All-Solid State Potentiometric Sensors for Desvenlafaxine Detection Using Biomimetic Imprinted Polymers as Recognition Receptors

Using single-walled carbon nanotubes (SWCNTs) as an ion-to-electron transducer, a novel disposable all-solid-state desvenlafaxine-selective electrode based on a screen-printed carbon paste electrode was created. SWCNTs were put onto the carbon-paste electrode area, which was protected by a poly (vinyl chloride) (PVC) membrane with a desvenlafaxine-imprinted polymer serving as a recognition receptor. Electrochemical impedance spectroscopy and chronopotentiometric techniques were used to examine the electrochemical characteristics of the SWCNTs/PVC coating on the carbon screen-printed electrode. The electrode displayed a 57.2 ± 0.8 mV/decade near-Nernstian slope with a 2.0 × 10−6 M detection limit. In 10 mM phosphate buffer, pH 6, the ODV-selective electrodes displayed a quick reaction (5 s) and outstanding stability, repeatability, and reproducibility. The usefulness of electrodes was demonstrated in samples of ODV-containing pharmaceutical products and human urine. These electrodes have the potential to be mass produced and employed as disposable sensors for on-site testing, since they are quick, practical, and inexpensive.


Apparatus
All potentiometric measurements were carried out using a bench pH/mV meter (PXSJ-216 INESA, Scientific Instrument Co., Ltd., Shangahi, China). The electrochemical impedance and chronopotentiometric measurements were investigated by Metrohm Auto lap (B.V., model 204, Utrecht, The Netherlands). The reference method was performed by using high-performance liquid chromatography (HPLC) coupled with UV/VIS detector (Series 200 Pump, Perkin Elmer, Waltham, MA, USA). A Millipore Milli-Q system was utilized for obtaining de-ionized water (18.2 MΩ. cm specific resistance) to prepare all solutions. All measurements were applied at room temperature (25 ± 2 • C).
The reference method was the reversed-phase liquid chromatographic method for the quantification of desvenlafaxine [6]. A total of 20 µL of the sample was injected into a column C 18 (5 µm, 250 mm × 4.6 mm i.d.). The amount of ODV was determined using a UV detector operated at a wavelength of 228 nm Stock solutions of 10 −2 M desvenlafaxine hydrochloride were prepared by dissolving 0.38 g of the salt drug in 100 mL of double distilled water. Serial dilutions occurred to obtain different concentrations of the drug (10 −2 -10 −7 M), using 10 mM phosphate buffer at pH 6.0. All solutions were kept at 4 • C.

MIPs and NIPs Synthesis
For this investigation, the bulk polymerization technique protocol was used for the polymer's synthesis and characterization [42,45]. A total of 20 mg of the template ODV were dissolved in 6 mL of dimethyl sulfoxide (DMSO) at 40 • C for 5 min, while being stirred during the synthesis. Following this, 2-mmol of methacrylic acid (MAA), a function monomer, was added to the mixture, and heated for an additional 10 min at 40 • C. A total Polymers 2022, 14, 4814 4 of 14 of 3-mmol of ethylene glycol dimethyl acrylate (EGDMA), a cross-linker, was added to the mixture after it had been heated to 60 • C. To remove oxygen from the reaction, the system was heated for 10 min before being purged with nitrogen gas for 5 min. In a closed reaction system, 50 mg of benzoyl peroxide (BPO), an initiator, were added, and after a steady increase in temperature to 100 • C, polymerization was completed in 1 h. After being transferred to a Petri dish, the MIP was briefly dried in a 70 • C oven for a few mins. In a 50-mL centrifuge vial containing 40-mL of methanol, the MIP particles were added to the vial to remove the templet and produce the cavities. The vial was put into the shaking incubator, and this process was performed six times, while maintaining a temperature of 35 • C and 150 rpm of shaking. The mixture was centrifuged for five minutes, and the supernatant was then examined, using HPLC, for the imprint molecules. Until the imprinted molecules could no longer be found, this procedure was repeated. The NIP was similarly synthesized and processed, but there was not a templet in the reaction mixture.

Sensor Fabrication and EMF Measurements
Commercial screen-printed devices were used, all of them purchased from Metrohm DropSens Screen-Printed Carbon Electrodes (L33 × W10 × H0.5 mm; ref. C11L). The working (4-mm diameter) and auxiliary electrodes were made of carbon, while the reference electrode was made of silver/silver chloride. Initially, the sensing membrane was prepared after dissolving 10.0 mg of either MIPs or NIPs, 2 mg of KpClTPB, 49.0 mg of PVC, and 49.0 mg of o-NPOE in 2.0 mL THF. The working electrodes were modified by dropcasting of 10-µL of 2-mg SWCNT/1 mL THF. After complete dryness, 10-µL of the sensing membrane cocktail was drop-casted over the modified conducting carbon orifices ( Figure 1). The solid-state Ag/AgCl reference electrode was fabricated after adding 10-µL of the reference membrane. This membrane consists of 75.0 mg PVB and 50.0 mg NaCl dissolved in 1 mL of methanol, which was cured at 60 • C for 30 min. After that, an insulator layer was then placed onto the printed electrodes except for the two orifices of working and reference spots and the area of the electrical contact. The fabricated electrodes were stored at 4 • C when not used and used directly in the potentiometric measurements. The fabricated electrodes were conditioned in 10 −2 mol/L ODV solution for 2 h prior to their use. The calibration plots were constructed after plotting the potential readings versus the logarithm [ODV]. The electrodes were stored in 10 −2 mol/L ODV solution when not in use.

Electrochemical Impedance and Chronopotentiometric Measurements
Electrochemical impedance spectroscopy (EIS) measurements were carried out using a three-electrode cell, including Ag/AgCl/KCl (3 mol/L) as a reference electrode (6.0729.100, Metrohm AG CH-9101 HERISAU, Switzerland) and Pt wire as an auxiliary. The practical frequency range was adjusted to between 100 kHz and 0.1 Hz, with a sinus-

Electrochemical Impedance and Chronopotentiometric Measurements
Electrochemical impedance spectroscopy (EIS) measurements were carried out using a three-electrode cell, including Ag/AgCl/KCl (3 mol/L) as a reference electrode (6.0729.100, Metrohm AG CH-9101 HERISAU, Switzerland) and Pt wire as an auxiliary. The practical frequency range was adjusted to between 100 kHz and 0.1 Hz, with a sinusoidal excitation signal, through an excitation amplitude of 10 mV. The measurements were applied by using a solution of 10 −3 mol/L of ODV in phosphate buffer (10 mmol/L) of pH 6.0, at an ambient temperature of 25 ± 1 • C.
For chronopotentiometric measurements, ±1 nA current was applied to the working electrode for 60 s, and the drift in potential was recorded against the time.

Analytical Applications
Drug stores provided some medications with desvenlafaxine, which is frequently used as an antidepressant. Ten pills were well ground before being sonicated in 5.0 mL of pH 6.0, 10 mM phosphate buffer. Using the developed all-solid-state ODV electrode, an aliquot equivalent to one tablet was employed for direct potentiometric measurements. A blank experiment was run in the same manner, and the recorded potential was compared with the calibration plot.
Adult males and females provided their urine samples, which were then spiked with various ODV doses and submitted to potentiometric drug assays. The goal of the experiment was to examine how a complex biological matrix might affect the performance of ODV-membrane-based sensors and drug recovery.

Sensors' Characteristics
The presented all-solid-state electrodes for ODV assessment were fabricated using synthesized main-tailored MIP beads as an active and selective recognition element. The non-modified (C/MIP-ODV) and the modified (C/SWCNTs/MIP-ODV) electrodes were labeled as electrode I, and electrode II, respectively. For the membrane-based sensor (C/SWCNTs/NIP-ODV) was labelled as electrode III. The performance characteristics of the presented electrodes were presented in Table 2. For electrodes I and II, they exhibited a Nernstian cationic slope of 56.4 ± 1.1, and a 57.2 ± 0.8 mV/decade with detection limits of 3.0 × 10 −6 , and 2.0 × 10 −6 M (R 2 =0.999), respectively. For electrode III, it exhibited a sub-Nernstian response with a slope of 38.3 ± 2.3 mV/decade over the linear range of 7.0 × 10 −5 -1.0 × 10 −3 M, and a detection limit of 5.0 × 10 −5 M. Table 2. Analytical performance of the presented electrodes in 10 mmol/L phosphate buffer, pH 6. From the data obtained, it was found that the best analytical performances were obtained using electrodes I and II. This can be attributed to the successful imprinting of the synthesized MIP beads and proves that the solid-contact SWCNTs layer has no effect on the potential response of the fabricated electrodes. The potentiometric plots of electrodes I and II are shown in Figure 2.

Parameter C/MIP-ODV C/SWCNTs/MIP-ODV C/SWCNTs/NIP-ODV
From the data obtained, it was found that the best analytical performances were obtained using electrodes I and II. This can be attributed to the successful imprinting of the synthesized MIP beads and proves that the solid-contact SWCNTs layer has no effect on the potential response of the fabricated electrodes. The potentiometric plots of electrodes I and II are shown in Figure 2.

Analytical Procedure Validation Study
To compare a defined characteristic of the drug substance or drug product to predetermined acceptance criteria for that characteristic, an analytical technique was designed. The choice of analytical instruments and methodology must be made early in the

Analytical Procedure Validation Study
To compare a defined characteristic of the drug substance or drug product to predetermined acceptance criteria for that characteristic, an analytical technique was designed. The choice of analytical instruments and methodology must be made early in the development of a new analytical procedure, depending on the desired purpose and scope of the analytical method. Specificity, linearity, limits of detection (LOD), limits of quantitation (LOQ), range, accuracy, and precision are among the parameters that may be assessed during method development. According to ICH guidelines [47], the presented method was verified for accuracy, precision, specificity, detection limit, quantitation limit, and robustness.

Accuracy and Precision of the Method
The recoveries of ODV using the method of standard additions were calculated to assess the method's accuracy. A pre-quantified sample (10 µg/mL) was mixed with known concentrations of ODV (0, 4, 7.5, and 12 µg/mL), and the concentrations were then determined. The recoveries ranged from 97.8 to 103.5%. The approach is accurate, as evidenced by the high recovery-values.
By repeatedly measuring a solution containing 10 µg/mL of ODV, the instrument's precision was assessed. Relative standard deviation was used to describe the results. The results of the intra-day and inter-day precision study of ODV are reported in terms of relative standard deviation. The study involved estimating the corresponding responses three times on the same day, and on three different days (first, second, and third day) for three different concentrations of DVX (1, 5 and 10 µg/mL). The potential measurement repeatability test was used to determine the instrument's precision, and it revealed that the ODV's RSD value was 1.1%. This was done to conduct intra-day and inter-day precision studies. RSD values for the ODV were determined to be 0.9-1.2% for the intra-day study and 1.1-1.5% for the inter-day precision study. The approach is precise, as evidenced by the low RSD readings.

Detection Limit and Quantification
The lowest concentration of an analyte at which background levels may be reliably distinguished is known as the detection limit (DL). The lowest amount of analyte that can be quantitatively measured with enough precision and accuracy is the limit of quantification (LOQ) of a specific analytical method. According to ICH recommendations, DL and LOQ were determined, using the following equation: LOD = 3.3 σ/S and LOQ = 10 σ/S where σ is the standard deviation of the y-intercepts of regression lines and S is the slope of the calibration curve. The limit of quantification (LOQ) for ODV was 5.0 × 10 −6 M, while the detection limit was 2.0 × 10 −6 M. The aforementioned information demonstrates that a drug's microgram quantity can be measured exactly and accurately. Table 2 provides an overview of the validation parameters.

pH Effect and Sensors' Selectivity
Robustness is the measure of an analytical method's level of repeatability or reliability when subjected to deliberate modification (external factors). However, the term robustness is defined as a metric assessing the stability of the results regarding too slight changes (internal factors) [48]. To test the method's robustness, the pH of the test solution was varied throughout a wide range, from 2 to 10. Following this, the potentiometric characteristics of an ODV membrane-based sensor were assessed at two drug concentrations (10 −4 and 10 −5 M), with respect to changes in pH values that were corrected using modest volumes of HCl and/or NaOH. Figure 3 depicts the effects of various pH values on the potentiometric response of the presented sensor, revealing the ranges of their stability over a pH range of 3.5-8.0 for the ODV membrane-based sensor. The pKa values of the ODV drug was 9.45 (amine) and 10.66 (phenol) [49]. At the above-mentioned pH range, ODV is completely ionized, and present in its cationic form. At pH > 8.5, the potential response dramatically declined, due to the formation of the non-cationic ODV form. Wide steady-potential reading ranges supported the investigation of highly robust and durable potentiometric sensors for ODV assessment. A phosphate buffer with a working pH of 10 mM (pH 6) was selected for additional potentiometric investigations of the applied sensor. steady-potential reading ranges supported the investigation of highly robust and durable potentiometric sensors for ODV assessment. A phosphate buffer with a working pH of 10 mM (pH 6) was selected for additional potentiometric investigations of the applied sensor. Investigations were conducted on the numerous interference behaviors that may manifest in the matrices during ODV determination. The modified separate solution method (MSSM) was used to evaluate the selectivity coefficient values [50]. To determine the impact of the transducer's nature on the electrode's selectivity behavior, the selectivity test was conducted on the presented electrodes (I and II). Since different species can coexist with the drug in either its pharmaceutical formulations or biological fluids, various interfering species were therefore chosen. These species included cationic salts (Na + , K + , Mg 2+ , and Ca 2+ ), carbohydrates (glucose and lactose), medicines, amino acids (alanine, ar- Investigations were conducted on the numerous interference behaviors that may manifest in the matrices during ODV determination. The modified separate solution method (MSSM) was used to evaluate the selectivity coefficient values [50]. To determine the impact of the transducer's nature on the electrode's selectivity behavior, the selectivity test was conducted on the presented electrodes (I and II). Since different species can coexist with the drug in either its pharmaceutical formulations or biological fluids, various interfering species were therefore chosen. These species included cationic salts (Na + , K + , Mg 2+ , and Ca 2+ ), carbohydrates (glucose and lactose), medicines, amino acids (alanine, arginine, and glycine), and amino acids (tramadol, venlafaxine, and aspirin). The selectivity coefficients (log K pot ODV,j ) are determined in Table 3 and indicate that the suggested potentiometric sensors were not significantly affected by the contaminated interfering ions. Additionally, neither the ODV applied sensors' selectivity behavior, nor the transducer's type have any impact. These sensors provided highly accurate and effective ODV matrix determination. Table 3. Selectivity coefficients (log K pot ODV,J ) of the presented ODV-membrane based sensors.

EIS and Chronopotentiometric Measurements
As shown in Figure 4A,B, the EIS measurements revealed the Nyquist relations (complex plane plots of -Z" vs. Z \ ) on the equivalent circuit models. The unveiled results of EIS plots are represented in Table 4. The results showed the significant effect of inserting SWCNTs as a transducing material between the ion-sensing membrane and the conducting substrate. The presence of the SWCNTs layer affected the bulk resistance (R b *), which decreased significantly from 0.1 ± 0.04 to 0.005 ± 0.0002 MΩ. On the other hand, the existence of this transducing material increased around fivefold the double-layer capacitances (C dl ) from 11.6 ± 0.6 to 91.7 ± 3.4 µF, which aided the electrical double-layer formation at the polymeric ISE membrane/solid contact interface [51]. In addition, it had a higher impact on increasing the geometric capacitances (C g ), which increased from 0.03 ± 0.001 to 0.88 ± 0.06 nF. All obtained data reflect the high lipophilicity of SWCNTs and the formation of high double-layer capacitance between the ion-selective membrane (ISM) and the electronic conducting transducer.
In line with Bobacka's method [52], the short-term potential stability of the presented electrodes was investigated. The chronopotentiograms for the presented electrodes in the absence and presence of the SWCNTs layer, are shown in Figure 4. The bulk membrane resistance (R b **) was calculated and decreased noticeably from 0.45 ± 0.04 to 0.15 ± 0.03 MΩ for the ODV membrane-based sensor in the absence and presence of the SWCNTs layer, respectively ( Table 4). The potential drift (∆E/∆t) showed a significant decrease from 57.8 ± 1.1 to 10.6 ± 2.1 µV/s in the presence of the SWCNTs conducting layer. These data reflect the incredible increase in potential stability in the presence of the SWCNTs layer. The double-layer capacitance (C L ) increased around threefold, from 13.2 ± 1.4 to 76.2 ± 1.6 µF in the presence of the SWCNTs conducting layer. The results confirm that the presence of the SWCNTs layers as transducing material between the ion-sensing membrane and the conducting substrate, exhibited high potential stability, conductivity, and high compatibility with the presented electrodes for a reliable determination of ODV in various complex matrices, without potential drift. In line with Bobacka's method [52], the short-term potential stability of the presented electrodes was investigated. The chronopotentiograms for the presented electrodes in the absence and presence of the SWCNTs layer, are shown in Figure 4. The bulk membrane resistance (Rb**) was calculated and decreased noticeably from 0.45 ± 0.04 to 0.15 ± 0.03 MΩ for the ODV membrane-based sensor in the absence and presence of the SWCNTs layer, respectively ( Table 4). The potential drift (∆E/∆t) showed a significant decrease from 57.8 ± 1.1 to 10.6 ± 2.1 μV/s in the presence of the SWCNTs conducting layer. These data reflect the incredible increase in potential stability in the presence of the SWCNTs layer. The double-layer capacitance (CL) increased around threefold, from 13.2 ± 1.4 to 76.2 ± 1.6 μF in the presence of the SWCNTs conducting layer. The results confirm that the presence of the SWCNTs layers as transducing material between the ion-sensing membrane and the conducting substrate, exhibited high potential stability, conductivity, and high compatibility with the presented electrodes for a reliable determination of ODV in various complex matrices, without potential drift.

Sensor's Durability
The durability of the proposed sensors for ODV determination was investigated after recording the potential (mV) vs. time (min). The test was performed by inserting the presented electrodes in 10 mM of phosphate buffer (pH 6) for 30 min. The solution was re-

Sensor's Durability
The durability of the proposed sensors for ODV determination was investigated after recording the potential (mV) vs. time (min). The test was performed by inserting the presented electrodes in 10 mM of phosphate buffer (pH 6) for 30 min. The solution was replaced by 10 −5 M of ODV, and the electrodes were immersed in this solution for another 30 min. Afterward, the solution was replaced by the phosphate buffer solution, and the electrodes were left in this solution for another 30 min. The potential response of the electrodes was recorded over these time intervals. The same steps were repeated for 10 -4 M ODV concentration. As shown in Figure 5, there is a noticeable potential drift for the electrode that does not contain the SWCNTs layer. This confirms the formation of a water-layer between the ion-sensing membrane and the electronic conductor substrate. The electrodes modified with SWCNTs exhibited high potential-stability over long time of the proceeding measurements.

Assessment of ODV by Direct Potentiometric Measurements
To assess the suitability of the suggested sensor for drug analysis, electrode II was used to measure the amount of ODV in different pharmaceutical goods. The potential of the solution was determined after the medication tablets were crushed, ground, dissolved in phosphate buffer solution of pH 6, sonicated, and filtered. The findings showed that 99.7 ± 1.7-100.2 ± 0.9% of the nominal values were recovered (Table 5). These outcomes were contrasted, using HPLC, with calculated data from the reference technique [6]. The two approaches were compared using F and t-Student tests; however, the results showed no discernible difference between them, confirming the applicability of the suggested methods for determining ODV in its solutions.

ODV Recovery from Spiked Urine Samples
To use this sensor in identifying overdose patients, particularly in situations where a quick and accurate evaluation diagnosis is necessary, monitoring ODV in urine samples was also evaluated. Following the addition of known ODV concentrations to various aliquots of human urine samples, potentiometric measurements were taken, using the suggested sensor. As shown in Table 6, the data revealed an average recovery of 98.2% with-

Assessment of ODV by Direct Potentiometric Measurements
To assess the suitability of the suggested sensor for drug analysis, electrode II was used to measure the amount of ODV in different pharmaceutical goods. The potential of the solution was determined after the medication tablets were crushed, ground, dissolved in phosphate buffer solution of pH 6, sonicated, and filtered. The findings showed that 99.7 ± 1.7-100.2 ± 0.9% of the nominal values were recovered (Table 5). These outcomes were contrasted, using HPLC, with calculated data from the reference technique [6]. The two approaches were compared using F and t-Student tests; however, the results showed no discernible difference between them, confirming the applicability of the suggested methods for determining ODV in its solutions.

ODV Recovery from Spiked Urine Samples
To use this sensor in identifying overdose patients, particularly in situations where a quick and accurate evaluation diagnosis is necessary, monitoring ODV in urine samples was also evaluated. Following the addition of known ODV concentrations to various aliquots of human urine samples, potentiometric measurements were taken, using the suggested sensor. As shown in Table 6, the data revealed an average recovery of 98.2% without any notable interferences from the species that are often present in samples of human urine-spiked medication. It has been reported that desvenlafaxine is the major active metabolite of venlafaxine (VEN). VEN is highly metabolized in humans, with urinary excretion of the unchanged compound between 1-10% of an administered dose [53]. Demethylation of o-desmethylvenlafaxine (ODV), is the main metabolite produced. Approximately 45% of desvenlafaxine is excreted unchanged in urine at 72 h after oral administration. Other metabolites, N,O-didesmethylvenlafaxine (16%), and N-desmethylvenlafaxine (1%), are biologically inactive [53] (Figure 6).  It has been reported that desvenlafaxine is the major active metabolite of venlafaxin (VEN). VEN is highly metabolized in humans, with urinary excretion of the unchange compound between 1-10% of an administered dose [53]. Demethylation of o-desmethy venlafaxine (ODV), is the main metabolite produced. Approximately 45% of desvenlafax ine is excreted unchanged in urine at 72 h after oral administration. Other metabolites N,O-didesmethylvenlafaxine (16%), and N-desmethylvenlafaxine (1%), are biologicall inactive [53] (Figure 6). The average half-life of VEN is 5.0 ± 2.0 h, whereas that of its main metabolite, ODV is 11.0 ± 2.0 h. According to the selectivity measurements, ODV can be measured in th presence of either venlafaxine and N,O-didesmethylvenlafaxine. The metabolite N desmethylvenlafaxine interferes during the measurements of ODV, but the amount pro duced is 1%. This backs up the usefulness of the sensor that was described for measurin ODV in actual samples, with little or no interference from the main drug metabolite.

Conclusions
Desvenlafaxine is monitored in some therapeutic drugs and spiked human urin samples using an affordable, miniature, all-solid state-based sensor that is very sensitiv and selective in its determination. The sensor works by using MIP beads dispersed in PVC membrane as a substance for detecting drugs. Ion-to-electron transducers and soli contact devices are both made of single-walled carbon nanotubes (SWCNTs). It consider The average half-life of VEN is 5.0 ± 2.0 h, whereas that of its main metabolite, ODV, is 11.0 ± 2.0 h. According to the selectivity measurements, ODV can be measured in the presence of either venlafaxine and N,O-didesmethylvenlafaxine. The metabolite N-desmethylvenlafaxine interferes during the measurements of ODV, but the amount produced is 1%. This backs up the usefulness of the sensor that was described for measuring ODV in actual samples, with little or no interference from the main drug metabolite.

Conclusions
Desvenlafaxine is monitored in some therapeutic drugs and spiked human urine samples using an affordable, miniature, all-solid state-based sensor that is very sensitive