A dual-detection HPTLC platform: combining smartphone-based imaging and densitometry for the analysis of diazepam, its metabolite and degradation product

TL;DR

This paper introduces a new HPTLC method using smartphone imaging to detect diazepam, its metabolite, and degradation product, offering a faster and eco-friendly alternative to traditional methods.

Contribution

A smartphone-based HPTLC method is developed for the simultaneous detection of diazepam, oxazepam, and ACB, validated against densitometric analysis.

Findings

The HPTLC/smartphone method showed linearity for DZP and ACB from 3.0 to 35.0 µg/spot and for OXP from 5.0 to 35.0 µg/spot.

The method was successfully applied to quantify DZP in pharmaceutical formulations and spiked human plasma samples.

Green analytical metrics confirmed the method's environmental compatibility and alignment with green and white analytical chemistry principles.

Abstract

In recent years, high-performance thin-layer chromatography (HPTLC) has gained prominence as a cost-effective, straightforward, and dependable analytical technique, particularly in forensic and pharmaceutical laboratories. With the rapid evolution of smartphone technologies, a novel dimension in analytical detection has emerged. Advent of smartphones with superior imaging modalities combined with simplicity of use and easy transition into the healthcare ecosystem has made the existing benchtop-based techniques much sleeker, cost-effective and rapid screening approaches. The developed and validated HPTLC method employing smartphone camera detection for simultaneous determination of Diazepam (DZP), its metabolite Oxazepam (OXP) and its degradation product 2-methylamino-5-chlorobenzophenone (ACB) was intended to meet this requirement and provide a useful replacement for the usual densitometric analysis. The chromatographic separation was performed on silica gel HPTLC plates with green mobile phase of heptane: ethyl acetate (7.0:3.0, v/v). After chromatographic development, the Dragendorff’s reagent was applied for visualization, and the plates were photographed with smartphone camera. Spot intensities were quantitatively analyzed using ImageJ software in the concentration range 3.0–35.0 µg/spot for both DZP and ACB, and 5.0–35.0 µg/spot for OXP. The data developed by the proposed method were compared with benchtop densitometric method. Linearity was established from 0.2 to 1.0 µg/spot for the three analytes detected at 230.0 nm using HPTLC/densitometry while the HPTLC/smartphone method exhibited linearity from 3.0 to 35.0 µg/spot for both DZP and ACB, and 5.0–35.0 µg/spot for OXP. The developed HPTLC/smartphone method showed that it could successfully be employed to quantify DZP in pharmaceutically marketed formulations in terms of speed, eco-friendly and simplicity. Additionally, the enhanced sensitivity of the densitometric technique enabled successful determination of DZP and OXP in spiked human plasma samples using ACB as an internal standard. To assess the environmental and practical merit of the methods, greenness and sustainability were evaluated using the Green Analytical Procedure Index (GAPI), Analytical Greenness Metric (AGREE), White Analytical Chemistry (WAC), and Blue Applicability Grade Index (BAGI). Results confirmed the developed methods’ analytical efficiency, environmental compatibility, and alignment with the principles of green and white analytical chemistry.

The online version contains supplementary material available at 10.1186/s13065-025-01711-y.

Figures

Chemical structures of (a): Diazepam (DZP), (b): Oxazepam (OXP), and (c): 2-Methylamino-5-chlorobenzophenone (ACB)

Zero-order absorption spectra of 5.0 µg/mL of DZP, ACB, and OXP each, using methanol as blank

Demonstration of Image J quantification after visualizing the developed plates using Dragendorff’s reagent

Full text

Introduction

The non-medical use of benzodiazepines has become a growing global public health concern. Nearly 60 countries report increasing unauthorized consumption of sedatives and hypnotics, particularly benzodiazepines and structurally related compounds [1]. In Africa, illicit mixtures such as nyaope—often containing uncontrolled pharmaceuticals (including benzodiazepines), alcohol, and industrial solvents—further intensify health risks due to their unregulated composition [2].

Addiction is defined as a persistent pattern of substance use causing psychological, social, and functional impairment. Despite serious health and behavioral consequences, affected individuals often continue substance misuse [3]. Medication misuse refers to using prescription drugs contrary to medical guidance, without a prescription, or for non-therapeutic purposes [4]. According to the World Health Organization, unsupervised psychoactive drug use increases the likelihood of developing drug use disorders, which are associated with significant morbidity, mortality, and economic burden. In 2022, an estimated 270 million people (5.5% of the global population aged 15–64) used drugs, of whom approximately 35 million suffered from drug use disorders. Substance use contributes to about half a million deaths annually and accounts for more than 42 million disability-adjusted life years (1.3% of the global disease burden) [5]. Drug-facilitated sexual assault (DFSA) involves administering substances to incapacitate victims, impair resistance, and reduce memory of the assault. DFSA cases have increased globally due to the accessibility and low detectability of the involved drugs. Non-medical use of prescription sedatives—including benzodiazepines, Z-drugs, neuroleptics, antihistamines, and related agents—has important clinical and forensic implications [6]. Diazepam (DZP), a widely misused benzodiazepine (Fig. 1a), is primarily prescribed for its anxiolytic effects [7]. Benzodiazepines enhance γ-aminobutyric acid (GABA) neurotransmission by increasing GABA binding to the GABA_A receptor, thereby amplifying inhibitory signaling [8]. Diazepam exhibits a high potential for abuse, supported by epidemiological data, clinical experience, and user reports [9]. Its metabolites persist for extended periods—up to 6 weeks in urine, 48 h in plasma, and 10 days in oral fluids—and tolerance often leads to dosage escalation and accumulation [10]. Given the health and legal implications of benzodiazepine misuse, rapid, reliable, and cost-effective analytical methods are needed.

High-performance thin-layer chromatography (HPTLC) is widely used for detecting and quantifying diazepam due to its simplicity, low cost, and capacity for simultaneous multi-sample analysis. When combined with UV/fluorescence detection and software, HPTLC achieves sensitivity comparable to HPLC [11, 12]. Densitometry remains central to TLC detection, and enhancing detection strategies continues to be important [13]. Smartphone-based analytical tools have recently gained attention due to their portability, affordability, and processing capabilities. These devices can acquire optical or electrochemical signals and analyze them through dedicated applications [14, 15]. TLC quantification using ImageJ, an accessible open-source NIH software, offers a practical approach for laboratories with limited resources and requires minimal technical expertise [16, 17]. Its integration into TLC imaging supports low-cost and reliable analysis [18, 19]. A thorough review of literature has revealed various TLC [12, 20–22], HPLC methods [23–26], GC [27–30], CE [31–34], spectroscopic techniques [35, 36], chemiluminescence [37, 38], potentiometry and voltammetry [10, 39–41] for determination of diazepam in different matrices. No chromatographic method has been proposed for determination of diazepam (DZP), its degradation product 2-methylamino-5-chlorobenzophenone (ACB) (Fig. 1c) and metabolite Oxazepam (OXP) (Fig. 1b).Fig. 1. Chemical structures of (a): Diazepam (DZP), (b): Oxazepam (OXP), and (c): 2-Methylamino-5-chlorobenzophenone (ACB)

Our goal was to develop and optimize an eco-conscious separation approach for the simultaneous analysis of DZP along with its degradation product, ACB and metabolite, OXP in spiked human plasma samples using the HPTLC technique. In addition to the traditional densitometric detection, efficient application of the often-utilized smartphone’s camera for quick on-site quantitation of the mentioned compounds was followed. Alongside this, digital photos were processed using ImageJ software, and a new, basic daylight illumination design was created.

Experimental

Chemicals, reagents and pharmaceutical samples

HPLC-grade ethyl acetate and methanol were procured from Sigma-Aldrich (Darmstadt, Germany), while heptane was obtained from RECTAPUR^®^ (France). Diethyl ether was sourced from Fluka Chemie GmbH (Switzerland). Diazepam United States Pharmacopeia (USP) reference standard, 2-methylamino-5-chlorobenzophenone, and oxazepam were all obtained from Sigma-Aldrich Chemie GmbH (Steinheim, Germany).

Dragendorff’s reagent was freshly prepared by combining 5 mL of a 1.7 g% w/v solution of basic bismuth nitrate in 20% v/v acetic acid, 5 mL of a 40 g% potassium iodide solution, 20 mL of glacial acetic acid, and 70 mL of distilled water, following established protocols [42].

Neuril^®^ ampoules (Batch No. 122305), intended for intramuscular or intravenous administration, were provided by Chemipharm Pharmaceutical Industries in collaboration with Memphis for Pharmaceuticals and Chemical Industries, Egypt. Each ampoule was labeled to contain 10 mg of diazepam per 2 mL solution.

Human plasma samples were supplied by the Holding Company for Biological Products and Vaccines (VACSERA, Giza, Egypt). Upon arrival, plasma samples were stored at 2–8 °C and processed within four hours of receipt. Aliquots were transferred into low-binding, single-use polypropylene cryovials and stored at − 80 °C for long-term preservation. All sample handling and processing were conducted under biosafety level 2 (BSL-2) conditions to ensure safety and integrity.

Stock standard solutions

Individual stock solutions of DZP, its degradation product 2-methylamino-5-chlorobenzophenone (ACB), and its principal metabolite oxazepam (OXP) were each prepared at a concentration of 1.0 mg/mL. Precisely 0.1 g of each compound was weighed and transferred into separate 100 mL volumetric flasks. The volume in each flask was then adjusted to the mark using methanol as the solvent. The resulting solutions were stored under refrigerated conditions at 8 °C and remained chemically stable for several months without signs of degradation.

Instruments and chromatographic conditions

Chromatographic separation was performed using normal-phase HPTLC on pre-coated silica gel 60 F254 aluminum plates (20 × 20 cm, 0.25 mm thickness; E. Merck, Darmstadt, Germany). Prior to analysis, the HPTLC plates were conditioned by overnight washing with methanol, followed by activation in a laboratory oven at 80 °C for 20 min.

Sample application was carried out using a CAMAG Linomat 5 autosampler equipped with a 100 µL CAMAG microsyringe (Muttenz, Switzerland), delivering 10 µL of sample solution as 4 mm bands, separated by 6 mm, and placed 1.5 cm from the bottom edge of the plate. Densitometric quantification was conducted using a CAMAG TLC Scanner 3 (S/N1302139) controlled by winCATS^®^ software. Detection was performed under UV light (254 nm) using a UV lamp (Vilber Lourmat, Marne-la-Vallée, France), operated within a ventilated cabinet. The scanner operated with a slit dimension of 3 × 0.45 mm and a scanning speed of 20 mm/s.

The mobile phase consisted of heptane and ethyl acetate in a volume ratio of 7:3 (v/v). This mixture was transferred into a twin-trough glass chamber and left for 30 min to allow for saturation and equilibration. Subsequently, the HPTLC plates were inserted into the chamber and developed to a height of 8.5 cm. Upon completion, the plates were air-dried at room temperature and scanned at 230.0 nm for quantification of analyte peak areas.

For visual detection via smartphone-assisted analysis, the developed plates were immersed in freshly prepared Dragendorff’s reagent for 30 s to visualize the analytes as intense, orange-colored spots. The plates were then left to dry for a minimum of five minutes. Imaging was carried out using a Samsung Galaxy A15 smartphone equipped with a 12 MP rear camera (f/1.8 aperture). The plates were placed in a cardboard box fitted with a DESAGA multipurpose UV lamp (254/366 nm) for uniform lighting. The lamp was pre-warmed for five minutes to stabilize illumination intensity and minimize variability. Images were captured in daylight mode from a distance of 15 cm using the following optimized settings: shutter speed 1/10 s, ISO 100, autofocus enabled, daylight white balance, and flash disabled.

Image analysis was performed using ImageJ software (version 153, NIH, USA). Captured images were loaded in RGB format, and the “Gels” analytical tool was employed. Using the rectangular selection tool, uniformly sized rectangles were drawn over each stained lane. The “Plot Lanes” function was used to generate densitograms, while the “Straight” and “Magic Wand” tools enabled precise identification and integration of peak areas corresponding to individual analyte spots.

Construction of calibration curve

HPTLC/densitometry

Calibration curves for DZP, ACB, and OXP were established by preparing serial dilutions from individual stock solutions (1.0 mg/mL). Aliquots corresponding to 0.2–1.0 mg were accurately transferred into separate 10 mL volumetric flasks and diluted to volume with methanol. Each concentration was applied in triplicate as 10 µL compact spots (equivalent to 0.2–1.0 µg/spot) on pre-conditioned HPTLC silica plates, as previously described in Sect. 2.3.

Following chromatographic development and densitometric scanning, the mean peak areas were plotted against the corresponding concentrations to generate calibration curves. Regression equations were calculated based on linear least-squares analysis.

HPTLC/smartphone

For the smartphone-based approach, stock solutions of DZP, ACB, and OXP (5.0 mg/mL each) were used to prepare working standards. Aliquots ranging from 0.6 to 7.0 mL of DZP and ACB were transferred into 10 mL volumetric flasks to yield final concentrations of 0.3–3.5 mg/mL (equivalent to 3.0–35.0 µg/spot at 10 µL application). For OXP, volumes ranging from 1.0 to 7.0 mL were used to obtain concentrations of 0.5–3.5 mg/mL (equivalent to 5.0–35.0 µg/spot). All solutions were diluted to volume with methanol.

Each solution was applied as a 10 µL compact spot on HPTLC plates. After development, color visualization, and drying, images were captured and analyzed using ImageJ software. The mean peak intensities were plotted against the corresponding concentrations to construct the calibration plots and derive the regression equations.

Solutions used for assessing robustness, accuracy, and precision were prepared accordingly, and all validation procedures were conducted in accordance with ICH guidelines to ensure analytical reliability and reproducibility.

Application to pharmaceutical dosage form

Five ampoules of Neuril^®^ (10 mg/2 mL) were transferred into a clean beaker. A volume of 0.60 mL was accurately measured and transferred into a 100 mL volumetric flask, and the volume was adjusted to the mark with methanol to yield a working solution of 0.03 mg/mL. For HPTLC/densitometric analysis, 10 µL of this solution was applied as a compact spot onto the HPTLC plate.

In parallel, the pharmaceutical dosage form was also analyzed using the HPTLC/smartphone-based method. For this purpose, a more concentrated solution of 0.5 mg/mL was prepared, and 10 µL of the sample was applied onto the plate for analysis.

TLC-densitometric method for assay of spiked human plasma samples

Stock standard solution of DZP and OXP

Primary stock standard solutions of DZP and OXP were prepared by accurately weighing 0.20 g of each pure compound into separate 100 mL volumetric flasks. The contents were diluted to volume with methanol to obtain concentrations of 2000.0 µg/mL. These solutions were mixed thoroughly and stored at 8 °C, remaining stable for several months. Secondary stock solutions at a concentration of 200.0 µg/mL were subsequently prepared by dilution with methanol.

Internal standard (ACB) stock standard solution

To prepare the internal standard (IS) primary stock solution of ACB (1000.0 µg/mL) in methanol, 0.1 g of pure ACB powder was accurately weighed into a 100 mL volumetric flask. The solution was diluted to the mark with methanol and mixed thoroughly. Secondary stock solution of 100.0 µg/mL for ACB was prepared by further dilution of the primary solution in methanol.

Working standard solutions (calibrators) of DZP and OXP

Working standard solutions for DZP and OXP were prepared by serial dilution of their 200.0 µg/mL secondary stock solutions in 10 mL volumetric flasks, resulting in nine calibrator concentrations: 4.0, 6.0, 8.0, 10.0, 12.0, 14.0, 16.0, 18.0, and 20.0 µg/mL. These calibrators were used to construct plasma calibration curves for quantitative analysis.

Internal standard (ACB) working solution

A working solution of ACB at 10.0 µg/mL was prepared by transferring 1 mL of the 100.0 µg/mL secondary stock solution into a 10 mL volumetric flask, diluting to the mark with methanol, and mixing thoroughly.

Calibration samples in human plasma

Fresh calibration samples were prepared on the analysis day, nine calibration samples were prepared by adding 25 µL of each drug working standard solution to 450 µL of blank human plasma in separate centrifuge tubes. The spiked samples were sealed and kept at − 20 °C, then thawed to room temperature just before analysis.

Extraction procedure of human plasma samples

A 500 µL aliquot of spiked human plasma was combined with 25 µL of the ACB internal standard working solution and vortexed for 1 min to ensure thorough mixing. Subsequently, 1 mL of diethyl ether was added, and the mixture was vortexed again for an additional minute. Phase separation was achieved by centrifugation at 4000 rpm in 10 min. The resulting upper organic layer was carefully collected and evaporated to dryness under a gentle stream of nitrogen gas. The dried residue was then reconstituted with 10 µL of methanol, and the entire volume was applied as a compact spot onto a pre-conditioned HPTLC plate. All subsequent analytical procedures followed the protocol previously described in Sect. 2.3.

Results and discussion

HPTLC method development and optimization

TLC is widely recognized as an effective analytical separation technique, particularly valuable for the screening and identification of analytes in diverse sample matrices. Among its notable advantages are its low operational cost, minimal sample cleanup requirements, and the ability to perform analyses with simplified sample preparation. TLC is especially recommended for the analysis of compounds that exhibit low detectability and require post-chromatographic derivatization for visualization. Historically, TLC has been extensively applied in the characterization of dyes and inks, as well as in monitoring contaminants in industrial chemicals. Its utility extends to the detection of toxic substances, pharmaceutical residues, and other trace-level analytes in biological fluids [43].

From an environmental standpoint, the adoption of the “three R” principle, Replace, Reduce, Reuse can enhance the eco-efficiency of high-performance TLC (HPTLC). This is achieved by substituting hazardous solvents such as chloroform, acetonitrile, and tetrahydrofuran with greener alternatives like methanol [44]. In the present study, various mobile phase compositions were evaluated using solvents such as methanol, ethanol, acetone, butanol, water, heptane, and ethyl acetate. These solvents were selected based on their favorable toxicity and environmental profiles, aligning with the principles of Green Analytical Chemistry. The final mobile phase composition was chosen not only for its effective separation but also because it minimized the use of the most hazardous components evaluated. A systematic evaluation of a number of solvent systems was performed to attain optimal analytes separation. Preliminary trials employed ethanol and ethyl acetate in various volumetric ratios (5:5, 3:7, 7:3, v/v); however, these mixtures failed to achieve adequate separation, even after pH adjustment using acetic acid or ammonia. Solvent systems such as (ethyl acetate: butanol: acetic acid) and (methanol: water: ammonia) at a ratio of 5:5:0.1 (v/v) were also evaluated, but they did not provide satisfactory separation of the analytes and produced tailing peaks. Consequently, alternative solvent systems composed of heptane and ethyl acetate were investigated across different ratios (5:5, 3:7, and 7:3, v/v).

Among the tested combinations, a mobile phase composed of heptane and ethyl acetate in a volumetric ratio of 7:3 (v/v) demonstrated optimal performance, enabling the clear resolution of DZP, ACB, and OXP as compact, well-defined spots. The addition of small volumes of 33% w/w ammonia solution or glacial acetic acid to the mobile phase did not significantly affect the separation quality or spot morphology. The measured retardation factor (Rf) values were 0.22 for DZP, 0.12 for OXP, and 0.69 for ACB (± 0.02), all yielding sharp, distinct peaks with excellent resolution, as shown in Fig. 2. The optimization of spot dimensions was another key factor to be accounted for and given due attention. The process of ordinary diffusion may induce minor dispersion of the applied spots. Furthermore, application of narrowed spot widths may provoke silica overloading, giving rise to peak tailing and faulty analytical outcomes. Accordingly, meticulous optimization of spot width and inter-spot spacing is indispensable to evade overlap between adjacent spots and confirm the accuracy of the results. An optimal spot width of 4 mm with inter-spot spacing of 6 mm was employed.Fig. 2a 2D HPTLC chromatogram showing separated peaks of DZP, OXP and ACB (1.0 µg/spot, for each) using a developing phase composed of heptane and ethyl acetate (7.0: 3.0, by volume) with detection at 230.0 nm. b. 3D scanning profile of the HPTLC chromatogram showing 0.2–1.0 µg/spot of DZP, OXP, and ACB

System suitability parameters were also evaluated and were found to comply with the requirements outlined by the United States Pharmacopeia (USP) guidelines [45], as detailed in Table 1.Table 1. Parameters required for system suitability testing of the proposed HPTLC methodParameterDZPOXPACBRetardation factor (R_f_)0.22 ± 0.020.12 ± 0.020.69 ± 0.02Resolution (R_S_)^a^2.007.00Capacity factor (K^’^)^b^3.557.330.45Selectivity (ɑ)^c^2.077.89Peak width0.700.600.90Tailing factor (T)^d^1.171.000.90^a^R_s_= 2*(R_f2_-R_f1_)/(W_1_ + W_2_)^b^K^’^= (1-R_f_)/R_f_^c^ɑ= K^’^2-K^’^1^d^T = W_0.05_/2f, where W_0.05_ is the width of the peak at 5% height and f is the distance from the peak maximum to the leading edge of the peak

Optimization of densitometric quantification

A scanning wavelength of 230.0 nm was selected for the quantification of the three components based on their UV absorption spectra, which are displayed in Fig. 3. The selection criteria depend on the maximized sensitivity, baseline resolution, spot integrity and noise reduction. Validation using triplicate plate scans demonstrated reproducible retention factors and consistent peak areas for all analytes, confirming the robustness of the optimized wavelength.Fig. 3. Zero-order absorption spectra of 5.0 µg/mL of DZP, ACB, and OXP each, using methanol as blank

Development of HPTLC/smartphone method

The widespread adoption of smartphones and their advancing technological capabilities have opened new avenues for their integration into analytical methodologies and pharmaceutical quality control. Leveraging the built-in camera of a smartphone offers a practical, portable, and cost-effective alternative to conventional analytical instruments and proprietary software. In this study, the smartphone-assisted HPTLC method was employed by capturing images of developed HPTLC plates using a smartphone camera, thereby enabling rapid and straightforward chromatographic data analysis.

To enhance the precision and reproducibility of this approach, several imaging configurations and key photographic parameters were systematically investigated and optimized. A critical aspect of the setup involved minimizing environmental variability such as fluctuations in light intensity and inconsistencies in imaging distance. To this end, a custom-designed cardboard enclosure was constructed to house the UV light source. The enclosure featured a rectangular opening (4 × 2 cm) aligned with the lamp’s surface and positioned exactly 15 cm from the HPTLC plate, ensuring uniform and stable illumination during image capture. This design effectively mitigated the influence of ambient light and stray shadows, which could otherwise compromise image quality and analytical accuracy.

Additionally, the vertical distance between the HPTLC plate and the smartphone camera was carefully considered during image acquisition. It was observed that increasing the distance beyond the optimized height led to a slight reduction in image resolution, potentially affecting spot definition and peak area determination. Therefore, precise control of imaging parameters was essential to achieving consistent and high-quality chromatographic results using the smartphone-based system.

Screening for a suitable staining reagent

Initial trials indicated that DZP and OXP lack intrinsic chromophores, making it impossible to directly quantify the plate using daylight illumination mode, while ACB showed bright, yellow-colored spots. To get around these restrictions, several TLC staining agents were investigated to get bright, stable spots for the two medications under study that would not fade over time. Iodine crystals were employed frequently as staining agent [46]. However, the intensity of generated color was found to be extremely changeable, time-sensitive and quickly fading away. The amino groups in the studied drugs indicated the use of ninhydrin, Dragendorff’s and phosphomolybdic acid as substitute staining agents. Unfortunately, ninhydrin and phosphomolybdic acid reagents required plate heating. Consequently, Dragendorff’s reagent was used to visualize the spots [47]. Orange- colored spots of the separated drugs were visible when the eluted plate was stained with Dragendorff’s reagent [42]. The color reaction between the studied analytes and dragendorff’s reagent is illustrated as per the supplementary information (Scheme S1).

Optimization of image shooting parameters and analysis

To standardize camera exposure across all images, the smartphone camera was manually configured by switching from automatic to professional (Pro) mode. Three key imaging parameters shutter speed, ISO sensitivity, and white balance were individually optimized. The flash was disabled to avoid glare, ISO sensitivity was set to 50 to minimize image noise, and a shutter speed of 1/10 s was maintained to ensure adequate exposure and sharpness. Under daylight illumination conditions, the white balance was manually set to 6000 K, which provided accurate color rendering for both analyte spots and plate background.

These carefully adjusted settings contributed to consistent image quality, enhancing resolution and analytical clarity. Following image acquisition, the captured photographs were imported into the desktop version of ImageJ software for analysis. The software automatically recognized each sample track and differentiated the spot signals from the background. The intensity of each spot was quantified and converted into a corresponding analytical parameter—namely, the peak area. This process enabled the transformation of visual data into measurable outputs. A demonstration of the ImageJ-based image analysis workflow is shown in Fig. 4.Fig. 4. Demonstration of Image J quantification after visualizing the developed plates using Dragendorff’s reagent

Method validation

According to ICH criteria, the proposed techniques were thoroughly validated in terms of linearity range, limit of detection (LOD), limit of quantification (LOQ), accuracy, precision and robustness [48].

Linearity and range

For HPTLC/densitometry, the integrated peak areas and the associated concentrations in a range of 0.2–1.0 µg/spot for DZP, OXP, and ACB were shown to be linearly correlated. However, the proposed HPTLC/smartphone approach demonstrated linearity across a concentration range equals to 3.0–35.0 µg/spot for DZP, ACB and 5.0–35.0 µg/spot for OXP. Various calculated regression parameters are summarized in Table 2.Table 2. Assay and validation parameters for quantification of DZP, ACB, and OXP by applying the proposed HPTLC/ densitometric and smartphone methodsParameterHPTLC/densitometric methodHPTLC/smartphone methodDZPACBOXPDZPACBOXPRange (µg/spot)0.2–1.00.2–1.00.2–1.03.0–35.03.0–35.05.0–35.0Slope9731.886538821219.37226.63116.64Intercept1521.11242.91742.76223.3710.22976.08Correlation coefficient (r)0.999950.999950.999950.99960.99960.9997Accuracy (mean ± SD)100.04 ± 0.64100.21 ± 0.49100.18 ± 0.53100.21 ± 0.836100.87 ± 1.15399.96 ± 1.096LOD0.010.010.011.011.000.78LOQ0.030.030.033.053.022.37Precision (± RSD%)Repeatability0.4920.7080.8990.7360.7250.630Intermediate precision0.6110.8350.7271.8651.5241.412Robustness1.3581.2941.5671.4041.5771.783

Accuracy

DZP, OXP, and ACB were evaluated in triplicate at five different concentration levels. The two suggested approaches’ excellent accuracy was demonstrated by the acceptable values of mean recoveries that were attained, which fell between 98 and 102% Table 2.

Precision

To assess the variability in scanning performance of each method, intra-plate precision (repeatability) was first evaluated. For the HPTLC/densitometric method, repeatability was determined by scanning the same spots three times at three different concentration levels: 0.6, 0.7, and 0.9 µg/spot for DZP; 0.6, 0.7, and 1.0 µg/spot for ACB; and 0.5, 0.7, and 0.9 µg/spot for OXP. In the case of the smartphone-assisted method, repeatability was assessed by calculating the relative standard deviation (%RSD) for three concentration levels: 10.0, 15.0, and 25.0 µg/spot.

Within-laboratory precision was further evaluated by analyzing the same concentrations either on three separate HPTLC plates or across two consecutive days. The %RSD values for all tested concentrations under both methods were calculated, with results consistently remaining below 2%, as presented in Table 2. Notably, the %RSD values obtained via the smartphone-based approach were in strong agreement with those acquired using the conventional benchtop densitometric method, demonstrating comparable precision and analytical reliability.

Robustness

After implementing an intended change in mobile phase ratio with a value of ± 0.1 for heptane and ethyl acetate, the proposed HPTLC chromatographic conditions were reevaluated. Low RSD% values indicated that no notable variation in the system suitability parameters was found, Table 2. A minor shift in the scanning wavelength (± 1 nm) was also used to assess the robustness of densitometric readings. The effect of making small adjustments to photography parameters, such as shutter speed and ISO sensitivity, was evaluated for the HPTLC/smartphone approach. Table 2. showed that both approaches’ RSD% values were satisfactory, indicating their strong resilience.

3.5 utilization of the two proposed methods for neuril® ampoules assay

The suggested HPTLC/densitometry and HPTLC/smartphone methods were successfully used to identify the presence of DZP in Neuril^®^ ampoules, yielding admissible percent recoveries which confirmed the validity of the proposed HPTLC methods as demonstrated in Table 3.Table 3. Quantitative determination of DZP in Neuril^®^ ampoules (batch No.122305) by applying the proposed HPTLC methodsNeuril^®^ ampoules                Found*% ± SDHPTLC/densitometryHPTLC/smartphone99.74 ± 0.88100.22 ± 0.867^*^Average of 3 determinations

Determination of DZP, OXP in spiked human plasma utilizing the proposed HPTLC/densitometric method

A critical step in the development of bioanalytical methods is sample preparation, which can lead to notable improvements in the sensitivity and selectivity of the target medicines’ quantification in various biological matrices. The simplest, fastest, and least expensive method of preparing biological samples is protein precipitation, particularly when many samples need to be examined [49]. Unfortunately, DZP and its metabolites have a high binding affinity for plasma proteins (DZP, 98%) [50]. Therefore, the extraction procedure using diethyl ether was followed instead. After the plasma samples were prepared, the proposed HPTLC/densitometric approach was implemented to determine their DZP and OXP contents. It is worth mentioning that the peak plasma concentration (C_max_) of DZP approaches 300 ng/mL after ten mg oral dose, while it is 407 ng/mL for OXP following 15 mg oral dose [51, 52]. The screening for a suitable internal standard was a challenging step. Several compounds (alprazolam, tianeptine, clonazepam & carbamazepine) that shared physicochemical properties, such as pK_a_, log partition coefficient, and their structures closely matched that of the studied drugs (DZP & OXP) were investigated as potential internal standards. However, as shown in Fig. 5a, b, ACB was chosen as our internal standard since it demonstrated optimal resolution from the molecules under investigation, independent of plasma components. The regression equation parameters of the proposed HPTLC/densitometric approach for determining DZP and OXP in spiked human plasma samples are compiled in Table 4. As stated in FDA bioanalytical method validation guidance for industry, the obtained values fell within the approved limits [53].Table 4. Linearity validation parameters obtained by applying the suggested HPTLC-Spectro densitometric approach for determining DZP and OXP in spiked human plasmaParameterDZPOXPRange (ng/mL)200.0–1000.0200.0–1000.0Correlation coefficient (r)0.99950.9995Slope1.26440.9886Intercept0.34510.3856Fig. 5a 2D HPTLC chromatogram of a human plasma sample spiked with 500.0 ng/mL 2-Methylamino-5-chlorobenzophenone (ACB) as an internal standard. b 2D HPTLC chromatogram of a human plasma sample spiked with 700.0 ng/mL DZP, OXP along with 500.0 ng/mL of ACB as an internal standard

Statistical comparison between the two proposed methods with the reported method

A statistical evaluation was performed to compare the analytical performance of the two proposed HPTLC methods with a previously reported method [54]. At a 95% confidence level, both Student’s t-test and F-test were applied to assess differences in accuracy and precision. The results indicated no statistically significant differences between the methods, as evidenced by the calculated t and F values, which fell within the acceptable range, as summarized in Table S1.

To further confirm the equivalence of the methods, one-way analysis of variance (ANOVA) was conducted. The computed F value (0.430) was found to be less than the critical tabulated F value (5.143) at p = 0.05. Moreover, the obtained p-value exceeded 0.05, supporting the null hypothesis and indicating that no significant difference exists between the proposed methods and the reported method. These results are detailed in Table S2.

Greenness assessment of the proposed methods

Currently, the adoption of the concept of green analytical chemistry (GAC), which is the establishment of safe and eco-conscious procedures and methodologies, is one of the principal trends in analytical chemistry [55]. Evaluating the environmental impact of various analytical techniques about their compliance with green analytical chemistry principles has been crucial [56, 57]. To affirm the sustainability of the proposed methods in contrast with the reported spectrophotometric method, green metric tools GAPI, AGREE, and WAC were exploited [58–60]. The detailed input data employed in AGREE and GAPI metric tools for all compared methods are mentioned in the supplementary information (Table S3) and (Table S4), respectively. AGREE tool reflects quantitatively and visually the extent to which an analytical procedure aligns with the twelve principles of green analytical chemistry. An overall score, varying from zero to one, is revealed via the center of a circular interface, with greener shades indicative of greater conformity. A clock-like pictogram encircles the score, revealing a color-coded estimate for each analytical step, promoting intuitive interpretation of the method’s environmental performance. Minimal sample treatment, streamlined and rapid workflow that demanded minimal procedural manipulation to resolve target analytes, less waste generated per sample, high throughput sample analysis, along with energy efficiency and employment of non-toxic solvents, explained the elevated AGREE score of the proposed methods over the reported method, in accordance with AGREE results. In the line of AGREE scores for both proposed methods, the HPTLC/densitometry outperformed the smartphone one due to absence of derivatization step. The GAPI tool promotes thorough assessment of the analytical workflow via evaluating different parameters as sampling procedure, hazard of reagents, energy consumption of instruments, and waste handling. The assessment is visually illustrated via a pentagram segmented into five domains, each color-coded: green, yellow, or red to imply minimal, intermediate, and significant ecological impact, respectively.

Following GAPI results, the proposed approaches outperformed the stated spectrophotometric method in terms of sample treatment in a more eco-conscious manner. Additionally, the smartphone method provided a lower energy consumption benefit than the densitometric one.

Whiteness assessment of the proposed methods

White analytical chemistry (WAC), has been promoted for a number of years, extending the idea of GAC and highlighting the need to strike the best possible balance between practicality and ecological sustainability. WAC stands for the Red-Green-Blue paradigm for a light hue, where one of its fundamental elements is greenness. Whiteness represents the overall quality of the method, while red is associated with analytical aspects such as trueness (formerly known as accuracy), precision, and sensitivity. The practicality and economics are represented by the blue. RGBfast, an optimized version of the RGB model that incorporates the ChlorTox Scale as one of its criteria, was used for assessment of the proposed methods. RGBfast’s significant features over earlier iterations include ease of use, authenticity, automated assessment, and restricted manipulation flexibility [55]. The densitometric approach received the greatest overall rank according to the whiteness assessment findings, particularly for the analytical performance and validation parameters. The smartphone method was more cost-effective, utilized less energy, and practical in terms of removing the costly detector with somewhat complex adjustment steps. Moreover, the smartphone-based approach distinguished itself via exploiting a notable eco-friendly favorable profile, in terms of minimal energy and hardware demand. The densitometric method was faster because the results were integrated simultaneously. In contrast to the reported spectrophotometric method, the suggested methods allowed simultaneous assay of several samples per a single analytical run, hence, the average per-sample analysis time was minimized. Less solvent was used and less waste was produced as a result, demonstrating the suggested techniques’ greater cost and time effectiveness. In comparison, the reported method revealed reduced greenness and whiteness scores, principally attributed to offline configuration, hazardous solvents, and the requirement for extraction. The suggested approaches demonstrated appropriate performance as presented, Table 5. The densitometric approach and a smartphone coupled with ImageJ software approach, were found to be similar in terms of sustainability and greenness when comparing the validation findings with the results of the greenness and whiteness assessment, Table 5. While the densitometric approach revealed marginally superior analytical efficiency, the smartphone one was remarkably valuable in terms of accessibility and minimal environmental footprint.Table 5. Greenness, whiteness, and blueness assessment profile comparison between the proposed HPTLC methods and the reported UV spectrophotometric one

Evaluation of the practicality of the proposed methods

For assessing the applicability of any analytical approach, a cutting edge, quick, and easy metric tool called Blue Applicability Grade Index (BAGI) was implemented, Table 5. Of the several benefits of the BAGI tool, the most significant is that it complements the current green metric tools, including AGREE and GAPI. It is also consistent with the sustainability of the environment [61]. This metric tool assesses the practicality of method via ten core attributes, involving the analysis procedure employed, number of analytes simultaneously assayed, sample throughput per hour, nature of reagents and materials utilized, requisite instrumentation, capacity of batch processing, preconcentration demand, degree of automation, protocol for sample preparation, and volume of sample. Upon evaluation of these attributes, the tool delivers a corresponding score with asteroidal pictogram that depicts the performance in line with the assessed criteria. The detailed input data employed in BAGI metric tools for all compared methods are mentioned in the supplementary information (Table S5). The results of BAGI demonstrated the superior performance of HPTLC/smartphone followed by HPTLC/densitometry to the reported method owing to their superior sample throughput, minimal sample preparation, and utilization of portable instrumentation.

Conclusion

Novel HPTLC-based approaches for the analysis of DZP in its pharmaceutical dosage form were successfully developed and validated. These methods incorporated a cost-effective and accessible alternative to traditional densitometric analysis by utilizing a smartphone camera in combination with open-source image processing software. The analytical performance and environmental impact of the proposed methods were assessed using three established metric tools for greenness and sustainability, in addition to the BAGI tool, which evaluated the practicality and operational feasibility of the methods. Statistical comparisons with a reported UV spectrophotometric method confirmed the validity and accuracy of the proposed techniques. Furthermore, method validation was performed in accordance with ICH guidelines, confirming their suitability for routine application. The methods demonstrated reliable performance in the quantitative analysis of DZP, ACB, and OXP—whether in their pure form, spiked human plasma samples, or finished pharmaceutical products. While the densitometric method exhibited slightly higher sensitivity and precision, the smartphone-based technique offered substantial advantages in terms of affordability, portability, operational simplicity, and reduced requirements for technical expertise and laboratory infrastructure. These benefits make it particularly attractive for resource-limited settings. The findings underscore the potential of smartphone-integrated analytical tools to serve as practical, sustainable, and efficient alternatives to conventional instrumentation for routine pharmaceutical analysis.

Supplementary Information

Below is the link to the electronic supplementary material.Supplementary Material 1.