Rapid On-Site Detection of Ginsenosides in Panax ginseng Using Portable Miniature Mass Spectrometry by Sodium Acetate-Enhanced Nanoelectrospray Ionization and a Same-Spectrum Same-Screening Quantification Strategy

Abstract

Ginsenosides are the major bioactive constituents in Panax ginseng (PG), exhibiting diverse pharmacological effects, including anticancer, neuroprotective, cardioprotective, and anti-inflammatory activities. On-site mass spectrometry (MS) analysis of ginsenosides plays a crucial role in PG quality control. However, owing to matrix effects, structural complexity, low volatility, poor proton affinity, and the presence of multiple glycosidic moieties, rapid analysis of ginsenosides using portable MS under ambient conditions remains challenging. Here, we developed a sodium acetate (NaOAc)-enhanced nanoelectrospray ionization (nESI) method enabling rapid, on-site ginsenoside profiling using miniature mass spectrometry (mMS). NaOAc-mediated cationization promotes efficient formation of [M + Na]⁺ ions, enhancing ion signal intensity. Ginsenosides were detected using full-scan MS and MS/MS, including Rb1, Rg1, Re, Rd, Rf, notoginsenoside R1, etc. For quantitative analysis of the three pharmacopoeial markers (Rb1, Rg1, and Re), a same-spectrum and same-screening (4S) strategy using structurally similar saponins as an internal standard was established. The method exhibited good linearity (5–100 μg·mL^–1, R ² > 0.99), acceptable precision (RSD < 13%), and recoveries. The limits of detection (LOD) for Rb1, Rg1, and Re were 0.10–0.20 μg·mL^–1, respectively. Application of the method to PG samples of different cultivation ages revealed that the Rb1/Re ratio strongly correlated with cultivation age. This work demonstrated a practical strategy to enhance ionization efficiency and quantitative robustness in mMS and provided a feasible approach for on-site chemical assessment of herb medicines.

1. Introduction

On-site, rapid, and reliable chemical analysis in field environments has become an increasingly important objective in analytical chemistry, , with significant applications in food safety, environmental surveillance, forensic science, clinical diagnostics, and quality control of Traditional Chinese Medicine (TCM). , Compared with laboratory analysis of field samples, on-site analysis allows for timely decision-making, saving time and resources, and avoiding sample degradation or contamination during transportation and storage. − Portable miniature mass spectrometry (mMS) is one of the most powerful analytical tools for this purpose, owing to its high portability, practical operability, and excellent chemical specificity. − In recent years, substantial progress has been made in the miniaturization of mass spectrometry systems. , Novel ion trap configurations and optimized ion transmission designs , have been developed, significantly enhancing sensitivity, fragmentation capabilities, and field applicability. Furthermore, artificial intelligence-driven automated data processing and intelligent control have expanded its potential for point-of-care testing applications. These advancements indicate that miniature mass spectrometry is transitioning from the laboratory to the field, providing a solid technological foundation for rapid and accurate on-site chemical analysis.

Panax ginseng (PG) has beneficial pharmacological effects in inflammatory and metabolic diseases, , and its economic value is positively correlated with growth age. These benefits are primarily attributed to ginsenosides, a unique class of triterpenoid saponins that serve as the principal bioactive constituents. − During plant development, both the composition and relative abundance of ginsenosides undergo dynamic changes. − These variations are driven by temporally regulated expression of key genes involved in the ginsenoside biosynthetic pathway and are further modulated by environmental factors such as temperature, photosynthetically active radiation (PAR), and soil water potential. − Such coordinated genetic and ecological influences contribute to the regional authenticity and characteristic chemical profiles of ginseng. Therefore, rapid and real-time discrimination of age-dependent ginsenoside accumulation patterns is of great importance for quality evaluation, market regulation, and value assessment of ginseng products.

However, ginsenosides are large and neutral triterpenoid saponins, lacking basic or acidic functional groups, and thus exhibit extremely poor proton affinity ([M + H]⁺) or deprotonation ([M – H]⁻) efficiency, , resulting in very low sensitivity that is often below the detection limits of portable instruments. Additionally, crude herbal matrices contain abundant endogenous components such as phospholipids, saccharides, and organic acids that induce substantial ion suppression. , These factors collectively hinder sensitive and reproducible detection of ginsenosides using mMS platforms.

For ionization enhancement, metal-adduct formation is a proven strategy to improve detection of neutral and weak acidic molecules, including saponins. − Sodium cationization can dramatically improve MS response. In this work, we introduce a NaOAc-enhanced nanoelectrospray ionization (nESI), enabling highly efficient sodium adduct formation of ginsenosides within a portable mMS workflow. This charge-directed or cationization-mediated ionization not only amplifies signals by several orders of magnitude but also suppresses competing ionization channels, thereby enhancing spectral simplicity and selectivity. The resultant diagnostic [M + Na]⁺ ions serve as markers, enabling the first successful profiling of ginsenosides directly from crude extracts using a miniature MS.

In addition, we established the same-spectrum and same-screening (4S) ginsenoside quantitative strategy incorporating internal standard normalization and optimized collision-induced dissociation conditions to ensure reproducible MS/MS-based quantification without chromatographic separation. The analytical performance of the developed method was systematically validated, including linearity, sensitivity, precision, and recovery. By applying the method to the rapid assessment of ginseng age, we verified its practical applicability and introduced compound ratio indicators to enhance the robustness against instrumental bias. Overall, this work provides a practical framework for enhancing ionization efficiency and quantitative reliability in mMS and demonstrates its potential for the rapid, on-site assessment of ginseng.

2. Materials and Methods

2.1. Chemicals and Materials

Ginsenoside standards (Rg1, Re, Rb1, Rb2, Rb3, Rc, Rd, Rf, 20S-Rg2, 20R-Rg2, 20S-Rh1, 20R-Rh1, 20S-Rg3, 20R-Rg3, Rg4, Rg6, and Noto-R2, purity >98%) were purchased from Chengdu Alpha Biotech Co., Ltd. (Chengdu, China). Gypenoside LXXV, vinaginsenoside R8, and mogroside IVa (purity >98%) were purchased from Yuanye Bio-Technology Co., Ltd. (Chengdu, China). Acetonitrile (ACN), methanol (MeOH), and formic acid (FA, MS grade) were purchased from Fisher Scientific (Pittsburgh, PA, U.S.A.). Sodium acetate (NaOAc, ACS grade) was purchased from Sigma-Aldrich (St. Louis, MO, U.S.A.). Nano-ESI capillary cartridges were obtained from PURSPEC Technology (China) Ltd. Different-aged PG roots were collected from Jilin and Liaoning Province. A systematic evaluation of additive types and concentrations, sample-to-solvent ratios, and ultrasonication times was conducted to determine optimal pretreatment parameters.

2.2. Preparation of PG Samples

Fresh ginseng roots were washed thoroughly, then placed in a forced-air drying oven and dried at 40 °C to constant weight. The dried samples were ground and passed through a No. 4 sieve to obtain PG powder, which was stored in a desiccator until analysis. PG powder (1 g) was extracted with 5 mL MeOH via 0.5 h ultrasonication. The mixture was filtered through a 0.45 μm membrane to obtain the crude extract. Taking an appropriate volume to combine with MeOH containing NaOAc in a suitable ratio to prepare the final analysis solution.

2.3. Miniature Mass Spectrometer for Identification

All experiments were performed on a Cell Portable Mass Spectrometer System (PURSPEC Technology (China) Ltd), which utilizes a linear ion trap (LIT) mass analyzer (8.5 kg). Mass axis calibration is performed using the nESI tip holder (high-mass calibration) (PURSPEC Technology (China) Ltd). The nESI tuning mix of high mass is supplied with the nESI tip holder. Ions were introduced through a discontinuous atmospheric pressure interface (DAPI). For sample loading, the sample solution was directly pipetted into the rear end of the nanoelectrospray (nESI) capillary using a micropipette. The solution automatically filled the emitter tip by capillary action, without the need for an external syringe pump or injection valve. The tuning mix was loaded into the integrated nESI tip, then the loaded tip was inserted into the tip holder. Finally, the assembled nESI cartridge was inserted into the instrument inlet to perform one-click mass axis calibration. Instrument parameters were positive ion mode; spray voltage, +2.0 kV; scan range, m/z 500–1300; injection volume was 80. For MS/MS analysis, precursor ions were isolated using SWIFT waveforms and fragmented using in-trap collision-induced dissociation (CID). The normalized collision energy (NCE) was systematically varied from 0 to 7 (step size = 0.5) to construct fragmentation profiles for each ginsenoside. Data was collected using PmsClientPro (v4.2.2.1), and mass axis calibration was performed before each set of experiments.

2.4. Miniature Mass Spectrometer for Quantification

Three pharmacopoeial marker ginsenosides (Rb1, Rg1, Re) were selected as target analytes. To correct signal fluctuation arising from ambient spray instability, structurally similar saponins with comparable molecular weight (mass difference < 50 Da) but distinct fragmentation behavior were used as internal standards (Table S1). Each internal standard was spiked at a fixed concentration into calibration standards and samples. The spray voltage was +2.0 kV, the scan range was m/z 500–1300, and the injection time was 80. The CID energy (V) was set to 3.5 for Rb1, 3.0 for Re, and 2.5 for Rg1. Quantification was performed using peak-area ratios of analyte vs internal standard in the MS² mode. Calibration curves were constructed using concentration levels. Methodological validation was conducted in accordance with the requirements of the Chinese Pharmacopoeia.

2.5. UPLC-MS Analysis

The data obtained from mMS were validated by using the UPLC-MS method previously established by our group. UPLC-MS analysis for detection was performed using a SHIMADZU CBM-30A UPLC system (Kyoto, Japan) coupled with an AB SCIEX Triple TOF 7600 mass spectrometer (CA, USA). UPLC-MS analysis for quantification was performed using a SHIMADZU CBM-30A UPLC system coupled with an AB SCIEX Triple Quad 4500 mass spectrometer (CA, USA).

3. Results and Discussion

3.1. Optimization of Ionization Conditions for Ginsenoside Detection

Efficient ion generation is the central determinant of MS performance, particularly in mMS systems, where the available ion current is inherently limited. Ionization has been shown to influence the intensity of analytes. Three representative ginsenosides (Rb1, Rg1, and Re) were selected as model analytes for method optimization.

First, paper capillary spray (PCS) and nanoelectrospray ionization (nESI) were compared. It was found that nESI provided higher ion transmission efficiency, superior stability, and lower signal drift (Figure S1). Subsequently, three compounds were subjected to MS1 scanning using nESI under both positive- and negative-ion conditions to determine the optimal ionization mode. Under negative-ion conditions, no characteristic ion peaks were detected. Under positive ion conditions, the [M + H]⁺ ion was observed, but its abundance was low (Figure S2).

To enhance the ionization efficiency, five commonly used additives (formic acid, acetic acid, ammonium formate, ammonium acetate, and sodium acetate (NaOAc)) were examined. Acidic and ammonium reagents produced negligible improvement in [M + H]⁺ and [M + NH4]⁺ abundance, confirming the intrinsically poor protonation efficiency of ginsenosides. In contrast, NaOAc generated abundant sodium adduct ions ([M + Na]⁺) for all three representative analytes, increasing MS1 signal-to-noise (S/N) ratios (Figure S3). Then, NaOAc concentrations ranging from 0.1 to 5 mM were investigated. Increasing NaOAc concentration beyond 1 mM did not improve ion abundance, consistent with saturation of available coordination sites on the ginsenoside molecules (Figure ). Considering both the analytical performance and instrument cleanliness, MeOH-containing 1 mM NaOAc was selected as the optimal ionization solvent. This solvent system provided strong and stable [M + Na]⁺ formation while minimizing salt deposition inside the mMS system.

1.

Effect of different concentrations of sodium acetate on the signal responses of ginsenosides Rb1, Rg1, and Re (A) and the corresponding visualization chart (B).

3.2. Detection of Ginsenosides in PG Samples

Under the optimized ionization, the mMS workflow was applied to the direct analysis of PG samples. A total of 11 ginsenosides were observed across the different PG samples. Their [M + Na]⁺ adduct ion and product fragments are shown in (Figure ). Fragmentation patterns were verified by UPLC-Q-TOF-MS/MS (Figures and S4). The same sample was analyzed consecutively 6 times, and the relative standard deviation (RSD) of the intensity for the target ginsenosides ranged from RSD 4.84–11.72% (Figure S5), demonstrating acceptable reproducibility for field-applicable mMS detection.

2.

Identification of ginsenosides in ginseng extract using mMS.

3.

Fragmentation patterns of ginsenosides Rb1, Rg1, and Re under the positive ion mode obtained by UPLC-QTOF-MS/MS, illustrating the major product ions.

Notably, structural isomeric ginsenosides that share identical molecular masses could be unambiguously differentiated by their distinct fragmentation signatures. For instance, Rg2 and Rg3, both yielding [M + Na]⁺ at m/z 807.4, exhibited markedly different product ion spectra. Rg2 generated a characteristic fragment ion at m/z 349.1 corresponding to the glucose-rhamnose group, whereas Rg3 produced a diagnostic aglycone fragment at m/z 365.1 arising from the glucose–glucose group (Figure S6). These structurally informative fragmentation differences enabled confident discrimination of co-occurring structural isomers without chromatographic separation, highlighting the capability of nESI–mMS for rapid and reliable characterization of structurally similar ginsenosides directly from complex PG matrices. However, this method could not distinguish stereoisomers (Figure S7).

3.3. Development of the 4S Quantitative Method

Quantification is essential for the PG quality assessment and clinical monitoring. The mMS system was configured to quantify three pharmacopoeia-listed marker compounds (Rb1, Rg1, and Re). To correct for variations in spray efficiency and environmental conditions, internal standards (ISs) with similar physicochemical properties and ionization behavior and with mass differences within 50 Da were selected to develop the same-spectrum, same-screening (4S) method. All IS produced stable [M + Na] ⁺ ions without overlapping fragment ions, enabling accurate co-detection and normalization (Figure S8). Using the 4S method, we precisely quantified Rb1, Rg1, and Re.

3.3.1. Selection of Ion Pairs and Energy

Ginsenosides Rg1 and Re have isomers such as Rf and Rd. Since the mMS system lacks chromatographic separation, the selectivity of MS/MS is crucial. As shown in Figure S9, analysis of the reference standards demonstrated distinct fragmentation patterns, with ginsenoside Re producing a characteristic fragment ion at m/z 789 and ginsenoside Rd yielding a characteristic fragment ion at m/z 365. Rg1 and Rf are isomeric ginsenosides sharing the same precursor ion at m/z 823. Further fragmentation showed that Rg1 readily produces an m/z 643 fragment, whereas Rf yielded only m/z 365 under the same energy conditions. Rb1 has no interfering isomers in ginseng, and its fragment ions include m/z 789 and m/z 365. After confirming the target ion pairs, stepwise CID (NCE 0–7) optimization was performed for the ginsenoside ions to obtain the optimal fragmentation energy. The fragmentation spectra of each compound are shown in Figure . The 4S method requires that the fragmentation energies of the compounds and their internal standards be identical. Considering this requirement and the instability at higher energy, the CID energy was finally set to 3.5 for Rb1, 3.0 for Re, and 2.5 for Rg1 (Table ).

4.

Fragmentation energy curves of ginsenosides Rb1, Rg1, Re and their corresponding internal standards.

1. Ion Pairs and Collision-Induced Dissociation (CID) Values for Ginsenoside Rb1, Re, Rg1, and Their Internal Standard Compounds.

name	MS1	MS2	IS	MS1	MS2	CID
Rb1	1131.3	789.3	Mogroside IVa	1146.6	805.5	3.5
Re	969.5	789.1	Vinaginsenoside R8	985.3	805.1	3.0
Rg1	823.2	643.3	Gypenoside LXXV	807.3	365.1	2.5

3.3.2. Method Validation

The ginsenosides exhibited good linearity within the range of 5–100 μg·mL^–1 (R ² > 0.99) (Figure ). The limits of detection (LOD) for Rb1, Rg1, and Re were 0.15, 0.10, and 0.20 μg·mL^–1, respectively, and the limits of quantification (LOQ) were 0.5, 0.5, and 1.0 μg·mL^–1, respectively (Table ). The stability of Rb1, Rg1, and Re (RSD, n = 6) was 6.5, 8.8, and 5.3. The intra-day precision (RSD, n = 6) ranged from 7.6% to 11.6%, and the inter-day precision (RSD, n = 3) ranged from 10.6% to 11.1%. The spiked recoveries based on the blank matrix ranged from 90.34% to 128.35%. Overall, the mMS meets the basic requirements for quantitative analysis of ginsenosides.

5.

Calibration curves of ginsenosides Rg1, Re, and Rb1 obtained by mMS.

2. Calibration Curves, Linear, LOD, LOQ, Stability, Precision, and Recovery for the mMS Method.

							precision (%)		recovery (%)
analyte	calibration curve	R 2	linear range (μg·mL–1)	LOD (μg·mL–1)	LOQ (μg·mL–1)	stability (%)	intra-day	inter-day	10 ppm	20 ppm	50 ppm
Rb1	y = 0.1356x – 0.3661	0.9991	5–100	0.15	0.5	6.5	11.6	12.1	128.35	90.34	96.41
Rg1	y = 0.011x – 0.0079	0.9909	5–100	0.10	0.5	8.8	8.1	10.6	111.43	95.14	98.99
Re	y = 0.0851x – 0.4172	0.9932	5–100	0.20	1.0	5.3	7.6	12.7	93.73	96.34	98.16

3.4. Optimization of PG Sample Pretreatment for Quantitative Analysis

Efficient intensity release of ginsenosides from the complex solid matrix of PG is a prerequisite for sensitive mMS quantitative analysis. Moreover, mMS suffers from prominent baseline drift and noise. To ensure an adequate signal-to-noise ratio and stable peak integration, the signal intensity should be at least one order of magnitude higher than the baseline (≥10⁴ counts). Signals below this threshold result in poor precision and high susceptibility to matrix interference. In addition, practical applications require rapid testing, meaning that complex extraction and concentration procedures are difficult to implement in real-world scenarios.

To address these needs, a methanol extraction method based on ultrasonication was selected, which allows simple, simultaneous, and high-throughput extraction of samples. Solid-to-solvent ratio and sonication time were systematically optimized to meet the quantitative requirements. The solid-to-solvent ratio (w/v) was screened at 1:5, 1:10, 1:20, and 1:50, and a ratio of 1:5 exhibited superior performance, with intensities above 1×10⁴, sufficient for reproducible quantification. Concurrently, sonication duration was evaluated across four levels (15, 30, 45, and 60 min). It was observed that signal intensities reached a relative equilibrium at 30 min, with no significant signal increase upon further sonication (Figure S10). These optimized conditions ensured robust MS signal intensities for the subsequent mMS quantitative analysis.

3.5. Quality Evaluation of PG Using mMS

Quality grading based on the cultivation age is a critical issue in the PG market. Younger and cheaper PG roots are frequently sold as older, higher-value materials. Traditional assessment relies on morphological inspection, which is subjective and error-prone. Therefore, establishing an objective and rapid chemical evaluation strategy is of considerable practical significance. Compared with UPLC-MS, which typically requires time-consuming sample preparation, mMS enables real-time and on-site detection, making it attractive for rapid ginseng age and quality screening.

Using the validated mMS method, PG samples cultivated for 3, 4, 5, and 6 years were analyzed. The measured concentrations of ginsenosides Rb1, Rg1, and Re are summarized in Table S2. One-way analysis of variance (ANOVA) was performed to assess the effect of the cultivation age on each ginsenoside. The analysis revealed that Rb1 content exhibited a statistically significant age-dependent increase (P < 0.05, Figure S11A). However, Re exhibited a minor downward trend, but Re and Rg1 showed no significant changes with age (P > 0.05, Figure S11B,C). Meanwhile, the results of ginsenosides measured by mMS show certain differences from the limits specified in the pharmacopoeias. This discrepancy is mainly attributed to the need for on-site rapid detection, which adopts a simpler sample pretreatment method and leads to an insufficient extraction of ginsenosides. However, experimental verification has confirmed that the extraction efficiencies of various ginsenosides are similar under different pretreatment methods and their relative proportions remain stable. Therefore, mMS can be used to identify PGs with different growth years based on the ratio of ginsenosides.

The distribution of ginsenosides within the epidermis, phloem, and xylem of ginseng follows a characteristic pattern. As the cultivation age increases, the phloem-to-xylem ratio changes, leading to a regular trend in ginsenoside accumulation across different ages. On the basis of this finding, we analyzed the ratios of different ginsenosides as measured by mMS. It was found that in higher quality six-year-old ginseng, the ratio of ginsenoside Rb1 to Re (Rb1/Re) was consistently above 1 (Figure A). Parallel analysis of the same batch of samples using UPLC-MS (Figure S12) confirmed the same trend in the Rb1/Re ratio, and Bland-Altman analysis showed that 96% of the points fell within the 95% limits of agreement (Figure D), indicating excellent reproducibility of the compound ratio measurement across platforms. Furthermore, UPLC-MS was employed to evaluate the Rb1/Re ratio under different liquid-to-solid ratios; the results demonstrated that the ratio is robust and independent of extraction (Figure S13). In summary, the ratio-based strategy using Rb1/Re enables the rapid and effective discrimination of ginseng cultivation ages. This evaluation approach offers a new paradigm for the on-site rapid quality assessment of traditional Chinese medicines.

6.

Rb1/Re (A), Rg1/Re (B), and Rb1/Rg1 (C) ratios in ginseng of different ages measured by mMS, and Bland–Altman analysis of Rb1/Re measured by mMS versus UPLC-MS (D).

4. Conclusion

Our study established an mMS method coupled with nESI and NaOAc as an ionization additive for the rapid analysis of ginsenosides in Panax ginseng. This method enabled the detection of 11 ginsenosides and successfully distinguished structural isomers based on characteristic fragmentation patterns, providing a portable, rapid, and low-resource solution for on-site analysis of ginseng products, significantly broadening the applicability of portable MS in this field. Meanwhile, we developed a 4S quantitation method for ginsenosides Rb1, Rg1, and Re, with validation demonstrating an acceptable analytical performance. Despite incomplete extraction during mMS rapid detection, the Rb1/Re ratio is stable under different sample-to-solvent conditions and across platforms, allowing for reliable age identification. This approach circumvents the quantification shortfall of rapid methods and offers a field-deployable, objective substitute for the conventional morphology-based evaluation of ginseng age. The current thresholds are derived from geographically limited cultivated ginseng samples; therefore, broader application will require systematic validation across diverse regions and cultivation types. Moreover, future improvements in ion source design, streamlined sample preparation, adoption of internal standard strategies, and incorporation of advanced calibration or data-driven algorithms are expected to further facilitate the use of mMS for on-site herbal assessment. Overall, this study establishes the feasibility of portable mass spectrometry for ginseng authentication and presents an innovative methodological framework for quality assessment of medicinal plants, thereby broadening the utility of mMS in herbal analyses.

Data will be made available on request.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.6c07119.

• Additional experimental details, compound information, mass spectrum characterization, instrumental condition optimization, sample pretreatment optimization, method validation, quantitative results, and correlation analysis of ginsenosides (PDF)

§.

L.L. and Y.L. contributed equally to this work. Longchan Liu: investigation, writing-original draft, writing-review and editing. Yongjin Li: formal analysis and writing-original draft. Haizhen Zhang: investigation. Haoyue Zhang: conceptualization. Wenxiang Fan: investigation. Linnan Li: writing-review and editing. Zhengtao Wang: conceptualization and funding acquisition. Li Yang: funding acquisition, writing-review and editing, and supervision of project administration.

The authors declare no competing financial interest.