Determination of Pharmaceuticals in Surface and Wastewater by Capillary Electrophoresis (CE): A Minireview

Abstract

Background: The detection of pharmaceuticals in surface and wastewater is becoming increasingly important for environmental monitoring. The release of pharmaceuticals from industry and human and animal excretion is causing devasting effects to the aqueous ecosystem and to the organisms that live in it. The various pharmaceuticals’ complex properties make it challenging to fully remove them from the wastewater prior to discharge. It is crucial that the concentration of the pharmaceuticals is at a safe limit before they are released into the ecosystem for the environment as well as human health. Scientists are currently developing methods to detect pharmaceuticals in surface and wastewater. High-performance liquid chromatography (HPLC) and gas chromatography (GC) are popular analytical methods. Method: Capillary electrophoresis (CE) is now gaining attention with its potential to detect pharmaceuticals in aqueous environments. CE coupled with the pretreatment procedures such as solid-phase extraction (SPE) provide excellent results for the determination of these contaminants in the water. Aim: The aim of this paper is to review the literature on capillary electrophoresis for the determination of pharmaceuticals in surface and wastewater.

Keywords:

• Capillary electrophoresis
• pharmaceuticals
• solid phase extraction (SPE)
• surface water
• wastewater

Introduction

Pharmaceuticals are consumed everyday by millions of people and animals across the globe for a variety of medical reasons. Pharmaceuticals are defined by the World Health Organization as “any substance or pharmaceutical product for human and veterinary use that is intended to modify or explore physiological systems or pathological states for the benefit of the recipient”. While they provide the necessary treatment for ailments in both humans and animals, the devasting effects they can have by entering the aqueous ecosystem are now being recognized. A paper published by Bielen et al. (2017) reported on their study of the embryotoxicity assay of effluents containing antibiotics on zebrafish. The paper indicated that exposure to those effluents caused high mortality rates as well as adverse effects such as heart and yolk edema, deformities, and delayed developments.

Researchers are now developing methods to efficiently detect and quantify active pharmaceuticals in surface and wastewaters. The most commonly used separation techniques involve preconcentration protocols such as solid phase extraction (SPE) coupled with high performance liquid chromatography (HPLC) (Peixoto et al. 2019; Sousa et al. 2020; Beldean et al. 2020; Li et al. 2021), gas chromatography – mass spectroscopy (GC-MS) (Gumbi et al. 2017), and capillary electrophoresis (CE). CE has generated interest as a favorable analytical approach for pharmaceuticals in surface and wastewater due to its numerous advantages. CE’s main advantage is its versality to separate out a wide range of molecules due to its several separation modes. Capillary zone electrophoresis (CZE) separates molecules based on their size to charge ratios. Micellar electrokinetic chromatography (MEKC) is used to separate neutral species. Capillary gel electrophoresis is employed to separate large molecules such as DNA based on size and charge. Capillary isoelectric focusing (CIEF) separates molecules based on their isoelectric points. The most uncommon form of CE is capillary isotachophoresis which separates molecules by sandwiching between two buffers of different electrophoretic mobilities. The most commonly used forms of CE for pharmaceuticals are CZE and MEKC. The aim of this paper is to review the literature that has used capillary electrophoresis for the determination of pharmaceuticals in water.

Capillary zone electrophoresis

A SPE-CE method was developed for the determination of salbutamol sulfate, cimetidine, and carbamazepine in influent and effluent wastewaters. An off-line SPE method was optimized using Oasis HLB cartridges. When the SPE method was tested on influent and effluent wastewaters, 89.7 to 102.0% of the analytes were recovered. Uncoated fused-silica capillaries with an internal diameter of 75 µm, a total length of 50 cm, and an effective length of 41.5 cm were used in the CZE method. The capillaries were conditioned with 1.0 M and 0.1 M NaOH for 20 minutes each, ultrapure water for 10 min, and running buffer for 30 min before the first use. The capillaries were rinsed with running buffer for 3 min after each analysis. To optimize the electrophoresis method, phosphate, tris, sodium tetraborate decahydrate, and boric acid buffers were investigated. Good resolution was observed with the phosphate and tris buffers. The tris buffer was eliminated from the selection because the solvent peak interfered with the analytes. Therefore, the phosphate buffer was selected. Experiments were carried out to determine the optimum pH of the phosphate buffer between 6.00 and 7.50. The pH study showed that by increasing the pH, cimetidine separated from salbutamol and moved closer to carbamazepine. When the pH had exceeded 6.70, salbutamol and cimetidine were well resolved but the separation between cimetidine and carbamazepine was poor. The best resolution of the analytes was achieved for pH values from 6.50 to 6.70. Therefore, a pH of 6.70 was selected to be optimum. A buffer concentration range from 10 to 30 mM was investigated to determine the best concentration for resolution. The study showed that an increase in the concentration resulted in an increase in migration time but the peak shape of carbamazepine was not satisfactory. When the concentration of phosphate buffer was higher than 20 mM, there was no improvement in the separation. The higher voltage instead caused the current to increase to a high value of 100 µA that created concern for joule heating. A phosphate buffer concentration of 20 mM was employed as it provided a stable ionic strength and minimal heating. To further adjust the solvent polarity and to help dissolve the analytes due to their low water solubility, 20% (v/v) methanol was added to the buffer. The method parameters were optimized using 20 mM phosphate buffer with 20% v/v methanol, a pH of 6.70, and a voltage of 20 kV. A diode array detector (DAD) detector was used at 214 nm which provided instrumental limits of detection between 2.0 and 3.2 µg/L. The reproducibility of this method was estimated to be less than 5% relative standard deviation. When this method was applied to determine the analytes in samples from a local wastewater plant, only carbamazepine was present in all samples (Wang, Sun, and Liu 2015).

Espina-Benitez et al. (2017) used an offline solid phase microextraction (SPME) as a preconcentration step. SPME is a rapid and simple technique that requires no solvents, and exhibits good linearity and sensitivity compared with SPE. The SPME procedure employed the needle of a commercial disposable syringe. Large volume sample stacking (LVSS) was also used for pretreatment to increase the sensitivity for the pharmaceuticals. By using both SPME and LVSS, the recoveries of clofibric acid, ketoprofen, and naproxen from a river water sample were 99.7 − 103.6%, 97.7-104.9%, and 97.9 − 101.6%, respectively. For the CZE method, a fused silica capillary with an internal diameter of 75 µm, a total length of 62 cm, and an effective length of 54 cm were used. The capillaries were flushed with 1 M and 0.2 M NaOH for 10 minutes each, and deionized water for 5 minutes for the first use. Each day the capillaries were rinsed with 0.2 M NaOH, deionized water, and buffer. Before each run, the capillaries were preconditioned with 1 M NaOH, deionized water, and buffer, and following each measurement, the capillary was rinsed with deionized water. Phosphate was selected to be the buffer based on the literature surrounding NSAIDS and pKa values of the buffer and the analytes. For the optimization of the buffer, a pH between 7.8 to 9.2 and a concentration range from 25.0 to 62.5 mM was analyzed. The study determined that a pH of 8.75 with a concentration of 31.25 mM provided the optimum results. The addition of 8% (v/v) acetonitrile to the buffer improved the resolution. A voltage of 20 kV was optimum as higher values did not improve the resolution but instead caused band broadening due to excessive heating. When the method was employed to analyze a river sample, the intra-day reproducibility for the analytes were from 12.8% in the 6 µg/L sample, 8.9 − 13.3% in 8 µg/L, 9.8 − 1.3% in 20 µg/L, and 8.5 − 9.8% in 50 µg/L. Only naproxen was detected in the 6 µg/L but was not present in the 8 µg/L sample.

CE with electrochemiluminescence detection coupled to hollow fiber - solid phase extraction (HF-SPE) was employed for the determination of 10 antibiotics in water samples from cattle farm drainage, chicken farm drainage, and a river. The determined antibiotics included β-lactams, cephalosporins, sulfonamides and fluoroquinolones. Hollow fiber solid-phase extraction (HF-SPE) was employed for preconcentration. The antibiotics were separated by CE. Phosphate buffer with a pH of 6.5 was employed with 50 mmol/L sodium chloride. The sodium chloride increased the ionic strength and improved the resolution of the peaks. When the method was tested, the authors noticed the peaks were not fully separated. Hence, an organic additive was added to the buffer. The addition of 35 mmol/L isopropanol to the buffer provided full separation. The authors did not report the buffer concentration used in the method. Voltages from 17 to 21 kV was tested with the method and 19 kV was determined to be the optimum. The antibiotics were determined using an electrochemiluminescence detector. Cyclic voltammetry and a potentiostatic method were conducted using a 3-electrode system composed of a 500 µm diameter platinum working electrode, a 300 µm diameter silver/silver chloride (Ag/AgCl) reference electrode, and a 1 mm diameter platinum wire auxiliary electrode. The electrodes were placed in a 40 mmol/L phosphate buffer and pH 5.8 6 mmol/L Ru(bpy)₃²⁺ (tris (bipyridine) ruthenium (II) chloride). The detector potential was reported to be 1.2 V compared to the Ag/AgCl electrode. All ten antibiotics were determined. The limits of detection for the analytes were between 0.4 –and1.2 µg/L. The recoveries of the antibiotics ranged from 81.6 to 107.8% in the cattle farm drainage, 82.3 to 108.4% in the chicken farm sample and 87.5 − 110.4% in the river water. (Zhang et al. 2020).

A CZE method with pressure-assisted electrokinetic injection (PAEKI) was developed for the determination of six sulfonamides: sulfamethazine, sulfmerazine, sulfamonomethoxine, sulfadizine, sulfamethoxazole, and sulfacetamide in milk, eggs, and pork and environmental waters. PAEKI involves first filling the capillary with the buffer, introducing a water plug under pressure, followed by electric injection under reverse voltage while applying assisted pressure. Optimization of the injection voltages, assisted pressure, injection time and the length of water plug showed the optimum enrichment parameters were −10 kV, 0.5 psi, 30 s, and 1.5 cm, respectively. The CZE method used bare fused-silica capillaries with an inner diameter of 75 µm, an outer diameter of 375 µm, a total length of 50.2 cm, and an effective length of 40 cm. A new capillary was rinsed with water for 10 minutes, 1.0 M NaOH for 40 minutes, water for 10 minutes, and the buffer for 30 minutes. The capillaries were also rinsed with water for 2 minutes, 1 M NaOH for 10 minutes, water, and the buffer solution for 5 minutes prior to each experiment. The optimization of the buffer included testing both borate and phosphate buffer solutions. Phosphate buffer was superior because borate did not completely separate the sulfonamides. Experiments were carried out to determine the optimum concentration of the buffer. The tested concentrations that were 10, 20, 30 and 40 mM. 20 mM was determined to be the optimum concentration. The optimum pH of 7.1 provided the best separation. The optimum concentration of organic additive acetonitrile was 10% (v/v) and provided the best separation with the shortest run time. The optimum voltage for the separation was 15 kV. The PAEKI-CZE method provided high recoveries of 89%. The limits of detection for the pharmaceuticals were from 0.0021 to 0.0152 µg/mL, 0.0104 to 0.0675 µg/mL, and 0.0097 to 0.0595 µg/mL in the tap, lake, and sea water. The limits of quantification were from 0.0072 to 0.0413 µg/mL, 0.0314 to 0.1752 µg/mL, and 0.0337 to 0.1981 µg/mL, respectively. (Yang et al. 2020).

Micellar electrokinetic chromatography

Sirén and Fellah (2017) developed a capillary zone electrophoresis (CZE) and a partial filling micellar electrokinetic chromatography (PF-MEKC) method for the determination of 11 hormones in influent and effluent wastewaters. An offline SPE preconcentration step was developed and validated using Strata X and amine silane cartridges. The amine silane cartridges were used to further concentrate the SPE filtrate from the Strata X and to monitor the efficiency for retaining steroid glucosides. The overall recoveries for the hormones were 36 to 98% for the influent and 23 to 79% for the effluent. While the SPE method provided good recoveries, testosterone-glucoside were not recovered by CZE. For both PF-MEKC and CZE, bare fused silica capillaries were used with a total length of 80 cm and an effective length of 71.5 cm. The capillary was conditioned with 0.1 M NaOH, deionized water, and buffer. The duration of the conditioning or cleaning was modified depending upon the sample, the duration of the run, or if a new capillary was used. The PF-MEKC mode of CE was used for nonionic androgen hormones such as androgen, estrogen, and progesterone. The PF-MEKC parameters were optimized with 20 mM ammonium acetate buffer with sodium dodecyl sulfate (SDS) and sodium taurocholate solution, a pH of 9.68, and a voltage of 25 kV. A diode array detector was used to separate the hormones with optimum sensitivity at 247 nm and 214 nm for the estrogens. A CZE method was optimized for the estradiol-glucosides and testosterone-glucoside with a 20 mM ammonium acetate buffer at pH of 9.68 and a 0.2 M 2-cyclohexylamino]-1-propane-sulfonic acids (CAPS) buffer with a voltage of 25 kV. The hormones were determined by a diode array detector at 200 nm for the estradiol-glucosides and 247 nm for testosterone-glucoside. CAPS buffer was reported to be more efficient than the ammonium acetate buffer as the former allowed the separation of the E3-glucoside hormone which was not detected using the latter. PF-MEKC was the most repeatable with precision from 0.6 to 5.1% compared to 2.3 to 6.4% and 5.9% for CZE using CAPS buffer and acetate buffer, respectively.

El Fellah, Duporté, and Sirén (2017) published another paper using a similar PF-MEKC method for the determination of estradiol glucoside, androstenedione, testosterone, and progesterone in water samples from hot and cold taps. An additional CZE method was reported for inorganic ions and botrydial mold in the water samples. Here only the PF-MEKC method for the hormones is reviewed. A vac master preconditioned with Strata-X and amino polar phase columns was employed for SPE. The recovery of the steroids was between 80 and 90%. The PF-MEKC parameters were the same as reported in Sirén and Fellah (2017). The identification of the steroids was performed by ultra-high-performance liquid chromatography coupled to electrospray ionization orbitrap high-resolution mass spectrometry (UHPLC-MS). The study reported that the migration times of the steroids using PF-MEKC were similar to their retention times by UHPLC. The limits of detection in tap water were 0.06, 0.10, 0.20, and 0.94 µg/mL for androstenedione, progesterone, estradiol-glucoside, and testosterone, respectively.

Rucins et al. (2018) developed a SPE - MEKC procedure for the determination of exemestane, toremifene, letrozole, anastrozole, and mifepristone in wastewater. Off-line SPE was carried out to precondition the wastewater using Strata SDB-L styrene-divinylbenzene cartridges. The recoveries of the hormones were between 72.5 and 86.1%. Bare fused silica capillaries with an inner diameter of 50 µm, a total length of 33.5 cm, and an effective length of 25.0 cm were used for MEKC. The capillaries were rinsed with 0.1 M NaOH and deionized water for 10 minutes each before the first use. Between each measurement, the capillaries were rinsed with 0.1 M NaOH, deionized water, and the buffer. A 50 mM borate buffer with a pH of 9.5 and 50 mM SDS was investigated to separate the hormones. The study showed that only two out of the five analytes were fully separated. The concentration of the buffer was changed from 25 mM to 100 mM which provided no improvement. There was also no enhancement in performance when the pH was changed from 8.5 to 10.0. Therefore, the addition of an organic solvent was investigated. Following the addition of the organic solvent, the analytes were less retained in SDS when a more nonpolar alcohol was added to the buffer. Methanol, ethanol, and acetonitrile did not fully separate the analytes. Full separation occurred when 1-propanol, 2-propanol, and 1-butanol were added to the buffer. The study determined that 1-propanol had the best resolution with the shortest analysis time. 10 to 20% (v/v) 1-propanol provided full separation. 15% (v/v) was deemed to provide the optimum separation and the most repeatable signal between replicate measurements. A voltage of 20 kV was shown to be optimum. The analytes were detected using a diode array detector. The limits of detection were between 1.2 and 7.7 µg mL⁻¹, while the limits of quantification were from 4.0 to 25.6 µg mL⁻¹.

A summary of the methods reviewed for each pharmaceutical is included in Table 1.

Table 1. Summary of the reviewed CE methods using ultraviolet (UV) and electrochemiluminescence (ECL) detectors for pharmaceuticals in environmental waters.

Compound	Drug Type	CE Mode	Buffer	Detector	Preconcentration Step	Recovery (%)	Limit of detection (µg/mL)	Limit of quantification (µg/mL)
17α-Hydroxyprogesterone	Hormone	MEKC	20 mM ammonium acetate with SDS and sodium taurocholate at pH 9.68	UV	SPE strata X and amine silane	–	0.30	0.90
17α-Methyltestosterone	Hormone	MEKC	20 mM ammonium acetate with SDS and sodium taurocholate at pH 9.68	UV	SPE strata X and amine silane	–	0.07	0.21
Amoxicillin	Antibiotic	CZE	phosphate buffer with a pH of 6.5 with 50 mM NaCl and 35 mmol/L isopropanol	ECL	HF-SPE	82.3 − 95.6	0.0012	–
Anastrozole	Hormone	MEKC	50 mM sodium borate pH 9.5 with 50 mM SDS	UV	SPE Strata SDB-L styrene–divinylbenzene	83.2	2.3	7.8
Androstenedione	Hormone	MEKC	20 mM ammonium acetate with SDS and sodium taurocholate at pH 9.68	UV	SPE strata X and amine silane	50 − 79	0.50	1.50
Benzylpenicillin	Antibiotic	CZE	phosphate buffer with a pH of 6.5 with 50 mM NaCl and 35 mmol/L isopropanol	ECL	HF-SPE	89.2 − 99.1	0.0009	–
Carbamazepine	Anticonvulsant	CZE	20.00 mM phosphate buffer at pH 6.7	UV	SPE Oasis HLB	108.0	0.0029	–
Cefalexin	Antibiotic	CZE	phosphate buffer with a pH of 6.5 with 50 mM NaCl and 35 mmol/L isopropanol	ECL	HF-SPE	91.1 − 108.4	0.0006	–
Cefotiam	Antibiotic	CZE	phosphate buffer with a pH of 6.5 with 50 mM NaCl and 35 mmol/L isopropanol	ECL	HF-SPE	83.3 − 101.9	0.0004	–
Cefmetazole	Antibiotic	CZE	phosphate buffer with a pH of 6.5 with 50 mM NaCl and 35 mmol/L isopropanol	ECL	HF-SPE	92.9 − 115.7	0.0005	–
Clofibric Acid	Fibrate	CZE	31.25 mM phosphate buffer at pH 8.75	UV	SPME & LVSS	99.7	0.0097 − 0.00127	0.0029-0.0038
Cloxacillin	Antibiotic	CZE	phosphate buffer with a pH of 6.5 with 50 mM NaCl and 35 mmol/L isopropanol	ECL	HF-SPE	86.9 − 105.4	0.0007	–
Estradiol-glucoside	Hormone	CZE	20 mM ammonium acetate at pH 9.68	UV	SPE strata X and amine silane	23 − 76	0.28	0.85
Estradiol-glucoside	Hormone	CZE	0.2 M 2-cyclohexylamino]-1-propane-sulfonic acids (CAPS)	UV	SPE strata X and amine silane	40 − 49	0.06 − 0.37	0.17 − 1.10
Exemestane	Hormone	MEKC	50 mM sodium borate pH 9.5 with 50 mM SDS	UV	SPE Strata SDB-L styrene–divinylbenzene	72.5	7.7	25.6
Fluxoymesterone	Hormone	MEKC	20 mM ammonium acetate with SDS and sodium taurocholate at pH 9.68	UV	SPE strata X and amine silane	–	0.50	1.50
Ibuprofen	Non-steroidal anti-inflammatory (NSAID)	CZE	20.00 mM phosphate buffer at pH 6.7	UV	SPE Oasis HLB	102.0	0.0013	–
Ketoprofen	Non-steroidal anti-inflammatory (NSAID)	CZE	31.25 mM phosphate buffer at pH 8.75	UV	SPME & LVSS	97.7	0.0097 − 0.0013	0.0029-0.0038
Letrozole	Hormone	MEKC	50 mM sodium borate pH 9.5 with 50 mM SDS	UV	SPE Strata SDB-L styrene–divinylbenzene	86.1	1.2	4.0
Mifepristone	Hormone	MEKC	50 mM sodium borate pH 9.5 with 50 mM SDS	UV	SPE Strata SDB-L styrene–divinylbenzene	74.7	4.3	14.5
Naproxen	Non-sterodial anti-inflammatory (NSAID)	CZE	31.25 mM phosphate buffer at pH 8.75	UV	SPME & LVSS	97.9	0.0097 − 0.00127	0.0029-0.0038
Norflaxcin	Antibiotic	CZE	phosphate buffer with a pH of 6.5 with 50 mM NaCl and 35 mmol/L isopropanol	ECL	HF-SPE	87.5 − 102.5	0.0006	–
Pefloxacin	Antibiotic	CZE	phosphate buffer with a pH of 6.5 with 50 mM NaCl and 35 mmol/L isopropanol	ECL	HF-SPE	97.6 − 110.4	0.0006	–
Progesterone	Hormone	MEKC	20 mM ammonium acetate with SDS and sodium taurocholate at pH 9.68	UV	SPE strata X and amine silane	51 − 60	0.11	0.32
Salbutamol Sulfate	Bronchodilator	CZE	20.00 mM phosphate buffer at pH 6.7	UV	SPE Oasis HLB	102.0	0.0018	–
Sulfaceamide	Antibiotic	CZE	20.00 mM phosphate buffer at pH 7.1	UV	–	89	0.0021 − 0.0675	0.0072-0.1981
Sulfadizine	Antibiotic	CZE	20.00 mM phosphate buffer at pH 7.1	UV	–	89	0.0021 − 0.0675	0.0072-0.1981
Sulfathiazole	Antibiotic	CZE	phosphate buffer with a pH of 6.5 with 50 mM NaCl and 35 mmol/L isopropanol	ECL	HF-SPE	88.3 − 107.8	0.0005	–
Sulfamethzine	Antibiotic	CZE	20.00 mM phosphate buffer at pH 7.1	UV	–	89	0.0021 − 0.0675	0.0072-0.1981
Sulfmerazine	Sulfonamide Antibiotic	CZE	20.00 mM phosphate buffer at pH 7.1	UV	–	89	0.0021 − 0.0675	0.0072-0.1981
Sulfamonomethoxine	Antibiotic	CZE	20.00 mM phosphate buffer at pH 7.1	UV	–	89	0.0021 − 0.0675	0.0072-0.1981
Sulfamethoxazole	Antibiotic	CZE	20.00 mM phosphate buffer at pH 7.1	UV	–	89	0.0021 − 0.0675	0.0072-0.1981
Sulfamethoxazole	Antibiotic	CZE	phosphate buffer with a pH of 6.5 with 50 mM NaCl and 35 mmol/L isopropanol	ECL	HF-SPE	81.6 − 101.1	0.0004	–
Testosterone	Hormone	MEKC	20 mM ammonium acetate with SDS and sodium taurocholate at pH 9.68	UV	SPE strata X and amine silane	–	0.38	1.15
Testosterone-glucoside	Hormone	MEKC	20 mM ammonium acetate with SDS and sodium taurocholate at pH 9.68	UV	SPE strata X and amine silane	98	0.50	1.50
Testosterone- glucoside	Hormone	CZE	20 mM ammonium acetate at pH 9.68	UV	SPE strata X and amine silane	–	0.27	0.80
Testosterone- glucoside	Hormone	CZE	0.2 M 2-cyclohexylamino]-1-propane-sulfonic acids (CAPS)	UV	SPE strata X and amine silane	36 − 71	0.08 − 0.47	0.24 − 1.41
Toremifene	Hormone	MEKC	50 mM sodium borate pH 9.5 with 50 mM SDS	UV	SPE Strata SDB-L styrene–divinylbenzene	78.9	3.6	11.9

Conclusion

The reviewed studies have demonstrated the potential of capillary electrophoresis for the determination of pharmaceuticals in environmental waters. While more work is needed to address issues in the CZE methods, the versatility has allowed a wide range of pharmaceuticals to be detected using ultraviolet and electrochemiluminescence detectors in the µg range. Although CE is known for its poor sensitivity, preconcentration by SPE, HF-SPE, and LVSS has enhanced the sensitivity of CE while providing high recoveries of the target pharmaceuticals.