Primary material used in this study consisted of young fruits of S. taccada collected from Trikora Beach, Bintan Island, Indonesia (1°07′50″ N, 104°37′35″ E). A total of 30 young fruit samples were collected for morphometric characterization. The collected samples were washed, cut into small pieces, air-dried, and subsequently oven-dried at 40 °C for 48 h using a Memmert oven to reduce the moisture content while preserving the thermolabile compounds. The dried samples were ground into a fine powder using a Miyako blender and stored in airtight containers at room temperature before extraction.

Powdered samples of young S. taccada fruits were subjected to bioactive compound extraction using a sequential maceration method. Sequential extraction was selected to fractionate metabolites according to their relative polarity, thereby reducing the complexity of individual extracts and facilitating comparative evaluation of solvent dependent metabolite distribution. This approach enables the progressive recovery of nonpolar, semipolar, and polar constituents, providing a broader representation of the phytochemical composition than singlesolvent extraction (Bitwell et al., 2023). Briefly, 50 g of each powdered sample was successively extracted with analytical grade n-hexane, ethyl acetate, and ethanol (200 mL; 1:4 w/v for each solvent), applied in order of increasing polarity, so that non-polar constituents were removed first, thereby minimizing their interference with the recovery of more polar metabolites in subsequent extraction steps. The 1:4 w/v ratio and 48 h maceration period were adopted based on established protocols for exhaustive metabolite recovery from plant matrices. Each extraction step was performed once for 48 h at room temperature with intermittent stirring to facilitate the solvent penetration and metabolite difusion. After each extraction cycle, the mixture was filtered using Whatman No. 42 filter paper to separate the filtrate from the residue. The residue was subsequently re-extracted using solvents of increasing polarity. All extractions were performed in triplicate to ensure reproducibility. The obtained filtrates were concentrated under reduced pressure using a vacuum evaporator (INTBUYING RE-201) at to remove the solvents and obtain the crude extracts. The resulting semisolid extracts were weighed to determine the extraction yield and stored at until further analysis. However, sequential extraction does not achieve complete separation of metabolite classes because compounds with intermediate polarity may be distributed across more than one solvent fraction. Therefore, the resulting fractions should be interpreted as relative enrichments of metabolites rather than chemically pure fractions.

Antioxidant activity (DPPH Assay) was determined following the method described by Gulcin & Alwasel (2023). The extracts were dissolved in methanol to obtain concentrations of 75-375 ppm, and ascorbic acid (positive control) was prepared at 5-25 ppm. A 1 mM DPPH solution was then prepared. Briefly, 4.5 mL of the sample or standard solution was mixed with 500 µL of DPPH and incubated at for 30 min (Memmert IN 55). The absorbance was measured at 517 nm using a UV-Vis spectrophotometer (UV-1800; Shimadzu). Methanol DPPH solution was used as a blank. All assays were performed in triplicate (n=3). Radical scavenging activity was calculated as the percentage of inhibition of DPPH.

values were determined from the linear regression of the percentage of inhibition versus concentration , where represents the concentration required to inhibit 50% of the DPPH radicals. The coeficient of determination (R²) was calculated for each regression model to evaluate the goodness of fit.

NMR Measurements and data analysis were performed following the procedure described by Kim et al. (2010). Because the ethanol extract exhibited the highest extraction yield and the strongest DPPH radical scavenging activity among the three solvent fractions, it was selected as the representative extract for NMR-assisted phytochemical characterization to obtain preliminary chemical information on metabolites potentially associated with antioxidant activity. NMR spectra were acquired on a JEOL JNM-ECZ500R/S1 spectrometer (¹H at 500 MHz, at 125 MHz) at 298 K. Each extract (ca. 20 mg) was dissolved in 600 µL methanol-d₄ (99.8% D, Sigma-Aldrich). ¹H NMR parameters: spectral width 15 ppm, 64 scans, relaxation delay 2.0 s, 32k data points. NMR parameters: spectral width 220 ppm, 10,000 scans, 1.0 s relaxation delay. Chemical shifts were referenced to residual solvent signals (δH 3.31, δC 49.0 ppm for methanol-d₄). Spectral processing was performed using the JEOL Delta v5.3 software with exponential line broadening (0.3 Hz for ¹H and 1.0 Hz fo . Metabolite assignments were based on a comparison of the and NMR spectral characteristics with previously reported data. Therefore, compound identification should be considered tentative and requires furthe validation using complementary analytica techniques such as 2D NMR or LC-MS/ MS. Because only one-dimensional 1H and NMR spectra were acquired, metabolite assignments were based on characteristic chemical shifts and comparisons with published spectral databases. Consequently, all reported metabolite identifications should be regarded as tentative annotations rathe than definitive structural identifications.

The experiment was arranged in a Completely Randomized Design (CRD) consisting of three treatments with three replicates. Before analysis, the data were tested for normality using the Shapiro–Wilk test and homogeneity of variance using Levene’s test. Diferences among treatments were evaluated using one-way analysis of variance (ANOVA).

When significant diferences were detected (p < 0.05), mean separation was performed using Duncan’s multiple range test (DMRT). All statistical analyses were performed using SPSS software (version 29; IBM Corp., Armonk, NY, USA). Statistical analyses were performed to compare the extraction yields and antioxidant activities of the solvent fractions. No formal correlation analysis was performed between metabolite composition and antioxidant activity because metabolite identification was qualitative and based on one-dimensional NMR profiling.

Young fruits of S. taccada were collected from Trikora Beach on Bintan Island in Indonesia. The morphological characteristics of the young fruits are shown in Figure 1. In this preliminary study, morphological observations focused on fruit diameter as the primary indicator of the developmental stage, whereas other morphometric parameters, such as fruit weight, length, peel thickness, and moisture content were not measured. These parameters are recommended for inclusion in future studies to provide a more comprehensive morphological characterization and improve reproducibility.

The young fruits of S. taccada were characterized by a green color, firm texture, and an average diameter of 0.77±0.12 cm, indicating an early developmental stage with high metabolic activity. At this stage, parenchymal tissues actively synthesize secondary metabolites, particularly phenolics, which commonly accumulate during early growth as part of the plant defense mechanism (Arunachalam, 2013; Yadav et al., 2021; Singh et al., 2022; Hilal et al., 2024). The predominance of polar secondary metabolites is particularly important because their polar nature strongly influences the extraction eficiency, which depends on solvent polarity. This green coloration further reflects active photosynthetic metabolism and the activation of the phenylpropanoid pathway, which regulates phenolic biosynthesis (Dai & Mumper, 2010; Singh et al., 2022; Herwibawa, 2025). In coastal environments, this pathway is often enhanced by salinity and UV stress, resulting in increased accumulation of antioxidant-related compounds (Ksouri et al., 2007; Mohammed et al., 2023). It should be noted that this interpretation of pigmentation, phenylpropanoid activation, and antioxidant accumulation is drawn from general plant physiological literature rather than from direct measurements of photosynthetic activity or phenylpropanoid pathway markers in the present study. These observations are consistent with the possibility that young S. taccada fruits contain relatively high proportions of polar metabolites, although their abundance was not quantitatively determined in the present study. The extraction yields of each fraction are shown in Figure 2.

Ethanol produced the highest extraction yield (1.173±0.003%), followed by ethyl acetate (0.727±0.003%), and n-hexane (0.227±0.0001%). This pattern indicates that the extractable metabolites in young S. taccada fruits are predominantly polar to semi-polar compounds, which exhibit greater solubility in ethanol because of their oxygenated functional groups (Dai & Mumper, 2010; Dirar et al., 2019; Nawaz et al., 2020; Ma et al., 2022; Mulyono et al., 2022; Emu et al., 2023). The predominance of ethanol-soluble metabolites observed in this study is consistent with previous findings for S. taccada. Nasution et al. (2026) reported that the methanolic extracts of S. taccada leaves and fruits contained higher levels of phenolic and flavonoid compounds than the less polar fractions, indicating that polar solvents are generally more efective for recovering antioxidant-associated metabolites from this species. Similar solvent-dependent extraction patterns have also been reported in coastal halophytes such as Ipomoea pescaprae and Cressa cretica, where polar solvents preferentially recover phenolic-rich fractions with higher antioxidant activity (Qasim et al. 2017; Nazir et al. 2018).

Importantly, this polarity-dependent yield may reflect diferences in the relative distribution of extractable metabolites among solvent fractions. The higher ethanol yield is consistent with the greater recovery of polar metabolites in the extractable fraction of young S. taccada fruits, potentially associated with antioxidant activity, in line with reports linking S. taccada activity to phenolic constituents (Apriandi et al., 2021; Fatmawati et al., 2021). In contrast, the low yield in n-hexane indicates a limited contribution of lipophilic compounds, whereas the intermediate ethyl acetate fraction indicates the presence of semi-polar metabolites, which may include phenolic constituents (Azmir et al., 2013; Do et al., 2014; Nawaz et al., 2020; Tripathi et al., 2025; Azzahra et al., 2025). This polarity-driven partitioning is further supported by the plant’s coastal adaptation, in which environmental stress enhances phenolic accumulation (Ksouri et al., 2007; Jacob et al., 2019; Shekhawat et al., 2021; Emura et al., 2022). Consequently, the observed yield diferences provide a meaningful basis for NMR assisted phytochemical characterization, as each fraction represents a distinct chemical profile that may provide a useful basis for subsequent comparison with antioxidant activity, although no formal statistical relationship was evaluated in the present study.

The antioxidant activity of extracts derived from the young fruits of S. taccada was evaluated using the DPPH radical-scavenging assay. Antioxidant performance was assessed using values as an indicator of radical scavenging potency (Table 1), whereas concentration-dependent inhibition profiles were expressed as percentage inhibition (% inhibition) at diferent concentrations of the extracts (Figure 3).

The antioxidant activity of extracts derived from young fruits of S. taccada was assessed using the DPPH radical scavenging assay, with values used as an indicator of free radical inhibition eficiency (Table 1). The concentration-dependent inhibition profiles are presented in Figure 3. Based on Table 1 and Figure 3, the antioxidant activity of young S. taccada fruit extracts difered significantly among the solvent fractions . Linear regression analysis of DPPH inhibition against extract concentration produced the following equations: for the n-hexane extract, 0.1567x+21.43 for the ethyl acetate extract, and y=0.1563x+32.20 for the ethanol extract. These high coeficients of determination indicate a strong concentrationdependent increase in radical scavenging activity across all solvent fractions. The ethanol extract exhibited the moderate antioxidant activity , followed by the ethyl acetate extract ppm), whereas the n-hexane extract exhibited weak antioxidant activity (2,965.25±1,966.56 ppm). This polarity-dependent trend is consistent with the preferential recovery of metabolite classes that are commonly associated with antioxidant activity.

Based on the DPPH classification (very strong <50 ppm, strong 50–100 ppm, moderate 100–150 ppm, weak150–200 ppm, and very weak >200 ppm) (Molyneux, 2004), the ethanol, ethyl acetate, and n-hexane extracts fell within the moderate, weak, and very weak categories, respectively.

A simple comparison between the extraction yield and antioxidant activity revealed an inverse relationship between the extraction eficiency and values. Fractions with higher extraction yields exhibited stronger antioxidant activity, with the ethanol fraction showing the highest yield and lowest value. Although formal correlation analysis was not performed because only three solvent fractions were evaluated this trend suggests that solvent polarity influences not only the extraction eficiency but also the recovery of antioxidant-associated metabolites. Nevertheless, this relationship should be interpreted cautiously because the total phenolic and flavonoid contents were not quantified in this study. Therefore, the observed relationship should be considered a qualitative trend rather than a statistically validated correlation.

The antioxidant activity observed in the present study is consistent with previous reports on S. taccada. Apriandi et al. (2021) reported an value of 130.71 ppm for young fruit extracts, whereas ethanol extracts of leaves and stems exhibited values of 109.81 and 87.15 μg/mL, respectively. These observations suggest that the antioxidant capacity of S. taccada varies among plant organs and may reflect diferences in metabolite accumulation patterns. Similar organ-dependent variation has also been reported by (Budiana et al. (2019) and Nasution et al. (2026), who observed values of and μg/ mL in leaf and stem extracts, respectively, accompanied by measurable phenolic and flavonoid contents. Comparable patterns have been reported in several medicinal halophytes, including Atriplex stocksii, Cressa cretica, and Ipomoea pes-caprae, where elevated phenolic and flavonoid contents are associated with stronger antioxidant activities (Qasim et al., 2017; Nazir et al., 2018).

The exceptionally large standard deviation observed for the n-hexane fraction likely reflects the very low inhibition values obtained across the tested concentration range. Because the inhibition percentages did not approach 50%, the estimate relied on substantial extrapolation from the regression model, reducing precision. Therefore, the reported should be interpreted mainly as an indicator of very weak radical scavenging capacity rather than a highly precise quantitative endpoint.

The concentration-dependent inhibition profiles shown in Figure 3 further support the influence of solvent polarity on the antioxidant performance. Increasing the extract concentration resulted in progressively higher DPPH inhibition percentages for all the solvent fractions. The ethanol extract consistently produced the highest inhibition values, increasing from 40.91±0.84% at 75 ppm to 87.94±2.15% at 375 ppm. Similarly, the ethyl acetate extract showed a marked increase from 27.05±7.88% to 76.90±1.52%, whereas the n-hexane extract exhibited limited inhibition, ranging from 7.29±6.77% to 13.84±6.43% across the tested concentrations. These findings demonstrate a clear dosedependent response and indicate that polar extracts possess substantially greater free radical-scavenging capacity than non-polar extracts. The stronger antioxidant activity of the ethanol fraction is likely associated with the preferential extraction of polar oxygenated metabolites, which are commonly reported as radical-scavenging constituents in plant extracts (Dai & Mumper, 2010; Nawaz et al., 2020; Ma et al., 2022; Azzahra et al., 2025; Tripathi et al., 2025). However, this interpretation remains inferential because the total phenolic and flavonoid contents were not quantified in the present study.

Previous studies on S. taccada fruits have also demonstrated the efects of developmental stage on antioxidant capacity. Fatmawati et al. (2021) reported that juice prepared from young fruits exhibited stronger antioxidant activity ppm) than that obtained from mature fruits , whereas mixed-stage fruits showed intermediate activity (35.52 ppm). The higher antioxidant activity in young fruits was accompanied by an elevated vitamin C content (20.24 mg), suggesting that antioxidant-related metabolites accumulate during early fruit development. Although these previously reported values. were substantially lower than those obtained in the present study, and several methodological diferences may explain this discrepancy. Fruit juice contains naturally occurring water-soluble antioxidants, including ascorbic acid, which may not be recovered in equivalent proportions after sequential solvent extraction. Therefore, diferences in sample preparation, extraction procedures, and analytical approaches may contribute to the observed variations in antioxidant activity.

Interestingly, preliminary qualitative phytochemical screening did not detect phenolics or flavonoids in these extracts. This apparent discrepancy may reflect the limited sensitivity of conventional qualitative screening methods when metabolites are present at low concentrations or in complex structural forms. The observed antioxidant activity, particularly in the ethanol fraction, suggests the presence of antioxidantassociated metabolites that require further chemical characterization to confirm their presence. Moreover, qualitative phytochemical screening generally exhibits lower sensitivity than instrumental approaches and may fail to detect compounds present at low concentrations or in their glycosylated forms.

To obtain preliminary information regarding metabolites potentially associated with the antioxidant activity observed in the ethanol fraction, NMR-assisted phytochemical characterization was performed on the ethanol extract, which exhibited the strongest DPPH radical scavenging activity and the lowest value among all solvent fractions. The ethanol extract exhibited a concentrationdependent increase in DPPH inhibition, reaching ppm, with an value of . These observations are consistent with the presence of polar metabolites in the ethanol fraction; however, the present data do not establish a causal relationship between these metabolites and their associated antioxidant activity. Therefore, ¹H and NMR analyses were conducted to obtain preliminary information regarding the chemical characteristics of the metabolites present in this bioactive fraction (Figure 4 and 5, respectively).

The ¹H NMR spectrum of the ethanol extract was dominated by intense resonances within the δH 3.1-4.0 ppm region, while the corresponding signals were observed at ppm. These chemical shift regions are commonly associated with oxygenated protons and carbon environments in carbohydrates and glycosylated metabolites. These metabolites frequently accumulate in developing fruits as carbon reserves and metabolic intermediates required for growth and cellular metabolism (Emwas et al., 2019; Chen et al., 2024; Dutta et al., 2025; Joo et al., 2025; Liu et al., 2025). In addition, a prominent signal observed at approximately δH 4.93 ppm is consistent with the proton environments commonly reported for carbohydrate-derived structures and glycosylated metabolites (Emwas et al., 2019; Kim et al., 2023).

The tentative assignments corresponding to the major ¹H and resonances in Figure 4 and 5 are listed in Table 2. An important observation from the NMR spectra is that the signals associated with carbohydrates were considerably more abundant than those associated with secondary metabolites. This finding suggests that primary metabolites constitute a substantial proportion of ethanol extracts. Therefore, although antioxidant-associated secondary metabolites may be present, their occurrence cannot be inferred solely from signal intensity because one-dimensional NMR does not provide suficient resolution to distinguish all overlapping constituents in complex plant extract. In addition to the carbohydrate-associated signals, several resonances were detected in the aliphatic region. Signals observed at δH 0.8-2.0 ppm and δC 14.57-39.04 ppm are characteristic of aliphatic proton and carbon environments commonly reported in sterol-, lipid-, and terpenoid-related metabolites (Dutta et al., 2025; Zhang et al., 2025). Similar resonances have been observed in plant extracts containing phytosterols and other hydrophobic secondary metabolites. Although these signals were less intense than those observed in the carbohydrate region, their presence indicates that the ethanol extract contains structurally diverse metabolite classes beyond primary carbohydrates.

The NMR spectrum also revealed signals at and 176.09 ppm.

Resonances within the 120-140 ppm region are commonly associated with unsaturated carbon environments, including olefinic and aromatic carbons, whereas signals near 170-180 ppm are typically attributed to carbonyl-containing functionalities, such as esters, carboxylic acids, or related oxygenated metabolites (Kim et al., 2010; Emwas et al., 2019; Dutta et al., 2025). The occurrence of these signals suggests the presence of metabolites possessing unsaturated and oxygenated structural features, although their exact identities cannot be determined from one-dimensional NMR data alone.

Unlike previous studies that primarily reported phytochemical classes through conventional phytochemical screening or crude extract analysis, the present study provides a solvent-specific NMR-based metabolite distribution in young fruits, highlighting the developmental stagespecific chemical characteristics. Previous phytochemical studies on S. taccada have reported the presence of flavonoids, flavonoid glycosides, terpenoids, alkaloids, and phenolic constituents in diferent plant organs, including leaves, stems, and fruits (El-Sayed et al., 2020; Jasna et al., 2025; Nasution et al., 2026). However, information on the chemical composition of young fruits is limited. The predominance of oxygenated signals observed in the present study is consistent with the tendency of developing fruits to accumulate carbohydrate-derived and glycosylated secondary metabolites during active growth. Therefore, the current findings contribute specifically to the understanding of metabolite distribution in young fruits, rather than extrapolating chemical characteristics from other plant organs. Overall, the NMR results provide a plausible chemical context for interpreting the antioxidant activity of ethanol extracts. However, because metabolite assignments were tentative and no quantitative phytochemical analysis or metabolite bioactivity correlation was performed, the contribution of individual metabolite classes should be regarded as a qualitative hypothesis requiring further confirmation (Dai & Mumper, 2010; Gulcin & Alwasel, 2023).

Nevertheless, the current data do not allow the direct attribution of antioxidant activity to specific metabolites because the spectra are dominated by carbohydrateassociated resonances and only onedimensional NMR data are available. The predominance of carbohydrateassociated resonances is consistent with the developmental stage of young fruits, during which primary metabolism remains highly active to support cell division, tissue diferentiation, and subsequent maturation. Consequently, dominant carbohydrate resonances are expected and should be regarded as one of the principal biochemical characteristics of young S. taccada fruits rather than a limitation of the extraction process (Yadav et al., 2021; Singh et al., 2022). Importantly, the predominance of carbohydrate signals does not preclude the presence of secondary metabolites. Additional resonances assigned to aliphatic, unsaturated, carbonyl-containing, and oxygenated compounds indicate that structurally diverse secondary metabolites were also present in the ethanol extract, although at a relatively lower abundance or partially overlapping with intense carbohydrate signals (Kim et al., 2010; Emwas et al., 2019). Within this biochemical context, these less abundant metabolites provide a plausible chemical basis for the observed radical-scavenging activity (El-Sayed et al., 2020; Nasution et al., 2026), reinforcing the qualitative interpretation noted above, rather than establishing a new conclusion.

Similar observations have been reported in Ipomoea pes-caprae and other coastal halophytes, where elevated phenolic and flavonoid contents are associated with stronger antioxidant activity (Qasim et al., 2017; Nazir et al., 2018; Budiana et al., 2019; Nasution et al., 2026). Although the total phenolic and flavonoid contents were not quantified in the present study, the stronger antioxidant activity of the ethanol fraction is consistent with the possibility that polar oxygenated metabolites contributed to the observed radical scavenging activity. However, this requires confirmation through quantitative phytochemical analyses and compound identification. Although metabolite assignments remain tentative, these findings extend our current knowledge of the chemical characteristics of this underexplored developmental stage and provide a basis for future metabolomic studies.

In addition, the DPPH assay represents a simplified chemical radical-scavenging model and does not directly reflect the biological antioxidant eficacy under physiological conditions. Therefore, the antioxidant activity reported in this study should be interpreted as an indication of in vitro radical scavenging potential rather than direct evidence of the in vivo antioxidant or pharmacological efects of the extracts. This study had several limitations. Because the 1D NMR spectrum is dominated by carbohydrate-associated resonances, metabolite assignments should be regarded as tentative chemical class annotations rather than definitive compound identifications. Furthermore, the total phenolic and flavonoid contents were not quantified, preventing a direct correlation analysis between the antioxidant activity and metabolite abundance. Consequently, the proposed relationship between solvent polarity, metabolite distribution, and antioxidant activity should be interpreted qualitatively. Future studies incorporating LC-MS/MS, 2D NMR, TPC/ TFC assays, and multivariate analyses are required to validate these observations.