Green UHPLC approach for the quantitative determination of tiopronin residues in cleaning validation processes

Abstract

This study presents the development and validation of a novel reverse-phase ultra-high-performance liquid chromatography (UHPLC) method for quantifying tiopronin residues on manufacturing equipment surfaces as part of cleaning validation. Tiopronin, a synthetic thiol compound with therapeutic applications, requires stringent manufacturing controls to ensure product safety and efficacy. Existing analytical techniques mainly focus on quantifying tiopronin in pharmaceutical formulations or biological matrices. However, no methods specifically address cleaning validation for tiopronin production equipment. Using an Analytical Quality by Design approach, this UHPLC method was optimized and validated by ICH Q2 (R2) guidelines. The technique employs a Waters ACQUITY UPLC H-Class PLUS C-18 column (100 mm × 2.1 mm, 1.7 µm) with a mobile phase of 88:12 (v/v) 0.1% v/v orthophosphoric acid (pH 2.1) and acetonitrile, achieving a tiopronin retention time of 1.3 min. The method demonstrated specificity, precision, accuracy, and linearity over a concentration range of 0.302 to 3.027 µg/mL. The limit of detection and quantification were 0.100 µg/mL and 0.301 µg/mL, respectively. The greenness of the method was assessed using AGREE, GAPI, and RGB 12 tools. The scores obtained were 0.67 for AGREE, 85.0 for BAGI, and 82.1 for RGB 12 suggesting proposed method is environment friendly. This novel method provides an efficient, eco-friendly solution for routine cleaning validation, addressing a significant gap in existing analytical techniques for tiopronin manufacturing.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13065-025-01656-2.

Introduction

Tiopronin, known as N-(mercaptopropionyl) glycine, is a synthetic thiol compound for multiple therapeutic purposes. It is a hepatoprotective agent, an antidote for heavy metal poisoning, and has radioprotective properties. Its primary application prevents kidney stones from elevated cystine levels in the urine, particularly in patients with severe homozygous cystinuria [1–3]. Additionally, tiopronin is being explored for its potential in treating rheumatoid arthritis [4]. However, despite its clinical benefits, tiopronin is associated with several toxicity risks. These include hypersensitivity reactions, hematological disorders such as leukopenia and agranulocytosis, nephrotoxicity, and dermatological effects. Adverse events such as rash, fever, arthralgia, and proteinuria have also been reported. Given the possibility of serious health effects stemming from even trace exposure, especially among sensitive individuals or through accidental cross-contamination, strict residue control during pharmaceutical manufacturing is warranted. Its delayed-release formulation, often utilizing polymeric coatings, further complicates the cleaning process. Its delayed-release formulation, with polymeric coatings, makes it harder to clean according to the FDA-approved label for tiopronin (THIOLA EC) (FDA, 2021) [5]. Due to its varied applications and potential toxicity, stringent control measures during its manufacturing are essential to ensure the safety and efficacy of the product. Effective cleaning of equipment surfaces used in tiopronin manufacturing is necessary to prevent cross-contamination, uphold product quality, adhere to regulatory requirements, protect the safety of operators, maintain equipment integrity, and avoid residue accumulation. Implementing robust cleaning procedures is vital for the safe and efficient production of pharmaceuticals [6–8].

The cleaning process is essential for ensuring product quality across different pharmaceuticals, preventing cross-contamination, and complying with guidelines set by the European Union for Good Manufacturing Practice and the United States Food and Drug Administration (USFDA) [9–11]. Cleaning validation encompasses two main activities: first, developing and validating the cleaning procedure to remove drug residues from manufacturing equipment surfaces, and second, developing and validating analytical methods to measure these residues on equipment surfaces [12, 13]. Assessing the sensitivity and specificity of the analytical method used for residue detection is crucial [14]. The analytical process must detect and quantify trace amounts of the drug substance on the surfaces of manufacturing equipment. Furthermore, the analytical procedure should be tested using the sampling method to demonstrate that residues can be recovered from equipment surfaces at specified levels, as verified in accuracy studies, before finalizing the sampling procedure [15–21].

There is an urgent need to develop a sensitive and selective method for analyzing tiopronin in swab samples for cleaning validation. Existing analytical techniques, including chemiluminescence flow-injection analysis [22], reverse-phase high-performance liquid chromatography [23, 24], electrospray ionization mass spectrometry [25], fluorometric methods [26], LC–MS/MS [27], and LC–ESI–MS/MS [28], have been reported for tiopronin quantitation in pharmaceutical formulations or human plasma. However, none of these methods specifically target cleaning validation for tiopronin manufacturing equipment. This study introduces an innovative reverse-phase ultra-performance liquid chromatography (UHPLC) method for quantifying tiopronin in swab samples for cleaning validation. Developed using an Analytical Quality by Design (AQbD) approach, this method adheres to stringent cleaning validation standards while focusing on eco-friendliness and reliability [29]. Including a greenness analysis based on Green Analytical Chemistry (GAC) principles further enhances the method's sustainability, making it suitable for routine laboratory quality control [30, 31]. By offering an efficient solution for detecting and removing tiopronin residues from manufacturing equipment, this research addresses a vital gap in current analytical methods for cleaning validation. This advancement helps prevent cross-contamination, ensures patient safety, and establishes a new standard for quality control, thus supporting enhanced product safety and regulatory compliance in pharmaceutical manufacturing.

Experimental

Drugs and solvents

The Tiopronin sample was manufactured by MSN Life Science Private Limited, Unit-II, in Telangana, India. Acetonitrile (ACN) with batch number T046E24 and HPLC-grade methanol with batch number R292E24, both manufactured by Rankem (Haryana), were used. Milli-Q water, prepared in-house, was utilized in the present investigation. Orthophosphoric acid 85% (OPA), batch number DK2D721841, manufactured by Merck, was used. The swab used in the study was Texwipe’s Alpha TX 714A.

Instrument setup and chromatographic optimization

Table 1 represents the UHPLC system setup and optimized chromatographic conditions.

Table 1.

Chromatographic setup and conditions

Parameters	Experimental conditions
Instrument	Waters ACQUITY UPLC H-Class PLUS
Stationary Phase	Acquity UPLC BEH C18 (100 mm × 2.1 mm; 1.7 µm)
Mobile Phase	0.1% v/v orthophosphoric acid (pH 2.1): Acetonitrile (88:12%, v/v)
Flow rate (mL/min)	0.3 mL/min
Injection Volume (µL)	6 µL
Run time (Minutes)	5 Minutes
Column oven temperature (°C)	40 °C
Sample cooler temperature (°C)	5 °C
Detector	UV–Vis
Detection Wavelength (nm)	230 nm
Diluent	Water and Acetonitrile in the ratio of 88:12 (% v/v)
Target concentration (µg/mL)	Tiopronin- 2 µg/mL

Preparation of diluent

Water and acetonitrile in the ratio of 88:12 (% v/v) was used as a diluent throughout the study.

Preparation of 0.1% v/v orthophosphoric acid solution (pH 2.1)

A volume of 1.0 mL of orthophosphoric acid was accurately measured and added to approximately 990 mL of purified water in a suitable container. The mixture was stirred thoroughly. The pH was checked and, if necessary, adjusted to 2.1 using dilute sodium hydroxide solution. After pH adjustment, the solution was diluted to a final volume of 1000 mL with purified water and mixed thoroughly. The resulting solution was then filtered through a 0.22 µm polyvinylidene fluoride (PVDF) membrane filter.

Preparation of mobile phase

The prepared 0.1% v/v orthophosphoric acid solution and acetonitrile were mixed in a volume ratio of 88:12 (v/v) in a clean container. The resulting mixture was filtered through a 0.22 µm membrane filter to remove any particulates. The filtered mobile phase was then sonicated for 20 min to degas it prior to use. This mobile phase composition was maintained consistently throughout the chromatographic analyses.

Preparation of Standard stock solutions

To ensure accurate measurement, a 50 mg standard of Tiopronin was transferred into a 50 mL volumetric flask and diluted up to the mark with diluent to achieve a concentration of 1000 µg/mL. From this solution, 5 mL was withdrawn and diluted with diluent in a 50 mL volumetric flask to prepare a 100 µg/mL solution (Stock solution). Subsequently, 2 mL of the 100 µg/mL solution was withdrawn and diluted with diluent to reach a final concentration of 2 µg/mL. This final solution, the working standard solution (WSS), containing 2 µg/mL Tiopronin, was used as the target concentration for analysis. Before injection into the UHPLC system, the solution was filtered using a 0.22 µm syringe filter.

Swab sampling and sample preparation

Swab pre-treatment

Swabs (Texwipe’s Alpha TX 714A) were pre-treated by dipping them into a container holding approximately 50 mL of diluent. The swabs were sonicated for 5 min to ensure thorough wetting, and the excess solvent was removed by gently pressing the swabs against the container’s wall.

Swab sampling

Using a pre-treated damp swab, the designated sampling area (2 × 2-inch square) on the equipment surface was vigorously scrubbed with firm strokes first in one direction, followed by perpendicular strokes, utilizing both sides of the swab to maximize residue collection. After sampling, the swab was transferred immediately into a clean glass vial.

Blank swab solution

To prepare blank controls, pre-treated swabs not used for surface sampling were placed into clean glass vials. The vials were left open for 30 min, after which 10.0 mL of diluent was added. The vials were then shaken for 2 min to extract any potential contaminants from the swabs. Prior to UHPLC analysis, the blank solutions were filtered through a 0.22 µm syringe filter.

Swab sample solution

Similarly, the swab samples collected from equipment surfaces were left uncapped in their respective glass vials for 30 min. Then, 10.0 mL of diluent was added to each vial, and the vials were shaken for 2 min to extract residues from the swabs. The extracted solutions were filtered through a 0.22 µm syringe filter before UHPLC injection.

Method development and optimization using AQbD framework

The Analytical Target Profile (ATP), defined within the Analytical Quality by Design (AQbD) framework, outlines the desired performance characteristics and acceptance criteria for the analytical method. This initial step involves identifying Critical Quality Attributes (CQAs) that directly influence the accuracy and consistency of the UHPLC procedure [29]. Key CQAs selected for this study included retention time, tailing factor, and theoretical plates, detailed in Supplementary Table 1.

In the early stages of AQbD method development, the Ishikawa (fishbone) diagram serves as a valuable tool for systematically identifying and categorizing all potential factors affecting UHPLC performance. By visually organizing variables into key categories—such as instrument, materials, environment, and procedures—it ensures a comprehensive assessment of method variability sources [32–34]. Critical method parameters (CMP) highlighted in the diagram guided focused risk evaluation via Failure Mode and Effects Analysis (FMEA) and informed experimental design. These CMPs were further evaluated through a quantitative risk assessment employing FMEA. The assessment utilized the Risk Priority Number (RPN), calculated as the product of Severity (S), Occurrence (O), and Detectability (D) considering scale of 1 to 5.

1

Parameters with RPN values greater than 25 were considered critical and subject to optimization[32, 34]. Based on the risk assessment, a Central Composite Design (CCD) was applied to investigate the effects of three critical variables—organic phase volume (X₁), flow rate (X₂), and column oven temperature (X₃)—on method performance indicators including retention time, theoretical plate count, and tailing factor. A total of 17 experimental runs were conducted using Design Expert Software v12, evaluating each variable at high (+ 1), intermediate (0), and low (− 1) levels.

Following response surface analysis using CCD, the method operable design region (MODR) was established to proactively minimize the risk of critical analytical failure and ensure the UHPLC method aligned with the defined analytical target profile. The predictive model was subsequently verified by conducting laboratory experiments, and the observed results were in good agreement with the model’s predictions.

Analytical method validation

The analytical method was subjected to a rigorous evaluation of its performance characteristics to establish its reliability in accordance with the guidelines outlined in ICH Q2(R2) [35]. The assessment encompassed linearity over the concentration range of LOQ to 150%, accuracy at three concentration levels (20%, 100%, and 150%) on different surfaces, and precision at the same levels considering working concentration of method 2 µg/mL of tiopronin. Specificity was demonstrated through the analysis of various matrices, including standard solution, diluent, placebo, blank swab, and surface swab samples from stainless steel (SS) coupons, silicon rubber, and acrylic sheets, as well as sample containing cleaning agents. Robustness was evaluated using a fractional factorial design to identify potential sources of variability like volume of ACN, flow rate, temperature, and injection volume. The study’s evaluation encompassed system suitability parameters such as tailing factor, theoretical plates, and peak area. Additionally, the method's sensitivity was established by determining the limits of detection and quantification.

Results and discussion

Initial method development

A novel UHPLC technique has been developed to precisely quantify Tiopronin residue collected from equipment surfaces after cleaning. The method development process involved a thorough review of chromatographic literature focusing on Tiopronin. Previous studies recommended blending buffers with organic solvents such as acetonitrile and methanol. The selection of the analytical column was crucial, emphasizing retention characteristics and selectivity. During the initial UHPLC method development, various mobile phase compositions were tested, varying acetonitrile and water ratios with and without buffers.

Acetonitrile was chosen over methanol due to its lower UV cut-off at 190 nm, making it suitable for applications requiring detection at lower UV wavelengths than methanol’s 205 nm. Standardized parameters across experiments included a flow rate of 0.3 mL/min, an injection volume of 6 µL, and a temperature of 40 ℃.

Initial experiments under different mobile phases and their different ratio conditions yielded unsatisfactory results, characterized by peak tailing, broadening, low-capacity factor, and inadequate theoretical plate counts (below 2000). However, successful outcomes were achieved using a Waters ACQUITY UPLC H-Class PLUS C-18 column (100 mm × 2.1 mm, 1.7 µm) with a mobile phase composed of 0.1% v/v orthophosphoric acid (pH 2.1) and acetonitrile in an 88:12 ratio. This combination produced well-defined peaks and met the required system suitability criteria. These preliminary findings provide valuable insights for further exploration.

Risk assessment studies

Various scientific databases were consulted to gather insights into UHPLC method development and factors influencing analysis efficiency. These findings were further enriched through discussions with colleagues and experts, enhancing comprehension. The resulting insights were organized into an Ishikawa-Fishbone diagram to visualize potential variables affecting UHPLC analysis (Fig. 1). The Ishikawa diagram was initially employed to visually map all potential method-related variables that could impact the UHPLC performance. Systematic transition from qualitative identification (Ishikawa) to quantitative risk evaluation FMEA, was performed. Figure 2, detailing the FMEA, employing Risk Priority Number (RPN) for quantification analysis, was created. This analysis identified vital parameters such as the percentage of the organic phase, flow rate, and column oven temperature as significant based on RPN, while others were deemed less critical. Subsequent method development focused on optimizing these parameters using a Central Composite Design approach.

Fig. 1.

The Ishikawa (fishbone) diagram to identify potential variables in UHPLC method development

Fig. 2.

Failure mode and effect assessment

Central composite design for factor optimization

Central Composite Design (CCD) is widely favored for optimizing UHPLC analysis because it can efficiently explore quadratic effects [36]. In our study, CCD was applied with three vital independent variables: volume of the organic phase, flow rate, and column oven temperature, all critical to chromatographic performance. Parameters such as retention time, theoretical plates and tailing factor were evaluated (Table 2). The resulted ANOVA data confirmed that retention time, theoretical plates, and tailing factor were all significantly influenced by the studied chromatographic parameters, with each response fitting a robust quadratic model (Table 3). For retention time, both organic phase volume and flow rate, along with their interaction and quadratic terms, were highly significant (F = 253.91, p < 0.0001), yielding an excellent model fit (R² = 0.9914). Theoretical plates were similarly modelled, with significant effects from organic phase volume, flow rate, column oven temperature, and their interactions (F = 730.19, p < 0.0001; R² = 0.9977), confirming the ability to fine-tune column efficiency across the design space. Tailing factor also demonstrated significant dependence on volume of organic phase, interaction terms, and nonlinear effects (F = 19.88, p < 0.0001; R² = 0.9393), although with slightly lower predictive power than the other responses. Notably, all models achieved suitable signal-to-noise ratios (Adeq Precision > 4), and lack-of-fit was not significant for theoretical plates and tailing factor, supporting model adequacy for method development.

Table 2.

Central-Composite Design (CCD) matrix with responses to the optimization of UPLC method

Run	Volume of organic phase (%)	Flow rate (mL/min)	Column oven temp. (ºC)	Retention time (min)	Theoretical plates (N)	Tailing factor
1	12	0.3	40	1.22	7767	1.57
2	12	0.468	40	0.78	7139	1.92
3	3.60	0.3	40	2.23	3258	1.20
4	20.40	0.3	40	0.96	16,065	1.58
5	12	0.3	48.4	1.18	7423	1.64
6	7	0.4	45	1.20	4207	1.83
7	12	0.3	40	1.21	7973	1.58
8	12	0.3	31.6	1.24	8214	1.49
9	17	0.2	35	1.57	12,678	2.08
10	12	0.131	40	2.75	7807	2.09
11	7	0.4	35	1.25	4628	1.57
12	7	0.2	45	2.37	4280	1.39
13	12	0.3	40	1.21	8027	1.62
14	17	0.4	35	0.79	11,394	1.65
15	7	0.2	35	2.49	4535	1.71
16	17	0.2	45	1.54	12,820	1.65
17	17	0.4	45	0.77	10,918	1.87

Table 3.

ANOVA data for retention time, theoretical plates and tailing factor

Source	Sum of squares	DF	Mean square	F-value	p-value (probability > F)	Significant if p-value < 0.05
Response 1—Retention time
Model	6.17	5	1.23	253.91	< 0.0001	Significant
A-Volume of organic phase	1.67	1	1.67	343.69	< 0.0001	Significant
B-Flow rate	3.87	1	3.87	797.08	< 0.0001	Significant
AB	0.0924	1	0.0924	19.02	0.0011	Significant
A2	0.2033	1	0.2033	41.84	< 0.0001	Significant
B2	0.4384	1	0.4384	90.22	< 0.0001	Significant
Residual	0.0535	11	0.0049
Lack of Fit	0.0534	9	0.0059	177.96	0.0056	Significant
Pure Error	0.0001	2	0.0000
Corresponding Total	6.22	16
Response 2 – Theoretical plates
Model	2.045E + 08	6	3.408E + 07	730.19	< 0.0001	Significant
A-Volume of organic phase	1.957E + 08	1	1.957E + 08	4193.76	< 0.0001	Significant
B-Flow rate	1.347E + 06	1	1.347E + 06	28.87	0.0003	Significant
C-Column oven temperature	4.010E + 05	1	4.010E + 05	8.59	0.0150	Significant
AB	1.285E + 06	1	1.285E + 06	27.53	0.0004	Significant
A2	4.459E + 06	1	4.459E + 06	95.55	< 0.0001	Significant
B2	3.693E + 05	1	3.693E + 05	7.91	0.0184	Significant
Residual	4.667E + 05	10	46,666.46
Lack of Fit	4.290E + 05	8	53,626.75	2.85	0.2857	Insignificant
Pure Error	37,650.67	2	18,825.33
Response 3 – Tailing factor
Model	0.7948	7	0.1135	19.88	< 0.0001	Significant
A-Volume of organic phase	0.1413	1	0.1413	24.74	0.0008	Significant
B-Flow rate	0.0028	1	0.0028	0.4922	0.5007	Significant
C-Column oven temperature	0.0000	1	0.0000	0.0040	0.9508	Significant
AB	0.0325	1	0.0325	5.69	0.0408	Significant
BC	0.1891	1	0.1891	33.12	0.0003	Significant
A2	0.0422	1	0.0422	7.39	0.0236	Significant
B2	0.3121	1	0.3121	54.66	< 0.0001	Significant
Residual	0.0514	9	0.0057
Lack of Fit	0.0500	7	0.0071	10.20	0.0921	Insignificant

The following regression equations generated provides a practical tool for predicting responses (retention time, theoretical plates and tailing factor) variations, supporting method robustness and operability within the validated design space.

The response surface plots illustrate the significant and interactive influences of key chromatographic variables on critical method parameters. Figure 3 depicts response surface, and contour plots and visually illustrates the correlation between CQAs and independent factors, highlighting the substantial impact of organic phase volume (ACN) compared to flow rate and column oven temperature. For retention time, the plots (Figs. 3a & d) reveal a marked nonlinear reduction with increasing organic phase volume and flow rate, where their main, interaction, and quadratic effects collectively govern analyte elution speed, enabling optimized method throughput while maintaining chromatographic integrity, as reflected by a strong model fit (R² = 0.9914). Regarding theoretical plates, three-dimensional surfaces (Figs. 3b & e) show maximum column efficiency at intermediate levels of organic phase and flow rate, with both factors and their interaction shaping a plateau of high plate counts; excessive adjustment leads to reduced efficiency, consistent with significant quadratic terms and excellent predictability (R² = 0.9977). For tailing factor, the response surfaces (Figs. 3c & f) highlight the dominant impact of organic phase volume, particularly at elevated flow rates, on peak symmetry, with significant quadratic and interaction effects delineating conditions where, minimal tailing is achieved, supporting robust peak shapes across the examined range (R² = 0.9393). Together, these surfaces provide an insightful framework for fine-tuning chromatographic conditions to balance speed, efficiency, and peak quality in the analytical method.

Fig. 3.

Contour plots viz. retention time (a), Theoretical plate counts (b), Tailing factor (c) and 3D response surface plots viz. retention time (d), Theoretical plate counts (e), Tailing factor (f)

Method operable design region and model validation

The Method Operable Design Region (MODR), as visualized in overlay plots (Figs. 4a & b), integrates the multivariate response surfaces of all critical quality attributes. The MODR is delineated as the yellow-shaded area where retention time, theoretical plates, and tailing factor simultaneously meet all ATP criteria. This operable region represents a “sweet spot” for method operation, within which robust analytical performance is assured despite minor fluctuations in chromatographic parameters. Defining this MODR not only ensures consistent performance but also provides operational flexibility, facilitating reproducibility and regulatory compliance in routine quality control settings. Model validation was performed by comparing predicted responses from the developed UHPLC method models with experimental results obtained at specific chromatographic conditions within the MODR (organic phase 12%, flow rate 0.3 mL/min, column oven temperature 40 ℃) showed in Table 4. The predicted and experimental values for retention time (1.20 vs. 1.21 min), theoretical plates (7840 vs. 7423), and tailing factor (1.59 vs. 1.55) demonstrated strong concordance. To statistically evaluate the agreement, a paired t-test was conducted on triplicate measurements of the experimental values compared to predicted values (considering predicted as theoretical means)[32]. The t-test results showed no significant differences for retention time (t = 0.87, p > 0.40), theoretical plates (t = 1.98, p > 0.10), or tailing factor (t = 1.13, p > 0.35), confirming the predictive models reliably estimate actual chromatographic responses within experimental variability. This strong agreement validates the robustness of the response surface models and supports their practical application for method control and optimization following AQbD principles.

Fig. 4.

Overlay plot showing the design space with optimal analytical conditions (a) flow rate vs volume of organic phase, (b) Column over temperature vs volume of organic phase

Table 4.

Results for model validation

Factors			Response	Tiopronin
Volume of organic phase (%)	Flow rate (mL/min)	Column oven temperature (℃)		Predicted value	Experimental value (n = 3)
12	0.3	40	Retention time (min)	1.20	1.21
			Theoretical plates (N)	7840	7423
			Tailing factor	1.59	1.55

Method validation

The proposed UHPLC method was validated following ICH guidelines and PDA Tech report on cleaning validation. System suitability assessment covered retention time, tailing factor, and theoretical plate count, all meeting acceptable criteria. Percentage relative standard deviations (RSDs) for peak area and retention time were within the acceptable range (< 10%), and theoretical plate counts exceeded 2000. Specificity was confirmed by the absence of interference at the Tiopronin peak retention time in chromatograms obtained from various solutions and swab samples (Fig. 5). A straight-line calibration curve for Tiopronin, along with the determination of LOD and LOQ, affirmed the method’s linearity (Supplementary Fig. 1). Method accuracy was further validated through recovery tests at three concentrations as described in methodology section. The developed UHPLC method for quantifying tiopronin residues demonstrated high accuracy across various surfaces, including stainless steel, silicon rubber, and acrylic sheets, consistent with PDA Tech Report 29 guidelines [37]. According to the report, swab recovery percentages of 70% or more are acceptable when qualifying the sampling method without corrections to limits or results. In this study, swab recovery from stainless steel and silicon rubber surfaces met these criteria, with recoveries ranging from 88.9% to 101.2%, indicating reliable performance for cleaning validation [38–40]. For acrylic sheets, lower recoveries were observed at the 20% spiking level, averaging 72.1%, which is still within the acceptable range. However, recovery at higher spiking levels (100% and 150%) improved, with averages of 94.6% and 96.1%, demonstrating the method’s accuracy (Table 5). These findings suggest that while recovery may vary by surface type, the method is effective for cleaning validation, ensuring residue detection aligns with regulatory standards and prevents cross-contamination. Overall, the UHPLC method meets the required swab recovery threshold for cleaning validation, ensuring it is suitable for routine use in pharmaceutical manufacturing, particularly on surfaces where high recovery rates are essential for accurate residue quantification. The precision of the developed UHPLC method for tiopronin quantification was evaluated through precision studies (Table 6), in line with PDA Tech Report 29 [37]. Intraday precision showed %RSD values of 0.34–3.84%, while interday precision ranged from 0.91–4.06%, demonstrating consistency across different days and concentration levels. These results confirm the method’s precision and suitability for routine cleaning validation, where pass/fail criteria are used, and precision at the residue limit is essential. Robustness was evaluated using a Fractional Factorial Design (FFD), allowing efficient assessment of factor combinations while minimizing experimental runs and maintaining the ability to assess factor effects and interactions (Table 7). This streamlined approach enabled the identification of critical factors affecting method robustness and optimization of method conditions. The Pareto chart derived from the FFD provides a visual representation of the relative influence of critical method parameters on system responses, such as peak area, theoretical plates, and tailing factor. Each bar in the chart corresponds to a factor or interaction effect, with its length reflecting the magnitude of its impact. The chart includes a reference line (t value limit) indicating statistical significance, enabling clear identification of parameters that have a meaningful effect on method performance. For the proposed UHPLC method, the Pareto chart highlighted flow rate, volume of ACN, injection volume and column oven temperature as the most influential factors, with interactions among these variables showing negligible or non-significant effects (Fig. 6). The results align with prior risk assessments, validating the robustness of the selected factors. By prioritizing these critical factors, the Pareto chart informed the refinement of method conditions to ensure robustness and reliability. The insights derived from the FFD Pareto analysis also support the method’s resilience to minor variations in operational parameters, making it suitable for routine application in cleaning validation. Overall, the Pareto chart serves as a powerful tool for systematically understanding and optimizing method performance within the AQbD framework.

Fig. 5.

Specificity studies, a) standard b) Diluent c) placebo d) blank swab e) Stainless steel coupon surface swab sample f) Silicon rubber surface swab sample g) Acrylic sheet surface swab sample h) Alconox sample and i) Liquinox sample

Table 5.

Accuracy studies

Accuracy level	Sample		Amt. spiked (µg/mL)	Amt. recovered (µg/mL)	% Recovery	Avg. % Recovery
Swab recovery
20%	1		0.404	0.394	97.5	98.2
	2			0.404	100
	3			0.392	97
100%	1		2.020	1.949	96.5	96.9
	2			1.967	97.4
	3			1.955	96.8
150%	1		3.030	2.952	97.4	96.9
	2			2.925	96.5
	3			2.934	96.8
Stainless-steel surface recovery
20%	1		0.404	0.38	94.1	95.4
	2			0.384	95
	3			0.392	97
100%	1		2.020	1.799	89.1	88.9
	2			1.796	88.9
	3			1.793	88.8
150%	1		3.030	2.926	96.6	96.6
	2			2.933	96.8
	3			2.924	96.5
Silicon rubber recovery
20%	1		0.404	0.414	98.6	99.8
	2			0.425	101.2
	3			0.418	99.5
100%	1		2.020	1.97	93.8	94.2
	2			1.979	94.2
	3			1.989	94.7
150%	1		3.030	3.009	95.5	95.5
	2			3.019	95.8
	3			2.996	95.1
Acrylic sheet recovery
20%	1	0.404		0.291	72.03	72.11
	2			0.289	71.53
	3			0.294	72.77
100%	1	2.020		1.9	94.1	94.6
	2			1.918	95
	3			1.912	94.7
150%	1	3.030		2.911	96.1	96.1
	2			2.914	96.2
	3			2.913	96.1

Table 6.

Intraday and Interday precision

Concentration Level (% of target)	Intraday Precision (%RSD) (n = 3)	Interday Precision (%RSD) (n = 3)
20% (0.4 µg/mL)	3.84%	4.06%
100% (2.0 µg/mL)	0.37%	0.91%
150% (3.0 µg/mL)	0.34%	1.07%

Table 7.

Fractional factorial design for robustness studies

Sr. No	Volume of organic phase (%)	Flow rate (mL/min)	Temperature (℃)	Injection volume (µL)	Area	Tailing factor	Theoretical plates
1	10	0.35	38	7	4864	1.62	5016
2	14	0.35	42	7	5043	1.87	7186
3	10	0.35	42	5	3477	1.63	6699
4	14	0.25	42	5	4805	1.52	10,281
5	14	0.35	38	5	3407	1.64	9249
6	14	0.25	38	7	6839	1.61	8302
7	10	0.25	42	7	6686	1.48	5200
8	10	0.25	38	5	4799	1.40	7630

Fig. 6.

Pareto charts for robustness studies

Greenness and sustainability assessment

Comprehensive greenness and sustainability (whiteness) evaluation was conducted for the proposed UHPLC method and compared against reported HPLC methods [23, 24] using three established tools: AGREE, BAGI, and RGB12. AGREE, grounded in the 12 principles of Green Analytical Chemistry, provides a visual greenness assessment and a score between 0 (poor) and 1 (ideal). As shown in Supplementary Table 2, our UHPLC method achieved an AGREE score of 0.67, substantially higher than those of the reported HPLC methods, primarily due to its lower solvent consumption, shorter analysis time, and reduced sample volume. In contrast, the reported methods displayed lower AGREE scores because of their greater solvent use, longer runtimes, and larger injection volumes. The detailed AGREE pictogram for our method further demonstrates adherence to the full spectrum of green analytical principles, emphasizing its minimal environmental footprint.

To further substantiate the environmental merit of our approach, we assessed the methods using the blue applicability grade index (BAGI) and the RGB12 tool. The proposed method’s BAGI score of 85.0, as compared to 57.5 and 60.0 for the reported methods, reflects its low reagent toxicity, reduced waste and energy requirements, and robust analytical performance. The RGB12 score (82.1 for the proposed method) likewise outperformed conventional HPLC approaches (66.5 and 67.8, respectively), mirroring advantages in analytical efficiency, environmental/safety attributes, and practical/economic feasibility—thanks to the method’s use of benign solvents, elimination of preconcentration or derivatization steps, and energy-efficient, low-volume operation.

Collectively, the AGREE, BAGI, and RGB12 results provide a multidimensional, evidence-based perspective confirming that the proposed UHPLC method is not only analytically robust, but also superior in environmental and operational sustainability. Supplementary Table 3 and 4 presents a side-by-side comparison of the greenness scores, highlighting the substantial ecological and practical benefits of adopting this UHPLC approach for routine pharmaceutical cleaning validation. This alignment with modern sustainability standards reinforces the method’s value amid the evolving demands for greener analytical practices in industry.

Conclusion

In conclusion, this study successfully developed and validated a novel reverse-phase ultra-performance liquid chromatography method for quantifying tiopronin residues in swab samples from manufacturing equipment as part of a cleaning validation process. Utilizing an Analytical Quality by Design approach, the method was optimized to meet stringent cleaning validation standards while ensuring eco-friendliness, as demonstrated by a greenness assessment based on Green Analytical Chemistry principles. The method exhibited excellent specificity, precision, accuracy, and linearity within a concentration range of 0.302 to 3.027 µg/mL, with a rapid retention time of 1.3 min. The detection limit and quantification were established at 0.100 µg/mL and 0.301 µg/mL, respectively. This UHPLC method fills a vital gap in cleaning validation for tiopronin manufacturing equipment and ensures compliance with regulatory requirements, such as those set by the FDA and EU guidelines. Its robustness and eco-friendly characteristics make it suitable for routine use in quality control laboratories. Ultimately, the method improves product safety by ensuring that residues are effectively removed, thus preventing cross-contamination and enhancing overall manufacturing efficiency in pharmaceutical production.

Author contributions

Jagdish Gohel – Main author – Planning and execution of concept and practical work Ajay Patel – Co-author – Helping hands in analytical method development and validation Rajendra Kotadiya – Corresponding author – Supervised and reviewed whole work/draft.

Funding

Open access funding provided by Parul University. None received.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.