Nanochannel Array on Electrochemically Polarized Screen Printed Carbon Electrode for Rapid and Sensitive Electrochemical Determination of Clozapine in Human Whole Blood

Rapid and highly sensitive determination of clozapine (CLZ), a psychotropic drug for the treatment of refractory schizophrenia, in patients is of great significance to reduce the risk of disease recurrence. However, direct electroanalysis of CLZ in human whole blood remains a great challenge owing to the remarkable fouling that occurs in a complex matrix. In this work, a miniaturized, integrated, disposable electrochemical sensing platform based on the integration of nanochannel arrays on the surface of screen-printed carbon electrodes (SPCE) is demonstrated. The device achieves high determination sensitivity while also offering the electrode anti-fouling and anti-interference capabilities. To enhance the electrochemical performance of SPCE, simple electrochemical polarization including anodic oxidation and cathodic reduction is applied to pretreat SPCE. The electrochemically polarized SPCE (p-SPCE) exhibits an enhanced electrochemical peak signal toward CLZ compared with bare SPCE. An electrochemically assisted self-assembly method (EASA) is utilized to conveniently electrodeposit a vertically ordered mesoporous silica nanomembrane film (VMSF) on the p-SPCE, which could further enrich CLZ through electrostatic interactions. Owing to the dual signal amplification based on the p-SPCE and VMSF nanochannels, the developed VMSF/SPCE sensor enables determination of CLZ in the range from 50 nM to 20 μM with a low limit of detection (LOD) of 28 nM (S/N = 3). Combined with the excellent anti-fouling and anti-interference abilities of VMSF, direct and sensitive determination of CLZ in human blood is also achieved.


Introduction
Schizophrenia is a common mental illness that places a great burden on society. Some 20-30% of patients suffer from treatment-resistant schizophrenia (TRS) [1]. Research has shown that clozapine (CLZ) is the only drug with good therapeutic effect on TRS [2][3][4]. The intake of CLZ plays a crucial role in the recovery of patients. Excessive CLZ results in a lot of side effects including excessive salivation [5], agranulocytosis [6], seizures [7], weight gain [8], and mumps [9], which will cause secondary damage to the patient's health [10]. Therefore, real-time determination of CLZ is of great significance to reduce the risk of relapse and readmission of patients.
with the excellent anti-fouling and anti-interference abilities of VMSF, the constructed sensor can realize rapid and sensitive determination of CLZ in human blood. This work demonstrates a new strategy for the construction of a reliable, miniaturized, disposable and integrated electrochemical sensor with an anti-fouling layer and high detection sensitivity and selectivity.

Electrochemical Polarization of SPCE
As illustrated in Figure 1, SPCE was first treated by simple electrochemical polarization to enhance its electrochemical performance. Electrochemical polarization is a simple and green method to prepare highly active carbon electrodes. This method usually involves the electrochemical oxidization and reduction of electrodes in conventional electrolyte solutions. Usually, the electrode is first electrochemically oxidized at a high potential and then electrochemically reduced at a low potential. Thus, no complex chemical reagents and tedious operations are needed. During the polarization process, the sp 2 -conjugated carbon on the surface of the carbon electrode is oxidatively etched at high potential, resulting in abundant edge carbon, defects and oxygen-containing functional groups. These groups serve as electrocatalytic active sites which can not only enhance the adsorption of organic electroactive molecules, but also facilitate interfacial electron transfer reactions. In addition, electrochemical polarization facilitates the formation of porous surfaces. Therefore, electrochemically polarized carbon electrodes tend to exhibit high electrochemical response and significant electrocatalytic activity. response to CLZ. Combined with the excellent anti-fouling and anti-interfe of VMSF, the constructed sensor can realize rapid and sensitive determina human blood. This work demonstrates a new strategy for the constructio miniaturized, disposable and integrated electrochemical sensor with an ant and high detection sensitivity and selectivity.

Electrochemical Polarization of SPCE
As illustrated in Figure 1, SPCE was first treated by simple electroch zation to enhance its electrochemical performance. Electrochemical po simple and green method to prepare highly active carbon electrodes. Thi ally involves the electrochemical oxidization and reduction of electrodes in electrolyte solutions. Usually, the electrode is first electrochemically oxid potential and then electrochemically reduced at a low potential. Thus chemical reagents and tedious operations are needed. During the polariz the sp 2 -conjugated carbon on the surface of the carbon electrode is oxidat high potential, resulting in abundant edge carbon, defects and oxygen-co tional groups. These groups serve as electrocatalytic active sites which ca hance the adsorption of organic electroactive molecules, but also facili electron transfer reactions. In addition, electrochemical polarization faci mation of porous surfaces. Therefore, electrochemically polarized carbon e to exhibit high electrochemical response and significant electrocatalytic act The cyclic voltammetry curves obtained in the electrochemical electr ard electrochemical probe solution (K3Fe(CN)6) on SPCE before and after e polarization are shown in Figure 2a. Compared with SPCE, p-SPCE sho cantly increased charging current (Inset in Figure 2a), indicating an increa troactive area of the electrode through electrochemical polarization. In the probe, p-SPCE exhibited higher peak current and lower peak-to-peak sep The cyclic voltammetry curves obtained in the electrochemical electrolyte or standard electrochemical probe solution (K 3 Fe(CN) 6 ) on SPCE before and after electrochemical polarization are shown in Figure 2a. Compared with SPCE, p-SPCE showed a significantly increased charging current (Inset in Figure 2a), indicating an increase in the electroactive area of the electrode through electrochemical polarization. In the case of a redox probe, p-SPCE exhibited higher peak current and lower peak-to-peak separation, showing an improved electron transfer rate after electrochemical polarization. This phenomenon was further confirmed by electrochemical impedance spectroscopy (EIS), as shown in Figure 2b. It is well known that the charge transfer resistance (R ct ) of an electrode is related to the diameter of the semicircular curve in the high frequency region. When the SPCE was electrochemically polarized, the p-SPCE displayed a reduced R ct owing to the increased electron transfer rate. ing an improved electron transfer rate after electrochemical polarization. This phenomenon was further confirmed by electrochemical impedance spectroscopy (EIS), as shown in Figure 2b. It is well known that the charge transfer resistance (Rct) of an electrode is related to the diameter of the semicircular curve in the high frequency region. When the SPCE was electrochemically polarized, the p-SPCE displayed a reduced Rct owing to the increased electron transfer rate.

VMSF Enquipment on the p-SPCE
After electrochemical polarization of the SPCE, VMSF was then electrodeposited on the electrode surface by the EASA method ( Figure 1). EASA is a convenient method by which to apply VMSF to conductive substrates; it can complete the rapid growth of VMSF within 10 s. When the p-SPCE was immersed in an acidic precursor solution containing siloxane (TEOS) and cetyltrimethylammonium bromide (CTAB) micelles, a negative current was applied to the electrode and OH-was generated. This local pH increase induced the self-assembly of surfactant micelles (SM) and the polycondensation of siloxane, leading to a hexagonally packed pore structure. The stable binding of VMSF with p-SPCE was attributed to two factors. First, the negatively charged oxygen-containing functional groups on the surface of p-SPCE facilitated the electrostatic adsorption of the cationic SM. Second, the OH groups on p-SPCE formed Si-O bonds with VMSF through a co-condensation reaction, bestowing the adhesion of VMSF on p-SPCE with high mechanical stability. Finally, the SM in the nanochannels was removed with HCl-EtOH solution to obtain a VMSF/p-SPCE with open nanochannels.
Transmission electron microscopy (TEM) was used to characterize the morphology of the nanopore/channel structure of VMSF. Figure 3 demonstrates top-view and cross-sectional TEM images of the VMSF nanochannel. As shown in Figure 3a, the VMSF had a well-ordered structure with uniform pore size. The diameter of the nanopores was between 2 nm and 3 nm. The cross-sectional view in Figure 3b clearly reveals the parallel nanochannel structure.

VMSF Enquipment on the p-SPCE
After electrochemical polarization of the SPCE, VMSF was then electrodeposited on the electrode surface by the EASA method ( Figure 1). EASA is a convenient method by which to apply VMSF to conductive substrates; it can complete the rapid growth of VMSF within 10 s. When the p-SPCE was immersed in an acidic precursor solution containing siloxane (TEOS) and cetyltrimethylammonium bromide (CTAB) micelles, a negative current was applied to the electrode and OH-was generated. This local pH increase induced the self-assembly of surfactant micelles (SM) and the polycondensation of siloxane, leading to a hexagonally packed pore structure. The stable binding of VMSF with p-SPCE was attributed to two factors. First, the negatively charged oxygen-containing functional groups on the surface of p-SPCE facilitated the electrostatic adsorption of the cationic SM. Second, the OH groups on p-SPCE formed Si-O bonds with VMSF through a co-condensation reaction, bestowing the adhesion of VMSF on p-SPCE with high mechanical stability. Finally, the SM in the nanochannels was removed with HCl-EtOH solution to obtain a VMSF/p-SPCE with open nanochannels.
Transmission electron microscopy (TEM) was used to characterize the morphology of the nanopore/channel structure of VMSF. Figure 3 demonstrates top-view and crosssectional TEM images of the VMSF nanochannel. As shown in Figure 3a, the VMSF had a well-ordered structure with uniform pore size. The diameter of the nanopores was between 2 nm and 3 nm. The cross-sectional view in Figure 3b clearly reveals the parallel nanochannel structure.  The integrity of VMSF was further proven through electrochemical characterization. The CV curves and Nyquist plots of the obtained on SM@VMSF/p-SPCE and VMSF/p-SPCE are shown in Figure 3c,d. As seen, SM@VMSF/p-SPCE exhibited almost no peak current while VMSF/p-SPCE had remarkable peaks ( Figure 3c). Thus, the filling of SM inside VMSF nanochannels inhibited the mass transfer of the redox probe through the nanochannels and then prohibited the subsequent electron transfer of between the electrode. EIS experiments also demonstrated the same phenomenon ( Figure 3d). An electrode containing micelles in nanochannels showed extremely high charge transfer resistance. On the other hand, the VMSF/p-SPCE with open nanochannels presented significantly lower charge transfer resistance. The above results demonstrate the successful modification of VMSF on a p-SPCE and the integrity of the nanofilm.

Dual Signal Amplification and Significantly Enhanced CLZ Response on VMSF/p-SPCE
The CV and DPV curves of CLZ on different electrodes are compared in Figure 4. It can be seen that CLZ exhibited a pair of reversible redox peaks, which were attributed to its redox reaction on the electrode surface ( Figure 4a). The response of bare SPCE to CLZ had the lowest peak current (Figure 4b). In contrast, the response of p-SPCE to CLZ was significantly enhanced. This was attributed to the enhancement of electrode performance by electrochemical polarization. On the one hand, electrochemical polarization increased the electroactive area, thereby increasing the interaction with the CLZ. On the other The integrity of VMSF was further proven through electrochemical characterization. The CV curves and Nyquist plots of the obtained on SM@VMSF/p-SPCE and VMSF/p-SPCE are shown in Figure 3c,d. As seen, SM@VMSF/p-SPCE exhibited almost no peak current while VMSF/p-SPCE had remarkable peaks ( Figure 3c). Thus, the filling of SM inside VMSF nanochannels inhibited the mass transfer of the redox probe through the nanochannels and then prohibited the subsequent electron transfer of between the electrode. EIS experiments also demonstrated the same phenomenon ( Figure 3d). An electrode containing micelles in nanochannels showed extremely high charge transfer resistance. On the other hand, the VMSF/p-SPCE with open nanochannels presented significantly lower charge transfer resistance. The above results demonstrate the successful modification of VMSF on a p-SPCE and the integrity of the nanofilm.

Dual Signal Amplification and Significantly Enhanced CLZ Response on VMSF/p-SPCE
The CV and DPV curves of CLZ on different electrodes are compared in Figure 4. It can be seen that CLZ exhibited a pair of reversible redox peaks, which were attributed to its redox reaction on the electrode surface (Figure 4a). The response of bare SPCE to CLZ had the lowest peak current (Figure 4b). In contrast, the response of p-SPCE to CLZ was significantly enhanced. This was attributed to the enhancement of electrode performance by electrochemical polarization. On the one hand, electrochemical polarization increased the electroactive area, thereby increasing the interaction with the CLZ. On the other hand, the active groups generated by electrochemical polarization enhanced the electron transfer rate of p-SPCE. Therefore, electrochemical polarization can enhance the sensitivity to CLZ. When VSMF was grown on the surface of p-SPCE, VMSF/p-SPCE showed the highest peak current, demonstrating the enrichment of CLZ by VMSF nanochannels. This was attributed to the electron sieving effect of the nanochannels. In the electrolyte solution, the silanol groups on the surface of VMSF were negatively charged (pK a = 2-3), which allowed electrostatic adsorption of CLZ (pK a = 7.6) to occur. Thus, the VMSF/p-SPCE had a dual signal amplification effect on the determination of CLZ, indicating its potential for as a sensitive CLZ detector. hand, the active groups generated by electrochemical polarization enhanced the electron transfer rate of p-SPCE. Therefore, electrochemical polarization can enhance the sensitivity to CLZ. When VSMF was grown on the surface of p-SPCE, VMSF/p-SPCE showed the highest peak current, demonstrating the enrichment of CLZ by VMSF nanochannels. This was attributed to the electron sieving effect of the nanochannels. In the electrolyte solution, the silanol groups on the surface of VMSF were negatively charged (pKa = 2-3), which allowed electrostatic adsorption of CLZ (pKa = 7.6) to occur. Thus, the VMSF/p-SPCE had a dual signal amplification effect on the determination of CLZ, indicating its potential for as a sensitive CLZ detector.

Optimization of Conditions for the Determination of CLZ
To improve the sensitivity to CLZ, conditions such as pH, enrichment time, and ionic strength were optimized. Figure 4c shows the DPV curves of CLZ on VMSF/p-SPCE at different pH. With the increase of pH, the oxidation peak potential of CLZ moved to the negative potential, and the peak current also changed greatly. The highest peak current was observed at pH 6. This was attributed to the effect of pH on the electrostatic interaction between the nanochannels with CLZ. An excessively low pH would reduce the negative charge on the surface of the VMSF nanochannels, which reduced the electrostatic force on CLZ. On the other hand, an overly high pH would reduce the positive charge of CLZ. Thus, pH = 6 was chosen for further investigation. Figure 4d displays the CV curves of CLZ at different scan rates. With the increase of scan rate, both the oxidation peak current and reduction peak current increased accordingly. Good linearity was observed between the peak current and the scan rate, indicating an adsorption-controlled process (inset in Figure 4d). Thus, the effect of enrichment time on the peak current of CLZ was investigate. As shown in Figure 4e, the peak current reached equilibrium after 60 s, indicating very short time for CLZ to reach mass transfer equilibrium on the electrode surface. This was attributed to the good permeability of VMSF and dual enrichment of the VMSF/p-SPCE. The ionic strength of the electrolyte solution affected the thickness of the electric double layer of the nanochannels, which changed the electrostatic interaction between the nanochannel and CLZ. Figure 4f shows the current responses of CLZ at different ionic strengths. When the ionic strength of the electrolyte solution was high, the electric double layer (EDL) did not overlap and the nanochannel exhibited electrostatic attraction to positively charged substances [44]. Thus, 0.1 M PBS was chosen as the electrolyte solution for the determination of CLZ.

Determination of CLZ Using VMSF/SPCE
The performance of the developed VMSF/p-SPCE sensor for the determination of CLZ using was investigated. Figure 5a displays the DPV curves obtained on the VMSF/p-SPCE in the presence of different concentrations of CLZ. As shown, the peak current gradually increased when the concentration of CLZ increased. A linear correlation was revealed between the oxidation peak current (I, µA) and concentration of CLZ (C, µM) in the range from 50 nM to 20 µM (I = 2.6 C − 0.05, R 2 = 0.996) (Figure 5b). The limit of detection (LOD) was 28 nM with a signal-to-noise ratio of 3. For comparison, linear detection of CLZ ranging from 1 µM to 15 µM was obtained on a bare SPCE (I = 0.647 C + 0.052, R 2 = 0.997). In the case of the p-SPCE, CLZ could be linearly detected from 0.5 to 20 µM (I = 1.44 C + 0.11, R 2 = 0.993). Thus, VMSF/p-SPCE exhibited the highest sensitivity. The LOD obtained on the VMSF/p-SPCE was lower than that on GCE modified by multiwall carbon nanotubes or on WO 3 nanoparticles hydride modified by α-terpineol [45], Bi-Sn nanoparticles applied to carbon aerogel-modified SPCE (Bi-Sn NP/CAG/SPCE) [46], ruthenium (IV) oxide nanoparticle-modified SPCE (RuO 2 NPs/SPE) [31] or a TiO 2 nanoparticlemodified carbon paste electrode (TiO 2 NP-MCPE) [47]. The low LOD of the developed sensor may be attributed to the dual amplification effects resulting from both the p-SPCE and VMSF nanochannels.

Anti-Interference and Anti-Fouling Properties of the VMSF/p-SPCE Sensor
Anti-interference and anti-fouling capabilities are important characteristics of electrochemical sensors, and are especially important for analyses of real samples with complex matrices. To study the anti-interference ability of the developed VMSF/p-SPCE sensor, the effects of common substances in human blood on the determination of CLZ were investigated. Briefly, electrolyte ions (K + , Na + , Mg 2+ , Ca 2+ ), common redox small molecules (ascorbic acid-AA, dopamine-DA, uric acid-UA) and metabolites (urea, glucose-Glu) were tested. As shown in Figure 6a, determinations of CLZ were not affected even when the concentrations of the above substances were 25 times higher than that of CLZ, proving the good anti-interference characteristic of the sensor. It is particularly noteworthy that three electroactive small molecules, i.e., AA, DA, and UA, did not interfere with the determination of CLZ, indicating that the constructed VMSF/p-SPCE sensor has excellent potential resolution. Blood samples or tablets associated with CLZ often contain large amounts proteins, hydroxypropyl methyl cellulose (HPMC) and starch. Therefore, the effects of bovine serum albumin (BSA), HPMC and soluble starch on the determination of CLZ were investigated. As shown in Figure 6b-d, these three species did not interfere with the determination of CLZ, further revealing the good anti-fouling ability of the electrode. In contrast, the responses changed significantly in presence or absence of these species (inset in Figure 6b-d). The excellent anti-fouling performance of the VMSF/p-SPCE sensor was attributed to the size and electrostatic exclusion effect of the VMSF nanochannels.

Anti-Interference and Anti-Fouling Properties of the VMSF/p-SPCE Sensor
Anti-interference and anti-fouling capabilities are important characteristics of electrochemical sensors, and are especially important for analyses of real samples with complex matrices. To study the anti-interference ability of the developed VMSF/p-SPCE sensor, the effects of common substances in human blood on the determination of CLZ were investigated. Briefly, electrolyte ions (K + , Na + , Mg 2+ , Ca 2+ ), common redox small molecules (ascorbic acid-AA, dopamine-DA, uric acid-UA) and metabolites (urea, glucose-Glu) were tested. As shown in Figure 6a, determinations of CLZ were not affected even when the concentrations of the above substances were 25 times higher than that of CLZ, proving the good anti-interference characteristic of the sensor. It is particularly noteworthy that three electroactive small molecules, i.e., AA, DA, and UA, did not interfere with the determination of CLZ, indicating that the constructed VMSF/p-SPCE sensor has excellent potential resolution. Blood samples or tablets associated with CLZ often contain large amounts proteins, hydroxypropyl methyl cellulose (HPMC) and starch. Therefore, the effects of bovine serum albumin (BSA), HPMC and soluble starch on the determination of CLZ were investigated. As shown in Figure 6b-d, these three species did not interfere with the determination of CLZ, further revealing the good anti-fouling ability of the electrode. In contrast, the responses changed significantly in presence or absence of these species (inset in Figure 6b Insets are the corresponding DPV curves obtained using single CLZ (black) or the binary mixture (red).

Determination of CLZ in Human Whole Blood with Low Sample Consumption
The excellent anti-fouling and anti-interference properties of the constructed VMSF/p-SPCE sensor show its great potential for use in direct electroanalyses of CLZ in complex samples. As a proof-of-concept, electrochemical determination of CLZ in human whole blood was investigated. Human whole blood was applied as the detection medium after dilution by a factor of 100. For comparison, determinations using a bare SPCE or p-SPCE were also undertaken. As shown in Figure 7a, the DPV peak current on the VMSF/p-SPCE increased with an increase in the concentration of CLZ. Although the detection sensitivity was reduced compared with that of the buffer system, the VMSF/p-SPCE sensor could still linearly detect CLZ from 0.5 μM to 12 μM (I = 0.563 C + 0.04, R 2 = 0.996) with a LOD of 110 nM (Figure 7b). In contrast, the CLZ signal on both the bare SPCE and p-SPCE showed a non-linear relationship with its concentration, indicating that the electrode was seriously contaminated. The regeneration and reusability of the constructed VMSF/p-SPCE sensor was investigated. The performance of the p-SPCE was also examined for comparison. The sensor can easily be regenerated by soaking in a hydrochloric acid-ethanol solution (0.1 M) for 2 min. As shown in Figure 7c,d, both electrodes can be easily regenerated, although the CLZ signals on the regenerated electrodes were almost undetectable in the electrolyte. In addition, the VMSF/p-SPCE sensor can be reused with no significant change in its response to CLZ (Figure 7c). However, the

Determination of CLZ in Human Whole Blood with Low Sample Consumption
The excellent anti-fouling and anti-interference properties of the constructed VMSF/p-SPCE sensor show its great potential for use in direct electroanalyses of CLZ in complex samples. As a proof-of-concept, electrochemical determination of CLZ in human whole blood was investigated. Human whole blood was applied as the detection medium after dilution by a factor of 100. For comparison, determinations using a bare SPCE or p-SPCE were also undertaken. As shown in Figure 7a, the DPV peak current on the VMSF/p-SPCE increased with an increase in the concentration of CLZ. Although the detection sensitivity was reduced compared with that of the buffer system, the VMSF/p-SPCE sensor could still linearly detect CLZ from 0.5 µM to 12 µM (I = 0.563 C + 0.04, R 2 = 0.996) with a LOD of 110 nM (Figure 7b). In contrast, the CLZ signal on both the bare SPCE and p-SPCE showed a non-linear relationship with its concentration, indicating that the electrode was seriously contaminated. The regeneration and reusability of the constructed VMSF/p-SPCE sensor was investigated. The performance of the p-SPCE was also examined for comparison. The sensor can easily be regenerated by soaking in a hydrochloric acid-ethanol solution (0.1 M) for 2 min. As shown in Figure 7c,d, both electrodes can be easily regenerated, although the CLZ signals on the regenerated electrodes were almost undetectable in the electrolyte. In addition, the VMSF/p-SPCE sensor can be reused with no significant change in its response to CLZ (Figure 7c). However, the peak currents for CLZ obtained on the p-SPCE were significantly reduced (Figure 7d). This phenomenon once again serves as evidence of the excellent anti-fouling performance of VMSF.
Molecules 2022, 27, x FOR PEER REVIEW 10 of 15 peak currents for CLZ obtained on the p-SPCE were significantly reduced (Figure 7d). This phenomenon once again serves as evidence of the excellent anti-fouling performance of VMSF. The corresponding calibration curve obtained on the VMSF/p-SPCE and DPV peak currents obtained on both the p-SPCE and bare SPCE. The regeneration and reuse performance of the VMSF/p-SPCE (c) and p-SPCE (d). The first peak current was obtained using the original electrode. Other peak currents were obtained in electrolyte (bottom) or CLZ solution (top) using the regenerated electrodes. The detected CLZ was 5 μM in diluted human blood (i.e., by a factor of 100).
The reliability of the device was further investigated by the standard addition method. Different concentrations of CLZ were artificially added to the human whole blood. Then, the whole blood was diluted by a factor of 10. As shown, the recovery of the concentration of CLZ ranged from 97.3% to 104%, with a relative standard deviation (RSD) of no more than 3.2% (Table 1), indicating good reliability for direct measurements of CLZ in complex samples. Since the SPCE features an integrated working electrode, a counter electrode and a reference electrode, a very small amount of sample (50 μL) can be dropped directly onto the surface of the electrode during tests. Therefore, the constructed VMSF/p-SPCE sensor can achieve rapid, direct and sensitive determinations of CLZ with very low sample consumption, indicating its great potential for use with fingertip blood.  The first peak current was obtained using the original electrode. Other peak currents were obtained in electrolyte (bottom) or CLZ solution (top) using the regenerated electrodes. The detected CLZ was 5 µM in diluted human blood (i.e., by a factor of 100).
The reliability of the device was further investigated by the standard addition method. Different concentrations of CLZ were artificially added to the human whole blood. Then, the whole blood was diluted by a factor of 10. As shown, the recovery of the concentration of CLZ ranged from 97.3% to 104%, with a relative standard deviation (RSD) of no more than 3.2% (Table 1), indicating good reliability for direct measurements of CLZ in complex samples. Since the SPCE features an integrated working electrode, a counter electrode and a reference electrode, a very small amount of sample (50 µL) can be dropped directly onto the surface of the electrode during tests. Therefore, the constructed VMSF/p-SPCE sensor can achieve rapid, direct and sensitive determinations of CLZ with very low sample consumption, indicating its great potential for use with fingertip blood.

Experiments and Instrumentations
The pore structure and vertically ordered nanochannels of the VMSF were characterized by transmission electron microscopy (TEM). Images were obtained on a transmission electron microscope (HT7700, Hitachi, Japan). The accelerating voltage was 100 kV. VMSF was gently scraped from p-SPCE, then dispersed in ethanol. Before characterization, the VMSF dispersion was dropped onto a copper grid. All electrochemical experiments, including cyclic voltammetry (CV), differential pulse voltammetry (DPV) and electrochemical impedance spectroscopy (EIS), were performed using an Autolab electrochemical workstation (PGSTAT302N, Metrohm, Herisau, Switzerland). The CV scan rate was 50 mV/s. During the DPV test, the step potential was 0.005 V, the pulse amplitude was 0.05 V, the pulse time was 0.05 s and the interval time was 0.2 s. EIS measurement was performed in K 3 Fe(CN) 6 /K 4 Fe(CN) 6 (2.5 mM) solution containing KCl (0.1 M) with a frequency range of 10 4 to 10 −1 Hz.

Electrochemical Polarization of SPCE
Electrochemical polishing was performed to remove impurities such as organics and polymers from the surface of SPCE electrodes. Briefly, H 2 SO 4 (0.05 M) was used as the medium and CV scans were then performed for 10 cycles between 0.4 and 1.0 V. After the electrode was rinsed with ultrapure water and dried with N 2 , it was placed in a phosphate buffer solution (PBS, 0.1 M, pH = 5). The subsequent electrochemical polarization included anodic oxidation and cathodic reduction. Firstly, the electrode was anodized by applying a constant potential of 1.8 V for 300 s. The oxidized SPCE was then cathodically reduced by CV scanning for three cycles (initial potential: 0 V; minimum potential: −1.3 V; maximum potential: 1.25 V; and scan rate: 100 mV/s). The obtained electrochemically polarized SPCE, that is, p-SPCE, was washed with ultrapure water and then dried with N 2 .

Preparation of VMSF-Modified p-SPCE
VMSF was electrodeposited on the surface of p-SPCE by the EASA method. To prepare the precursor solution for VMSF deposition, ethanol (20 mL) was mixed with an equal volume of NaNO 3 (0.1 M, pH = 2.6). Then, CTAB (1.585 g) was added under magnetic stirring until the solution became clear. After TEOS (3050 µL) was added, the obtained solution was stirred at room temperature for 2.5 h. To grow VMSF, a p-SPCE electrode was placed in the precursor solution and a galvanostatic current (−52.2 µA) was applied for 10 s. As surfactant micelles (SM) were present within VMSF nanochannels, the obtained electrode was denoted as SM@VMSF/p-SPCE. After thorough rinsing with ultrapure water, the SM@VMSF/p-SPCE was aged at 80 • C for 10 h. Then, SM was removed by immersing the SM@VMSF/p-SPCE in 0.1 M HCl-EtOH solution (v:v = 1:1) for 5 min with stirring. An electrode with open nanochannels was finally obtained and named VMSF/p-SPCE.

Electrochemical Determination of CLZ
PBS (0.1 M, pH = 6) was used as the detection electrolyte. To investigate the effect of a complex medium on detection, human whole blood was also applied as the detection medium after diluting by a factor of 100 using PBS electrolyte. Electrochemical determination of clozapine was carried out by DPV. After adding different concentrations of CLZ, a DPV curve was recorded. For the real sample analysis, different concentrations of CLZ were added to the human whole blood to simulate changes in CLZ levels in patient blood after taking the drug. Then, human whole blood with added CLZ was diluted by a factor of 10 using PBS electrolyte. Next, 50 µL of the obtained solution was dropped on the surface of VMSF/p-SPCE electrode. After static enrichment for 1 min, a DPV curve was recorded.

Conclusions
In summary, using simple equipment, a miniaturized, integrated and disposable electrochemical sensing platform was constructed comprising a vertically ordered mesoporous silica nanomembrane film (VMSF) on electrochemically polarized SPCE (p-SPCE). The p-SPCE was characterized by increased active area and improved electron transfer rate, thereby improving the determination sensitivity of CLZ. The nanochannels of the VMSF also achieved electrostatic enrichment toward CLZ. Due to the excellent size/charge exclusion effect of the nanochannels and the high potential resolution ability of p-SPCE, the constructed VMSF/p-SPCE sensor had excellent anti-interference and anti-fouling properties, allowing it to realize rapid, direct and sensitive determinations of CLZ in human whole blood. In addition, the developed VMSF/p-SPCE sensor has the advantage of very low sample consumption (50 µL), indicating great potential for use with fingertip blood. The nanochannel array on SPCE established here also provides a new strategy for rapid, direct and sensitive electroanalyses of complex samples.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.