Environmental Monitoring of Celecoxib, Ketoprofen, and Meloxicam in Pharmaceutical Wastewater by SPE-Assisted Micellar Electrokinetic Chromatography

Abstract

The continuous discharge of pharmaceutical residues into aquatic environments has raised significant environmental concerns due to their persistence and incomplete removal during wastewater treatment. Non-steroidal anti-inflammatory drugs (NSAIDs) are among the most frequently detected pharmaceutical contaminants in industrial effluents. In this study, a sensitive and selective analytical method was developed for the simultaneous determination of ketoprofen (KTP), meloxicam (MEL), and celecoxib (CEL) in pharmaceutical wastewater using micellar electrokinetic chromatography (MEKC) combined with off-line solid-phase extraction (SPE). A high-volume SPE procedure (1000 mL sample) followed by evaporation and reconstitution provided a theoretical enrichment factor of approximately 10,000. Under optimised conditions, complete separation was achieved in less than 10 min. The method exhibited excellent linearity over a range of 0.5–20 µg/mL (r² > 0.999), with limits of detection in wastewater ranging from 14 to 18 ng/L. Accuracy and precision complied with ICH Q2(B) guidelines, and recoveries from spiked wastewater samples ranged from approximately 99% to 101%, indicating efficient extraction and minimal analyte loss. The validated method was successfully applied to real pharmaceutical wastewater samples, demonstrating its suitability for the routine monitoring of trace-level NSAIDs in complex industrial matrices.

1. Introduction

Non-steroidal anti-inflammatory drugs (NSAIDs) are among the most extensively consumed pharmaceutical compounds worldwide due to their analgesic and anti-inflammatory properties. Their widespread use, combined with incomplete removal during wastewater treatment, has resulted in their frequent detection in aquatic environments. In addition to municipal discharges, pharmaceutical manufacturing facilities represent a direct and potentially concentrated source of NSAID contamination in industrial effluents. Continuous release of these compounds raises environmental concerns and highlights the need for sensitive and reliable analytical methods capable of monitoring trace levels in complex wastewater matrices [1,2]. Celecoxib (CEL), meloxicam (MEL), and ketoprofen (KTP) represent therapeutically important NSAIDs with different cyclooxygenase selectivity profiles. While KTP is a classical non-selective NSAID, CEL and MEL exhibit preferential or selective COX-2 inhibition. The chemical structure of the studied drugs is given in Figure 1.

Due to their widespread consumption and incomplete removal during wastewater treatment processes, these pharmaceuticals are continuously discharged into aquatic environments through effluents originating from households, hospitals, and pharmaceutical manufacturing facilities [3].

Despite their therapeutic efficacy, NSAIDs are associated with toxicological risks, including gastrointestinal, renal, and cardiovascular adverse effects, particularly for COX-2 selective inhibitors [4,5,6]. In addition, their persistence and resistance to degradation contribute to environmental contamination and potential ecotoxicity [7]. Therefore, sensitive monitoring of these pharmaceuticals in wastewater effluents is essential.

The quantification of the investigated drugs in various sample forms was conducted utilising a range of analytical techniques, including liquid chromatography coupled with mass spectrometric detection (LC-MS/MS) [8,9], voltametric methods [10,11], UV-spectrophotometry [12], and thin-layer chromatography (TLC) [13]. A variety of techniques for assessing NSAIDs in aquatic environments have been established, including LC-MS/MS [14,15] and gas chromatography mass spectrometry (GC-MS/MS) [16]. Conversely, the capillary electrophoresis (CE) technique, when combined with various pre-concentration methods, has been utilized to analyze these compounds in water samples [17,18,19].

Capillary electrophoresis (CE) has emerged as a significant analytical tool for the analysis of NSAIDs in recent years, attributed to its high separation efficiency, rapid analysis, minimal sample and reagent consumption, and, consequently, reduced costs. The introduction of micellar electrokinetic capillary chromatography (MEKC) by Terabe et al. [20] has significantly expanded the application scope of CE. The MEKC technique has demonstrated efficacy in analysing NSAIDs across various sample types [21,22,23,24,25].

This work integrates high-volume off-line SPE (1000 mL) with MEKC, achieving significant preconcentration and enhanced sensitivity for complex pharmaceutical industrial wastewater. Unlike earlier CE/MEKC methods, which focus on on-column stacking and are typically used for surface or mineral waters, this approach enables ng/L-level detection with cost-effective instrumentation, reduced solvent use, and short analysis times. It offers a practical alternative to LC–MS/MS for routine industrial monitoring.

The exposure of the local environment to industrial effluents containing NSAIDs poses a significant threat to the well-being of all living organisms. This assertion indicates the importance of monitoring and quantifying the studied NSAIDs in pharmaceutical industrial effluent. This study aims to develop a reliable analytical method for the accurate, sensitive, and selective determination of pharmaceuticals in industrial wastewater.

2. Materials and Methods

2.1. Chemicals, Reagents and Standard Solutions

All chemicals utilized were of analytical grade, and the solvents employed were of HPLC quality. The chemicals listed below were supplied by Merck, Darmstadt, Germany. Acetonitrile (ACN), 1 M hydrochloric acid, methanol (MeOH), tris(hydroxymethyl)aminomethane (Tris), sodium hydroxide (NaOH), and sodium octanesulfonate (SOS). All work was conducted using ultrapure water sourced from the Milli-Q Plus system, Millipore Bedford, MA, USA.

The NSAIDs under examination were supplied by Bio Vision (Milpitas, CA, USA). We received a certificate indicating that the purity levels of CEL, KTP, and MEL were 99.75%, 100.15%, and 99.91%, respectively. The appropriate amount of each drug was dissolved in MeOH to prepare stock standard solutions (1000 μg/mL) of each NSAID, which were subsequently stored in the dark at 4 °C. Working standard solutions were freshly prepared by diluting the stock solutions with 50% ACN in water to achieve a concentration of 100 μg/mL. The target background electrolyte (BGE) consists of a 10 mM Tris buffer with 60 mM SOS and 20% (v/v) acetonitrile. It was adjusted to a pH of 11.0 using sodium hydroxide. The prepared BGE underwent degassing in an ultrasonic bath for a minimum of 15 min prior to utilization.

2.2. Instrumentation

The experiments were carried out by utilizing an Agilent CE system (Waldbronn, Germany) that was outfitted with a diode array and the multiple wavelength detectors G1315 C/D and G1365 C/D connected to a computer loaded with Agilent ChemStation Software (version B.04.03, Agilent Technologies, Waldbronn, Germany).

An uncoated fused silica capillary with a 50 µm inner diameter and a length of 70 cm was utilised for separation purposes. The effective length measured was 60 cm. Peak areas, migration times, peak heights, peak widths, resolution parameters, signal-to-noise ratios, and diode array spectral data were analysed using Agilent ChemStation software. The pH of the solutions was measured using a Mettler Toledo pH meter (Greifensee, Switzerland).

2.3. Real Samples Collection

Real wastewater samples free of the studied drugs, as well as other real wastewater samples suspected to contain these drugs, were collected from a pharmaceutical plant situated in the Al-Kharj district (Riyadh, Saudi Arabia). To prevent sample deterioration, the samples were stored in opaque glass vials at 4 °C until analysis to prevent analyte degradation.

2.4. Capillary Preconditioning

The preconditioning method for new capillaries consisted of a 15 min flush with a 0.1 M NaOH solution, followed by a 5 min flush with deionized water, and finally a 10 min flush using a BGE solution. A pre-injection flush lasting 3 min with a BGE solution was conducted prior to each injection.

2.5. Electrophoretic Conditions

To achieve optimal separation, a voltage of +25 kV was applied at room temperature for the separation of the studied drugs. The sample solution was injected hydrodynamically for a duration of 5 s at a pressure of 34.5 mbar (0.5 psi). The detection of the studied drugs using DAD was carried out at 214 nm.

2.6. Method Validation

ICH-Q2B criteria were utilized for the assay validation [26].

2.6.1. Linearity

Various aliquots of the investigated NSAIDs (5–200 µg) were transferred to a series of 10 mL volumetric flasks with accuracy and independence. To reach a concentration of 0.5–20 µg/mL for each medication under investigation, the volume of each flask was then adjusted with 50% ACN in water. The BGE served as the elution solution in the electrophoretic analysis of the samples.

2.6.2. Accuracy

The recovery percentage of an analyte from a defined quantity serves as a measure of accuracy. The results from nine samples containing 1, 5, and 15 µg/mL of each drug under investigation were analyzed using the procedure under linearity.

2.6.3. Precision

For multiple statistically significant experiments, precision is shown as the percentage relative standard deviation of the inter- and intra-day precision. Each drug was tested at three concentrations (1, 5, and 15 µg/mL) three times, either on the same day (intra-day) or over three consecutive days (inter-day).

2.6.4. LOD and LOQ

The limit of quantification (LOQ) is defined as the minimum quantity of a drug that can be effectively and reliably determined, while the limit of detection (LOD) refers to the minimum amount that can be identified above background noise. The LOQ and LOD are determined by identifying concentrations that yield peaks with heights ten and three times the baseline noise, respectively. The S/N ratio was measured by the CE software (D version) according to the analyte standard in solvent and matrix.

2.6.5. Robustness

Minor changes to the proposed technique can be utilized to assess its robustness. The modification of the acetonitrile concentration (±1%) in the BGE facilitated this achievement. A ±1 kV variation in the applied voltage was implemented.

2.7. Applications

For the SPE procedure, 200 mg, 6 mL Oasis HLB cartridges from Waters, Milford, MA, USA, were utilized. The SPE conditions were adjusted in this study with respect to sample pH, loading volume, and elution solvent to maximize analyte recovery and minimize matrix interferences prior to MEKC analysis. The cartridge was prepared by the gradual addition of five millilitres of MeOH. Subsequently, it was equilibrated with five milliliters of high purity water, which was adjusted to a pH of 2.0 through the careful addition of 1 M HCl dropwise.

Prior to spiking experiments, an unspiked wastewater sample (1000 mL) was subjected to the complete SPE–MEKC procedure to experimentally verify the absence of the target NSAIDs. No detectable peaks were observed at the characteristic migration times of KTP, MEL, or CEL. The electropherograms were further evaluated to ensure the absence of co-migrating interferences. Subsequently, one liter of the verified wastewater sample was adjusted to pH 2.0 using 1 M HCl. After that, it was spiked with 0.5 µg of each drug under investigation and loaded onto the cartridge at a flow rate of 5 mL/min. Subsequently, it was washed with 5 mL at pH 2.0 of 5% MeOH solution in water dropwise, followed by air drying for 5 min. The analytes were subsequently eluted using 2 × 5 mL of MeOH in a dropwise manner. The eluate was evaporated using a rotary evaporator at 40 °C until the solvent was completely removed. The reconstitution was performed using 100 μL of 50% ACN in water, followed by vortexing for 30 s to ensure complete dissolution.

The wastewater sample suspected of containing NSAIDs (unspiked) underwent the same pre-treatment procedures, excluding the spiking of the sample with the studied NSAIDs. It was subsequently analyzed using capillary electrophoresis under identical conditions. The calibration in the matrix was performed by spiking five blank wastewater samples, free of NSAIDs, with varying amounts of NSAIDs to achieve concentrations ranging from 50 to 2000 ng/L. The prepared samples undergo the previously described SPE procedure. The processed samples are injected following reconstitution, and a calibration curve was plotted using the initial, un-concentrated levels in relation to the final peak areas.

The overall analytical workflow employed in the present study is summarized in Scheme 1.

3. Results and Discussion

3.1. Optimizing the Method

The quality of separation in MEKC is significantly affected by various experimental parameters, such as the pH of the background electrolyte (BGE), surfactant concentration, organic modifier content, and applied voltage. These parameters together affect electro-osmotic flow (EOF), ionization of analytes, micelle production, and interactions between analytes and micelles. In the end, they control migration behavior, resolution, and peak efficiency.

3.1.1. BGE pH’s Impact

The pH of the background electrolyte is very important for capillary electrophoretic separations because it controls both the ionization state of the analytes and the magnitude of the EOF. In this study, the pH of BGE was varied over the range of pH 10.0 12.0 to study its effect on separation efficiency. The BGE was made up of 10 mM Tris buffer, 60 mM sodium octanesulfonate (SOS), and 20% (v/v) acetonitrile (ACN). The concentration of each analyte was kept at 5 µg/mL. Figure 2 shows that raising the pH level increase peak area and improve separation efficiency up to pH 11.0. There was no noticeable improvement in performance at higher pH levels (>11.0), and the peak shape got a little worse. Consequently, pH 11.0 was designated as the ideal BGE pH for ensuing analyses.

The influence of pH on separation can be rationalized based on the acid dissociation constants of the analytes. KTP (pKa ≈ 4.0) and MEL (pKa ≈ 4.1) are fully ionized at pH 11.0 and therefore migrate as anions. In contrast, CEL (pKa ≈ 11.1) remains largely neutral under these alkaline conditions. This difference in ionization state affects their effective electrophoretic mobility and interaction with the negatively charged SOS micelles, contributing to the observed resolution. Moreover, SOS micelles remain stable under strongly alkaline conditions, ensuring consistent micellar partitioning and reproducible MEKC separation.

3.1.2. Surfactant Concentration’s Effect

The concentration of surfactant is a vital parameter in MEKC, as it regulates micelle formation, dimensions, and quantity, thus affecting analyte distribution between the aqueous phase and the micellar pseudostationary phase. The concentration of SOS was checked to study its effect on separation performance. It was varied between 50 and 70 mM, keeping everything else the same. As illustrated in Figure 3, increasing SOS concentration resulted in a progressive enhancement of peak area up to 60 mM, suggesting improved analyte solubilization and partitioning into the micellar pseudostationary phase. Beyond this concentration, a decline in peak area was observed. This reduction may be attributed to increased buffer viscosity and stronger micelle–analyte interactions at higher surfactant concentrations, which can decrease effective electrophoretic mobility and broaden peaks. As a result, a surfactant concentration of 60 mM SOS was chosen as the best for the procedure being presented.

The critical micelle concentration (CMC) of SOS in pure aqueous solution at 25 °C is reported to be approximately 120–150 mM [27]. Nevertheless, the CMC of ionic surfactants is significantly influenced by solvent composition and ionic strength. The presence of 20% (v/v) acetonitrile and 10 mM Tris buffer reduces the effective CMC due to decreased solvent polarity and enhanced hydrophobic interactions. Therefore, the selected concentration of 60 mM SOS was sufficient to promote stable micelle formation under the applied mixed-solvent alkaline conditions, ensuring effective micellar electrokinetic separation.

3.1.3. Influence of Organic Modifier and Applied Voltage

In MEKC, adding an organic modifier like acetonitrile has a big effect on both the characteristics of the micelles and the way in which the analyte is distributed. ACN affects the stability, viscosity, and solubility of micelles, which in turn affects the shape of the peaks and the effectiveness of the separation. ACN concentrations between 15% and 25% (v/v) was checked, and the best peak symmetry and efficiency were found at 20% (v/v) ACN. The voltage that is delivered is also very important for electrophoretic separations. Increasing the voltage usually speeds up migration times, but too much voltage might cause Joule heating, which makes peaks wider and lowers resolution. The voltage levels between 20 and 30 kV was checked. A voltage of 25 kV was the best balance between analysis time, resolution, and peak efficiency.

3.2. Separation of the Analyzed NSAIDs

After optimizing all experimental parameters, a standard mixture of CEL, KTP, and MEL, each at a concentration of 5 µg/mL, was evaluated under the chosen MEKC conditions. Figure 4 shows the ensuing electropherogram, which shows that the three analytes are separated at the baseline and have clear, symmetrical peaks.

The resolution of the separated peaks is determined in relation to the adjacent peak, as outlined in Equation (1):

$${\mathrm{R}}_{\mathrm{A},\mathrm{B}} = \backslash frac\left\{2\right({\mathrm{t}}_{\mathrm{B}} - {\mathrm{t}}_{\mathrm{A}}\left)\right\}\{{\mathrm{W}}_{\mathrm{A}} + {\mathrm{W}}_{\mathrm{B}}\}$$

(1)

In this context, R_A,B represents the resolution of adjacent pair peaks, while t_A and t_B denote the migration times of these peaks. Additionally, W_A and W_B refer to the widths of the respective peaks. The resolution factor values indicate acceptable resolution (

Table 1. Validation results of the proposed capillary electrophoretic method.

Parameter	KTP	MEL	CEL
Accuracy (Mean * ± SD)	99.79 ± 0.31	99.75 ± 0.65	100.20 ± 0.81
Resolution factor (Rs)	-	RKTP/MEL = 3.20	RMEL/CEL = 9.70
Precision: Repeatability * (RSD%) Intermediate precision * (RSD%)	0.64 0.99	0.73 1.11	0.73 1.09
Robustness: Elution liquid composition Applied voltage	100.23 ± 0.29 99.22 ± 0.79	99.28 ± 0.38 99.27 ± 0.85	100.47 ± 1.03 98.91 ± 1.14
Linearity range (µg/mL) Slope Intercept Correlation coefficient (r)	0.5–20 1.0038 0.175 0.9996	0.5–20 0.7622 −0.11 0.9996	0.5–20 0.8879 0.195 0.9998
LOD in solvent (µg/mL) LOD in matrix (µg/mL) LOQ in solvent (µg/mL) LOQ in matrix (µg/mL)	0.13 0.14 × 10−4 (14.00 ng/L) 0.42 0.45 × 10−4 (45.00 ng/L)	0.15 0.18 × 10−4 (18.00 ng/L) 0.50 0.60 × 10−4 (50.00 ng/L)	0.14 0.15 × 10−4 (15.00 ng/L) 0.48 0.50 × 10−4 (50.00 ng/L)

* Average of three readings.

).

3.3. Method Validation

The ICH Q2(B) guidelines [26] were used to check the method performance. For all of the medications examined, there was a linear relationship between the peak area and the analyte concentration in the range of 0.5–20 µg/mL. This is how the regression equations were found:

PA (KTP) = 1.0038 C + 0.175 r = 0.9996

(2)

PA (CEL) = 0.8879 C + 0.195 r = 0.9999

(3)

PA (MEL) = 0.7622 C 0.16 r = 0.9996

(4)

where PA is the peak area before multiplication by 10⁴, C is the concentration (µg/mL) and r is the correlation coefficient. As shown in Table 1, the approach showed great repeatability, accuracy, and moderate precision, with relative standard deviation values that were within acceptable ranges. Robustness testing, which involved small changes in applied voltage and ACN content, showed that these adjustments did not have significant impacts on how well the analysis worked. No noticeable changes in migration times or peak shapes were observed throughout repeated analytical runs, indicating stable capillary performance under the selected alkaline conditions (pH 11.0).

The LOD and LOQ results showed that the approach is sensitive enough to be used to analyze NSAIDs in the environment.

3.4. Method Application

3.4.1. Quantification of NSAIDs in Wastewater Samples

Because wastewater matrices are so complicated and MEKC is so sensitive to matrix interferences, it is very important to choose a sample pre-treatment method carefully. Solid-phase extraction (SPE) was chosen because it cleans up better, uses less solvent, and can enrich samples more than liquid–liquid extraction and protein precipitation methods. The theoretical volumetric enrichment factor (~10,000) was calculated based on the ratio between the initial wastewater sample volume (1000 mL) and the final reconstitution volume (100 µL) after SPE and solvent evaporation. The obtained enrichment factor made it possible to accurately measure CEL, KTP, and MEL at very low levels (in the µg/L and ng/L ranges), which is very important for environmental monitoring. Oasis HLB cartridges were used because they had a good blend of hydrophilic and lipophilic characteristics, which made them better at recovering than other examined sorbents. Optimizing the SPE parameters, such as the sample’s pH, loading flow rate, washing composition, and elution solvent, led to the effective removal of matrix interferences including salts and humic compounds, while keeping the analytes of interest. Given the broad retention characteristics of Oasis HLB sorbent, the proposed extraction protocol is expected to be adaptable to other environmental water matrices with higher organic content following matrix-specific validation.

Table 2. Determination of studied NSAIDs in spiked wastewater.

Sample	Added Before Extraction (ng/L)	Found KTP (ng/L)	Found MEL (ng/L)	Found CEL (ng/L)	KTP Recovery% ± S.D.	MEL Recovery% ± S.D.	CEL Recovery% ± S.D.
Sample 1	100.0	101.0 100.0 103.0	102.0 100.0 100.0	100.0 103.0 102.0	101.33 ± 1.25	100.67 ± 0.94	101.78 ± 1.25
Sample 2	200.0	198.0 201.0 203.0	202.0 200.0 203.0	201.0 199.0 198.0	100.33 ± 1.03	100.83 ± 0.62	99.66 ± 0.62
Sample 3	300.0	301.0 302.0 297.0	299.0 297.0 300.0	302.0 299.0 303.0	99.99 ± 0.72	99.56 ± 0.42	100.44 ± 0.56

shows the analytical results from spiked wastewater sample.

It is important to distinguish between overall method accuracy and SPE recovery efficiency. The accuracy results presented in Table 1 were obtained using standard solutions prepared in solvent and reflect the intrinsic performance of the MEKC separation and quantification system, in accordance with ICH Q2(B) guidelines. In contrast, the recovery values in Table 2 were derived from spiked wastewater samples processed through the complete SPE–evaporation–reconstitution procedure, thereby reflecting the extraction efficiency of the sample preparation step. The recoveries (approximately 99–101%) indicate efficient analyte retention and minimal loss during sample processing. Real wastewater samples were analysed by the proposed method, and the results are showed in

Table 3. Determination of studied NSAIDs in real wastewater samples (unspiked).

Sample	Found KTP * (ng/L)	Found MEL * (ng/L)	Found CEL * (ng/L)
Real wastewater sample 1	400	300	600
Real wastewater sample 2	520	910	400
Real wastewater sample 3	700	560	480

* Average of three readings.

and Figure 5.

The electropherogram of the real wastewater samples exhibited a stable baseline with minimal background interference. The target NSAIDs were well resolved with symmetrical and sharp peaks under the optimised MEKC conditions. No significant matrix co-migration was observed, confirming the selectivity of the proposed SPE–MEKC method in complex industrial wastewater matrices.

The selectivity of the method was further supported by the absence of interfering peaks at the migration times of the investigated NSAIDs in matrix blank samples processed under identical SPE–MEKC conditions. In addition, peak identification in wastewater samples was established by comparing the migration times of the detected peaks with those obtained from reference standard solutions analyzed under the same electrophoretic conditions. The well-resolved peaks and consistent migration times indicate that the developed SPE–MEKC method provides adequate selectivity for the determination of ketoprofen, meloxicam, and celecoxib in the investigated pharmaceutical wastewater samples.

3.4.2. Statistical Comparison with a Reference Method

A statistical comparison was made between the suggested SPE-MEKC method and reference LC-MS/MS methods to see how reliable it was. As shown in

Table 4. Statistical comparison between the determination of standard solutions of NSAIDs by the proposed method and the reference-reported methods.

Item	KTP	MEL	Reference Method [14] *	CEL	Reference Method [15] ∞
Mean ± SD	99.79 ± 0.31	99.75 ± 0.65	99.97 ± 0.53	100.20 ± 0.81	99.81± 0.59
RSD	0.31	0.65	0.53	0.81	0.59
Variance	0.10	0.43	0.28	0.65	0.35
n	5	5	5	5	5
F-value (6.39)	2.99	1.52	-	1.90	-
Student’s t-test (2.31)	0.66	0.58	-	0.92	-

* The reference form of liquid chromatography coupled to quadrupole time-of-flight mass spectrometry (LC-QTOF MS) was performed on using the C₁₈ reversed-phase column (4.6 mm × 200 mm, 5 μm) as the stationary phase. The mobile phase was acetonitrile (A) and 20 mM ammonium acetate in water (adjusted to pH 3.5 by the addition of acetic acid) (B) in ratio 85:15, v/v at isocratic elution conditions with flow rate of 1.0. mL/min. ^∞ The reference LC-QTOF MS was performed on using C18 (2.1 × 100 mm, 1.7 µm) as stationary phase and Water + 0.1% formic acid (A) and acetonitrile + 0.1% formic acid (B) as mobile phase with flow rate 0.5 mL/min.

, the computed Student’s t and F values were lower than the equivalent tabular values at the 95% confidence level. This means that there was no statistically significant difference in accuracy or precision between the two procedures.

To further contextualise the analytical performance and practical applicability of the proposed SPE–MEKC method, a comparative evaluation with representative LC–MS/MS and CE-based approaches reported in the literature is presented in

Table 5. Comparison of the proposed SPE–MEKC method with representative LC–MS/MS and CE-based methods for NSAID determination in water samples.

Method	Matrix	LOD (ng/L)	Analysis Time (min)	Sample Volume (mL)	Organic Solvent Consumption per Run	Instrumental Cost
LC-QTOF-MS (Ref. [14])	Wastewater	1–10	15–20	100–500	High (mL-scale mobile phase)	Very high
LC-QTOF-MS (Ref. [15])	Environmental water	0.5–5	10–15	200–500	High (mL-scale mobile phase)	Very high
MEKC + on-column preconcentration (Ref. [21])	River water	50–200	12–18	≤50	Very low (µL-scale BGE)	Moderate
MEKC + stacking strategies (Ref. [22])	Mineral water	20–100	10–15	≤50	Very low (µL-scale BGE)	Moderate
Proposed SPE–MEKC (This work)	Pharmaceutical wastewater	14–18	<10	1000	Low (limited SPE elution + µL-scale BGE)	Moderate

.

The comparative evaluation demonstrates several practical advantages of the proposed SPE–MEKC method. Compared with LC–MS/MS approaches, the present method requires substantially lower organic solvent consumption, involves simpler and more accessible instrumentation, and reduces operational cost, while still achieving ng/L-level detection limits through high-volume preconcentration. Unlike previously reported CE/MEKC methods that primarily rely on on-column stacking and are mainly applied to surface or mineral waters, the present work integrates high-volume off-line SPE (1000 mL) with MEKC to achieve substantial preconcentration and improved sensitivity in complex pharmaceutical industrial wastewater. The method combines cost-effective instrumentation, reduced solvent consumption, and short analysis time with ng/L-level detection capability, thereby providing a practical alternative to LC–MS/MS for routine industrial monitoring.

Certain limitations should also be considered. The reliance on an off-line extraction step increases total sample preparation time relative to direct injection or purely on-column stacking methods. In addition, operation under alkaline conditions (pH 11.0) necessitates appropriate capillary conditioning to maintain long-term reproducibility. Despite these considerations, the developed methodology offers a balanced combination of sensitivity, cost-effectiveness, environmental sustainability, and practical applicability for routine monitoring of pharmaceutical effluents.

4. Conclusions

A sensitive, selective, and cost-effective SPE–MEKC method was successfully developed and validated for the simultaneous determination of ketoprofen, meloxicam, and celecoxib in pharmaceutical industrial wastewater. The integration of high-volume solid-phase extraction (1000 mL) with micellar electrokinetic chromatography provided a theoretical enrichment factor of approximately 10,000, enabling ng/L-level detection limits (14–18 ng/L) with short analysis time (<10 min) and low solvent consumption. The method demonstrated excellent linearity (0.5–20 µg/mL, r² > 0.999), high precision (RSD < 1.2%), satisfactory robustness, and near-quantitative recovery (99–101%) in spiked wastewater samples.

Application to real effluent samples confirmed its reliability for routine trace-level environmental monitoring in complex industrial matrices. Compared with LC–MS/MS techniques, the proposed approach offers reduced operational cost, simpler instrumentation, and improved sustainability, making it particularly suitable for routine screening laboratories. Future studies may extend this strategy to additional emerging pharmaceutical contaminants to further support environmental risk assessment efforts.