Sustainable and rapid ATR-FTIR fingerprinting coupled with a chemometric approach for quality-based batch differentiation of marketed septilin syrup formulations

Septilin syrup is a widely used Ayurvedic polyherbal product known for its role in supporting the immune system. Since it is prepared from plant-based ingredients, natural variation in raw materials can affect the consistency of different production batches. Maintaining uniform quality is therefore an ongoing challenge. This study aims to explore batch-to-batch variation from an analytical perspective, without linking it to clinical outcomes. For this purpose, attenuated total reflectance–Fourier transform infrared (ATR–FTIR) spectroscopy was used to generate characteristic chemical fingerprints of marketed samples. Twelve samples from five different batches were examined, with each individual sample analysed in triplicate. ATR–FTIR, a quick and non-destructive technique, was employed to record spectral profiles. The collected data were then interpreted using Principal Component Analysis (PCA) and Hierarchical Cluster Analysis (HCA) to identify similarities, differences, and clustering patterns among the batches. The PCA and HCA results revealed distinct clustering trends, suggesting clear variability in spectral patterns across batches. These differences reflect underlying compositional changes, possibly due to variations in raw materials or manufacturing processes. Certain spectral regions were particularly important in distinguishing batches. Additionally, green assessment tools confirmed that the method is environmentally friendly, efficient, and resource conscious. Overall, ATR–FTIR combined with chemometric analysis provides a fast, reliable, and sustainable way to monitor batch consistency. While chemical differences were observed, further studies are needed to determine their impact on therapeutic performance.

Herbal medicines have been used for centuries as an integral component of healthcare systems worldwide and continue to gain widespread acceptance in both traditional and modern therapeutic settings (Patel and Kanaki, 2020). These plant-derived formulations represent a rich source of bioactive compounds, including alkaloids, flavonoids, phenolics, and terpenoids, which contribute to their diverse pharmacological activities such as immunomodulatory, anti-inflammatory, antimicrobial, and antioxidant effects (Latif and Nawaz, 2026). In particular, polyherbal formulations—combinations of multiple medicinal plants—are a hallmark of traditional systems such as Ayurveda, where they are designed to exert synergistic therapeutic effects through multi-target interactions (Patel and Kanaki, 2021).

Septilin is a well-established Ayurvedic polyherbal formulation that is widely used as a supportive agent in the management of recurrent infections and immune-related disorders. It is composed of a blend of medicinal plants including Tinospora cordifolia (Guduchi), Glycyrrhiza glabra (Yashtimadhu), Rubia cordifolia (Manjistha), Emblica officinalis (Amlaki), and Commiphora wightii (Guggulu), among others (Ahmed et al., 2022). These botanicals are traditionally valued for their roles in enhancing immunity, reducing inflammation, and promoting overall health. Owing to this diverse composition, Septilin is commonly prescribed as an adjuvant in respiratory tract infections, ENT disorders, and dental infections, where it contributes to improved clinical outcomes and reduced recurrence (Nigade et al., 2023).

Although the therapeutic properties of Septilin provide important background for its widespread use, the present study does not evaluate pharmacological efficacy or clinical outcomes. Rather, the focus is on analytical assessment of batch-to-batch compositional consistency using ATR-FTIR fingerprinting and chemometric analysis. The therapeutic relevance of Septilin is closely linked to its broad spectrum of biological activities. The formulation has been reported to exhibit antimicrobial effects against commonly encountered pathogens, thereby supporting its role in infection control. In addition, its anti-inflammatory properties help alleviate tissue inflammation and associated symptoms in conditions such as tonsillitis, bronchitis, and sinusitis. From an immunological perspective, Septilin plays a significant role in modulating host defence mechanisms. Key constituents such as T. cordifolia are known to enhance antibody production and stimulate macrophage activity, while G. glabra contributes antiviral and immunostimulatory activities. Furthermore, the presence of antioxidant-rich components, including E. officinalis and C. wightii, enables the formulation to mitigate oxidative stress, which is often associated with inflammation and immune dysfunction. Collectively, these biological properties contribute to enhanced immune resilience, faster recovery, and reduced recurrence of infections, particularly when the formulation is used alongside conventional therapies (Nigade et al., 2023; Shruthi and Vijayalaxmi, 2016; Shruthi et al., 2022).

Despite its well-recognised therapeutic importance, the complex and multi-component nature of Septilin presents significant challenges in quality assurance. Unlike synthetic pharmaceuticals with defined chemical entities, herbal formulations contain numerous phytoconstituents whose concentrations can vary considerably due to factors such as geographical origin, seasonal variation, cultivation practices, harvesting time, and processing conditions (Dhami and Mishra, 2015; Ncube et al., 2012; Pant et al., 2021; European Medicines Agency EMA, 2025). This inherent variability often leads to batch-to-batch inconsistencies, which can influence product reproducibility and reliability. Since consistent chemical composition is closely associated with consistent therapeutic performance, ensuring batch uniformity is a critical requirement for maintaining product quality and safety (Wang et al., 2023; Xiong et al., 2013).

Consequently, quality control of herbal formulations has emerged as a central concern in pharmaceutical research and regulatory practice. Standardisation strategies are essential to ensure that herbal products deliver reproducible efficacy and meet safety expectations (Pawar et al., 2026). However, conventional analytical methods may not adequately capture the holistic chemical profile of polyherbal systems and are often time-intensive and resource-demanding. In this regard, there is increasing emphasis on adopting rapid, comprehensive, and non-destructive analytical approaches capable of assessing the overall chemical integrity of such formulations.

Fingerprinting techniques have gained prominence as effective tools for evaluating complex herbal matrices. Among these, Fourier transform infrared (FTIR) spectroscopy provides characteristic spectral profiles that represent the collective chemical composition of a sample (Patel et al., 2024). This approach aligns with the concept of using holistic chemical fingerprints rather than relying solely on individual marker compounds, which may not fully represent the activity of multi-component systems (Patel et al., 2025). When combined with chemometric techniques such as Principal Component Analysis (PCA), spectral data can be effectively processed to identify patterns, classify samples, and detect subtle compositional variations between batches, while Hierarchical Cluster Analysis (HCA) enables the assessment of similarities and differences among samples and facilitates quality-based classification. Therefore, the present study employed HCA alongside ATR–FTIR fingerprinting to further investigate the compositional relationships among marketed Septilin syrup batches and to complement the findings obtained from PCA analysis (Patel et al., 2026).

In addition to analytical efficiency, there is a growing emphasis on sustainability in pharmaceutical analysis. Green analytical chemistry principles advocate the development of environmentally friendly methods that minimise solvent usage, reduce waste, and lower operational impact. ATR–FTIR spectroscopy aligns well with these principles due to its rapid analysis, minimal sample preparation, and non-destructive nature, making it an attractive tool for routine quality assessment.

In this context, the present study focuses on the evaluation of batch-to-batch variability in marketed Septilin syrup using ATR–FTIR spectroscopy coupled with chemometric analysis. Twelve samples representing five different batches were analyzed, with three replicates for each individual. By employing a rapid and sustainable fingerprinting approach, this work aims to provide a reliable method for monitoring compositional consistency in complex polyherbal formulations. While detectable spectral variations may indicate differences in chemical composition between batches, their direct implications on therapeutic performance remain beyond the scope of the present investigation and warrant further study.

Materials

For the present investigation, a total of twelve commercially available Septilin syrup samples were collected from licensed pharmaceutical outlets. These samples represented five distinct manufacturing batches, allowing both intra-batch and inter-batch evaluation. To ensure the reliability and reproducibility of observations, two to three independently procured samples were analysed from each selected batch, with each sample scanned in triplicate by ATR-FTIR.

All samples were carefully examined for labeling compliance, and expiry date to ensure suitability for analytical testing. Prior to analysis, the syrups were stored under recommended conditions, protected from direct light and temperature fluctuations, to prevent any degradation or alteration of their phytochemical composition.

Unlike conventional phytochemical investigations that often rely on solvent extraction or fractionation techniques, the present study was deliberately designed to evaluate Septilin syrup in its native marketed form. This approach offers two important advantages: firstly, it preserves the inherent chemical complexity of the formulation; and secondly, it reflects real-world use conditions more accurately. Furthermore, avoiding extraction aligns with the concept of developing rapid and environmentally sustainable analytical methods.

ATR–FTIR spectroscopic analysis

The chemical fingerprinting of Septilin syrup was carried out using Attenuated Total Reflectance–Fourier Transform Infrared (ATR–FTIR) spectroscopy, a non-destructive and highly informative analytical technique suitable for complex multi-component systems.

For analysis, a small and uniform amount of syrup sample was carefully placed directly onto the ATR crystal surface. No pre-treatment, dilution, or extraction was required, making the process simple and reproducible. Spectral acquisition was performed over the mid-infrared range of 400–4,000 cm -1, which encompasses most vibrational frequencies associated with organic functional groups commonly found in herbal constituents. Each spectrum was recorded using 16 co-added scans at a resolution of 4 cm -1, providing a balance between spectral clarity and acquisition time. Prior to sample analysis, background spectra were recorded under identical conditions and automatically subtracted to eliminate atmospheric interference such as CO 2 and moisture.

To ensure methodological consistency and accuracy, three replicate spectra were recorded for each sample. The spectral profiles were visually inspected to confirm reproducibility and absence of significant instrumental noise or artefacts. The resulting data were then exported for advanced chemometric processing. Thus, five commercially available production batches of Septilin syrup were evaluated. Each batch comprised two to three independently procured samples, yielding a total of twelve samples. For each sample, ATR-FTIR spectra were recorded in triplicate. Prior to chemometric analysis, replicate spectra were averaged to generate a single representative spectrum for each sample. Consequently, PCA and HCA were performed using twelve averaged spectra.

Chemometric analysis

Given the complexity of ATR–FTIR data and the overlapping nature of spectral bands in polyherbal formulations, multivariate statistical techniques were employed to extract meaningful patterns and relationships. Prior to analysis, spectral data were subjected to essential preprocessing steps, including baseline correction, normalization, and scaling. These steps help reduce noise, eliminate variability due to sample thickness or instrument conditions, and enhance comparability between spectra.

Principal component analysis (PCA)

Principal Component Analysis (PCA) was employed as the principal chemometric technique to investigate underlying patterns and relationships within the spectral dataset. PCA is a widely used unsupervised multivariate statistical method that reduces the complexity of large datasets by transforming numerous correlated variables into a smaller number of independent variables known as principal components. These newly generated components retain most of the information present in the original dataset while minimizing redundancy among variables. Each principal component accounts for a certain proportion of the total variance, with the first few components generally explaining the majority of the variability observed among the samples.

By projecting the data into a lower-dimensional space, PCA facilitates the visualization of sample distribution, clustering behavior, and overall similarities or differences within the dataset. This approach is particularly useful for identifying natural groupings, detecting trends, and highlighting potential outliers that may arise due to variations in sample composition. In the present study, score plots derived from the principal components were generated to evaluate the grouping of Septilin syrup samples according to their respective batch identities and to assess batch-to-batch consistency. The graphical representation of the PCA results provided valuable insights into the chemical relationships among the analysed samples and enabled a clearer interpretation of the spectral data. All statistical analyses were performed using Minitab® software (Version 17), ensuring reliable data processing, robust interpretation, and enhanced confidence in the observed classification patterns.

To aid visual interpretation, PC1 and PC2 loading profiles were smoothed using a Savitzky–Golay filter (window width 25 data points, third-order polynomial) prior to plotting. This step reduces point-to-point spectral noise while preserving the position and shape of genuine absorption features, and was applied for display purposes only; the underlying PCA model and sample scores were computed from the unsmoothed spectral data.

To assess whether batch discrimination could be attributable to instrumental effects such as ATR contact pressure or sample-thickness variation rather than compositional differences, PCA was additionally performed on Standard Normal Variate (SNV)-normalized spectra, which remove multiplicative scaling and additive baseline offsets from each spectrum.

Hierarchical cluster analysis (HCA)

Hierarchical Cluster Analysis (HCA) was performed on the pre-processed ATR–FTIR spectral dataset to evaluate the similarity patterns among Septilin syrup samples. Prior to analysis, spectral data were subjected to standard preprocessing procedures including baseline correction and normalization. HCA was conducted using Euclidean distance as the similarity measure and complete linkage as the clustering algorithm. The resulting dendrogram was used to visualize the hierarchical relationships among samples and to identify clustering trends corresponding to specific manufacturing batches. This unsupervised chemometric technique provides an additional means of assessing batch-to-batch variability and verifying sample classification obtained through PCA.

Greenness, blueness, and whiteness evaluation

In addition to demonstrating strong analytical performance, the long-term sustainability and practical relevance of the proposed method were carefully examined through a comprehensive, multi-dimensional assessment framework. This framework integrates the concepts of greenness, blueness, and whiteness, offering a balanced evaluation that not only considers environmental impact but also practical usability and overall methodological efficiency. By adopting this holistic approach, the study ensures that the method aligns with modern expectations of sustainable and responsible analytical practices.

To begin with, the environmental sustainability of the method was assessed using the AGREE (Analytical GREEnness) metric. This tool provides a structured and quantitative evaluation based on the twelve guiding principles of green analytical chemistry. These principles emphasize minimizing the use of hazardous substances, reducing waste generation, improving energy efficiency, and promoting safer analytical procedures. By applying AGREE, the method’s compliance with these principles was systematically quantified, providing a clear and reliable indication of its environmental friendliness.

In parallel, a more detailed investigation of environmental impact was carried out using the Complex Green Analytical Procedure Index (Complex GAPI). Unlike simpler tools, Complex GAPI offers a broader perspective by examining the entire analytical workflow from start to finish. This includes considerations related to sample collection, transportation, storage, preparation, and final measurement. Each stage is critically evaluated in terms of its environmental burden, allowing for the identification of potential areas for improvement. As such, the use of Complex GAPI ensures that sustainability is not only assessed at the measurement stage but is embedded throughout the entire analytical process.

Beyond environmental considerations, the practical applicability of the method was evaluated using the Blue Applicability Grade Index (BAGI). This metric focuses on the operational side of analytical methods, ensuring that they are not only sustainable but also feasible and efficient in real-world laboratory settings. The BAGI assessment takes into account several key factors, including sample throughput, which reflects how many samples can be analyzed within a given time frame, as well as runtime efficiency, which relates to the speed of analysis. Additionally, the complexity of sample preparation is considered, as simpler procedures generally enhance usability and reduce resource consumption. Instrument accessibility is another important factor, ensuring that the method can be implemented without requiring highly specialized or prohibitively expensive equipment.

To provide a well-rounded sustainability evaluation, the RGB12-based White Analytical Chemistry (WAC) framework was applied. This model integrates three core dimensions: environmental impact (green), analytical performance (red), and practical usability (blue). By combining these aspects into a single system, WAC delivers a more realistic and balanced assessment than relying on individual metrics alone.

The ATR–FTIR method showed strong performance across all three dimensions, confirming that it is not only environmentally responsible but also reliable and efficient in practice. Its high whiteness score highlights its suitability for routine analysis of Septilin syrup samples. Moreover, integrating AGREE, Complex GAPI, BAGI, and RGB12 helps overcome the limitations of using a single tool, offering a more complete picture of method sustainability. This combined strategy reflects a robust and contemporary approach to developing sustainable analytical methods. (Patel et al., 2026; Patel et al., 2025).

ATR–FTIR spectral analysis and comparative interpretation

The ATR–FTIR spectra obtained for all Septilin syrup samples exhibited complex yet consistent fingerprint profiles typical of polyherbal formulations (Figure 1). The overall similarity in spectral patterns confirms the presence of a common formulation matrix; however, careful examination revealed subtle variations in peak intensities and minor shifts across different batches, indicating underlying compositional differences.

A broad absorption band in the region of 3200–3500 cm -1 was consistently observed across all samples and can be attributed to O–H stretching vibrations associated with phenolic compounds, alcohols, and hydrogen-bonded constituents. Similarly, the 2800–3000 cm -1 region displayed characteristic C–H stretching vibrations of aliphatic groups. Slight differences in peak intensities in this region indicate variation in lipophilic components within the formulation. The presence of an absorption band near ∼1700 cm -1, corresponding to carbonyl (C=O) stretching vibrations of esters, organic acids, or ketones, further supports the chemically diverse nature of the formulation. Minor differences in this peak among batches suggest variability in such constituents.

The most significant distinctions were observed within the fingerprint region (1000–1300 cm -1), which represents complex vibrational modes of carbohydrates, glycosides, and other bioactive compounds. Variations in peak patterns and intensities in this region were particularly evident when spectra from different batches were compared. When interpreted alongside the batch details in Table 1, these spectral differences indicate that samples belonging to the same batch exhibit highly similar profiles, whereas inter-batch comparisons reveal detectable variation. This highlights the sensitivity of ATR–FTIR spectroscopy in capturing subtle compositional changes in multi-component herbal systems.

It should be noted that Septilin is a syrup-based formulation containing water and carbohydrate-rich excipients. Therefore, the spectral variations observed in the O-H stretching region (3200–3500 cm -1) and the fingerprint region (1000–1300 cm -1) may arise from both phytochemical constituents and formulation matrix components. Consequently, the observed batch discrimination reflects differences in the overall chemical fingerprint of the formulation rather than exclusively the herbal active constituents.

Principal component analysis (PCA) and batch-to-batch variation

To further investigate these spectral differences, Principal Component Analysis (PCA) was applied to the FTIR dataset. The PCA score plot (Figure 2) provided an effective visualization of similarities and differences among the samples based on their spectral characteristics.

The first principal component (PC1) accounted for the majority of the variance (94%) within the dataset, while the second principal component (PC2) contributed additional discriminatory information (variance:4.3%). Together, these two components captured the most significant variations (98.3%) among the samples, enabling clear interpretation in reduced dimensional space.

A well-defined clustering pattern based on batch identity was observed. Samples belonging to the same batch formed tight clusters, indicating high intra-batch consistency and reproducibility in manufacturing. This confirms that, within individual batches, the formulation is highly uniform. Batch separation along PC1 was retained following SNV normalization (Supplementary Figure S1), supporting a compositional rather than purely instrumental basis for the observed discrimination. While SNV normalization supports a compositional basis for the observed batch differences, residual instrumental variation (e.g., sample thickness, ATR contact pressure) cannot be fully excluded and should be assessed further using controlled-pathlength measurements in future work.

In contrast, distinct separation between clusters corresponding to different batches was evident. For example, samples from earlier manufacturing batches showed slight positional shifts compared to more recent batches, indicating gradual compositional variation over time. These differences are consistent with the known variability associated with herbal raw materials, where factors such as geographical origin, seasonal variation, harvesting conditions, and processing techniques can influence phytochemical composition.

Importantly, no extreme outliers were observed in the PCA plot. The results indicate detectable and reproducible batch-to-batch variation alongside strong intra-batch consistency, as samples belonging to the same manufacturing batch clustered closely together while remaining distinguishable from other batches. The present study was not designed to establish predefined acceptance limits for batch conformity; therefore, the observed variation should be interpreted descriptively rather than as evidence of “acceptable” or “unacceptable” product quality. Consequently, PCA confirms the presence of systematic batch-to-batch differences while demonstrating good reproducibility within individual batches.

Correlation between spectral features and PCA findings

The variations observed in the PCA score plot (Figure 2) can be directly correlated with the spectral differences identified in the FTIR analysis. Differences in key regions, particularly the O–H stretching region (3200–3500 cm -1), the carbonyl region (∼1700 cm -1), and the fingerprint region (1000–1300 cm -1), contribute significantly to the separation of samples along the principal component axes.

Among these, the fingerprint region plays a dominant role due to its high chemical specificity and sensitivity to minor compositional changes. While these variations may not be immediately apparent through visual inspection alone, chemometric analysis effectively amplifies and quantifies these differences, enabling clear discrimination between batches. This highlights the complementary nature of ATR–FTIR and PCA in the analysis of complex herbal formulations.

It is important to emphasize that, although these variations were detected analytically, they do not automatically translate into meaningful differences in therapeutic effectiveness or biological activity. In other words, changes at the chemical or compositional level may not necessarily alter the overall health benefits or intended function of the product. Nevertheless, systematically monitoring such variability remains a critical aspect of quality control. Consistency in composition is essential for ensuring batch-to-batch uniformity and meeting regulatory expectations.

The PC1 and PC2 loading profiles (Figure 3), smoothed by Savitzky-Golay filter (window = 25 points, third-order polynomial) as can be seen Figures 3A,B help to identify which spectral regions drive these separations. The PC1 loading is dominated by two regions: a broad feature centred around 3200–3400 cm -1, consistent with O–H stretching vibrations, and a set of sharper, higher-magnitude features between 1300 and 950 cm -1 within the fingerprint region—the latter contributing most strongly to the component, in agreement with the visual clustering pattern in the PCA score plot. The PC2 loading shows a distinct profile, with a pronounced negative contribution around 3500–3700 cm -1 and a strong positive peak near 1000–1050 cm -1, indicating that PC2 captures a different combination of O–H and fingerprint-region features than PC1, consistent with its role in separating batches that PC1 alone does not fully resolve (for example, batches 111,230,952 and 111,230,238). Together, the two loading profiles support the earlier observation that the fingerprint region and the O–H stretching region are the primary contributors to inter-batch spectral variation.

Hierarchical cluster analysis (HCA) and batch-to-batch variation

The HCA dendrogram (Figure 4) provided a clear picture of the relationships among the twelve Septilin syrup samples. Samples 1, 3, and 12 from batch 111,241,072 clustered closely together, indicating a high level of similarity in their ATR–FTIR fingerprints. Likewise, samples 2 and 11, belonging to batch 111,240,082, formed a neighbouring subgroup and later joined the cluster containing samples from batch 111,241,072. This close association suggests that these two batches share a very similar overall chemical composition.

A second major cluster was formed by samples 4–10. Within this group, samples 4 and 5 (batch 111,230,952) showed strong similarity and clustered together at an early stage. A similar pattern was observed for samples 9 and 10 (batch 111,230,238) and samples 6, 7, and 8 (batch 111,230,233), demonstrating good consistency within each manufacturing batch. Interestingly, batches 111,230,952 and 111,230,238 were more closely related to each other than to batch 111,230,233, as evidenced by the order in which the clusters merged in the dendrogram.

Overall, the clustering pattern showed that samples originating from the same batch grouped together before joining samples from other batches, highlighting excellent intra-batch uniformity. At the same time, the separation between the major clusters indicates the presence of moderate batch-to-batch variation, likely reflecting natural differences in herbal raw materials and processing conditions. The HCA results closely supported the PCA findings and further confirmed the capability of ATR–FTIR fingerprinting combined with chemometric analysis for monitoring the quality consistency of marketed Septilin syrup formulations.

Sustainability assessment and method evaluation

In addition to its analytical performance, the sustainability of the ATR–FTIR method was comprehensively evaluated using greenness, blueness, and whiteness metrics. The results demonstrate that the proposed method aligns well with modern analytical and environmental requirements.

The AGREE assessment indicated strong compliance with green analytical chemistry principles, primarily due to the elimination of solvent use, minimal sample preparation, and reduced energy consumption. The Complex GAPI evaluation further supported these findings, confirming a low environmental impact across the analytical workflow.

From an operational perspective, the BAGI assessment highlighted excellent method applicability, characterized by rapid analysis time, simplicity, and minimal technical requirements. These features make the method highly suitable for routine quality control of herbal formulations.

Furthermore, the RGB-based whiteness evaluation demonstrated a balanced integration of environmental sustainability, analytical reliability, and practical applicability. This confirms that the method not only performs effectively but also meets broader criteria of sustainable analytical practice.

When compared with previously reported HPTLC methods (Figure 5), the ATR–FTIR approach offers several advantages, including reduced analysis time, absence of hazardous solvents, and non-destructive analysis. While HPTLC provides compound-specific separation, it is relatively labour-intensive and less aligned with green analytical principles. In contrast, ATR–FTIR provides a rapid, holistic fingerprinting approach that is better suited for routine monitoring of complex polyherbal formulations.

The present study establishes ATR–FTIR spectroscopy coupled with chemometric analysis as a powerful and efficient sustainable approach for evaluating batch-to-batch variability in Septilin syrup.

The method successfully captured subtle compositional differences between batches while confirming high internal consistency within individual batches. Its ability to detect such variations in a rapid and non-destructive manner makes it particularly valuable for quality control of complex polyherbal formulations.

Compared to conventional analytical techniques, the proposed approach offers significant advantages, including reduced analysis time, elimination of solvents, minimal sample preparation, and lower operational costs. Additionally, its strong performance across sustainability metrics aligns with current trends in green analytical chemistry.

While the observed spectral variations indicate measurable differences in chemical composition, their biological significance remains unexplored and warrants further investigation through pharmacological and clinical studies. Future studies should focus on the development and validation of supervised chemometric models, including SIMCA and PLS-DA, using larger sample sets encompassing multiple manufacturing lots. Cross-validation and external validation should be performed to determine classification accuracy, sensitivity, and specificity. Furthermore, statistically defined acceptance limits should be established for routine batch qualification. Experimental designs incorporating storage conditions and product ageing will also be necessary to distinguish manufacturing-related variability from post-production changes.

Overall, this work provides a practical, reproducible, and environmentally conscious strategy for detecting and monitoring batch-to-batch variation in herbal pharmaceutical products. While the study demonstrates consistent clustering patterns within manufacturing batches, it does not establish formal acceptance criteria for batch conformity. Such criteria would require additional validation studies and predefined quality specifications.