly atractylodin, atractylone, and hinesol) was higher in group D. The contents of substances in profile 4 were higher in groups Y and D, and were mainly β-pinene, citronellyl acetate, and α-guaiene. The contents of substances in profile 5 were higher in groups Y and D, and were mainly geranyl acetate, γ-elemene, and elemol. The content of substances in profile 6 was higher in groups S and D, and were mainly limonene, elixene, and diethyl phthalate. Fig. 3K-means cluster analysis of VOCs. The x-axis shows the different samples and y-axis shows the standardized amount of each metabolite. Sun drying (S), shade drying (Y), hot air drying (H) microwave drying (W) and freeze drying (D) K-means cluster analysis of VOCs. The x-axis shows the different samples and y-axis shows the standardized amount of each metabolite. Sun drying (S), shade drying (Y), hot air drying (H) microwave drying (W) and freeze drying (D) The box plot shows the 14 main VOCs that differ between different groups (Fig. 4). Among them, the contents of β-elemene, β-maaliene, and menthofurane in group S were higher. The contents of β-elemene, β-maaliene, 4-heptylpheno, velleral, and linderene in group Y were higher. The contents of elemol and ethylpropiophenone were higher in group H. The contents of α-pinene, bornyl acetate, menthofurane, and 4-heptylpheno in group W were higher, while the contents of atractylodin, atractylone, hinesol, and β-eudesmol were higher in group D. Fig. 4Box plots of 14 different VOCs, with the upper and lower quartiles representing the upper and lower limits of the data, the upper and lower quartiles being the values at 75% and 25% of the points in the dataset, and the data points outside the lower quartiles being considered the lower outliers. Sun drying (S), shade drying (Y), hot air drying (H) microwave drying (W) and freeze drying (D) Box plots of 14 different VOCs, with the upper and lower quartiles representing the upper and lower limits of the data, the upper and lower quartiles being the values at 75% and 25% of the points in the dataset, and the data points outside the lower quartiles being considered the lower outliers. Sun drying (S), shade drying (Y), hot air drying (H) microwave drying (W) and freeze drying (D) Functional annotation and enrichment of VOCs The KEGG annotation results of VOCs are shown in Fig. 5. The annotated pathways are all associated with metabolism. The biosynthesis of secondary metabolites (6), biosynthesis of terpenoids and steroids (3), monoterpenoid biosynthesis (3), and sesquiterpenoid and triterpenoid biosynthesis (3) were highly enriched. Fig. 5KEGG classification of VOCs KEGG classification of VOCs Correlation of odor response intensity with differential VOCs The correlations between the odor response intensity of A. chinensis and VOCs are shown in Fig. 6. W5S was strongly significantly positively correlated with β-maaliene, atractylodin, atractylone, hinesol and β-eudesmol (P < 0.01), and significantly positively correlated with velleral (P < 0.05). W1W was strongly positively correlated with atractylodin, atractylone, hinesol and β-eudesmol (P < 0.01), and significantly positively correlated with β-maaliene and velleral (P < 0.05). W2W was strongly positively correlated with atractylodin, atractylone and hinesol (P < 0.01), and significantly positively correlated with β-maaliene, β-eudesmol, and velleral (P < 0.05). Fig. 6Correlation between odor response intensity and differential VOCs. Red represents positive correlation, blue represents negative correlation, * represents P < 0.05, ** represents P < 0.01, *** represents P < 0.001 Correlation between odor response intensity and differential VOCs. Red represents positive correlation, blue represents negative correlation, * represents P < 0.05, ** represents P < 0.01, *** represents P < 0.001 Discussion The analysis of functional components of Atractylodes chinensis showed that drying method had a significant influence on the quality of A. chinensis. The freeze drying process led to higher functional component content because dehydration under low temperature conditions avoids the degradation or inactivation of components that may occur under high temperature [19]. At the same time, freeze drying is carried out in a vacuum environment, which can effectively remove water molecules while maximizing the retention of volatile components [20]. The ice crystals formed during the freeze drying process can play a protective role to prevent the functional components from being damaged or oxidized during the dehydration process, and the ice crystals sublimate to form water vapor and quickly discharge, avoiding the infiltration of water and associated impacts on the functional components [21]. The content of functional components of A. chinensis after the hot air drying and microwave drying treatments were low, because the high temperature easily causes the loss or degradation of volatile components in A. chinensis [22]. At the same time, A. chinensis exposed to air is prone to oxidation reactions, resulting in the inactivation or deterioration of some functional components [23]. In the process of water evaporation, some water-soluble nutrients may be removed, resulting in a decrease in the overall functional component content, which is consistent with the results of Oliveira et al. [24], who found that drying increases the loss of water-soluble vitamins in dried vegetables. The use of slow drying methods, such as sun drying and shade drying, can better retain the functional substances in A. chinensis, because it is usually carried out under natural conditions and low temperature, and thus can avoid the damage caused by high temperature to functional components [25]. Slow drying also allows the cell structure and components in A. chinensis to remain relatively intact, avoiding the shrinkage or deformation caused by rapid drying. According to Bettaieb Rebey et al. [26], the contents of total phenols and flavonoids in fennel were highest when it was dried in the shade and lowest when it was dried in hot air, which is consistent with the results of this study. The analysis of the odor response intensity of A. chinensis showed that the odor characteristics of different drying methods were significantly different. There were significant differences in the response intensity and in terpenoids, inorganic sulfides, organic sulfides, nitrogen oxides, and methyl compounds in A. chinensis after different treatments. Terpenes in A. chinensis, such as atractylone and atractylolide, are important functional components. Wu et al. [27] used an electronic nose combined with GC-MS to determine that terpenes, such as caryophyllene and α-pinene, are characteristic odor volatiles of chrysanthemum. These compounds are usually highly volatile and prone to volatilization losses during drying. Different drying methods will affect its volatilization rate, and high temperature drying will accelerate the volatilization of odor substances, resulting in reduced response strength, while low temperature sun or shade drying can better retain them [28]. The duration of drying will also affect functional content. A slow drying process can allow odor components to escape slowly, while rapid drying may lead to a large loss of these components [29]. The results of OPLS-DA showed that the response of the electronic nose to odor characteristics could effectively distinguish A. chinensis dried using different methods. In particular, the odor characteristics of shade dried, hot air dried, and freeze dried A. chinensis were significantly different, which provided information to support a new classification and quality characterization of A. chinensis products. Similarly [30], used seven algorithms with electronic noses to classify 12 types of herbs, among which SVM and LDA showed higher performance and accuracy. The analysis of VOCs showed that terpenoids, esters, ketones, and phenols were the main substances in A. chinensis. Terpenoids are important secondary metabolites of plants and usually have a strong aroma. Depending on their specific structure, they can emit various aroma types, such as floral, citrusy, loose, or herbal. The smell characteristics of esters are generally sweet, fruity, and floral, often pleasing, and they play an important role in pollination and fruit spread of plants. The smell of ketones and phenols is aromatic, spicy, or bitter [31, 32]. The results of GC-MS analysis showed that atractylodin, atractylone, hinesol, β-eudesmol, and other volatile compounds were important functional components of A. chinensis, and these substances together formed the odor characteristics of A. chinensis. Among them, atractylodin has a slightly sweet and woody aroma, and has anti-inflammatory, antibacterial, and antioxidant properties [33]. The smell of hinesol can be described as slightly sweet, spicy, and bitter, and it has anti-inflammatory, immunomodulatory, and has antitumor effects. β-eudesmol is characterized by a warm, woody, and herbaceous aroma with anti-inflammatory and antioxidant effects [34]. Xiang et al. [35] found that the types of VOCs of ginseng dried using different drying methods were basically the same, but the VOC content was significantly different, which was similar to the results of this study. The results of OPLS-DA analysis of VOCs were similar to those of the electronic nose, indicating that the odor response intensity was in good agreement with the composition of VOCs. This suggests that that the A. chinensis samples dried using different methods could be distinguished by odor analysis. The results of K-means and box diagram analysis showed that the contents of β-elemene, β-maaliene, and menthofurane were higher in group S. Among them, the odor of β-elemene is a spicy fennel odor. Lu et al. [36] studied Lycium barbarum L. and determined that β-elemene is a characteristic aroma volatile, and its content first increases and then decreases during fruit development. Adam et al. [37] showed that β-maaliene and other sesquiterpenoids provided a woody aroma for Aquilaria malaccensis. Menthofurane has a distinct minty aroma, which is fresh and cool [38]. The contents of β-elemene, β-maaliene, 4-heptylphenol, velleral, and linderene in group Y were higher. Sanahuja et al. [39] found in celery that the content of 4-heptylphenol and other phenols in the stem was higher than that in the root, which had medicinal, spicy, and smoky odors. Velleral is often found in medicinal plants and has antibacterial activity and a bitter taste [40]. Linderene has a sour, musky, woody odor and also has antibacterial and antioxidant activity [41]. The contents of elemol and ethylpropiophenone were higher in group H. Elemol has the odor of camphor, turpentine, and pine, which Letchamo et al. [42] found in plants such as knotgrass, pepper, and agarwood. The contents of α-pinene, bornyl acetate, menthofurane, and 4-heptylphenol in group W were higher. Among them, α-pinene and bornyl acetate have the scent of pine and camphor [43]. The contents of atractylodin, atractylone, hinesol, and β-eudesmol were higher in group D. Atractylodin and atractylone have a punchy, bitter taste, hinesol, and β-eudesmol has a woody odor, and these high content substances together make up the odor of A. chinensis. The results of KEGG annotation and enrichment showed that VOC composition is mainly related to the biosynthesis of secondary metabolites, the biosynthesis of terpenoids and steroids, monoterpenoid biosynthesis, and sesquiterpenoid and triterpenoid biosynthesis metabolic pathways. Secondary metabolites are usually related to a plant’s defense mechanism, attracting pollinators, inhibiting competitors, and other functions. In A. chinensis, the synthesis of VOCs is affected by the secondary metabolic synthesis pathway. Plants typically synthesize compounds such as phenylalanine and tyrosine through metabolic pathways of aromatic amino acids, which are further converted into volatile organic compounds [44]. The activity and expression of key enzymes in the secondary metabolic synthesis pathway will also directly affect the synthesis of VOCs. When the expression of these enzymes increases, the synthesis of related VOCs also increases [45]. Abbas et al. [46] suggested that the substances produced by the secondary metabolic synthesis pathway are related to the formation of the fragrance of many plants. Monoterpenes are usually synthesized from precursors such as isoprene pyrophosphate (IPP) and phenylalanine through multiple reactions in plants, and monoterpene synthase converts these precursors into monoterpenes through a catalytic process [47]. Sesquiterpenoids are mainly derived from monoterpenoids, and triterpenoid synthesis usually involves more complex enzyme-catalyzed reactions to produce a variety of biologically active compounds [48]. In A. chinensis, terpenoid volatile compounds, such as atractylone and menthol, have specific aromatic odors and biological activity, which affect plant odor and pollinator attraction and may also participate in plant defense mechanisms. Sesquiterpenoids, such as terpinol, also play important roles in plant growth, development, and defense. These compounds not only affect the smell of A. chinensis but also have a profound impact on its ecological niche and medicinal value, which is similar to the findings of Jiang et al. [49] for chrysanthemum. Correlation analysis showed that the intensity of W5S, W1W, and W2W sensors were significantly correlated with atractylodin, atractylone, β-eudesmol, and other main volatile components of A. chinensis. Further verification of these correlation coefficients shows that the odor response intensity of the electronic nose can effectively distinguish the odor of A. chinensis dried using different methods and can be used for rapid determination and differentiation. Xu et al. [50] found that by analyzing the correlation between aroma and VOCs, an electronic nose could predict the levels of α-pinene oxide, α-maaliene and other terpenoids in different celery seed oils. In the drying process for ginger, Pei et al. [51] found a strong correlation between terpene compounds and the response values of W1W and W2W sensors. In this study, the strong correlation between the electronic nose and GC-MS odor analysis results provided a new insight to inform quality characterization and differentiation of A. chinensis. Conclusion In this study, the effects of different drying methods on Atractylodes chinensis functional components, odor response intensity, and VOC composition were investigated using an electronic nose and GC-MS. Results showed that the content of functional components was higher in groups D and Y, and lower in group H. The electronic nose had the highest response values in the W1W (terpenes), W2W (organic sulfides), and W5S (nitrogen oxides) channels. The odor response intensity of groups D and Y was higher, while group H was lower, and odor response intensity had a strong positive correlation with VOC composition. The VOCs of A. chinensis were mainly terpenoids, esters, phenols, and ketones. The contents of differential VOCs in group D were higher, and were mainly atractylodin, atractylone, bornyl acetate, and other substances with woody, camphor, and bitter and hot odors. The biosynthesis of secondary metabolites and the biosynthesis of terpenoids and steroids were the main metabolic pathways. Overall, hot air drying lost more of these functional components and odors, while more were retained through freeze drying.