Unveiling the influence of temperature treatments on the structural and Metabolomic profile of Pandan leaves (Pandanus amaryllifolius) instant powder: a multi-analysis approach Authors: Chengxu Liu, Zhen Zhang, Yuqi Xin, Yuanyuan Zhang, Yiyun Wu, Chuanrui Luo, Peishi Liu, Yewei Huang, Cunchao Zhao, Ya Wang Published: 2025 ABSTRACT AbstractPandan leaves (Pandanus amaryllifolius), widely used in Southeast Asia, hold significant promise for natural product and fragrance industries, though underutilized due to limited processing technologies. This study investigated the impact of three temperature-based extraction methods—room temperature (RT), low temperature (LT), and high temperature (HT) on the structural and flavor properties of instant pandan leaf powder. Quality characteristics were assessed via UV–Vis, FT-IR, XRD, and SEM analyses. Metabolite profiles were characterized using GC–MS and UPLC-MS/MS. HT treatment produced powder with a porous structure and significantly enhanced sweetness and umami (P < 0.05), while RT retained the highest levels of volatile compounds (44.02 %) and was rich in alcohol-based non-volatile metabolites. LT treatment showed moderate effects. Overall, HT improved solubility and flavor, whereas RT better preserved aromatic compounds. These findings help optimize processing methods to improve pandan leaves' functional and sensory properties, supporting broader applications in food products and fragrance industries.Keywords: Pandanus amaryllifolius, Instant pandan leaf powder, Temperature treatment, Metabolomics, Flavor profile Introduction Pandan leaves (Pandanus amaryllifolius Roxb.), native to Indonesia, were introduced to China in the 1950s and are now widely cultivated in regions including Hainan, Yunnan, Guangdong, and Fujian (Sinicae, 2000). This plant is primarily propagated through stem cuttings, and both its cultivation and processing involve relatively low technical and economic barriers (Bhuyan & Sonowal, 2021). Pandan leaves are renowned for their rich aromatic profile, primarily due to the presence of 2-acetyl-1-pyrroline, a compound responsible for their distinctive fragrance (Wakte et al., 2012). Recent studies have highlighted the presence of phenolic compounds and flavonoids in pandan leaves, which exhibit significant antioxidant and anti-inflammatory properties (Ningrum et al., 2015). Moreover, pandan leaves offer not only a unique aromatic quality but also considerable potential health benefits (Saenthaweesuk et al., 2016). For instance, Pandan leaves are commonly used to support digestive health. They exhibit potential in regulating blood sugar levels. Pandan leaves also contribute to promoting blood circulation, thereby improving cardiovascular health. Additionally, they contain several essential nutrients, including vitamins and minerals, which help enhance immune function and boost the body's resistance (Bhuyan & Sonowal, 2021). Given these attributes, pandan leaves have promising applications that extend beyond their traditional use as a food flavoring to include functional foods and even medicinal resources. Pandan leaves, a representative herbaceous plant of Hainan, combine a distinctive tropical flavor with notable health benefits (Huang et al., 2025), positioning them as a potential local signature specialty food. In line with policy initiatives such as “Green Hainan,” “Healthy Hainan,” and “Tropical Specialty Efficient Agriculture,” pandan leaves are progressively transitioning from a traditional folk ingredient to an industrialized product, emerging as a key feature in Hainan's tropical specialty food sector (Z. Zhu et al., 2022). Their natural green quality, unique aroma, and abundant health-promoting compounds align with the growing consumer demand for natural, functional, and regionally characteristic foods, highlighting their strong development potential. Hainan, China's only tropical province, offers an ideal environment for pandan leaf cultivation. However, the industry faces challenges due to seasonal variability (L. Xu et al., 2025), with cultivation currently characterized by dispersed farming practices, inconsistent management standards, and limited standardization. Additionally, extreme weather events, such as typhoons, significantly impact yields. Post-harvest processing of pandan leaves also presents challenges, as the leaves are prone to browning and the loss of their distinctive aroma, complicating preservation efforts (Routray & Rayaguru, 2010). As such, promoting the deep processing of pandan leaves is essential to enhancing their added value and improving their commercial viability. In recent years, primary processed products such as pandan powder and pandan juice have gained increasing attention, leading to the emergence of specialty foods, including pandan cakes and ice cream (Bhattacharjee et al., 2005). As the markets for new-style tea beverages and health foods continue to expand, pandan leaves have become increasingly popular among consumers due to their natural green color and distinctive aroma. The development of instant pandan leaf powder products that maintain desirable flavor profiles and storage stability holds significant potential for addressing preservation challenges, thereby creating opportunities for broader market expansion. Instant powder is a processed product derived from plant materials through extraction, concentration, and drying, resulting in a powder that is easily dissolvable. This form is widely utilized in the food and health supplement industries (Ni et al., 2020). The extraction temperature plays a crucial role in determining the quality of the instant powder (Naji et al., 2023). The impact of temperature on the quality of instant powder is reflected in multiple aspects: high temperatures can enhance the dissolution rate, but excessively high temperatures may damage flavor compounds, leading to poor taste or the development of off-flavors. Conversely, low temperatures may result in incomplete dissolution of the powder, affecting the user experience. High temperatures can also cause particle clumping, which impairs flowability and may destroy nutrients such as vitamins, thus reducing nutritional value. Furthermore, an increase in temperature can lead to color changes, potentially triggering browning reactions and affecting the appearance. While high-temperature extraction can enhance the release of certain bioactive components, it may also lead to the degradation of heat-sensitive compounds (Antony & Farid, 2022). Temperature variations during the extraction process can lead to distinct flavor profiles in the resulting instant powders. Consequently, it is essential to systematically investigate the impact of different temperature treatments on the structural and flavor characteristics of pandan leaf instant powder. This study aimed to prepare pandan leaf instant powder using three distinct temperature treatments and to evaluate its chemical properties and flavor components through various analytical techniques. The findings are intended to provide a scientific foundation for optimizing the processing technology of pandan leaf instant powder. Materials and methods Materials and reagents Fresh pandan leaves (leaf length 30–80 cm, green to dark green) were purchased from Huishuyuan Agricultural Products Co., Ltd., Haifeng County, Hainan Province. Concentrated hydrochloric acid, copper sulfate, potassium sulfate (Beijing Bailinwei Technology Co., Ltd., analytical grade); potassium iodide, petroleum ether, Folin-Ciocalteu reagent, gallic acid standard (Shanghai Yuanye Biotechnology Co., Ltd., analytical grade). Methyl red indicator, bromocresol green indicator, methylene blue indicator (Nanjing Chemical Reagents Co., Ltd., analytical grade); anhydrous ethanol, anhydrous ether (Tianjin Shuangchuan Chemical Reagent Factory, analytical grade); sulfuric acid, boric acid (Guangdong Guanghua Technology Co., Ltd., analytical grade). Preparation of Pandan leaf instant powder Fresh pandan leaves were rinsed 1–2 times with clean water and soaked for 2–3 min. Excess water was removed using absorbent paper, followed by indoor air-drying for 10–15 min. The indoor temperature is 25 °C, with humidity ranging from 45 % to 60 %. The remaining surface moisture was eliminated with a blower. The dried leaves were sliced into thin strips (1.0–2.0 cm wide) using a slicer, then ground and passed through a 100-mesh sieve to produce pandan leaf powder, its moisture content is approximately 10 %. A total of eighty grams of the powder was mixed with 2400 mL of purified water and divided into three groups: RT (25 °C for 5 h), LT (4 °C for 60 h), and HT (85 °C for 45 min) (He et al., 2023). Each extract was centrifuged at 4500 rpm for 15 min and filtered to obtain a clear solution. This filtrate was concentrated via rotary evaporation to reach ∼10 % solid content. Determined according to the national standard GB/T 12295–2008 ‘Determination of Soluble Solids Content in Fruit and Vegetable Products — Refractometric Method (Mai & Liang, 2018). The concentrate was pre-frozen at −30 °C for 24 h and freeze-dried under vacuum for 72 h. Final drying below the material's maximum temperature produced three types of instant powders: RT, LT, and HT. Determination of chemical composition in instant powder The moisture, ash, protein and fat content were determined according to CNIS guidelines specified in GB 5009.3–2016 (Li et al., 2023), GB 5009.4–2016 (Yuanyuan Zhang et al., 2024), GB 5009.5–2016 (Fu et al., 2025), and GB 5009.6–2016 (Zhou et al., 2024). UV spectroscopy determination UV–Vis absorption spectra were recorded using a UV-3600 spectrophotometer equipped with a 10 mm quartz cuvette. Instant powder samples were dissolved in deionized water to prepare a 1.0 mg/mL solution, and the wavelength scan was conducted over the range of 200–800 nm (Mandru et al., 2023). Infrared spectroscopy determination Fourier-transform infrared (FT-IR) spectra were recorded using a Nicolet 6700 FT-IR spectrometer (Thermo Fisher Scientific, USA). The pandan leaf instant powder was prepared as KBr pellets and analyzed via infrared spectroscopy. The scan range was set between 500 and 4000 cm−1, with a resolution of 8 cm−1 and 16 scans per sample (J. Xu et al., 2023). X-ray diffraction (XRD) analysis The crystal structure of the instant powder was characterized using an Empyrean X-ray diffractometer (Malvern Panalytical X'Pert3 Powder, Netherlands). The instrument was operated at a voltage of 40 kV and an incident current of 40 mA. The scanning speed was set to 10°/min, and the diffraction angle (2θ) was recorded over a range of 5° to 65° (Xiao et al., 2020). Scanning Electron microscopy (SEM) analysis The surface morphology of the instant powder was analyzed following the method outlined by (Chen et al., 2022), with slight modifications. Observations were made using a Zeiss Sigma 360 scanning electron microscope (SEM) equipped with an Oxford Ultim Max 40 EDS detector. Coarse instant powder particles were mounted on the sample stage using conductive carbon tape, sputter-coated with a 50 nm gold layer, and examined under an accelerating voltage of 4–5 kV. Electronic nose and tongue analysis Two grams of sample were weighed and placed into a 10 mL glass vial. The vial was allowed to stand at 25 °C for 30 min before measurement. The sampling interval was set to 1 s per group, with a sensor self-cleaning time of 100 s, a baseline reset time of 5 s, and a sample preparation time of 5 s. The sample injection flow rate was maintained at 400 mL/min, and the total analysis time was 60 s. Twelve grams of the sample were dissolved in 120 mL of water at 80 °C to create a 1:10 homogenized solution. The mixture was shaken for 30 min and centrifuged at 4800 rpm for 10 min. The supernatant was collected and cooled to room temperature. A 105 mL portion was used to assess six taste attributes: sourness, bitterness, astringency, umami, saltiness, and sweetness. A reference solution of KCl and tartaric acid was used for calibration. The sensor was cleaned using specific solutions for positive (KCl, ethanol, KOH) and negative (water, ethanol, HCl) electrodes, and taste measurements were recorded and repeated for accuracy. GC–MS determination of volatile compounds in instant powder Chromatographic analysis was performed using a LECO Pegasus BT 4D GC × GC-TOF MS system (LECO, St. Joseph, MI, USA), which included an Agilent 8890 A gas chromatograph (Agilent Technologies, Palo Alto, CA, USA), a dual-stage jet modulator, and a split/splitless injection module. Separation was carried out on a DB-Heavy Wax column (30 m × 250 μm × 0.5 μm; Agilent, USA) using high-purity helium as the carrier gas, at a constant flow rate of 1.0 mL/min. The temperature program for the DB-Heavy Wax column was as follows: the initial temperature was set to 50 °C and held for 2 min, then ramped at 5 °C/min to 240 °C, where it was held for an additional 5 min. Mass spectrometric detection was conducted with a LECO Pegasus BT 4D mass spectrometer (LECO, St. Joseph, MI, USA). The transfer line and ion source temperatures were both maintained at 250 °C. Data acquisition was performed at a rate of 10 spectra per second. Electron impact ionization was employed at 70 eV, with a detector voltage set to 1960 V. The mass scan range was m/z 35–550. Phytochemistry of Instant Powder by UPLC-MS/MS Ultra-performance liquid chromatography (UPLC) analysis was conducted using a Thermo Vanquish system (Thermo Fisher Scientific, USA) with an ACQUITY UPLC® HSS T3 column (2.1 × 100 mm, 1.8 μm; Waters, USA). The flow rate was 0.3 mL/min, the column temperature was maintained at 40 °C, and the injection volume was 2 μL. In positive ion mode, the mobile phase comprised 0.1 % formic acid in water (A2) and 0.1 % formic acid in acetonitrile (B2). The gradient elution was: 0–1 min at 10 % B2, 1–5 min ramping to 98 % B2, 5–6.5 min at 98 % B2, 6.5–6.6 min back to 10 % B2, and 6.6–8 min at 10 % B2. For negative ion mode, the mobile phase used 5 mM ammonium formate in water (A3) and acetonitrile (B3), following the same gradient. Mass spectrometry was performed using a Thermo Orbitrap Exploris 120 with an electrospray ionization (ESI) source in both positive and negative modes. Spray voltages were 3.50 kV (positive) and − 2.50 kV (negative). Sheath and auxiliary gas flow rates were 40 and 10 arbitrary units, respectively. The capillary temperature was 325 °C. Full MS scans were acquired from m/z 100–1000 at 60,000 resolution. MS/MS analysis used higher-energy collisional dissociation (HCD) at 30 % collision energy and 15,000 resolution, with a data-dependent method targeting the top four ions and dynamic exclusion to reduce redundancy. ROAV calculation To evaluate the contribution of volatile compounds to the overall aroma profile, the Relative Odor Activity Value (ROAV) was calculated. Compounds with ROAV values greater than 1 are classified as key volatile components (Yi et al., 2021). The formula for calculating ROAV is as follows: where OAV max represents the highest odor activity value among the volatile compounds, and OAV i is the odor activity value of a specific volatile compound. C i represents the concentration percentage of the volatile compound in the sample, and T i represents the odor threshold of that compound in water. Statistical analysis All results are presented as the me232: an ± standard deviation of three replicates. The mean values, standard deviations, and radar charts were generated using Microsoft Office 2010. One-way analysis of variance (ANOVA) followed by Duncan's multiple range test was performed using SPSS version 20.0 to evaluate significant differences (P < 0.05). Multivariate analysis was conducted using SIMCA 14.1, with variables deemed statistically significant when the variable importance in projection (VIP) score was greater than 1.0 and P < 0.05. Results Chemical composition of instant powder The main component contents of instant powders prepared by three different extraction methods (RT, LT, and HT), were systematically compared (Table 1), revealing significant effects of extraction temperature on the nutritional composition and physicochemical properties of the powders. The moisture content of the instant powder obtained from the three groups does not differ significantly, as the variation in moisture content primarily depends on the drying process of the instant powder, rather than the temperature in the extraction method. Therefore, although different extraction temperatures may affect the composition and properties of the extracts, the moisture content of the instant powders remains similar. This reduction is advantageous for inhibiting microbial growth and enhancing the dryness and stability of the product. In terms of nutritional components, the HT group exhibited the highest protein content (13.6 g/100 g), significantly surpassing the RT group and slightly exceeding the LT group. This suggests that high temperatures facilitate the release and solubilization of protein structures. The LT group also demonstrated a relatively high protein content (13.3 g/100 g) and low moisture content (6.9 %), striking a good balance between nutritional value and stability. Ash content was highest in the RT group (8.19 g/100 g), indicating that minerals and inorganic salts are better preserved under non-high-temperature extraction conditions. No significant differences in fat content were observed among the three extraction methods, with all samples showing low fat levels (0.65–0.70 g/100 g). This suggests that fat extraction is not particularly sensitive to the temperature used during the extraction process. The extraction yield of pandan leaf instant powder obtained by different extraction methods is approximately 20 %, which is consistent with the chemical composition findings in existing research on pandan leaves. (Routray & Rayaguru, 2010). Table 1Composition of major components in the instant powder samples.Table 1Moisture content(%)Ash content(g/100 g)Protein(g/100 g)Fat(g/100 g)RT9.9 ± 0.09a8.19 ± 0.19a12.5 ± 0.23b0.69 ± 0.11aHT9.0 ± 0.28b7.58 ± 0.28b13.6 ± 0.21a0.70 ± 0.10aLT9.6 ± 0.27a7.18 ± 0.23b13.3 ± 0.50a0.65 ± 0.09aNote: Each value represents the mean ± standard deviation (n = 3); different letters within the same column indicate significant differences according to Duncan's test (p < 0.05). Room temperature (RT), low temperature (LT), and high temperature (HT). Composition of major components in the instant powder samples. Note: Each value represents the mean ± standard deviation (n = 3); different letters within the same column indicate significant differences according to Duncan's test (p < 0.05). Room temperature (RT), low temperature (LT), and high temperature (HT). Analysis of UV–visible-near infrared spectra results UV–Vis-NIR spectroscopy was employed to systematically investigate the optical properties of instant powder extract prepared under different temperature treatments. As depicted in Fig. 1A, the samples from all three temperature treatments exhibited nearly identical absorption profiles in the ultraviolet-visible region, suggesting that the extraction temperature had minimal impact on the main photoactive components of the powders. Structural characterization of the three instant powder samples.(A) UV–Vis spectra of the three instant powders;(B) infrared spectra;(C) X-ray diffraction patterns. Within the wavelength range of 360–400 nm, all samples exhibited a characteristic weak absorption peak, likely associated with the π → π* electronic transitions of polyphenols and flavonoids, which were well-preserved in the samples (Picollo et al., 2019). Additionally, the samples displayed weak absorption in the visible light region (450–600 nm) (Rajeshkumar & Lakshmi, 2021). This optical characteristic is consistent with the light appearance of the sample, indicating that the freeze-drying method used in the preparation of the instant powder effectively reduced the temperature-induced degradation of pigment substances, thereby preserving the product's excellent color and sensory quality. Infrared spectroscopy analysis Fourier-transform infrared (FT-IR) analysis was conducted on instant powder samples subjected to three different temperature treatments, with the results presented in Fig. 1B. The infrared spectra of the samples were largely similar, with the main characteristic peak positions being consistent across the three treatments, suggesting that the core chemical composition of the samples remained largely unchanged by the temperature treatment. Weak absorption peaks in the range of 2990.51–2997.07 cm−1 were attributed to the C—H stretching vibrations of methyl (−CH₃) and methylene (−CH₂) groups, which are commonly found in proteins and lipids. These features suggest the presence of aliphatic hydrocarbons in the samples, likely originating from residual lipids in plant cell membranes or hydrophobic metabolites. A weak absorption peak near 2030 cm−1 was observed, which is typically associated with the stretching vibrations of C Created by potrace 1.16, written by Peter Selinger 2001-2019 C (alkyne) or C Created by potrace 1.16, written by Peter Selinger 2001-2019 N (nitrile) groups (Obasi et al., 2024). These peaks may result from secondary metabolites with distinct structures or trace compounds formed during processing. The absorption peak in the range of 1483.19–1487.57 cm−1 typically corresponds to C—H bending vibrations, which are commonly found in fatty chain structures. This absorption peak is more likely related to glycolipids (Pasieczna-Patkowska et al., 2025). The relatively strong absorption peak at 1176.26 cm−1 was attributed to the stretching vibrations of C—O bonds, in combination with the vibrational effects of sugar backbones. This peak is typically associated with glycosidic bonds (C–O–C) in plant polysaccharides such as cellulose, hemicellulose, or starch (Liu, Sun, et al., 2022). Additionally, the absorption peak near 905.82 cm−1 was attributed to the C Created by potrace 1.16, written by Peter Selinger 2001-2019 O stretching vibration (Yujie Zhang et al., 2022), likely originating from trace amounts of carboxylic acids, esters, or ketone compounds, including polyphenol oxidation products or organic acid derivatives. FT-IR analysis primarily aims to identify the main components and trace chemical constituents in the instant powder, while also determining whether the temperature processing conditions induce changes in surface functional groups and molecular structures. In conclusion, FT-IR analysis confirmed the presence of major components such as polysaccharides, proteins, and lipids in the instant powder, while demonstrating that the different temperature treatments did not significantly alter the molecular structural characteristics of these components. X-ray diffraction analysis In this study, X-ray diffraction (XRD) analysis was performed on the three Pandan instant powder samples, with the results shown in Fig. 1C. All samples exhibited typical A-type crystalline diffraction patterns, characterized by prominent diffraction peaks around 2θ ≈ 17°, 20°, and 22°. These diffraction features closely resemble those observed in plant polysaccharides, such as starch and cellulose (Yui & Ogawa, 2005), suggesting that the primary crystalline components in the instant powders likely originate from plant cell wall polysaccharides or storage polysaccharides. Notably, significant differences in crystallinity were observed among the samples treated at different temperatures. The sample treated at high temperature (HT) exhibited the highest crystallinity, with a value of 68.3 ± 2.1 %, which was significantly higher than that of the low-temperature-treated (LT) sample, which had a crystallinity of 52.4 ± 1.8 % (P < 0.05). This enhancement in crystallinity may be attributed to the increased molecular thermal motion under high-temperature conditions, which facilitates the rearrangement and stacking of polysaccharide chains, leading to the formation of a more ordered crystal structure (Du et al., 2022). Structural characterization of instant powder The microstructures of the instant colorful leaf powders prepared at different temperatures were examined to understand how processing conditions affect their surface characteristics. At magnifications of 2000× and 5000×, the surface of the room temperature (RT) sample appears relatively smooth, with uniform particle sizes, and exhibits several tiny pores and cracks (Figs. 2A,B). These features are likely a result of moisture evaporation during the drying process. In contrast, the low-temperature (LT) sample shows a rough surface with uneven particle sizes and irregular morphology (Figs. 2C,D), which can be attributed to the lower shear forces exerted on the particles during cold water extraction, leading to more irregular particle shapes. The high-temperature (HT) sample presents a porous surface with fragmented particles that form a distinct porous structure (Figs. 2E,F). This characteristic is likely due to the rapid evaporation of moisture during the hot water extraction process, which contributes to the formation of a more porous surface. Overall, the morphological differences observed across the samples are attributed to variations in the processing temperature. These structural differences may influence the dissolution rate and flowability of the instant powders (Ren et al., 2024). The increased porosity of the HT sample facilitates rapid dissolution, while the denser surface of the RT sample may contribute to better flow stability. Scanning electron microscopy (SEM) analysis. (A) RT,2000✖;(B) RT,5000✖;(C) LT,2000✖;(D) LT,5000✖;(E) HT,2000✖;(F) HT,5000✖. Analysis of electronic nose results The electronic nose sensors generate distinct signal responses that allow for the differentiation of volatile components in food samples. As illustrated in the radar chart (Fig. 3-A), all three pandan leaf instant powder samples prepared at different temperatures demonstrate relatively high sensor responses to W1W (inorganic sulfides), W2W (aromatics and organic sulfides), W5S (nitrogen oxides), W1S (short-chain alkanes), and W2S (alcohols, aldehydes, and ketones). These results suggest that the volatile aroma components in the samples are largely similar across the different temperature treatments. Notably, the HT sample exhibits higher sensor response values compared to the LT sample, while the RT sample shows lower response values for sensors W3S, W1C, W3C, W6S, and W5C. To further analyze the data, principal component analysis (PCA) was applied to reduce the dimensionality and classify the electronic nose signals (Fig. 3-B). The first (PC1) and second (PC2) principal components together account for 97.3 % of the total variance, effectively distinguishing the odor characteristics among the samples. While the three instant powder samples appear to have similar principal component distributions, a degree of separation is still evident along the PC1 axis, indicating that extraction temperature plays a role in influencing the distribution of key volatile aroma components in the pandan leaf instant powders. The RT sample exhibits lower response values across multiple non-polar sensors, which suggests that it may be more effective in preserving highly volatile aromatic compounds. Electronic nose and electronic tongue analysis. (A) electronic nose radar chart;(B) electronic nose PCA plot;(C) electronic tongue radar chart;(D) electronic tongue PCA plot. Analysis of electronic tongue results As shown in the radar chart (Fig. 3-C), significant differences were observed among the three pandan leaf instant powder samples in terms of sourness and saltiness. The HT sample exhibited higher levels of sweetness and umami compared to both the LT and RT samples. In contrast, the HT sample showed lower levels of bitterness, astringency, astringent aftertaste, and bitter aftertaste than the LT and RT samples. These variations in taste characteristics are attributed to the changes in the composition of substances within the raw materials, which are influenced by the temperature treatment during the extraction process. Among the three instant powder samples, the values for umami and umami aftertaste were relatively the highest. The umami taste is primarily associated with the content of amino acids and nucleotides, which are products of protein degradation (Ramesh et al., 2025), while the bitterness is linked to the presence of certain ketones and organic acid compounds in the samples (Shi et al., 2022). Richness, also referred to as umami aftertaste, represents the persistence of the umami flavor (Ninomiya, 2002). The results suggest that the RT sample has a more complex taste profile, characterized by a milder and better-balanced overall flavor. To further analyze the differences in taste characteristics, principal component analysis (PCA) was performed (Fig. 3-D). PC1 and PC2 together explained 78.3 % of the total variance, effectively distinguishing the three groups of samples. The RT and LT samples were positioned on the negative side of PC1, while the HT sample was located on the positive side of PC1. This indicates that the HT extraction process has a positive effect on enhancing the sweetness and umami of the instant powders. The observed differences in taste may be attributed to the production of umami compounds, such as amino acids and nucleotides, during the protein degradation process under heat extraction. Additionally, Maillard reaction products formed during high-temperature treatment may contribute to the enhanced sweet and aromatic perception of the HT sample (Liu, Wang, et al., 2022). UPLC-MS/MS phytochemical analysis of instant powder Metabolite profiling of the three pandan leaf instant powder samples was performed using gas chromatography–mass spectrometry (GC–MS), resulting in the identification of a total of 266 metabolites. These metabolites included 98 alcohols, 36 esters, 19 ketones, 18 aldehydes, 12 acids, 12 amines, 12 hydrocarbons, 7 ethers, 7 alkenes, 6 pyrroles, 5 phenols, 4 pyridines, 1 alkaloid, 1 amino acid, 1 furan, and 27 other compounds (Figs. 4-A,E). GC–MS Analysis of pandan instant powder: (A) bar chart of compound content;(B) peak area chart;(C) relative content chord chart;(D) clustered heatmap of differential volatile compounds;(E) clustered heatmap of volatile compounds. The RT sample exhibited significantly higher peak areas than the LT and HT samples, suggesting that it contained the highest concentration of volatile compounds. As illustrated in Fig. 4-B, all three instant powder samples showed the highest peak intensities for alcohols, with relatively high levels of esters, ketones, and acids. Alcohols, which are the predominant metabolites, play a key role in influencing taste and aroma, and serve as precursors for the synthesis of many aroma-related compounds. These compounds typically impart floral and fruity aromas, contributing to the pleasant scent profile of the powders (Ma Qi et al., 2019). Ester compounds, formed through esterification reactions between acids and alcohols, typically have fruity aromas and also contribute to the overall flavor profile (Guo et al., 2022), Although aldehydes generally have low odor thresholds, their relatively low content does not diminish their significant impact on the overall aroma of the powders. Hydrocarbons, particularly saturated hydrocarbons, tend to have relatively higher odor thresholds and lower concentrations, thereby exerting a weaker influence on flavor. In contrast, unsaturated hydrocarbons possess lower odor thresholds, leading to a more pronounced effect on flavor (J. Xu et al., 2023). Pyrrole compounds, which contribute caramel and roasted aromas, increase in concentration during the Maillard reaction, which is particularly prominent at higher processing temperatures. This reaction enhances the flavor and aroma of the instant powders by producing a series of compounds, including pyrroles, pyrazines, and furans, from the degradation of amino acids (Xiang et al., 2023). The relative content of key metabolites in the samples is presented in Fig. 4-C. The RT sample exhibited the highest relative content of alcohol (44.02 %), while the HT sample also showed a similarly high alcohol content (44.06 %), indicating that high temperature did not significantly affect the alcohol compounds. In contrast, the LT sample contained a significantly lower proportion of alcohol (30.08 %) compared to the other two samples. The ester content in the RT sample was 8.30 %, compared to 2.97 % in the LT sample and 1.02 % in the HT sample. Ketones accounted for 36.53 % in the RT sample, 20.49 % in the LT sample, and 30.69 % in the HT sample. Acids comprised 9.48 % in the RT sample, 14.22 % in the LT sample, and 22.68 % in the HT sample. These findings suggest that alcohol, esters, ketones, and acids are the major volatile compounds present in pandan leaf instant powders. To further elucidate the differences in metabolite composition among the samples, differential metabolites were identified based on VIP values greater than 1 and P values less than 0.05. A clustered heatmap of these metabolites is presented in Fig. 4-D. The clustering results revealed that the RT and HT samples shared a similar compound composition, while the LT sample formed a distinct cluster. Notably, key differential metabolites such as 2-methyltridecane, 3-methylpentadecane, and undecyl ester were significantly enriched in the RT and HT groups, suggesting that these compounds may contribute significantly to the aroma profile. The LT sample, on the other hand, was rich in polar compounds such as formamide, furfurylamine, and isovaleric acid, indicating that the LT method is more favorable for preserving certain polar metabolites. In summary, the RT treatment was found to better preserve the original aroma characteristics of pandan leaves, whereas the HT treatment, despite a loss of some volatile compounds, resulted in the generation of complex new aroma compounds via the Maillard reaction. These reactions enriched the flavor complexity of the product, enhancing its overall sensory profile. LC-MS metabolomics results analysis To further investigate the impact of extraction temperature on the non-volatile metabolic profile of pandan leaf instant powders, a non-targeted metabolomics analysis was performed using ultra-performance liquid chromatography coupled with tandem mass spectrometry (UPLC-MS/MS). The classification of the three distinct pandan leaf instant powder samples was assessed through principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA), which allowed for a comprehensive comparison of metabolic composition across the different temperature treatments. The results of the multivariate analyses revealed significant differences in the metabolic profiles of the samples, with clear distinctions between the three groups, suggesting that variations in processing temperature substantially influenced the flavor characteristics and non-volatile components of the pandan leaf instant powders. The PCA model, shown in Fig. 5-A, demonstrated that the first and second principal components (PC1 and PC2) accounted for 82.3 % and 7.8 % of the variance, respectively. The LT sample formed a distinct cluster, while the RT and HT samples each formed separate, well-defined clusters. These results confirm that the extraction temperature played a crucial role in shaping the metabolic and flavor profiles of the instant powders. Additionally, OPLS-DA analysis, which focused on the content of volatile compounds, provided further insight into the metabolic differences between the samples. Fig. 5-B illustrates a clear separation of the LT, RT, and HT samples based on their volatile compound content, with the model showing a high explanatory variance (R2Y = 0.999) and strong predictive ability (Q2 = 0.7), indicating a reliable classification. To assess the robustness and statistical significance of the model, permutation tests were conducted, yielding a Q2 value of −0.26, which further confirmed the validity of the supervised model (Fig. 5-C). LC-MS analysis of pandan instant powder. (A) PCA analysis;(B) OPLS-DA analysis;(C) permutation test;(D) clustered heatmap of differential non-volatile compounds. Differential metabolites were identified based on VIP values greater than 1.9 and P values less than 0.01, and a total of 88 non-volatile differential metabolites were found across the three samples. As illustrated in Fig. 5-D, the HT group showed a significant increase in amino acid derivatives, which are strongly associated with umami taste and aftertaste, aligning with the results obtained from the electronic tongue analysis. The heatmap analysis of these differential metabolites highlighted variations in their relative abundance, with red indicating metabolites that were present in higher levels than the average, and blue denoting those present in lower levels. Notably, while the overall abundance of non-volatile metabolites in the HT sample was slightly lower than that of the RT and LT samples, this suggests that high-temperature treatment may lead to the degradation or structural changes of some heat-sensitive compounds. Despite this, the HT extraction process significantly enhanced the levels of flavor-related metabolites, particularly amino acids, which contribute to the development of richer and more complex flavor profiles. Metabolomics analysis offers valuable insights into the molecular mechanisms underlying the effects of different extraction temperatures on the flavor and functional components of pandan leaf instant powders. High-temperature (HT) processing was found to promote the accumulation of taste-active metabolites, such as amino acids and nucleotides, which contribute to the enhancement of the umami flavor and overall sensory profile of the product. On the other hand, room-temperature (RT) extraction was more effective in preserving heat-sensitive functional compounds, providing a balance between flavor preservation and functional component retention. Overall, these findings demonstrate that varying extraction temperatures significantly influence both the flavor and nutritional composition of pandan leaf instant powders, allowing for the optimization of processing conditions based on desired sensory and functional outcomes. ROAV analysis of volatile flavor compounds in instant powder To further elucidate the contribution of key flavor components in pandan leaf instant powder, a quantitative analysis of the major volatile compounds was performed, based on their Relative Odor Activity Values (ROAV) (Y. Zhu et al., 2020). The ROAV values range from 0 to 100, with higher values indicating a greater contribution of the compound to the overall aroma profile of the instant powder. Volatile compounds with ROAV values ≥1 are generally considered to be key aroma-active compounds in the samples. Table 2 presents the five compounds with ROAV values ≥1, alongside their corresponding odor thresholds: 3-methylbutanal, 2,3-pentanedione, ethanol, acetic acid, and nonanal. Among these, acetic acid exhibited the highest ROAV of 100 in all three processing conditions (LT, HT, and RT), making it the most prominent aroma compound in the pandan leaf instant powder. Acetic acid contributes a refreshing sour and subtly fruity note to the product's overall flavor. These aroma compounds collectively shape the distinctive aroma profile of pandan leaf instant powder. The results of the ROAV analysis were consistent with the trends observed in the GC–MS quantification, further reinforcing the superior aroma-retention performance of the RT sample. In contrast, HT processing appears to enhance the complexity of the flavor profile, suggesting that the unique flavor characteristics of the instant powder are shaped by the synergistic interaction of multiple components. Table 2Characteristic aroma compounds identified in Pandan leaf instant powder.Table 2NameRIthreshold(ug/Kg)LTROAVHTRT3-methylbutanal926.72222220.00111.511.743.362,3-pentanedione1071.2666670.00035.136.638.72ethanol935.78888890.623.052.868.22Acetic acid1461.80.013100.00100.00100.00nonanal1407.2333330.00311.791.642.45Note: Room temperature (RT), low temperature (LT), and high temperature (HT). Characteristic aroma compounds identified in Pandan leaf instant powder. Note: Room temperature (RT), low temperature (LT), and high temperature (HT). Conclusion This study investigated the impact of three temperature treatments on the structural characteristics and flavor profiles of pandan leaf instant powder. The high-temperature (HT) treatment resulted in greater crystallinity (68.3 %) and a porous microstructure, potentially enhancing both the dissolution rate and the overall stability of the powder. In contrast, the room-temperature (RT) treatment produced smoother particles, which improved flowability. Sensory evaluation using an electronic tongue showed that the HT sample had significantly higher levels of sweetness and umami, along with reduced bitterness and astringency, resulting in a more favorable taste profile compared to other groups. Data from the electronic nose and Relative Odor Activity Value (ROAV) analysis indicated that the RT sample preserved more of the pandan leaf's typical volatile aromatic compounds. Untargeted metabolomics revealed 266 metabolites, with the HT sample showing a 40 % higher abundance of non-volatile compounds, particularly alcohol linked to sweetness and umami. Meanwhile, the RT sample retained 44.02 % of the volatile compounds, maintaining the characteristic pandan aroma. This aroma retention is likely due to the suppression of enzymatic reactions at lower temperatures. Overall, the HT method is better for flavor and solubility, ideal for functional food development, while RT treatment excels in aroma preservation, suitable for aromatic-focused products. This study provides a theoretical foundation and technical support for the deep industrial processing, quality optimization, and utilization of pandan leaf instant powder and other aromatic plants. CRediT authorship contribution statement Chengxu Liu: Data curation. Zhen Zhang: Methodology. Yuqi Xin: Conceptualization. Yuanyuan Zhang: Investigation. Yiyun Wu: Investigation. Chuanrui Luo: Investigation. Peishi Liu: Investigation. Yewei Huang: Supervision. Cunchao Zhao: Supervision. Ya Wang: Supervision. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements This study received support from the Wuzhishan Municipal Government. Footnotes Supplementary data to this article can be found online at https://doi.org/10.1016/j.fochx.2025.103327. Contributor Information Yewei Huang, Email: lichuangyewei100@163.com. Cunchao Zhao, Email: zcc272744372@163.com. Ya Wang, Email: wangya9188@126.com. Supplementary data Supplementary material Data availability The authors do not have permission to share data. References Associated Data Supplementary Materials Supplementary material Data Availability Statement The authors do not have permission to share data.