Comparative Characterization of Volatile Flavor Profiles and Key Markers in Early‐, Mid‐, and Late‐Ripening Honey Peach (Prunus persica L. Batsch) Cultivars Using GC‐IMS and HS‐GC‐MS

ABSTRACT

This study evaluated the nonvolatile and volatile flavor profiles of six honey peach (Prunus persica L. Batsch) cultivars across early‐ (‘Ri Chuan’ and ‘Jing Hong’), mid‐ (‘Bai Feng’ and ‘Xia Hui’), and late‐ripening (‘Zhong Hu Jing’and‘Wan Hu Jing’) stages at comparable maturity. Late‐ripening cultivars exhibited higher fruit weight, soluble solids content, and total sugars (95.34 vs72.01 mg/g in early cultivars). Although total acidity showed no significant group‐level differences, organic acid compositions varied markedly: quinic and citric acids predominated in early cultivars, whereas malic acid was higher in mid‐ and late‐ripening groups. Sucrose was the predominant sugar, while glucose, fructose, and sorbitol drove cultivar variation. GC‐IMS and HS‐GC‐MS identified 51 and 71 volatile organic compounds (VOCs), respectively. Early‐ripening cultivars were characterized by green/fatty aldehydes and ketones; mid‐ripening cultivars by green‐note alcohols and fruity esters; and late‐ripening cultivars by terpenoids, lactones, and esters imparting floral, fruity, creamy, and ripe‐peach aromas. Integrating VIP values, univariate testing, and odor‐activity evaluation identified 12 key differential aroma‐active markers (e.g., ethyl isobutyrate, hexyl acetate, nonanal, and γ‐decalactone). Overall, peach flavor differentiation is governed by coordinated shifts in sugar–acid and volatile compositions, highlighting the complementary power of dual‐platform profiling.

1. Introduction

Honey peach (Prunus persica L. Batsch) is highly appreciated by consumers for its tender flesh, rich nutrition, and attractive aroma. Aroma is a core attribute determining fruit flavor quality and market acceptance, arising from the combined perception of a complex mixture of volatile organic compounds (VOCs). In peach fruit, VOC composition changes dynamically during fruit development, ripening, and postharvest handling, driven by genotype, maturity stage, and environmental factors such as temperature (Liu et al. 2025; Lombardo et al. 2011). Several compound classes are repeatedly associated with characteristic peach aroma. Aldehydes and alcohols contribute green, fatty, and fresh notes; hexyl acetate and related esters contribute fruity notes; γ‐ and δ‐lactones contribute characteristic peach‐like and creamy notes; and linalool imparts floral notes (Li et al. 2023. Biosynthetic studies have further elucidated the plausible biochemical contexts for these compounds. For example, acyl‐CoA oxidase activity has been associated with lactone production, the enzyme encoded by PpAAT1 catalyzes reactions related to ester and γ‐decalactone formation, and PpbHLH1 regulates linalool biosynthesis in peach (Xi et al. 2012; Peng et al. 2020; Wei et al. 2021). Therefore, elucidating the intrinsic relationship between fruit ripening and volatile metabolic profiles is essential for understanding the formation and divergence of peach flavor.

In addition to developmental stage, cultivar genetic background and ripening type are key factors shaping the volatile metabolic profile of peach fruit. A recent analysis of 114 peach germplasms identified 41 VOCs belonging to nine chemical categories, including aldehydes, esters, lactones, alcohols, terpenols, ketones, phenols, alkanes, and ethers, demonstrating that germplasms harvested in different seasons could be clearly distinguished by principal component analysis (Su et al. 2025). Another study involving 42 commercial peach cultivars demonstrated pronounced inter‐cultivar differences in volatile composition, in which esters, lactones, alcohols, aldehydes, and terpenoids jointly constituted the main aroma‐related chemical classes, and γ‐decalactone, δ‐decalactone, benzaldehyde, and linalool were recognized as important contributors to typical peach aroma (Mohammed et al. 2021). A recent review further summarized that the key aroma‐active compounds in peach include lactones, esters, terpenoids, and aldehydes, which are markedly affected by variety, maturity, and storage/handling temperature (Liu et al. 2025). However, studies based solely on the developmental sequence of a single cultivar or on postharvest ripening cannot fully separate the effects of developmental progression from those of cultivar or ripening type, thereby limiting our understanding of how aroma‐style differences form among early‐, mid‐, and late‐ripening peach cultivars at comparable physiological maturity.

On the analytical front, accurate characterization of volatile compounds depends strongly on advances in analytical techniques. In recent years, fruit flavor research has moved from single‐platform analysis toward multidimensional and complementary identification strategies. Headspace gas chromatography‐ion mobility spectrometry (GC‐IMS) offers rapid analysis, simple sample preparation, high sensitivity, and intuitive fingerprint visualization, making it suitable for capturing the overall volatile profile. In contrast, headspace gas chromatography‐mass spectrometry (HS‐GC‐MS), particularly when combined with headspace solid‐phase microextraction (HS‐SPME), provides stronger compound identification and quantitative capability, and can be further coupled with odor activity value (OAV) or relative odor activity value (ROAV) analysis to evaluate the aroma contribution of key volatiles. The combined use of GC‐IMS and HS‐GC‐MS has been successfully applied to peach fruit, allowing comprehensive profiling of volatile constituents and screening of main aroma contributors (Sun et al. 2022). Therefore, integrating the global fingerprinting advantage of GC‐IMS with the compound‐level resolution of HS‐GC‐MS provides a multilayered analytical framework for peach flavor evaluation.

Based on these considerations, the present study therefore compared six honey peach cultivars representing three ripening categories at comparable cultivar‐specific maturity. The objectives were to (1) characterize selected nonvolatile quality traits (sugars, organic acids, and phenolics); (2) compare volatile fingerprints obtained by GC‐IMS and HS‐GC‐MS; and (3) identify compounds with potential sensory relevance using OAV and multivariate statistical analysis. Cultivar was treated as the primary biological unit, and ripening categories were used only for descriptive summaries within the sampled cultivars.

2. Materials and Methods

2.1. Plant Materials and Sample Preparation

Six representative commercial honey peach (Prunus persica L. Batsch) cultivars with distinct ripening characteristics were evaluated in this study, comprising two early‐ripening cultivars (‘Ri Chuan’ and ‘Jing Hong’), two mid‐ripening cultivars (‘Bai Feng’ and ‘Xia Hui’), and two late‐ripening cultivars (‘Zhong Hu Jing’and ‘Wan Hu Jing’). To eliminate geographic location and orchard management confounders, all fruits were harvested during the 2025 growing season from the same representative commercial orchard in Changzhou, Jiangsu Province, China. Fruits of early‐, mid‐, and late‐ripening cultivars were harvested at their respective peak commercial maturity stages on June 16, July 2, and July 25, 2025, respectively.

Harvest selection was strictly standardized based on objective commercial quality indices, including cultivar‐specific harvest windows, fruit background color, soluble solids content (SSC), fruit weight, and uniform visual appearance, ensuring the exclusion of fruits with disease, pest damage, or mechanical injury. Samples were explicitly defined as possessing comparable cultivar‐specific commercial maturity.

For each cultivar, three independent biological replicates were prepared, with each replicate consisting of six representative fruits. Fruits were pitted, sliced into small pieces, immediately flash‐frozen in liquid nitrogen, and stored at −80°C until further analysis. Prior to analytical determinations, frozen fruit tissues were homogenized into a fine powder under liquid nitrogen using a cryogenic grinder. The prepared powder was utilized for the analysis of organic acids, soluble sugars, volatile flavor compounds, and physicochemical attributes.

2.2. Determination of Physicochemical and Nutritional Quality Attributes

Fruit weight was measured using an electronic balance (JA2003, Shanghai Sunny Hengping Scientific Instrument Co., Ltd., Shanghai, China). SSC was determined using a handheld digital refractometer (PAL‐1, Atago Co., Ltd., Tokyo, Japan). Moisture content was determined using a modified gravimetric drying method as described by Yuan et al. (2022). Total sugar content was measured using the anthrone colorimetric method with some modifications according to Yuan et al. (2022). Total acid content was evaluated by potentiometric titration, and vitamin C content was determined using the 2,6‐dichloroindophenol titration method as described by Liu et al. (2025). Total phenolic content was analyzed using the Folin‐Ciocalteu method and expressed as mg gallic acid equivalents (GAE) per 100 g fresh weight (Raposo et al. 2024).

2.3. Determination of Organic Acids

Organic acid components were determined using an Agilent 1260 liquid chromatography system (Agilent Technologies, USA) (Jiang et al. 2023). The analytical method was modified from Zheng et al. (2023). Quantification was performed using the external standard method. Briefly, 1.0 g of powdered sample was extracted with 5 mL of 20 mmol L⁻¹ sodium dihydrogen phosphate solution at pH 2.0. The mixture was ultrasonically extracted for 10 min and then centrifuged at 6100 × g for 8 min at 4°C. The supernatant was collected and filtered through a 0.22 µm polyethersulfone (PES) aqueous membrane before HPLC analysis.

Chromatographic separation was performed on an Agilent ZORBAX SB‐C18 column (4.6 mm × 250 mm, 5.0 µm). The column temperature was maintained at 35°C, and the flow rate was 0.3 mL min^{− 1}. Detection was performed at 210 nm using a UV detector. The mobile phases consisted of methanol as solvent A and 20 mmol L^{− 1} sodium dihydrogen phosphate solution at pH 2.0 as solvent B. The gradient elution program was as follows: 0–33 min, 100% B; 33–34 min, 100%–10% B; 34–37 min, 10% B; 37–38 min, 10%–100% B; and 38–53 min, 100% B. Organic acid concentrations were calculated by substituting the peak areas into the corresponding standard curves.

2.4. Determination of Soluble Sugars

Soluble sugar profiles were determined using an ultra‐performance liquid chromatography‐tandem mass spectrometry system (UPLC‐MS/MS; SCIEX Triple Quad 4500, SCIEX, USA), with modifications based on the method described by Gao et al. (2024). Quantification was performed using the external standard method. Briefly, 0.1 g of powdered sample was diluted to 50 mL with ultrapure water, ultrasonically extracted, and centrifuged. A 1 mL aliquot of the supernatant was then vortexed with 1 mL of acetonitrile. To fit within the linear range of the instrument, 20 µL of this solution was further diluted with 1.98 mL of 50% (v/v) aqueous acetonitrile. The mixture was passed through a 0.22 µm nylon membrane syringe filter before LC‐MS/MS analysis.

Chromatographic separation was performed on an ACQUITY UPLC BEH Amide column (2.1 mm × 100 mm, 1.7 µm). The column temperature was maintained at 30°C, the flow rate was 0.4 mL min^{− 1}, and the injection volume was 5 µL. The mobile phases consisted of 0.1% ammonia solution as solvent A and acetonitrile as solvent B. The gradient elution program was as follows: 0–14 min, 90%–70% B; 14–15 min, 70%–50% B; 15–16 min, 50% B; 16.0–16.1 min, 50%–90% B; and 16.1–18.0 min, 90% B. Soluble sugar concentrations were calculated using the corresponding standard curves.

Mass spectrometric detection was performed in negative electrospray ionization mode using multiple reaction monitoring. The spray voltage was set at −4500 V, the source temperature was 550°C, the curtain gas was 30 psi, and both nebulizer gas and auxiliary gas were set at 55 psi. The target compounds together with their qualitative ion pairs, quantitative ion pairs, declustering potential (DP), and collision energy (CE) are listed in Table S1.

2.5. GC‐IMS Analysis of Volatile Compounds in Honey Peaches

Volatile compounds were analyzed using headspace GC‐IMS (FlavourSpec 8100, G.A.S., Germany) according to the method described by Shi et al. () with minor modifications. 5.0 g of peach powder was mixed with 2.0 mL of saturated sodium chloride solution in a 20 mL headspace vial. The sample was incubated at 45°C for 15 min with agitation at 500 r/min. The syringe temperature was maintained at 85°C, and 500 µL of headspace gas was injected into the instrument. Volatile compounds were separated on an MXT‐5 capillary column (15 m × 0.53 mm, 1 µm film thickness) maintained at 60°C. High‐purity nitrogen (≥99.999%) was used as the carrier gas. The carrier‐gas flow rate was maintained at 2 mL/min for the first 2 min and then increased linearly to 100 mL min⁻¹ from 2 to 15 min. The total analysis time was 15 min.

High‐purity nitrogen (≥99.999%) was also used as the drift gas at a flow rate of 150 mL/min. The IMS detector temperature was maintained at 45°C. The drift tube was 9.8 cm in length and operated at an electric field strength of 500 V/cm. GC‐IMS data were processed using Laboratory Analytical Viewer (LAV) software supplied with the instrument. Volatile compounds were tentatively identified by comparing their retention indices and ion‐mobility drift times with reference data in the NIST and IMS libraries incorporated into the GC × IMS Library Search software. Difference plots and fingerprint plots were generated to visualize variations in VOCs among the peach cultivars.

2.6. HS‐GC‐MS Analysis of Volatile Compounds

Volatile compounds were further analyzed using headspace solid‐phase microextraction coupled with gas chromatography‐mass spectrometry (HS‐GC‐MS) according to the method of Wang et al. (2026), with appropriate modifications. Briefly, 5.0 g of ground peach flesh was placed into a 20 mL headspace vial containing a magnetic stir bar. Then, 3 mL of saturated sodium chloride solution and 30 µL of 3‐nonanone (0.2463 mg/L) as the internal standard were added. The vial was equilibrated at 40°C for 30 min under magnetic stirring at 600 r/min. Volatile compounds were then extracted using a 65 µm PDMS/DVB fiber for 30 min prior to HS‐GC‐MS analysis.

GC separation was performed on an HP‐5MS capillary column (60 m × 0.25 mm, 0.25 µm). The inlet temperature was maintained at 250°C, and the carrier gas flow rate was 1.0 mL min^{− 1}. Samples were injected in splitless mode. The oven temperature program was as follows: the initial temperature was held at 35°C for 2 min, increased to 220°C at 4°C min^{− 1} and held for 2 min, and then increased to 245°C at 15°C min^{− 1}.

For MS analysis, electron ionization was used. The transfer line, ion source, and quadrupole temperatures were set at 280°C, 250°C, and 150°C, respectively. Volatile compounds were identified by comparing their mass spectra with those in the NIST17 mass spectral database and authentic standards, together with retention index information. Quantification was performed using 3‐nonanone as the internal standard.

2.7. Statistical Analysis and Aroma Contribution Analysis

To evaluate the contribution of individual volatile compounds to overall peach aroma, OAV, and ROAV calculations were performed separately for HS‐GC‐MS and GC‐IMS datasets (Lou et al. 2026; Wang et al. 2026). For HS‐GC‐MS quantitative data, OAV was calculated as:

$$\mathrm{OA}{\mathrm{V}}_i=C_i/T_i$$

where C_i is the semi‐quantitative concentration (ng/g) and T_i is the published odor threshold of compound i in water. Concentrations and thresholds were standardized to matching units prior to calculation. For GC‐IMS data, as signal intensities were relative, ROAV was calculated as:

$$\mathrm{ROA}{\mathrm{V}}_i=\left[\left(I_i/T_i\right)/\left(I_k/T_k\right)\right]\times 100$$

where I_i and T_i represent the relative peak intensity and odor threshold of compound i, respectively, and k represents the compound contributing maximum I/T value in a given sample (ROAV _k = 100).

Data preprocessing and primary calculations were performed using Microsoft Excel 2021 (Microsoft Corp., Redmond, WA, USA). Multivariate statistical analyses, including principal component analysis (PCA) and partial least squares–discriminant analysis (PLS‐DA), were performed using MetaboAnalyst 5.0. Differences among cultivars were evaluated by one‐way analysis of variance (ANOVA) followed by Tukey's post‐hoc test (p < 0.05) in IBM SPSS Statistics 26.0 (IBM Corp., Armonk, NY, USA). Pearson correlation analysis was conducted to examine pairwise variable associations. Correlation heatmaps and expression networks were generated using ChiPlot and GraphPad Prism 10 (GraphPad Software, Boston, MA, USA).

3. Results

3.1. Physicochemical and Nutritional Quality Attributes of Six Honey Peach Cultivars

The physicochemical and nutritional quality attributes varied significantly across the six honey peach cultivars (Table 1). Fruit weight showed the most pronounced cultivar‐dependent variation, ranging from 216.74 ± 8.48 g to 282.83 ± 3.07 g. The two early‐ripening cultivars, ‘Ri Chuan’and ‘Jing Hong’exhibited the lowest fruit weights (216.74 ± 8.48 g and 223.93 ± 2.85 g, respectively), which were significantly lower than those of the mid‐ and late‐ripening cultivars (p < 0.05). ‘Bai Feng’and ‘Xia Hui’displayed intermediate fruit weights of 260.62 ± 7.26 g and 266.94 ± 2.89 g, respectively, whereas ‘Zhong Hu Jing’and ‘Wan Hu Jing’reached 276.63 ± 4.42 and 282.83 ± 3.07 g, respectively. The highest value was observed in ‘Wan Hu Jing ’which was 30.49% higher than that of ‘Ri Chuan’ Thus, among the six cultivars examined, fruit weight showed a gradual increase from the early‐ to the late‐ripening groups.

TABLE 1.

Basic nutritional quality parameters of different honey peach cultivars.

Cultivar group	Cultivar name	Fruit weight /g	Soluble solids /%	Moisture content /%	Vitamin C content mg/100g	Sugar content mg/g
Early‐ripening cultivar	'Ri Chuan''	216.74 ± 8.48d	9.78 ± 0.63b	89.09 ± 2.83b	6.43 ± 1.04a	70.72 ± 7.95c
	'Jing Hong'	223.93 ± 2.85d	9.96 ± 0.73b	88.65 ± 1.79ab	6.77 ± 0.49a	73.29 ± 6.20bc
Mid‐ripening cultivar	'Bai Feng'	260.62 ± 7.26c	12.49 ± 0.76a	88.56 ± 0.42ab	6.36 ± 1.15a	77.52 ± 0.54abc
	'Xia Hui'	266.94 ± 2.89bc	11.56 ± 0.94ab	91.35 ± 0.86a	7.15 ± 0.27a	75.31 ± 4.37abc
Late‐ripening cultivar	'Zhong Hu Jing'	276.63 ± 4.42ab	13.03 ± 0.61a	91.57 ± 0.92ab	6.52 ± 0.22a	94.41 ± 10.99ab
	'Wan Hu Jing'	282.83 ± 3.07a	12.78 ± 1.50a	91.64 ± 1.06ab	6.88 ± 0.42a	96.27 ± 10.97a

Note: Values are expressed as mean ± standard deviation (n = 3). Different lowercase letters indicate significant differences among cultivars (p < 0.05). Means sharing at least one common letter are not significantly different, whereas means with no letters in common differ significantly according to Tukey's multiple comparison test; Early‐ ripening, mid‐ ripening, and late‐ ripening cultivars were harvested on June 16, July 2, and July 25, 2025, respectively.

A similar cultivar‐dependent pattern was observed for SSC, which ranged from 9.78% ± 0.63% to 13.03% ± 0.61%. The early‐ripening cultivars ‘Ri Chuan’and ‘Jing Hong’showed relatively low SSC values of 9.78% ± 0.63% and 9.96% ± 0.73%, respectively. In contrast, ‘Bai Feng’ ‘Zhong Hu Jing’and ‘Wan Hu Jing’exhibited elevated values of 12.49% ± 0.76%, 13.03% ± 0.61%, and 12.78% ± 1.50%, respectively. ‘Xia Hui’ showed an intermediate SSC level of 11.56% ± 0.94% and did not differ significantly from cultivars sharing the same significance letters. ‘Zhong Hu Jing’ showed the highest SSC, representing an increase of 3.25 percentage points, or approximately a 33.23% increase, relative to ‘Ri Chuan’These results indicate that several mid‐ and late‐ripening cultivars had greater soluble‐solids accumulation than the two early‐ripening cultivars.

Moisture content varied within a comparatively narrow range of 88.56% ± 0.42% to 91.64% ± 1.06%. ‘Bai Feng’semi‐ had the lowest value, whereas ‘Wan Hu Jing’had the highest. Among the cultivars, ‘Ri Chuan’ (89.09% ± 2.83%) differed significantly from ‘Xia Hui’ (91.35% ± 0.86%), while most other pairwise comparisons were not significant (p > 0.05). Therefore, in contrast to fruit weight and SSC, moisture content showed relatively limited cultivar‐to‐cultivar variation.

Vitamin C content was also stable, ranging from 6.36 ± 1.15 to 7.15 ± 0.27 mg/100 g FW. ‘Xia Hui’showed the highest numerical value and ‘Bai Feng’ the lowest; however, all six cultivars shared the same significance letter, indicating no significant differences in vitamin C content were observed among cultivars (p > 0.05). Thus, the present data do not support a distinct ripening‐stage‐associated pattern for vitamin C accumulation.

Total sugar content showed greater variation among cultivars, ranging from 70.72 ± 7.95 to 96.27 ± 10.97 mg/g FW. The early‐ripening cultivars ‘Ri Chuan’ and ‘Jing Hong’ contained 70.72 ± 7.95 and 73.29 ± 6.20 mg/g, respectively, whereas the mid‐ripening cultivars ‘Bai Feng’ and ‘Xia Hui’ contained 77.52 ± 0.54 and 75.31 ± 4.37 mg/g, respectively. The two late‐ripening cultivars exhibited the highest values, reaching 94.41 ± 10.99 mg/g in ‘Zhong Hu Jing’ and 96.27 ± 10.97 mg/g in ‘Wan Hu Jing’. ‘Wan Hu Jing’ had significantly higher sugar content than ‘Ri Chuan’ and ‘Jing Hong’ whereas ‘Zhong Hu Jing’ was significantly higher than ‘Ri Chuan’ (p < 0.05). The difference between ‘Wan Hu Jing’ and ‘Ri Chuan’ was 25.55 mg/g, corresponding to an increase of 36.13%.

When the cultivars within each ripening category were summarized descriptively, the early‐ripening cultivars had mean fruit weight, SSC, and sugar content values of 220.34 g, 9.87%, and 72.01 mg/g, respectively, whereas the corresponding values for the late‐ripening cultivars were approximately 279.73 g, 12.91%, and 95.34 mg/g. These values were 26.96%, 30.75%, and 32.41% higher, respectively, in the late‐ripening group. In comparison, moisture content changed only from 88.87% to 91.61%, and vitamin C content remained similar at 6.60 and 6.70 mg/100 g. Collectively, these results show that, among the six cultivars evaluated, the late‐ripening cultivars were characterized by greater fruit weight, SSC, and sugar content, whereas moisture and vitamin C contents were comparatively similar among cultivars.

3.2. Physicochemical Characteristics and Nonvolatile Flavor Components of Six Honey Peach Cultivars With Different Ripening Periods

Representative appearances of the six honey peach cultivars are shown in Figure 1A. Noticeable differences in fruit color and external morphology were observed among the early‐, mid‐, and late‐ripening cultivars, indicating considerable phenotypic variation among the cultivars examined. To further characterize their basic flavor and nutritional properties, total acidity, total phenolic content, individual organic acids, and soluble sugars were evaluated.

FIGURE 1.

Basic nutritional quality attributes of different fresh peach cultivars. (A) Representative photographs of peaches from different cultivars. Early‐ripening cultivars: ' Ri Chuan' and ' Jing Hong'; mid‐ripening cultivars: ' Bai Feng' and ' Xia Hui'; late‐ripening cultivars: ' Wan Hu Jing' and ' Zhong Hu Jing.' Scale bar = 1 cm; (B) Total acid content among different peach cultivars; (C) Total phenolic content among different peach cultivars; (D) Comparison of organic acid composition among different peach cultivars. Different lowercase letters indicate significant differences among cultivars; (E, F) Changes in soluble sugar composition among different peach cultivars. (n = 3, p < 0.05).

Total acid content was approximately 4–5 mg/g FW across the three ripening groups (Figure 1B). Although the early‐ripening group showed a numerically higher mean total acid content and the late‐ripening group a lower value, no significant differences were detected among the three groups (p > 0.05). In contrast, total phenolic content differed markedly among groups (Figure 1C). The early‐ripening group showed the highest total phenolic content, followed by the late‐ripening group, whereas the mid‐ripening group had the lowest value. These results indicate that phenolic accumulation varied more strongly among the six cultivars than total acidity.

Despite the absence of significant differences in total acid content, the profiles of individual organic acids differed substantially among the ripening groups (Figure 1D). Quinic, malic, and citric acids were the predominant organic acids and occurred at substantially higher levels than acetic, tartaric, oxalic, shikimic, and fumaric acids. Quinic acid was highest in the early‐ripening group, intermediate in the mid‐ripening group, and lowest in the late‐ripening group, with significant differences among all three groups (p < 0.05). Malic acid showed an opposite pattern, with a significantly lower level in the early‐ripening group than in the mid‐ and late‐ripening groups, whereas the latter two groups did not differ significantly. Citric acid exhibited a distinct distribution pattern, being highest in the early‐ripening group, lowest in the mid‐ripening group, and intermediate in the late‐ripening group, with significant differences among all three groups.

The minor organic acids also showed compound‐specific variation. Acetic, tartaric, and oxalic acids were significantly lower in the early‐ripening group than in the mid‐ and late‐ripening groups, whereas no significant differences were observed between the latter two groups. Shikimic acid showed the reverse pattern and was significantly higher in the early‐ripening group than in the mid‐ and late‐ripening groups. Fumaric acid was likewise lower in the early‐ripening group than in the other two groups. Collectively, these results show that differences in acid‐related composition among the cultivars were more clearly reflected in the relative distribution of individual organic acids than in total acid content alone.

The major soluble sugar profile also differed among ripening groups (Figure 1E). Among the four major soluble sugars and sugar alcohols quantified, sucrose was the predominant component; however, its concentration did not differ significantly among the early‐, mid‐, and late‐ripening groups. Glucose and fructose exhibited similar distribution patterns, with both being significantly higher in the early‐ripening group than in the mid‐ripening group. The late‐ripening group showed intermediate levels and did not differ significantly from either of the other two groups. Sorbitol showed a clearer group‐associated pattern, with a significantly higher concentration in the early‐ripening group than in the mid‐ and late‐ripening groups, whereas no significant difference was observed between the latter two groups. Thus, although sucrose represented the major soluble sugar fraction, variation in glucose, fructose, and particularly sorbitol contributed more strongly to differences in soluble sugar composition among the groups.

Minor sugars and sugar alcohols also displayed distinct distribution patterns (Figure 1F). Arabinose was highest in the late‐ripening group and lowest in the mid‐ripening group, resulting in a significant difference between these two groups, whereas the early‐ripening group did not differ significantly from either. Maltitol and xylitol showed no significant differences among the three groups. Rhamnose also varied only moderately, with overlapping significance‐group letters among the groups. In contrast, raffinose was markedly enriched in the late‐ripening group and was significantly higher than in both the early‐ and mid‐ripening groups. Erythritol occurred at relatively low levels and did not differ significantly among groups.

Taken together, the six honey peach cultivars exhibited distinct physicochemical and nonvolatile flavor profiles. Total acid content remained relatively similar among the ripening groups, whereas the distributions of quinic, malic, citric, and several minor organic acids varied considerably. Likewise, sucrose remained comparatively stable, whereas glucose, fructose, sorbitol, arabinose, and raffinose showed group‐dependent differences. Total phenolic content provided particularly clear discrimination among the three groups. Therefore, variation in nonvolatile flavor quality was associated primarily with differences in the composition and relative distribution of individual organic acids, sugars, sugar alcohols, and phenolic compounds rather than coordinated changes in total acidity or the predominant sugar component.

3.3. Analysis of Volatile Compounds in Honey Peaches From Different Cultivars

3.3.1. Volatile Fingerprints and Characteristic Compounds Based on GC‐IMS

The GC‐IMS fingerprint profiles revealed pronounced differences in the relative signal distributions of volatile compounds among peach cultivars representing different ripening groups. A total of 51 volatile compounds were retained after GC‐IMS data curation, comprising 13 aldehydes, 12 ketones, 10 esters, 8 alcohols, 6 compounds classified as others, and 2 acids (Figure 2A). The fingerprint profiles were highly consistent among biological replicates of the same cultivar, indicating high analytical repeatability.

FIGURE 2.

Analysis of volatile compounds in different peach cultivars based on GC‐IMS and HS‐GC‐MS. (A) GC‐IMS fingerprint of volatile compounds. Color intensity represents the relative signal intensity of each compound. (B) Correlation chord diagram showing the distribution and chemical classification of volatile compounds detected by HS‐GC‐MS. Colors represent different chemical classes and volatile compounds.

Early‐ripening cultivars exhibited relatively strong signals for heptanal, hexanal, (E)‐2‐heptenal, (E)‐2‐octenal, and 1‐octen‐3‐one, indicating a volatile profile associated predominantly with green, fresh, fatty, and mushroom‐like notes. In the mid‐ripening cultivars, ethyl propanoate, propyl acetate, (E)‐2‐hexen‐1‐ol, pyrazine, and 2‐n‐butylfuran showed comparatively higher signal intensities, suggesting a more composite profile characterized by fruity and green notes with minor roasted nuances. Late‐ripening cultivars showed relatively strong signals for fenchol, 1‐hexanol, 2‐heptanone, 2‐octanone, benzaldehyde, and ethyl isobutyrate, indicating that their volatile profiles were shaped by contributions from esters, aromatic aldehydes, alcohols, and ketones.

Chemical‐class analysis showed that aldehydes, ketones, esters, and alcohols accounted for 25.49%, 23.53%, 19.61%, and 15.69% of the detected compounds, respectively (Figure 3A). Carbonyl compounds (aldehydes and ketones) collectively accounted for 49.02% of the total detected volatiles, constituting a major component of the volatile profile.

FIGURE 3.

Chemical‐class distribution of volatile compounds detected in honey peach cultivars using different analytical platforms. (A) Volatile compounds detected by GC‐IMS; (B) volatile compounds detected by HS‐GC‐MS. Different colors represent distinct chemical classes, and the size of each petal corresponds to the number and proportion of compounds within that class.

3.3.2. Volatile Compounds Detected by HS‐GC‐MS

As shown in Figure 2B, a total of 71 volatile compounds were identified by HS‐GC‐MS, comprising 21 esters, 20 aldehydes, 9 alcohols, 9 miscellaneous compounds, 8 ketones, 3 furans, and 1 acid. The correlation chord diagram revealed extensive positive and negative relationships among the detected volatiles. Esters, aldehydes, alcohols, and ketones correlated not only within their respective chemical classes but also across classes, indicating a complex network of coordinated volatile variation.

Chemical‐class analysis showed that esters and aldehydes accounted for 29.58% and 28.17% of the detected compounds, respectively, together representing 57.75% of the total volatiles detected by HS‐GC‐MS (Figure 3B). Esters are commonly associated with sweet and fruity notes, whereas aldehydes contribute green, fatty, fresh, and almond‐like characteristics. Alcohols, ketones, and miscellaneous compounds further contributed to the diversity of floral, fruity, and peach‐like aroma attributes.

3.3.3. Differential Volatile Profiles Evaluated by Multivariate Statistics

Supervised discriminant analysis (PLS‐DA) showed a clear separation among the early‐, mid‐, and late‐ripening peach samples in both the GC‐IMS and HS‐GC‐MS datasets (Figure 4A,C). Samples from the same ripening category clustered tightly together, while the three groups occupied distinct regions in the score plots. For GC‐IMS (Figure 4B), the model yielded R ² Y and Q ² values of 0.909 and 0.857, respectively (difference of 0.052), supporting a good fit and satisfactory internal predictive performance. The HS‐GC‐MS model (Figure 4D) showed even sharper group separation, with R ² Y and Q ² values of 0.982 and 0.976, respectively. The tighter clustering observed in HS‐GC‐MS data may be attributed to its broader analytical coverage of esters, terpenoids, lactones, and other structurally diverse volatiles, whereas GC‐IMS is particularly sensitive to low‐molecular‐weight and highly volatile carbonyl compounds.

FIGURE 4.

PLS‐DA analysis of early‐, mid‐, and late‐ripening peaches using volatile profiles determined by GC‐IMS and HS‐GC‐MS. (A) PLS‐DA score plot; (B) Permutation test plot of the PLS‐DA model (R ² Y = 0.909 and Q ² = 0.857). Component 1 and Component 2 accounted for 30.6% and 38.2% of total variance, respectively; (C) PLS‐DA score plot; (D) Permutation test plot of the PLS‐DA model (R ² Y = 0.982 and Q ² = 0.976). Component 1 and Component 2 accounted 46.1% and 44.1% of total variance, respectively.

Permutation testing confirmed the robustness of both models, as permuted R ² and Q ² values were consistently lower than those of the original models. These results indicate that the two analytical platforms captured distinct but complementary aspects of volatile variation among the six cultivars.

3.3.4. Screening and Distribution Patterns of Differential Volatile Compounds

Differential volatile compounds were screened based on variable importance in projection (VIP) values derived from PLS‐DA models in combination with univariate analysis (VIP > 1.0 and p < 0.05). A total of 24 and 14 differential volatile compounds were identified from the GC‐IMS and HS‐GC‐MS datasets, respectively.

In the GC‐IMS dataset (Figure 5A), (E)‐2‐pentenal exhibited the highest VIP value, followed by n‐nonaldehyde, 2‐methyltetrahydrofuran‐3‐one, and 2‐acetylfuran. Overall, aldehydes accounted for a major proportion of high‐VIP compounds, including (E)‐2‐pentenal, 3‐methyl‐2‐butenal, 2‐methylbutanal, heptanal, hexanal, benzaldehyde, and (E)‐2‐heptenal. The relative abundance heatmap revealed that most small‐molecule aldehydes, ketones, and furans were enriched in early‐ripening samples. Conversely, benzaldehyde, 5‐methylfurfural, and fenchol showed higher signal intensities in late‐ripening samples. Ethyl isobutyrate was enriched in mid‐ and late‐ripening samples, whereas hexyl acetate and thiophene were highest in the mid‐ripening group.

FIGURE 5.

Variable importance in projection (VIP) scores and relative‐abundance heatmaps of volatile compounds in early‐, mid‐, and late‐ripening honey peach cultivars. (A) Volatile compounds detected by GC‐IMS; (B) volatile compounds detected by HS‐GC‐MS. Bars represent the VIP scores derived from the PLS‐DA models, and the heatmaps show the row‐wise standardized relative abundance of each compound among the three ripening‐period groups. Red and green indicate relatively high and low abundance, respectively. Compounds with VIP > 1 were considered candidate discriminant volatiles.

In the HS‐GC‐MS dataset (Figure 5B), undecanal showed the highest VIP value, followed by dimethylallyl acetate and 4‐methylvalerophenone. Tridecane, n‐propyl acetate, ethyl acetate, undecanal, octanal, 4‐methylvalerophenone, cyclohexanecarboxaldehyde, and dimethylallyl acetate were more abundant in early‐ripening samples. In contrast, ethyl caprylate, 2‐hexenal, a‐terpinene, ethyl trans‐4‐decenoate, and D‐limonene were enriched in late‐ripening samples. Overall, HS‐GC‐MS analysis revealed that late‐ripening cultivars possessed greater abundance of esters, while relatively higher concentrations of aldehydes were detected in early‐ripening cultivars.

3.3.5. Candidate Differential Aroma‐Active Compounds

To screen key volatile compounds with potential sensory importance, VIP values were integrated with univariate statistical significance and odor activity evaluation (OAV / ROAV).

For the GC‐IMS dataset, four candidate differential aroma‐active compounds were identified using the combined criteria of VIP > 1.0, p < 0.05, and ROAV > 1.0: 3‐penten‐2‐one, ethyl isobutyrate, n‐nonaldehyde, and pentyl acetate (Table 2). Among them, 3‐penten‐2‐one and n‐nonaldehyde were enriched in early‐ripening cultivars. Pentyl acetate was highest in mid‐ripening cultivars, whereas ethyl isobutyrate remained high in early‐ and late‐ripening cultivars.

TABLE 2.

Characteristic volatile compounds in different honey peach cultivars.

Detection method	Volatile compound	Threshold/(mg·kg−1)	Different varieties			Aroma description
			Early‐ripening	Mid‐ripening	Late‐ripening
GC‐IMS(ROAV)
1	Ethyl isobutyrate	0.0000081	100.000	91.613	100.000	Sweet, fruity
2	3‐Penten‐2‐one	0.0015	10.590	1.555	3.673	Fruity
3	n‐Nonaldehyde	0.00032	3.070	1.445	1.221	Rose, citrus, strong oily
4	Pentyl acetate	0.00001	46.319	100.000	63.510	Pineapple, grape peel
HS‐GC‐MS(OAV)
1	7,8‐Dihydro‐β‐ionone	0.001	40.837	3.047	47.152	Woody
2	Benzaldehyde	0.024	9.198	8.011	12.981	Bitter almond odor
3	Decanal	0.00008	191.411	40.570	120.848	Sweet orange note
4	(E)‐2‐Nonenal	0.00008	305.012	72.371	217.263	Fresh cucumber and fatty odors
5	Nonanal	0.00032	236.527	66.196	223.275	Rose, citrus notes
6	γ‐Decalactone	0.0007	82.785	10.226	283.489	Fruity, peach aroma
7	(E)‐2‐Octenal	0.00034	58.401	27.836	65.435	Fruity, peach aroma
8	Hexyl acetate	0.002	102.877	22.437	47.653	Fruity, green, apple, banana, sweet
9	(E,E)‐2,4‐Nonadienal	0.0000017	1248.990	483.445	2211.593	Fresh cucumber and fatty aroma

Note: Odor thresholds were obtained primarily from Compilations of Odor Threshold Values in Air, Water, and Other Media (2nd edition), supplemented with relevant original literature where necessary. Flavor characteristic descriptions were sourced from the online ingredient library of the Flavor & Extract Manufacturers Association (FEMA), available at https://www.femaflavor.org.

For the HS‐GC‐MS dataset, nine candidate differential aroma‐active compounds were identified (VIP > 1.0, p < 0.05, and OAV > 1.0): benzaldehyde, hexyl acetate, (E)‐2‐octenal, nonanal, decanal, γ‐decalactone, (E)‐2‐nonenal, (E, E)‐2,4‐nonadienal, and 7,8‐dihydro‐β‐ionone (Table 2). Hexyl acetate showed higher concentrations in early‐ripening cultivars, contributing to green/fruity notes. In contrast, benzaldehyde, γ‐decalactone, (E, E)‐2,4‐nonadienal, and 7,8‐dihydro‐β‐ionone were enriched in late‐ripening cultivars. Notably, the pronounced enrichment of γ‐decalactone in late‐ripening cultivars highlights its role as a key contributor to characteristic peach‐like and creamy aromas, whereas benzaldehyde and 7,8‐dihydro‐β‐ionone enhanced almond‐like and floral notes, respectively.

4. Discussion

4.1. Differences in Sugar and Organic Acid Composition Contribute to Nonvolatile Flavor Differentiation Among Peach Cultivars With Different Ripening Periods

Sugar and organic acid composition constitute the fundamental chemical basis of fruit flavor. The balance among sweetness, acidity, and volatile aroma—rather than any single component, ultimately dictates the sensory quality of stone fruits (Aslam et al. 2026). Recent studies emphasize that sugar, acid, and aroma traits are metabolically interconnected via central carbon metabolism, although their coordination is strongly modulated by genotype and developmental stage (Xue et al. 2025; J. Zheng et al. 2019).

In the present study, late‐ripening cultivars accumulated significantly higher levels of soluble solids and total sugars, whereas early‐ripening cultivars exhibited lower sugar contents. However, individual sugar dynamics did not completely mirror total sugar trends. While sucrose was the dominant soluble sugar, glucose, fructose, sorbitol, and raffinose contributed more strongly to cultivar‐ and ripening‐stage‐associated differences. Sucrose accumulation in peach is governed by sucrose synthesis, cleavage, and transport. Recent studies demonstrated that the transcription factor PpNAP4 activates PpSUS1 and PpSPS2 to promote sucrose accumulation (Dai et al. 2026), while sugar transporters such as PpPMT1 modulate sugar allocation (Li et al. 2026). The elevated sugar levels observed in late‐ripening cultivars likely reflect differences in source‐sink carbon partitioning and enzymatic regulation of sugar metabolism (Vimolmangkang et al. 2016; Dai et al. 2026; Lombardo et al. 2011; Zanon et al. 2015).

Organic acid profiles also provided clear discrimination among cultivars. Although total acidity varied within a narrow range, quinic, malic, and citric acids displayed striking group‐specific patterns. Quinic and citric acids were enriched in early‐ripening cultivars, whereas malic acid predominated in late‐ripening cultivars. Malate accumulation in peach is regulated by both metabolic processes and vacuolar storage, whereas citrate accumulation appears to be controlled predominantly at the metabolic level (Zheng et al. 2021; Zhou et al. 2023). Thus, flavor differences among these cultivars are governed by the specific composition and ratios of individual sugars and organic acids rather than total sugar or acid content alone.

4.2. Ripening‐Period‐Associated Differences in Volatile Composition Reflect Coordinated Variation Among Multiple Aroma Metabolic Pathways

Peach aroma arises from a complex mixture of volatiles derived from fatty acid metabolism, amino acid pathways, terpenoid biosynthesis, carotenoid cleavage, and phenylpropanoid metabolism (Liu et al. 2025). Characteristic compounds such as γ‐decalactone, hexanal, benzaldehyde, (E)‐2‐hexenal, β‐ionone, linalool, and 1‐hexanol impart green, fruity, floral, and peach‐like aroma characteristics (Liu et al. 2025).

The present study demonstrated that early‐ripening cultivars were enriched in C6‐C8 aliphatic aldehydes and ketones (e.g., heptanal, hexanal, (E)‐2‐heptenal, and 1‐octen‐3‐one), imparting fresh and green notes. These volatiles are predominantly synthesized via the lipoxygenase (LOX) pathway from unsaturated fatty acids (Wang et al. 2016; Su et al. 2026). Conversely, late‐ripening cultivars exhibited significant enrichment of lactones (γ‐decalactone), aromatic aldehydes (benzaldehyde), norisoprenoids (7,8‐dihydro‐β‐ionone), and monoterpenes (linalool, α‐terpineol). γ‐Decalactone, catalyzed by alcohol acyltransferase PpAAT1 (Peng et al. 2020), is the key contributor to the desirable sweet, peach‐like, and creamy aroma. Meanwhile, terpenoid biosynthesis is transcriptionally regulated by factors such as PpbHLH1 during fruit ripening (Wei et al. 2021).

Interestingly, mid‐ripening cultivars exhibited distinct volatile signatures enriched in esters such as pentyl acetate and propyl acetate, rather than acting as a simple intermediate transition between early and late cultivars. These findings suggest that volatile differences reflect distinct genotype‐ and ripening‐stage‐specific metabolic programs across different pathways.

4.3. Potential Metabolic Links Between Sugar‐Acid Composition and Volatile Aroma Formation

Fruit flavor perception integrates nonvolatile taste components (sugars and acids) with volatile aroma attributes. At the biochemical level, primary carbohydrate metabolism supplies precursor pools (e.g., acetyl‐CoA, pyruvate, phosphoenolpyruvate, and amino acids) for secondary volatile synthesis (Aslam et al. 2026).

In the investigated cultivars, late‐ripening peaches exhibited concurrent enrichment of soluble sugars (raffinose and total sugars) and secondary volatile metabolites such as lactones (γ‐decalactone), esters, and norisoprenoids. Conversely, early‐ripening cultivars accumulated higher organic acids (quinic and citric acids) alongside LOX‐derived green‐note aldehydes. These parallel variations suggest a coordinated flux allocation of central carbon metabolism between nonvolatile flavor precursors and volatile biosynthetic pathways during fruit development and ripening (Li et al. 2026). Further studies utilizing isotopic flux analysis and transcriptomics will help elucidate the exact regulatory mechanisms connecting primary sugar‐acid metabolism to aroma formation.

4.4. Complementary GC‐IMS and HS‐GC‐MS Analysis Improves Screening of Candidate Aroma‐Active Compounds

Due to the vast chemical diversity and broad volatility range of peach volatiles, no single analytical platform can achieve complete coverage. GC‐IMS excels in rapid headspace analysis, high sensitivity toward low‐molecular‐weight volatiles, and visual fingerprinting. In contrast, HS‐GC‐MS provides robust spectral identification, broader compound coverage (including lactones, sesquiterpenes, and higher esters), and accurate semi‐quantification (Sun et al. 2022; Wang et al. 2023).

In this study, GC‐IMS captured clear differences in small aldehydes, short‐chain esters, and ketones, identifying candidates such as 3‐penten‐2‐one, ethyl isobutyrate, nonanal, and pentyl acetate (ROAV > 1.0). Meanwhile, HS‐GC‐MS successfully identified lactones (γ‐decalactone), long‐chain aldehydes, and norisoprenoids (7,8‐dihydro‐β‐ionone) as key discriminants (OAV > 1.0). Combining VIP > 1.0, statistical significance (p < 0.05), and odor activity thresholds (OAV/ROAV > 1.0) established a rigorous framework for screening candidate aroma‐active compounds. The integration of these two complementary analytical techniques offers a comprehensive strategy for dissecting complex fruit aroma profiles and identifying key sensory discriminants.

5. Conclusions

In this study, GC‐IMS and HS‐GC‐MS were integrated to characterize the volatile profiles and establish the key differential aroma components across early‐, mid‐, and late‐ripening peach cultivars at comparable physiological maturity. Multivariate statistical analysis demonstrated clear separation among the three ripening groups, highlighting the complementary advantages of GC‐IMS (high‐sensitivity small‐molecule fingerprinting) and HS‐GC‐MS (broad spectral identification and semi‐quantification).

The volatile profiles exhibited a distinct ripening‐stage‐dependent evolution pattern. Early‐ripening cultivars were characterized by lipid oxidation‐derived C6/C8 aldehydes and ketones, imparting fresh, green, and fatty notes. In mid‐ripening cultivars, the aroma profile transitioned toward enhanced fruity characteristics driven by increased ester diversity. Late‐ripening cultivars displayed the highest aroma complexity, marked by the enrichment of floral terpenes, norisoprenoids, and key lactones (γ‐decalactone), which together defined the mature and classic peach flavor profile.

Overall, this study elucidates the ripening‐dependent volatile evolution in honey peaches and provides a robust multi‐platform framework for flavor quality assessment, germplasm evaluation, and harvest management. Future integration of multi‐omics approaches with sensory evaluation will help decipher the underlying molecular mechanisms and guide the breeding of high‐aroma peach cultivars.

Author Contributions

Rui Wang: data curation, software, investigation, writing – original draft, writing – review and editing, visualization, formal analysis. Yahui Li: writing – review and editing. Zhiyong Zhang: writing – review and editing. Lu Shi: conceptualization, writing – review and editing, supervision. Shulin Wang: writing – review and editing, supervision. Ying Liang: writing – review and editing, supervision, funding acquisition, conceptualization.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supplementary Material: jfds71469‐sup‐0001‐TableS1‐S3.docx