Recent Advances in Benzodiazepine Electroanalysis

Abstract

Benzodiazepines are psychoactive drugs with wide clinical applications. Unfortunately, due to their sedative effects, benzodiazepines are frequently used as date rape drugs or in drug-facilitated crimes. Considering the electroactive nature of benzodiazepines and the unique advantages of electrochemical techniques, this review presents a critical discussion of the state of the art of benzodiazepine electroanalysis. Aspects related to sample preparation as well as electrodes (from mercury electrodes to bare or modified solid electrodes and to disposable sensors) and techniques (mainly voltammetry) used for the quantification of benzodiazepines in different matrices (pharmaceuticals, body fluids, alcoholic and soft drinks) were discussed. Considering the actual achievements in the field, some general suggestions for possible further research were given.

1. Introduction

BZDs are psychoactive agents that act as CNS depressants. Their rapid entry into the brain, enabled by their high lipid solubility, allows them to enhance the activity of GABA, the brain’s primary inhibitory neurotransmitter. BZDs quickly won favor in clinical practice because they reduce anxiety within minutes and at the same time act as anticonvulsants, hypnotics, skeletal–muscle relaxants, and general sedatives. In both generalized anxiety disorder and panic disorder, BZDs produce outcomes that match or surpass other drug classes while causing fewer side-effects [1] and outperform tricyclic antidepressants in preventing panic attacks [2].

BZDs are widely used for acute symptom relief across a range of conditions: insomnia, anxiety disorders, seizures, spasticity, and muscle tension. Primary indications are: anxiolysis (ALP, CDZO, DZP, LZP, OZP with an elimination half-time ranging between 6 and 100 h) and hypnosis (FNZP, NZP, TMP, with an elimination half-time ranging between 2 and 100 h) [3].

Because BZDs amplify GABA A-receptor signaling, it has been proposed that they might partially modify the expression of AD. Brain tissue analyses and spectroscopy studies consistently show that GABA levels fall in AD. However, while BZDs theoretically compensate for the GABA shortfall in AD, their long-term impact on memory and neuropathology has not been settled: evidence points simultaneously to potential benefits, neutral effects, and harm, depending on the compound, dose, duration, and clinical context [4].

Pharmacologically, BZDs have different actions, where, in addition to anxiolytic effects, various analogs have demonstrated anticancer, anticonvulsant (CLZP), anesthetic (MDZ), antipsychotic (OLP, clozapine), muscle-relaxing, antitubercular, and antimicrobial properties.

However, BZDs are frequently implicated in polydrug abuse, especially among users of opioids (oxycodone and heroin). Studies report that over half of those abusing oxycodone also misuse BZDs (most commonly DZP). This combination increases the risk of overdose due to compounded respiratory and CNS depression [5]. BZDs’ use has expanded among individuals, especially among those with substance dependence, and has become increasingly common in drug-facilitated assaults and road-traffic accidents. Enhancing analytical methods for detecting these drugs in biological samples is crucial in toxicology, as it enables the detection of exposure or intoxication [6].

A 34% surge in anxiolytic prescriptions has been seen concurrently with COVID-19-related stress and with the wider mental-health awareness movement. In polysubstance abuse, BZDs are commonly combined with opioids or alcohol to enhance euphoria or ease comedowns, making them the second-most common prescription drug class implicated in fatal overdoses, after opioids [7]. Survey data shows stark age gradients: about half (~51%) of individuals aged 18–25 admit to non-medical BZD use, whereas misuse falls to ~4% in those individuals aged ≥ 65 [8].

Every benzodiazepine has a 1,4-benzodiazepine core, with sidechain substitutions generating several subfamilies (2-keto, 3-hydroxy, 7-nitro, triazolo- and imidazo- derivatives) (Figure 1). These drugs bind to the GABA A receptor, a ligand-gated chloride channel, which is allosterically modulated. The neurotransmitter GABA allows Cl⁻ to enter the cell, hyperpolarizing neurons and decreasing excitability. BZDs boost this effect allosterically, yielding an even larger chloride current [1].

The relationship between chemical structure and biological activity is illustrated in

Table 1. Chemical structure-biological activity relationship [9].

Structure:	A benzene ring A fused to a seven-membered diazepine ring B.
	In most compounds, this scaffold is complemented by a third substituent, an aryl ring C.
	All three ring systems are essential for the pharmacological activity of BZDs.
Benzene ring A:	Introducing a strong electron-withdrawing group (for example, -Cl or -NO2) to position 7 yields optimal activity.
	Attaching groups to other positions on the same ring diminishes potency.
Diazepine ring B:	Indispensable for proper receptor binding through the amide or N-alkyl groups.
	Attaching a methyl group to position 1 augments biological activity.
	Voluminous groups attached to position 1, such as tert-butyl, decrease activity.
Aryl group C:	Supports receptor binding through hydrophobic and steric interactions.
	Electron-attracting groups in position 2′ enhance activity.
	Groups attached to position 4′ reduce biological activity.

[9].

BZDs can be classified based on their chemical structure in 2-keto BZDs (for example: CDZO, DZP), 3-hydroxi BZDs (for example: LZP, OZP, TMP), 7-nitro BZDs (for example: CLZP, FNZP, NZP), triazolo- BZDs (for example: ALP), imidazo- BZDs (for example: MDZ, loprazolam) [1].

According to their half-life, they can be divided into three groups: short (MDZ, OZP), intermediate (ALP, BZP, LZP, TMP), and long (CDZO, CLZP, DZP) [1], with half-lives ranging from minutes or hours (short) to days (long). Additionally, certain BDZs form active metabolites, such as DZP and desmethyldiazepam, whereas others, like oxazepam, do not [1].

In recent years, “designer” analogs have appeared and are sold online with minimal information provided as to their toxicology. One example is flubromazolam (a triazolo-benzodiazepine: 8-bromo-6-(2-fluorophenyl)-1-methyl-4H-triazolo[4,3-a][1,4]benzodiazepine) prized by recreational users for deep hypnosis, prolonged anterograde amnesia, and a steep tolerance curve. Deschloroetizolam (a thienodiazepine) is another example of a “designer” analog, which augments GABA A signaling, like its classical relatives. A third example, clonazolam, the triazolo derivative of clonazepam, is suggested to have twice the potency as alprazolam. Meanwhile, meclonazepam is structurally similar to clonazepam but presents additional anti-schistosomal activity and sedation, keeping it off the pharmaceutical market [7].

Oral preparations are generally well-absorbed, while the intravenous route is rapid but may irritate local tissues. Intramuscular administration is preferred for DZP. High lipid solubility confers large volumes of distribution. Absorption rates vary widely, with DZP, for example, reaching peak plasma levels rapidly. Lipid solubility governs blood–brain-barrier penetration and onset of action [1].

BZD therapy can occasionally trigger paradoxical behavioral effects. Patients may become irritable, uninhibited, unusually outspoken or inappropriate, and may even display anger or aggression. Regardless of class, the dose-limiting adverse effects remain confusion, daytime somnolence, anterograde amnesia, and ataxia [10].

Although therapeutic BZDs carry low intrinsic toxicity, they are not negligible. Typical acute overdose produces somnolence, slurred speech, ataxia, and altered consciousness, yet pupils remain reactive and vital signs are usually stable [7]. When outright abuse is documented, it almost invariably occurs in individuals already using other depressant substances, most commonly being alcohol, opioids, or additional sedative-hypnotics, rather than abusing only BZDs. Nonetheless, evidence does not place them among the main “gateway” or standalone drugs of abuse [11].

Considering all the above-mentioned aspects, it is clear that BZDs’ determination is essential for pharmaceutical, clinical, and forensic applications, such as purity checks, counterfeit detection, biological fluid drug level monitoring for adequate medical treatments, or to detect their presence in incidents like drug-facilitated sexual assaults or traffic accidents. There are reviews that briefly summarize the methodologies reported in the literature regarding sample preparation methods [12] and BZDs (bio)analysis in biological matrices with emphases on toxicological or clinical purposes [13,14,15,16,17]. Due to BZDs’ increasing use as date-rape drugs, especially by spiking different types of beverages, their detection and quantification in matrices other than biological ones is also important. In this respect, chromatographic techniques have usually been employed [18,19,20,21], with some fluorometric methods [22]. More recently, a portable, microfluidic device based on a colorimetric reaction was developed for the rapid naked-eye detection of DZP and MDZ in drinks [23]. Electrochemical methods were also reported for BZDs analysis in beverages [24], presenting the advantage that they allow determinations in colored or turbid solutions.

Another aspect of the widespread use of BZDs is permeation into the environment, mainly through wastewater, due to their slow degradation. BZD concentrations from 0.01 μg/L to several μg/L have been found in different types of natural waters [25], a factor that can affect aquatic life, as BZDs can be harmful to living species. This pharmaceutical class can be considered CECs. It was reported that among BZDs, lorazepam had the highest concentration (89.9 μg/L) in aquatic environments [26]. While chromatography is the technique of choice for the analysis of complex mixtures of structurally related species, for example, using the UHPLC-MS/MS method for the simultaneous determination of 14 BZDs residues in aquaculture water and sediments [27], voltammetry has also been applied for the quantification of various CECs, with BZDs also among them. Compared to LC-MS or LC-MS/MS, voltammetry has lower material consumption and reagents, and reduced analysis time, mostly without the need for sample pretreatment, using cheaper instruments and less qualified personnel, but is less sensitive and has lower selectivity [28].

However, there is an increasing need to develop portable, cost-effective, rapid, sensitive, and selective BZD detection devices for various samples (Figure 2) [28]. Electrochemical sensors are often the right choice in this respect because they have rapid response times, are usually quite easy to prepare, have various preparation methods, and can be miniaturized. Additionally, they can have various designs [29] and can be coupled with different techniques to obtain low detection limits and linear ranges wide enough to be applied for the accurate analysis of multiple types of samples.

It has already been established that electrochemical techniques can be important tools for the successful assessment of drugs, for both their quantification and the investigation of their redox properties, which can help in understanding some of their biological actions. This paper aimed to review and discuss the electrochemical sensors and methods developed for the analysis of BZDs and reported in the literature over the last 10 years. A similar review was published in 2019, discussing the evolution of BZD electroanalysis from the early 1960s [30]. However, there has been recent significant progress, especially regarding modified and disposable electrodes. More recent reviews dealing with biosensing of illicit drugs [31,32], emphasizing point-of-care testing [33] or beverage screening [24] include BZDs electroanalysis, among other analysis techniques. The originality of this review paper lies in its structure, with the discussion conducted for each analyte separately, to facilitate an almost index-like approach. Since in an analytical process the sample preparation step is as important, if not more important, than the actual analysis stage, this review focuses on the sample preparation techniques employed based on sample type, and on presenting the analysis methods.

2. Benzodiazepines Analysis

Most BDZs, except for MDZ, have poor solubility at physiological pH, but can be solubilized in acidic environments. However, without the use of organic co-solvents or surfactants, most BDZs were shown to form unstable protonated forms [34]; thus, depending on the sample type and target calibration, pH can play an important role in stock and sample preparation protocols.

There are three typical analytical starting positions governing the analysis of 1,4-benzodiazepines: physiological fluids (blood or serum, urine, saliva, or other), pharmaceutical formulations (tablets, drops, or injections), or beverages (alcoholic or soft drinks). These, in turn, align with the purposes of pharmacokinetics, quality control, or forensic investigations. A fourth, less commonly explored analytical path is that of analyzing surface water and sewage; this is mainly performed for environmental purposes or community drug-use monitoring. Often, BZDs concentrations in natural waters (especially during rainy periods) are below the LoDs of even stripping voltammetric methods, so that pre-concentration techniques should be applied before the actual analysis [25].

BZDs are electroactive compounds due to the existence in their molecular structure of the 4,5-azomethine group, which undergoes a reduction process, generating a cathodic signal (Scheme 1).

Depending on the presence of additional functional moieties, supplementary cathodic and/or anodic peaks can be observed in the voltammograms of the corresponding analytes. Voltammetric (CV and SWV) investigations of DZP, CLZP, and ALP at graphite SPE in standard and deoxygenated solutions coupled with LC-HRMS analysis of their degradation products allowed the association of characteristic voltammetric peaks of each analyte with the corresponding redox active group. Thus, based on the different electrochemical behavior, BZDs were classified into three groups, and the approach was then applied to classify illicit pills with unknown BZD content [35].

Voltammetric techniques are important tools not only for quantitative analysis, but also for the investigation of interactions between drugs and other bioactive molecules. CV and SWV studies indicated that DZP and CDZO formed an electro-inactive complex with bovine serum albumin and allowed the assessment of the molar ratio (m) and the binding constants (βs) of the formed biocomplexes, which were 1.23 and 2.00 × 10⁶ L/mol for the DZP-complex and 1.04 and 7.76 × 10⁴ L/mol for the CDZO-complex [36].

BZD electroanalysis has been performed with conventional unmodified or modified electrodes (GCE, SPE, CPE), as well as with newly developed designs, like PADs or wearable sensors, most of them being used in voltammetric detection. Amperometry, EIS, and potentiometry were also applied.

2.1. Alprazolam (ALP)

For biological samples, ALP has been studied in neat, spiked urine [37]. Given the high likelihood that illicit ALP tablets contain other controlled substances [38], tablets have been prepared in small amounts of alcohol, filtered and diluted with (pH 10.00), and analyzed against other potential BZDs (DZP, CLZP, and the like), opioids (fentanyl, U-4770), cannabinoids (5F-ADB), stimulants (cocaine, amphetamine, 3,4-MDMA), anesthetics (ketamine, lidocaine) with the main intent of discriminating between Xanax (commercial name for ALP) and fake Xanax tablets. For beverages [38,39,40,41], samples have been either spiked and then 25 to 100-fold diluted with the working buffer [39], diluted with water (1:1, v/v) and then spiked [40], or diluted tenfold with buffer, then spiked [41]. Studying its potential as a surface water contaminant, samples were filtered and directly analyzed [28].

Cyclic voltammetric investigation of ALP at HMDE indicated a cathodic peak at about −0.95 V (PBS pH 7.00) attributed to the irreversible, adsorption-controlled reduction of the azomethine group involving 2e⁻ and 2H⁺ according to the reaction presented in Scheme 1(a) [28]. Similar behavior was reported on BDDE [42] and CPE [43], but in this case, the ALP reduction process was diffusion-controlled (Scheme 1(b)). Among the solid electrodes, the most widely used is the GCE. At bare GCE, ALP did not show any voltammetric signal, but after the electrochemical pretreatment of the electrode (Figure 3), a well-defined cathodic peak was observed (Figure 4). This difference between ALP’s electrochemical behavior at the two investigated working electrodes was attributed to a better adsorption of the analyte on the EPGCE due to its increased electroactive surface area and the presence of oxygen-containing groups that could interact with the analyte through hydrogen bonds (Figure 4). EPGCE was applied to ALP determination from six non-alcoholic and alcoholic spiked drinks without interference from substances such as ascorbic acid, citric acid, sucrose, and acetaminophen, which are commonly contained in beverages used in sexual assault cases. Applying the standard addition method, the recoveries obtained were between 82.0 ± 0.2 and 109.0 ± 0.3% [39].

Electrode surface modification with different types of (nano)materials or combinations of them usually results in more sensitive and selective determinations, due to improved conductivity, larger electroactive surface area, and electrocatalytic properties of the modifier(s). For example, a gold nanoparticle decorated reduced graphene oxide nanocomposite, obtained in a single-step laser-assisted procedure, was drop-casted on a GCE, generating a sensor with enhanced electrochemical performances towards ALP detection. In this case, the electrode surface modification resulted in a decrease in the charge transfer resistance and an increase in the electroactive surface from 3884 kΩ and 0.080 cm², for GCE, to 20.6 kΩ and 0.126 cm², for the LI-AuNP-rGO/GCE. The situation of adding two 1 mg ALP tablets to a 300 mL beverage was simulated. After a ten-fold dilution of the spiked drink, the analyte concentration was within the linear range of the voltammetric method developed using this sensor, and ALP could be reliably detected without interference from other substances found in spiked drinks used in these types of scenarios [41]. The steps followed for the fabrication of AuNUs-Fe-Ni@GO/GCE developed for ALP determination are illustrated in Figure 5. The nanomaterials increased the surface area and the conductivity of the electrode, while electrochemical pretreatment enhanced the number of oxygen-containing groups, which favored ALP adsorption at the electroactive surface. This biosensor presented high selectivity in the presence of compounds normally found in blood (ascorbic acid, citric acid, and glucose) and a good long-term stability, preserving 92.2% of the original signal value after two weeks of storage in a refrigerator [44].

It is worth mentioning that the above-described sensors enabled ALP analysis in the presence of other BZDs, like CLZP, DZP [28,39,40], LZP [41], DZP, CDZO, and OZP [44].

An interesting protocol for the detection of counterfeit Xanax tablets was by coupling electrochemical measurements with data analysis [38]. Using an optimized pH of 10, the voltammetric behavior of ALP and other BZDs, as well as that of common adulterants, like 5F-ADB, fentanyl, acetaminophen, ketamine, amphetamine, lidocaine, caffeine, 3,4-MDMA, cocaine, U-47700, were investigated and the corresponding oxidation and reduction peaks were attributed to different functional moieties belonging to the investigated species (e.g., reduction of thiophene ring, of fluoride atom, oxidation of secondary and tertiary amines). Based on the correlation between peak potential and functional groups, the authors developed a flow-chart system for the on-site analysis of falsified Xanax tablets.

PADs represent a good alternative for drug analysis, due to advantages, such as low cost and ease of preparation. Improved performance characteristics for ALP detection were obtained using an electrochemical microfluidic PAD having a methylene blue-doped Ag@Pd nanohybrid modified working electrode, due to the synergistic effects of the nanohybrids, which generated an increased electroactive surface area, and the redox mediator, which allowed an increased rate of electron transfer [37]. Another electrochemical microfluidic PAD, employing three working electrodes modified with AgNPs, each with an aptamer specific for ketamine, methamphetamine, or ALP, enabled the on-site, rapid (30 s), simultaneous quantification of these illicit drugs. However, the authors mention that due to the large dimensions of the platform-transducer system, it has the disadvantage of not being portable, its miniaturization being necessary [40].

The performance characteristics and the analytical applications of all these methods developed for ALP determination are summarized in

Table 2. Performance characteristics of electrochemical methods reported in the literature for ALP quantification.

Electrode	Technique/Conditions/Ep (V)	Linear Range (mol/L)	LoD (mol/L)	Sample	Ref.
HMDE	DPAdCSV/Eacc −0.80 V; tacc 120 s; PBS pH 7.00/~−0.95	1.62 × 10−8–6.48 × 10−8	2.27 × 10−10	River water	[28]
MB-Ag@Pd/EμPAD	CV/0.10 V, 35 °C and 10 s of incubation time; PBS pH 7.00/~−0.50	3.51 × 10−9–1.05 × 10−6	8.78 × 10−11	Urine	[37]
CNFs/SPE	SWV/B-R buffer pH 10.00/~−1.30	1.00 × 10−5–3.00 × 10−4	2.00 × 10−7	Xanax tablets	[38]
EPGCE	DPAdCSV/Eacc −0.10 V; tacc 480 s; B-R buffer pH 9.00/~−1.10	3.24 × 10−7–1.29 × 10−5 1.29 × 10−5–6.48 × 10−5	9.72 × 10−9	Beverages	[39]
Apt-AgNPs-EPMAD	CV/5.00 × 10−4 mol/L [Fe(CN)6]3−/4− in 0.10 mol/L PBS/~0.40	3.24 × 10−8–1.62 × 10−5	3.24 × 10−8	Beverages	[40]
LI-AuNP-rGO/GCE	DPAdCSV/Eacc −0.10 V; tacc 200 s; B-R buffer pH 9.00/~−0.95	3.24 × 10−9–3.24 × 10−7 3.24 × 10−7–2.59 × 10−5	9.81 × 10−10	Beverages	[41]
AuNUs-Fe-Ni@GO/GCE	DPV/B-R buffer pH 9.00/~−1.05	1.30 × 10−8–1.62 × 10−6 3.24 × 10−6–1.62 × 10−4	3.24 × 10−9	Diluted blood serum	[44]

. It can be observed that the LODs obtained for BDDE, CPE, and CNFs/SPE are not as low as those obtained with Hg and modified electrodes, but they are satisfactory for the analysis of pharmaceutical samples [38,42]. Although stripping voltammetry at HMDE obtained very low detection limits, this is not sufficient for BZDs detection in natural waters, their concentration in this situation being lower. Moreover, humic acid interfered with the determination [28].

2.2. Bromazepam (BZP)

BZP was determined from tap water and wastewater effluents from a pharmaceutical industrial plant, after pH adjustment to 4.00 and filtration, if necessary [45].

CV studies of BZP at BDDE emphasized that, only in the direct scan, one peak due to the 2e⁻ and 1H⁺ diffusion-controlled, irreversible reduction of the azomethine group, most probably with the generation of a radical anion [42]. At CPE, the BZP reduction process proceeded according to Scheme 1(a) [43].

Two membrane ISEs containing either BZP-PTA or BZP-TPB as ion-pair complexes were reported for rapid (10 s–20 s) BZP potentiometric determination. The electrodes presented Nernstian responses of 54 ± 0.321 mV/decade (BZP-PTA) and 57 ± 0.223 mV/decade (BZP-TPB), a linear concentration range of 1.00 × 10⁻⁶ mol/L–1.00 × 10⁻³ mol/L, and a LoD of 7.94 × 10⁻⁷ mol/L BZP [45].

2.3. Clonazepam (CLZP)

Another widely analyzed BZD, CLZP, has been systematically studied along the three main paths. For biological samples, serum has either been simply diluted with PBS pH 7.00 [46,47], then spiked and analyzed, or processed via ethyl acetate LLE, which was then dried and reconstituted with electrolyte solution [48]. Urine was diluted with PBS [46], then spiked and analyzed, or centrifuged, filtered, then diluted with PBS [47], and analyzed via standard addition. Human blood serum [49] has been prepared by methanol precipitation, centrifugation, and filtration, with the supernatant being used for further analyses. Saliva [50] has been analyzed as-is, and human plasma [51] has been prepared by spiking, then processing via organic solvent deproteinization with acetonitrile, centrifugation, and analysis.

The standard protocol of weighing a number of tablets to determine their average mass, homogenizing all to a powder and then sampling a certain amount has been used [47,49] with dissolving either in a 2:8 methanol/water (v/v) mixture, followed by dilution with PBS pH 7.00 [47], or dissolving with just methanol, filtration, and dilution with PBS pH 7.20 [49].

Alcoholic [48] and non-alcoholic [48,52] drinks have been prepared by spiking, then diluting to a certain phosphate concentration [48] or analyzed as-is [52].

Due to the existence of a nitro group in its molecular structure, CLZP presented in the first forward cathodic scan only one reduction peak (R1) at about –0.6 V to –0.8 V, which was attributed to the irreversible, 4e⁻ and 4H⁺ reduction of the –NO₂ moiety to hydroxylamine. In the reversed anodic scan, an oxidation signal appeared (O2), which formed a redox peak pair with the new reduction peak (R2) recorded during the second scan in the negative direction. Peaks (O2) and (R2) (situated at less negative potentials) were associated with the oxidation of the –NHOH group to –NO and the reverse reduction of –NO to –NHOH, respectively, in processes involving 2e⁻ and 2H⁺ (Scheme 2) [47,49,53].

Some investigations explore more negative potentials, recording the cathodic signal (R3) generated by reduction of the 4,5-azomethine group (Scheme 1) [48,50]. Moreover, one paper also indicated a cathodic signal, at about 0.3 V, generated by the reduction of the >C=O group (Scheme 3). However, for quantitative purposes, the -NO₂ reduction signal was exploited the most [50].

With mercury electrodes in a supporting electrolyte of pH 7.00, CLZP presented two irreversible, adsorption-controlled cathodic peaks associated with the reduction of the nitro group (E_p~−0.40 V) and of the azomethine group (E_p~−1.05V), respectively [25]. Using Cotrell’s relation and chronoamperometric measurements data, CLZP diffusion coefficient in PBS pH 7.00 was calculated to be 1.12 × 10⁻⁶ cm²/s [51] or 1.14 × 10⁻⁶ cm²/s [52].

Owing to their good electrical conductivity, catalytic properties, and large surface area, as-is metal or carbon-based (nano)materials or in composites were widely used to change the electrochemical surface properties of various sensors [49]. For example, the GCE modification with silver fibers and ionic liquid (1-butyl-3-methylimidazolium tetrafluoroborate) nanocomposite resulted in a 3.5-fold amplification of CLZP reduction peak and its shift by about 100 mV towards less negative potentials, in comparison to the bare GCE. The improved electrochemical properties of the modified electrode were attributed both to the Ag nanofibers, which had electrocatalytic properties and increased the electroactive surface area, and to a better electron transfer due to the ionic liquids, which enhanced the conductivity of the sensor [46]. On the other hand, the CLZP reduction peak was 3.8 times higher at the C₃N₄/CNH-MSN/GCE, compared to the bare electrode, due to doubling of the surface area, the high N content of the nanocomposite, which ensures a low charge-transfer resistance, and the intrinsic catalytic activity of copper. The electrode showed a stable response even after being stored for 2 weeks in ambient conditions, without being used. The results obtained for CLZP voltammetric quantification in real samples were in good agreement with those obtained by HPLC [47].

Other interesting inorganic materials for sensor construction are LDHs and MXenes, which have common features, like high surface area, electrocatalytic activity, and non-toxicity. A FeCu-LDH@MXene nanocomposite obtained by a hydrothermal synthesis involving a co-precipitation step was applied to modify a GCE. Chronocoulometric measurements emphasized that the surface area of the resulting electrode (0.267 cm²) was almost ten-fold larger than that of the bare electrode (0.029 cm²). The enhanced electroactive surface area and the synergistic electrocatalytic effects of FeCu-LDH and MXene resulted in amplified CLZP reduction signals and, consequently, a sensitive detection. Moreover, the electrode presented a good stability for at least one month [51].

pMela/MWCNT-COOH/GCE was obtained by casting onto the GCE surface and drying of a MWCNT-COOH suspension, and subsequent melamine electropolymerization by cyclic voltammetry. At the modified electrode, the CLZP main reduction peak was shifted by 140 mV in the positive direction, and its intensity increased 14 times compared to the unmodified electrode. These effects were assigned to the enhanced adsorption of the analyte on the polymer film, the catalytic activity of the MWCNT-COOH, and the higher effective surface area. Adsorptive stripping analysis of CLZP at this electrode involved the adsorption and subsequent reduction of the analyte, followed by the oxidation of the reduction product and recording of the voltammogram during the anodic stripping step. The electrode surface was cleaned between measurements by successively cycling the applied potential in the supporting electrolyte. The response of the electrode stored at room conditions decreased by only 8% after 15 days, suggesting good stability [53].

An alternative to the classical solid electrodes is represented by disposable electrodes, among them the best-known being SPEs with various types of working electrodes, which were also applied to CLZP quantification either as such [48] or modified [50,52]. The mammalian glycoprotein lubricin was used to modify the rGO surfaces of SPEs, due to its ability to act as a size-selective molecular sieve, separating the large (bio)molecules from the smaller ones (analyte) and preventing thus the electrode surface fouling. Moreover, lubricin also enabled the CLZP detection in unprocessed saliva in the presence of other similar compounds, like DZP and MDZ, suggesting that this device could be applied in POC testing [50].

The performance characteristics and the analytical applications of all these methods developed for CLZP determination are summarized in

Table 3. Performance characteristics of electrochemical methods reported in the literature for CLZP quantification.

Electrode	Technique/Conditions/Ep (V)	Linear Range (mol/L)	LoD (mol/L)	Sample	Ref.
HMDE	DPAdSV/Eacc −0.30 V; tacc 120 s; PBS pH 8.00/−0.40	2.06 × 10−9–4.78 × 10−8	6.02 × 10−10	Natural water	[25]
AgFs-IL/GCE	DPV/PBS pH 7.00/~−0.53	1.00 × 10−7–2.50 × 10−4	6.60 × 10−8	Pharmaceuticals, urine, serum	[46]
C3N4/CNH-MSN/GCE	DPV/PBS pH 7.00/~−0.72	8.20 × 10−7–7.69 × 10−5	2.50 × 10−9	TPharmaceuticals, urine, human serum	[47]
SPE	DPAdSV/Eacc −1.50 V; tacc 60 s; PBS pH 11.00/Oxidation~−0.50	6.50 × 10−6–2.53 × 10−5	6.50 × 10−6	Fortified wine, serum	[48]
Fe3O4/R-SH/Pd/GCE	DPV/PBS pH 7.20/~−0.72	1.00 × 10−8–1.00 × 10−6	3.02 × 10−9	Pharmaceuticals, human serum	[49]
LUB-rGO/SPE	SWV/PBS pH 7.4/~−0.70	2.50 × 10−8–2.50 × 10−7	1.67 × 10−9	Saliva	[50]
FeCu-LDH@MXene/GCE	DPV/PBS pH 7.00/~−0.62	6.60 × 10−7–4.18 × 10−4	9.00 × 10−8	Pharmaceuticals, human plasma	[51]
CoOOH-rGO/SPCE	DPV/PBS pH 7.40/0.11	1.00 × 10−7–3.50 × 10−4	3.80 × 10−8	Alcoholic and non-alcoholic beverages	[52]
pMela/MWCNT-COOH/GCE	SWAdASV; Eacc −0.63 V; tacc 100 s; B-R buffer pH 7.00/Oxidation ~−0.08	5.00 × 10−8–1.00 × 10−5	6.30 × 10−8	Serum	[53]

.

2.4. Diazepam (DZP)

By far, one of the most investigated BZDs is DZP, a 1,4-benzodiazepine with an azomethine group with a pK_a of 3.4 [54]. Blood, serum, or urine samples were collected from volunteers after diazepam oral administration [54,55], or from non-user volunteers [56,57,58,59,60] with sample spiking post-collection. In either case, prior to analysis, the biological matrices have been deproteinized by use of organic solvent [54], diluted with the buffer solution [55,56,57], or used without pretreatment post-spiking [56,58,59,60].

Pharmaceutical formulations are usually tablets but may also be found as injection solutions or drops. For tablets, the most common sample preparation [54,55,56,57,61,62,63] is grinding one or more tablets, after determining a tablet’s average mass, weighing a certain amount of that combined powder, usually the average weight of one tablet, and dissolving it to a certain concentration, then potentially filtering and volumetrically achieving the desired target concentration depending on the tablet’s active substance content. Various solvents have been used for the dissolution of diazepam, or for final volume diluents, such as methanol:water (3:2, v/v) [54] or (1:4, v/v) [55], water [56], buffer [57], the target beverage [61], alcohol [62,63], buffer (phosphate, pH 7.00 [54,56] or pH 4.00 [57]; acetate, pH 4.80 [55]), or acidic aqueous solutions (0.01 mol/L HCl [62]). For liquid and drop pharmaceutical formulations [63], the choice solvent was alcohol, followed by a diluent consisting of alcohol:electrolyte.

A special application encompassing testing of solid residues was the use of agarose hydrogels with 0.10 mol/L PBS [64], with direct sampling of the specimens.

Beverage analysis entailed the use of both alcoholic and soft drinks, or a mixture of those, to mimic bar settings where illicit drink spiking may occur. For analysis, samples were either diluted with buffer and water, sonicated, then spiked [61], spiked with a crushed diazepam tablet [63], followed by dilution with alcohol:electrolyte for analysis using the agarose hydrogels [64] dipped into the spiked and non-spiked drinks, or by spiking the drinks and analyzing without other treatment steps [58].

A thin three-electrode flow-cell was applied to the CV investigation of DZP reduction at GCE and its subsequent quantification by FI_DPA [61]. DZP irreversible reduction at various tested electrodes was diffusion-controlled [56,58] and pH-dependent, involving an equal number of electrons and protons [54,55,65]. An adsorption-controlled reduction of DZP was reported at SPCE [63]. Most papers stated a 2e⁻ and 2H⁺ reduction of the 4,5-azomethine group from the benzodiazepine ring (Figure 1, Scheme 1) [54,63,65,66]. However, there are also some exceptions. At the BiPPGE, the cathodic process involved one electron, and the transfer coefficient, α, was found to be 0.34. The diffusion coefficient, estimated from chronoamperometric data, was 6.82 × 10⁻⁵ cm²/s [55]. In another study [54], a much lower diffusion coefficient of 3.06 × 10⁻⁶ cm²/s was obtained from CV data. It was reported that in acidic medium, DZP reduction at MWCNT/GC/CPE DZP proceeds via 1e⁻ and 1H⁺ transfer. The proposed mechanism involved two chemical steps during which the benzodiazepine ring is cleaved, followed by the irreversible reduction, which finally leads to a pinacol derivative as reaction product [58].

Some literature reports indicated that DZP’s highest reduction signal was obtained at pH below its pK_a, which is about 3.0, where the analyte is protonated at the N⁴ atom of the azomethine group. However, most analyses were performed at pH near the physiological value [54,56,59,64,65].

Mercury electrodes were the first sensors used for BZDs analysis, exploiting their azomethine group reduction peak. DPAdSV at the HMDE method for DZP quantification presented limited selectivity because other BZDs could interfere with the determination, CLZP being an example. However, quantifying CLZP based on its -NO₂ reduction signal (situated at less negative potentials) allowed DZP determinations by subtracting the CLZP quantity from the common peak result [25].

GCE surface modification with various (nano)composites (e.g., by drop-casting C₆₀–CNT/IL [56], graphene nano-sheets, and subsequent potentiostatic electrodeposition of Ag nanodendrites [54], or by drop-casting the graphene oxide and lysine on the potentiostatic pretreated GCE [65]) resulted in a lower peak potential and an amplified reduction signal of DZP due to the synergistic effects of the components [54,56]. The obtained modified electrodes presented good stabilities for two weeks [54] or more [56], but unfortunately, other BZDS could interfere in DZP determination [54]. Due to DZP’s electro-reduction being pH-dependent, and the tested beverages having various pH values, it was observed that in alcoholic drinks, with pH varying from 5 to 8.5, the analytical signal was shifted. Therefore, preliminary beverage pH measurements were necessary [65].

SPEs bare [63] or modified electrochemically ex situ with an Sb film [62] or by drop-casting SiO₂ encapsulated QDs on the potentiodynamic pretreated electrode and subsequent covering with an electrogenerated MIP [60] were also reported for DZP analysis. When using SPE [63] for the analysis of beverages containing Coca Cola, the sensitivity, linear concentration range, and LoD previously established in the supporting electrolyte were altered. Therefore, for DZP determination in these kinds of samples, a matrix-matched calibration curve was necessary. Even though SPEs are disposable electrodes, it was reported that on such an electrode, Sb could be deposited three times, with an intermediate potentiostatic cleaning step before each deposition. Moreover, an Sb/SPE could be used for up to 20 measurements [62]. The MIP@SiO₂-MPA-AuZnCeSeS-QDs/SPCE was applied to DZP quantification either by EIS or by SWV, using the redox probe K₃[Fe(CN)₆]/K₄[Fe(CN)₆]. The EIS analysis was based on the linear increase of the charge transfer resistance with DZP concentration, while the SWV determination relied on the linear decrease in the anodic signal of the redox probe with increasing analyte concentrations. EIS measurements were carried out using both a portable potentiostat connected to a laptop and a hand-held potentiostat connected to a smartphone. The authors stated that the portable potentiostat was about seven times more rapid but four times more expensive than the hand-held smartphone-based one. However, the use of SWV allowed obtaining a larger linearity range and a lower LoD [60].

CPEs modified with MIP grafted on functionalized MWCNTs [66] or with glassy carbon and MWCNTs [58] were also reported. The electrode containing glassy carbon presented a high stability of 15 months, while the presence of the MIP offered better performance characteristics but the analysis was more time-consuming, involving more steps, such as the MIP-modified electrode being incubated after the template extraction for 12 min in B-R buffer solution pH 4.00 containing DZP, followed by electrode washing before the actual voltammetric measurement [66].

A disposable wearable glove-based multi-sensor array was developed for the on-site detection of DZP, the explosive picric acid, and the pollutant 4-nitrophenol in liquid and powder samples. The three-electrode configuration was stencil-printed on the fingertips of a nitrile glove. The working electrodes were subsequently covered with MWCNTs/PEDOT:PSS. The thumb finger of the glove was covered with an agarose containing an ionic gel diluted in buffer, which acted as a semi-solid electrolyte. The electrical circuit was closed by contacting the thumb finger and the sensing fingertip, containing the analyte. The electrode system was connected to a portable potentiostat coupled with a smartphone. These wearable sensors cost USD 1.00, and their electrochemical stability lasted at least ten days [64].

POC testing needs cost-effective, miniaturized devices that allow the rapid, sensitive, and selective analysis of small sample volumes (20 μL). Microfluidic electrochemical PADs meet these requirements by featuring two integrated, paper-patterned electrodes that have a high surface-to-volume ratio, thus enabling low LoDs. Such a microdevice obtained by Si@GNRs drop-casted on the working electrode of the PAD was developed for DZP detection in urine spiked with the drug from tablets and injections with a recovery of 92.0% [59].

A disposable solid-state potentiometric sensor for DZP detection in biological fluids was also developed [57], consisting of a microfabricated Cu electrode on a printed circuit board. A conductive polymer, namely poly (3-octylthiophene), was drop-casted on the Cu surface to improve the stability of the potential drift. Further, the electrode surface area was covered with a PVC-based ion-sensitive membrane containing calix[4]arene as the ionophore. The sensor presented Nernstian response (55.0 ± 0.4 mV/decade), short response time (11 ± 2 s), and long-term stability (3 months). The potentiometric selectivity constants (log K^pot_DZP,i) of −2.26 and −3.89 for the possible interfering compounds OZP and 2-(N-methylamino)-5-chlorobenzophenone, which are a metabolite and degradation product, respectively, emphasized that DZP can be detected in the presence of these species. Other performance characteristics of this sensor are listed in

Table 4. Performance characteristics of electrochemical methods reported in the literature for DZP quantification.

Electrode	Technique/Conditions/Ep (V)	Linear Range (mol/L)	LoD (mol/L)	Sample	Ref.
HMDE	DPAdSV/Eacc −0.60 V; tacc 120 s; PBS pH 6.00/−1.00	9.48× 10−10–7.06 × 10−8	2.81 × 10−10	Natural water	[25]
AgNDs/GNs/GCE	DPV/PBS pH 7.00/~−1.08	1.00 × 10−7–1.00 × 10−6 1.00 × 10−6–2.00 × 10−5	8.56 × 10−9	Pharmaceuticals, human plasma	[54]
BiPPGE	DPV/ABS pH 4.80/~−0.9	1.40 × 10−6–1.67 × 10−5	1.10 × 10−6	Pharmaceuticals, urine	[55]
C60–CNT/IL/GCE	DPV/PBS pH 7.00/~−1.00	3.00 × 10−7–7.00 × 10−4	8.70 × 10−8	Pharmaceuticals, serum, urine	[56]
Cu/POT/ISM sensor	Potentiometry	5.00 × 10−7–1.00 × 10−2	1.20 × 10−7	Plasma, urine, saliva, human milk	[57]
MWCNT/GC/CPE	SWV/B-R buffer pH 1.00/~−0.70	1.49 × 10−6–2.15 × 10−5	3.30 × 10−7	Alcoholic drinks, synthetic serum	[58]
Si@GNRs/EμPAD	CV/PBS pH 7.00/−1.50	3.50 × 10−9–3.50 × 10−3	1.50 × 10−9	Spiked urine	[59]
MIP@SiO2-MPA- AuZnCeSeS-QDs/SPCE	EIS	1.00 × 10−7–1.00 × 10−4	4.90 × 10−6	Electrolyte Saliva Electrolyte	[60]
		1.00 × 10−7–1.00 × 10−4 (I)	2.30 × 10−6 (I)
		1.00 × 10−6–1.00 × 10−3 (II)	2.70 × 10−6 (II)
	SWV/NaTKBor/0.25	1.00 × 10−7–1.00 × 10−3	1.00 × 10−6
GCE	FI-DPA/HCl/KCl buffer pH 1.00	2.00 × 10−5–2.50 × 10−4	1.34 × 10−5	Energetic drink, beer	[61]
SbFSPE	LSV/0.10 mol/L HCl/~−0.90	5.00 × 10−7–1.00 × 10−5	3.30 × 10−7	Pharmaceuticals	[62]
SPE	BIA-DPAdSV/Eacc −1.20 V; tacc 150 s; B-R buffer pH 6.00/~−1.35	5.00 × 10−6–4.00 × 10−5	2.00 × 10−6	Pharmaceuticals; fortified rum	[63]
MWCNTs/PEDOT:PSS/glove	SWV/B-R buffer pH 7.00/~−1.30	5.00 × 10−7–1.00 × 10−5	6.00 × 10−8	Alcoholic and non-alcoholic beverages	[64]
GO/Lys/GCE	DPV/PBS pH 7.00/−0.86	1.70 × 10−7–6.90 × 10−7 6.90 × 10−7–3.50 × 10−5	4.60 × 10−8	Alcoholic and non-alcoholic beverages	[65]
MIP-MWCNTs-COOH/CPE	SWV/0.40 mol/L HCl/~−0.68	8.00 × 10−9–1.00 × 10−6	3.70 × 10−9	Pharmaceuticals, human serum	[66]

(I) a portable potentiostat connected to a laptop and (II) on a hand-held potentiostat connected to a smartphone device.

. The sensor is affordable, sensitive, specific, user-friendly, and rapid; it necessitates only portable devices for on-site analysis and is easily deliverable to the end-user. All these parameters make it adequate for point-of-care diagnosis.

The performance characteristics and the analytical applications of all these sensors developed for DZP determination are summarized in Table 4.

2.5. Flunitrazepam (FNZP)

In biological samples, FNZP has been either studied from plasma [67,68] or serum [69]. Plasma samples were either centrifuged, diluted with B-R buffer (1:10 v/v, pH 3.0) and analyzed via standard addition [67], or analyzed as-is [68], while serum was processed via centrifugal filtration, dilution with PBS (1:5 v/v, pH 6.50), and spiked.

Alcoholic [70] and soft drinks [70,71] were spiked, without further treatment or dilution, or in the case of soft drinks, a 1:1 B-R buffer dilution [68] was employed. In a special application of a wearable electrode detector [72], FNZP was analyzed alongside scopolamine and ketamine spiked in alcoholic drinks.

Like the other BZDs, FNZP presented a cathodic peak at more negative potentials due to the azomethine group reduction (Scheme 1), and the characteristic signals attributed to the –NO₂ group reduction (Scheme 2), as it was previously presented for CLZP. The processes were pH-dependent, and diffusion-adsorption controlled [67,69] or only diffusion-limited [68]. The charge transfer coefficient (α) and k_cat for the oxidation-reduction signals corresponding to the redox pair –NH–OH/–N=O at the TiO₂@CuO-N-rGO/p(L-Cys)/GCE were estimated to be 0.61 and 2.92 s⁻¹, indicating a high electron transfer rate, while the value of the diffusion coefficient was found to be 2.26 × 10⁻⁶ cm²/s [67]. Chronoamperometric measurements allowed the estimation of FNZP diffusion coefficient in B-R buffer pH 7.00 and k_cat at AGO-CuNPs/SPCE as 2.62 × 10⁻⁶ cm²/s and 0.88 s⁻¹, respectively [71].

Voltammetric quantification of FNZP at nanomolar levels was achieved using GCEs modified with a conductive polymer (poly-L-cysteine or poly-β-cyclodextrin) and reduced graphene oxide doped with TiO₂@CuO-N or with boron, respectively. Compared to GCE, the effective surface area of the modified electrodes was increased by about 7 and 27 times, while the charge transfer resistance decreased by 12 and 1.4-fold for TiO₂@CuO-N-rGO/p(L-Cys)/GCE and Eβ-CD/B-rGO/GCE, respectively. In both situations [67,69], at AGO-CuNPs/SPCE [71] and at AuNPs/MnFe₂O₄NPs/CPE [68], the peak corresponding to the oxidation of the –NHOH group was used for FNZP quantitative determination. The stability of the electrochemical response of these sensors towards FNZP ranged from 8 days for the Eβ-CD/B-rGO/GCE up to one month [67,68] or longer [71] for the others.

A SPE with an iron-sparked graphite working electrode modified with GOx and GluHD was reported for FNZP analysis in drop-volume samples. The Fe-sparking of the graphite electrode tripled the electrode response while the immobilized enzyme performed an in situ deoxygenation of the solution, necessary for the investigation of FNZP reduction processes. The GluHD was added to provide the glucose necessary for the enzymatic reaction [70].

The performance characteristics and the analytical applications of all these sensors developed for FNZP determination are summarized in

Table 5. The performance characteristics of voltammetric methods developed for FNZP electrochemical quantification.

Electrode	Technique/Conditions/Ep (V)	Linear Range (mol/L)	LoD (mol/L)	Sample	Ref.
TiO2@CuO-N-rGO/p(L-Cys)/GCE	DPV/B-R buffer pH 3.00; tacc 30 s/Oxidation~0.65	1.00 × 10−9–5.00 × 10−5	3.00 × 10−10	Plasma	[67]
AuNPs/MnFe2O4NPs/CPE	DPV/B-R buffer pH 4.00/Oxidation~0.38	1.00 × 10−7–1.00 × 10−4	3.30 × 10−7	Human plasma	[68]
Eβ-CD/B-rGO/GCE	DPV/PBS pH 6.50/ Oxidation~0.00	2.00 × 10−9–5.00 × 10−7 5.00 × 10−7–2.00 × 10−5	6.00 × 10−10	Human blood serum	[69]
GOx/GluHD/FeSPC	DPV/PBS pH 7.00/~−0.55	1.00 × 10−6–1.00 × 10−5	7.00 × 10−7	Alcoholic and soft drinks	[70]
AGO-CuNPs/SPCE	DPV/B-R buffer pH 7.00/ Oxidation~0.25	4.00 × 10−7–1.40 × 10−4	1.30 × 10−7	Fruit juice	[71]

.

A new screening tool, which may be applied in forensic science, consisting of a cheap ($0.19 cost) in-house 3D printed electrochemical finger (e-finger), was produced for the rapid, direct multi-analyte testing of FNZP, ketamine, scopolamine, and paracetamol in undiluted alcoholic drinks. The sample was deoxygenated with an effervescent vitamin C tablet. The device, consisting of three carbon black-based plastic electrodes, was connected to a mini-potentiostat coupled to a smartphone controlled via an Android application. FNZP was detected based on its reduction signal, while the other analytes presented oxidation peaks at well-separated potentials. The FNZP was determined in the concentration range 2 mg/L–16 mg/L, and the LoDs in different beverages were 0.07 mg/L (whisky), 0.08 mg/L (vodka), 0.09 mg/L (gin), and 0.011 mg/L (beer) FNZP [72].

Chen et al.’s [73] review on analytical methods applied for FNZP detection in biological samples, which includes a small chapter dedicated to electroanalysis, is worth mentioning here.

2.6. Lorazepam (LZP)

LZP was analyzed from pharmaceutical injection liquids and drinking water, where both were diluted by PBS pH 7.00 and analyzed by the standard addition method using SWV at NiO/SWCNTs/CPE. LSV measurements indicated that LZP oxidation was a diffusion-controlled process, and the diffusion coefficient determined by chronoamperometry using Cotrell’s equation was 9.05 × 10⁻⁶ cm²/s. LZP oxidation peak (PBS pH 7.00, E_p~0.90 V) varied linearly with the analyte concentration in the range 1.00 × 10⁻⁷ mol/L–2.80 × 10⁻⁴ mol/L. The obtained LoD was 5.00 × 10⁻⁸ mol/L LZP [74].

A MWCNTs/NADES/CPE was applied for the simultaneous determination of LZP and noscapine in pharmaceutical samples. LZP diffusion coefficient calculated in this work was 7.83 × 10⁻⁷ cm²/s. DPV analysis of LZP at MWCNTs/NADES/CPE presented a linear range of 1.00 × 10⁻⁶ mol/L–2.22 × 10⁻³ mol/L and a LoD equal to 6.90 × 10⁻⁷ mol/L LZP [75].

2.7. Midazolam (MDZ)

In biological samples, MDZ has been analyzed in plasma after organic deproteinization and dilution with acetate buffer (pH 5.60), followed by spiking [76], or in urine by spiking, centrifuging, and B-R buffer dilution [77]. Pharmaceutical liquid dosage samples were studied post-dilution with either acetate [76] or B-R buffer [77]. Finally, alcoholic beverages were studied following both pre- and post-spike B-R dilution [78].

Cyclic voltammograms of MDZ recorded at different working electrodes [76,77,78] are characterized by at least an irreversible cathodic peak, typical for BZDs (Scheme 1). When using a MIP-CPE in the cathodic potential window, due to the structural similarity between MDZ and ALP, DZP and OZP, it was observed that these BZDs were also bound to the electrode surface, but to a much lesser degree, so using this sensor allowed for the selective detection of MDZ [77].

By applying the direct potential scan in the cathodic direction at BDDE in B-R buffer pH 2.00, MDZ presented the BZDs’ common electrochemical behavior (Scheme 1). However, in these conditions, if the forward scan was oriented towards positive potentials, an anodic peak was observed, attributed to an irreversible, diffusion-controlled electrode process. Based on voltammetric data (CV and SWV) and density functional theory (DFT) calculations, the following electro-oxidation mechanism was proposed (Scheme 4) [78].

The use of MDZ oxidation offers the advantage of higher selectivity, thus enabling its determination in the presence of other BZDs, like CLZP, DZP, and FNZP [78].

A more recent paper dealing with MDZ electrochemistry at SiNPs modified CPE covered by chronoamperometry with AgNPs emphasized that this analyte presented four reduction signals. The first two peaks were assigned to the irreversible reduction of the 4,5–C=N– with a hydroxylamine derivative as intermediate, and the last two to the >C=C< and –N=C< from the imidazole ring. This sensor was suitable for reliable MDZ detection in human plasma of healthy volunteers, but in the case of patients treated with anesthetic drugs, like ketamine and structurally related compounds, there was an interference of approximately 24% [76].

The performance characteristics and the analytical applications of all these sensors developed for MDZ determination are summarized in

Table 6. The performance characteristics of voltammetric methods developed for MDZ electrochemical quantification.

Electrode	Technique/Conditions/Ep (V)	Linear Range (mol/L)	LoD (mol/L)	Sample	Ref.
AuNPs/SiNPs-CPE	DPV/ABS pH 5.60/~−1.35	4.00 × 10−7–2.90 × 10−4	2.28 × 10−8	Pharmaceuticals, human plasma	[76]
MIP-CPE	SWV/0.10 mol/L HCl/~−0.55	5.00 × 10−10–1.00 × 10−7 1.00 × 10−7–1.00 × 10−6	1.77 × 10−10	Pharmaceuticals, urine	[77]
BDDE	DPV/B-R buffer pH 2.00/ Oxidation ~1.60	1.00 × 10−6–3.00 × 10−6 4.00 × 10−6–2.50 × 10−5 1.00 × 10−6–1.00 × 10−5 1.00 × 10−6–1.50 × 10−5	7.40 × 10−8 4.60 × 10−7 4.30 × 10−7 3.30 × 10−7	Aqueous Vodka Whisky Red wine	[78]

.

2.8. Nitrazepam (NZP)

In biological samples, NZP has been studied in spiked human serum [79] diluted 1:100 with PBS (pH 7.40), or in urine samples [80], following N,N-dipropylamine LLME, and evaporating onto the electrode, then read in buffer solution. Tablets [79,81] were finely ground, dissolved in either ethanol [79] or methanol [81], and analyzed following buffer dilution. In the case of both alcoholic and non-alcoholic beverages [81], crushed tablets were spiked into the target liquid and analyzed as-is.

Because NZP contains in its molecular structure the same electroactive groups as CLZP, its voltammetric behavior is similar (Scheme 1 and Scheme 2). Thus, in the first scan, NZP cyclic voltammograms emphasized, in the cathodic forward direction, two peaks (R1 at −0.60 to −0.75 V and R3 at −1.20 to −1.30 V due to the reduction of the –NO₂ and 4,5-azomethine group, respectively) and, in the reversed scan, an anodic signal (O2 at −0.10 V to −0.20 V). In the subsequent scan, a new cathodic peak appears at about −0.20 to −0.3 V (R2), which forms a peak pair with (O2), which was assigned to the nitroso/hydroxylamine redox couple [79,80,81]. Peak R1 was mostly used for NZP quantification [80,81].

NZP detection was improved by using a GCE modified with GO-HNT by drop-casting, due to the synergetic effects of both components, namely, HNT increased the electroactive surface area (from 0.06 to 0.263 cm²) and GO provides electron-rich functionalities, increasing the electrocatalytic properties of the electrode [81]. Another sensor reported for NZP analysis was obtained by drop-casting at the GCE surface of an aqueous solution of rGO and drying, followed by potentiodynamic electrodeposition of AgNPs. GCE modification with AgNPs, rGO, and AuNPs-rGO increased NZP peak current about 2, 11, and 15 times, respectively [79].

NZP was voltammetrically determined after the homogeneous liquid–liquid microextraction employing N,N-dipropylamine as switchable solvent. The NZP containing extractant was dropped onto the surface of a GCE, which was previously pretreated by CV. After the evaporation of the solvent, the electrode was immersed in PBS pH 7.00, and differential pulse voltammograms were recorded [80].

The performance characteristics and the analytical applications of all these sensors developed for NZP determination are summarized in

Table 7. The performance characteristics of voltammetric methods developed for NZP electrochemical quantification.

Electrode	Technique/Conditions/Ep (V)	Linear Range (mol/L)	LoD (mol/L)	Sample	Ref.
GCE AuNPs/GCE rGO/GCE AuNPs-rGO/GCE	DPV/PBS pH 7.40/Oxidation −0.12	2.00 × 10−5–2.20 × 10−4 5.00 × 10−6–3.20 × 10−4 5.00 × 10−6–1.00 × 10−4 5.00 × 10−7–4.00 × 10−4	6.66 × 10−6 1.66 × 10−6 1.66 × 10−6 1.66 × 10−7	Pharmaceuticals, serum	[79]
Electroactivated GCE	SS-HLLME/DPV/PBS pH 7.00/~−0.75	1.07 × 10−10–7.11 × 10−10 7.11 × 10−10–1.60 × 10−8	3.20 × 10−12	Urine	[80]
GO-HNT/GCE	CV/PBS pH 7.00/~−0.72	1.60 × 10−7–1.50 × 10−4	7.90 × 10−7	Beverages	[81]

.

2.9. Olanzapine (OLP)

OLP has been primarily studied in urine [82,83,84], serum [83,84,85], and from pharmaceutical tablets [83,84,85]. Urine samples have been analyzed by the standard addition method as-is [82,83], or either urine or serum [84] samples were spiked, followed by organic solvent deproteinization, centrifugation, and dilution with PBS pH 7.00. Serum was also prepared by organic solvent deproteinization pre- and post-spiking [83], and dilution with B-R buffer, or by a 1:10 pre-spike dilution with acetate buffer solution (pH 5.50). Pharmaceutical tablets were ground, and a certain weight was either dissolved in methanol followed by water dilution [83] or directly dissolved into buffer [84,85], prior to analysis.

Some studies emphasized an anodic signal for OLP at modified GCEs [82,84,86], which was assigned to the oxidation of the secondary amino group (>¹NH) from the diazepine ring B (Figure 1). Somewhat different mechanisms were proposed for this process (Scheme 5).

In a study regarding OLP’s possible oxidative removal from contaminated aqueous samples, the CV recordings at BDDE revealed two anodic signals and a cathodic one. The anodic peak situated at a less positive potential (0.82 V, pH 2.00) was considered to form a pair with the reduction signal from about 0.20 V, associated with a quasi-reversible redox couple involved in surface-confined electrode processes. The more anodic signal (1.61 V) was due to an irreversible, diffusion-controlled oxidation process. Using chronoamperometric data and Cotrell’s equation, a diffusion coefficient of 6.50 × 10⁻⁴ cm²/s was estimated for OLP [87].

At AuNPs/p(L-Ala)/pre-PGE, OLP presented two anodic and one cathodic signals due to irreversible processes. It was considered that the adsorption-controlled oxidations took place at the benzodiazepine ring and involved 1e⁻ and 1H⁺ (similar to Scheme 5(c)) and 1e⁻ and no H⁺, respectively, while 2e⁻ and 2H⁺ participated in the diffusion-controlled reduction at the imine group (similar to Scheme 1(b)). The authors have also proposed a mechanism for OLP electrochemical behavior [83].

A detailed voltammetric study showed that at GCE, OLP participated in four electrochemical processes. The first anodic signal (process I) appeared at about 0.20 V and was a reversible process, because a counter cathodic peak (process II) was observed. Two additional, ill-defined oxidation signals were observed at about 0.85 V (process III) and 1.10 V (process IV). The studies revealed that process I was followed by a chemical step so that it follows an EC mechanism. Thus, processes I/II are proton-coupled electron transfer processes involving the nitrogen oxidation with the formation of a radical which can either dimerize or participate in a reaction involving H₂O addition. Process III involved 1e⁻ and 1H⁺ and led to the formation of an ammonium radical in the piperazine ring. In fact, processes III and IV formed together a two-step oxidation involving 2e⁻ and 1H⁺ at the piperazine group, generating as a final product an imine. In that paper, the authors proposed a mechanism for OLP electrooxidation [88].

QD-based nanocomposites were used to modify the GCE surface by drop casting to improve its electroanalytical performance towards OLP analysis. Enhanced current responses for OLP oxidation was due to the increase in the specific electroactive area (by about 1.5 and 3 times for S-CQDs/Fe₂O₃/GCE and NC@N,S@GQDs/GCE, respectively [82,86]) in comparison to bare GCE, the decrease in the charge transfer resistance (from 435.6 Ω for GCE to 64.5 Ω for NC@N,S@GQDs/GCE [86]) and to the synergistic effects of the modifier’s components. For example, better adsorption of OLP molecules was ensured by the H-bonds formed between the diazepine ring and the functional groups of the S-CQDs, while the Fe₂O₃ NPs were responsible for the better electrocatalytic oxidation of the analyte [82]. Similar results were obtained for PGEs, whose electroactive surface area increased by 1.4 and 3 times, respectively, after modification with AuNPs/p(L-Ala) [83] and N,P-CNOs/GNP [84]. At chronoamperometric pretreated PGE, OLP presented two oxidation peaks, whose intensity increased 2.5 times and 6.72 times, respectively, after AuNPs/p(L-Ala) deposition at the pretreated PGE surface [83]. Regarding the selectivity of the presented electrodes, valine, cysteine and some common sugars, like glucose, fructose and sucrose, did not present any signal at AuNPs/p(L-Ala)/pre-PGE [83] and S-CQDs/Fe₂O₃/GCE, while small anodic responses were observed for ascorbic acid and uric acid, but they did not influence OLP determination at the modified GCE [82]. The previously mentioned possible interferents, as well as other compounds, like metformin, paroxetine [84], or clozapine, asenapine, aripripazole, even when present in 100-fold excess [86], did not influence the OLP anodic signal recorded at N,P-CNOs/GNP/PGE and NC@N,S@GQDs/GCE, respectively.

A gold microelectrode modified with platinum black by chronopotentiometry was developed for rapid and sensitive OLP analysis in microliter volumes of undiluted serum samples. When compared to the unmodified electrode, this sensor presented a 6.7 times higher sensitivity and 5.1 times lower LoD [89]. As is obvious from the literature and from the data presented in this review, most electrochemical investigations were conducted applying different voltammetric methods. However, a hydrodynamic amperometric method, at a constant potential of 0.13 V and using a gold modified electrode, was described for the sensitive OLP determination [90].

Potentiometric analysis of OLP was also reported using either a carbon paste electrode modified with the ion-exchanger OLP-TP [85] or an ion-selective electrode with a PVC membrane containing the ion pair OLP-TPB [91]. Both electrodes presented Nernstian responses (59.2 mV/decade and 60 mV/decade) and a short response time of 4 and <10 s, for the OLP-TP/CPE [85] and the OLP-TPB/DOP/PVC, respectively [91]. OLP-TP/CPE was used in the potentiometric titration of OLP with phosphotungstic acid as well as in the study of its complexation with β-cyclodextrin. For the OLP-β-cyclodextrin complex, a molar ratio of 1:1 was found, and the inclusion constant was found to be 973.8 [85].

The performance characteristics and the analytical applications of all these sensors developed for OLP determination are summarized in

Table 8. The performance characteristics of voltammetric methods developed for OLP electrochemical quantification.

Electrode	Technique/Conditions/Ep (V)	Linear Range (mol/L)	LoD (mol/L)	Sample	Ref.
S-CQDs/Fe2O3/GCE	CV/PBS pH 7.00/Oxidation ~0.20	1.00 × 10−7–3.50 × 10−5	6.00 × 10−9	Urine	[82]
AuNPs/p(L-Ala)/pre-PGE	DPV/B-R buffer pH 4.50/ Oxidation~0.50	2.71 × 10−7–1.00 × 10−3	8.10 × 10−8	Pharmaceuticals, serum, urine	[83]
N,P-CNOs/GNP/PGE	DPV/PBS pH 7.00/ Oxidation ~0.15	4.50 × 10−10–4.50 × 10−8	4.40 × 10−11	Pharmaceuticals, serum, urine	[84]
OLP-TP/CPE	Potentiometry titration	7.50 × 10−7–5.60 × 10−4	5.00 × 10−7	Pharmaceuticals, blood	[85]
NC@N,S@GQDs/GCE	SWAdSV/Eacc 0.10 V; tacc 120 s; B-R buffer pH 7.00/Oxidation ~0.60	1.50 × 10−8–9.00 × 10−7	5.00 × 10−9	Pharmaceuticals, plasma, urine	[86]
PtB/Au	DPV/PBS pH 5.00/Oxidation ~0.20	2.00 × 10−8–1.20 × 10−7	1.56 × 10−8	Serum	[89]
BMBPBP/CdS-QDs/MWCNTs/Au	HA/RDE/PBS pH 7.00	2.00 × 10−8–1.00 × 10−4	6.00 × 10−9	Serum	[90]
OLP-TPB/DOP/PVC	Potentiometry	4.00 × 10−6–1.00 × 10−2	2.02 × 10−6	Pharmaceuticals, urine	[91]

.

2.10. Oxazepam (OZP)

Biological samples (serum or urine) were spiked and analyzed as-is, while OZP tablets were analyzed by the standard addition method, but without an indication of sample preparation [92].

OZP presented only an irreversible reduction peak (Scheme 1), which was exploited for its quantitative determination using modified GCEs. The OZP cathodic signal (PBS pH 7.00, E_p~−0.60 V) was increased 4.1 times and shifted by 0.22 V in the positive direction when the GCE surface was covered with (Ag-Pt)NPs/GNs nanocomposite. DPV enabled OZP quantification in the concentration range 5.00 × 10⁻⁸ mol/L–1.50 × 10⁻⁴ mol/L with a LoD 4.20 × 10⁻⁸ of mol/L. The electrode presented good selectivity from common ions and molecular species (sugars, urea, acid uric, acetaminophen, folic acid), and its response was stable over four weeks. The applicability of the developed method was tested by OZP analysis in pharmaceuticals, serum, and urine samples [92]. A B-rGO/p(ASP)/GCE was developed for OZP determination by DPV (E_p~−1.10 V) in B-R buffer pH 7.00 after a preconcentration time of 60 s. The method presented a LoD of 3.00 × 10⁻¹⁰ mol/L OZP and two linear ranges, 1.00 × 10⁻⁹ mol/L–1.00 × 10⁻⁶ mol/L and 1.00 × 10⁻⁶ mol/L–8.00 × 10⁻⁴ mol/L, with different slopes, which were explained by a monolayer adsorption at low concentrations, followed by a multi-layer adsorption at higher analyte concentrations. Using this electrode, OZP was determined from plasma without interference of CLZP, LZP, fentanyl, zolpidem, and common biologically important species [93].

2.11. Tetrazepam (TZP)

TZP was analyzed by using human serum in 0.10 mol/L PBS electrolyte. DPV measurements were performed at a CPE modified with nitrogen-doped carbon nanoparticles. TZP reduction signal (PBS pH 7.40, E_p −0.83 V) was used for its quantitative determination in the concentration range 0–2.25 × 10⁻⁴ mol/L. The method presented a LoD of 1.73 × 10⁻⁸ mol/L TZP and was applied to the analyte determination in blood serum [94].

2.12. Multidrug Studies

Certain publications intended the development of similar yet standardized methods or electrodes to analyze multiple benzodiazepines, and as such, these are presented below, grouped by the target matrix.

2.12.1. Biological Samples

DZP, NDZP, BZP, and ALP were individually studied from pure weight or tablet samples with solvation in methanol. DZP was specifically determined by spiking a certain amount from a pharmaceutical tablet into synthetic urine mixed with 0.10 mol/L NaOH [95].

Urine samples were centrifuged and filtered, then diluted with ABS pH 6.00 prior to spiking. Analysis of DZP and OZP at a MIP-MMWCNTs-COOH/PGE, presented two well-separated DPV reduction peaks, thus allowing their concomitant quantification (Table 9). The central composite design method was used to establish the optimum conditions for the preparation of the modified PGE which involved the following steps: (i) the potentiodynamic activation of a 0.7 mm HB type PGE in 0.10 mol/L H₂SO₄, (ii) modification of the activated PGE with magnetic carboxylic MWCNTs by dip-coating; (iii) covering the whole system with a thin layer of MIP prepared using methacrylic acid as functional monomer, ethylene glycol dimethacrylate and 2,2-azobisisobutyronitrile as crosslinker and initiator, respectively, and (iv) removing of DZP and OZP from the polymeric matrix by washing with a methanol:acetic acid (9:1) mixture. Similar compounds, like ALP, CDZO, LZP, and TMP, did not interfere in DZP and OZP voltammetric determination for this electrode. The high selectivity conferred by the MIP was also demonstrated by the high imprinting factors of 2.9 and 2.6 for DZP and OZP, respectively [96].

Whole blood and urine spiked with NZP and 7-aminon-NZP were processed via microwave-assisted extraction using a pH 9.00 borax buffer and ethyl acetate, followed by two series of centrifugation, evaporation of the extractant, and a final small-volume aqueous reconstitution of the sample residue. Subsequently, NZP and 7-amino-NZP were analyzed by ITIES using ITV. In this technique, the analytical signals are represented by ionic currents generated by the ion transfer between two immiscible solutions due to an applied Galvani potential difference. The electrochemical behavior of the analytes is controlled by their partition characteristics between the two phases. Thus, this technique has the advantage that it can be applied to non-redox species. Voltammetric information obtained from measurements performed at eLLI at macroscopic and microscopic ITIES was exploited to investigate the interfacial behavior of the two analytes and for their quantitative analysis (Table 9). The diffusion coefficients of the two compounds in aqueous (D_NZP,aq, D_7a-NZP,aq) and organic (D_NZP,o, D_7a-NZP,o) media were established to be 1.41 × 10⁻⁵ cm²/s, 1.12 × 10⁻⁵ cm²/s, 1.31 × 10⁻⁵ cm²/s, and 1.11 × 10⁻⁵ cm²/s, respectively [97].

OLP, acetaminophen, and tramadol were targeted for simultaneous analysis by spiking into human plasma or urine samples previously diluted ten-fold with PBS pH 7.00. DPV at a CuO-rGNR/IL/CPE enabled the simultaneous determination of OLP, acetaminophen, and tramadol based on their well-resolved cathodic signals situated at 0.20 V, 0.44 V, and 0.90 V, respectively. The tested mixtures contained the analytes in the concentration ranges 2.00 × 10⁻⁵ mol/L–2.00 × 10⁻⁴ mol/L OLP, 2.00 × 10⁻⁴ mol/L–4.00 × 10⁻⁴ mol/L acetaminophen, and 5.00 × 10⁻⁵ mol/L–9.50 × 10⁻⁴ mol/L tramadol [98].

CLZP, ALP, DZP, OZP, and CDZO were targeted in spiked human serum samples analyzed as-is [99]. Human plasma was spiked with tenfold volumes of CLZP, DZP, ALP, CDZO, or OZP prepared in 0.10 mol/L NaOH [100]. The electrodes employed in these studies for the individual determination of BZDs were GCE electrochemically modified by layer-by-layer deposition of poly-dopamine, electrodeposition of chitosan, and subsequently chronoamperometric synthesis of AuNPs on the polymeric layers [99], as well as a similar poly-dopamine-coated electrode onto which poly-folic acid was additionally electropolymerized [100] (Table 9).

In an HPLC-ED [101] application, quetiapine, OLP, clozapine, fluphenazine, promazine, promethazine, levomepromazine, chlorpromazine, thioridazine, and perphenazine were spiked into control serum and human plasma samples, followed by precipitation with acetonitrile and centrifugation. Suspect patient samples were also diluted with water (1:1) prior to analysis. Thus, using this method, phenothiazines and BZDs belonging to the first generation of antipsychotic drugs were quantified using their oxidation signals recorded at a PGE. The method presented for all tested drugs a linear range up to 5.00 × 10⁻⁷ mol/L and LoDs in the nmol/L range.

A gold electrode covered electrochemically with poly-chitosan and subsequently by layer-by-layer deposition of a conductive ink based on AgNPs-N doped graphene QDs was developed for the quantification of several BZDs in plasma samples without any pretreatment step [102] (Table 9).

A LSG multiplex sensing platform was developed for simultaneously detecting amphetamine, cocaine, and BZDs (via metabolite detection), by using saliva extracted by centrifugation from oral pad swabs, testing both spiked and neat healthy saliva and spiked and neat patient saliva. This sensing platform with one reference, one counter, and three working electrodes, connected to a custom-made potentiostat and a smartphone running a custom application, represented a POC testing device applied to detect the mentioned analytes in saliva provided by healthy volunteers and methamphetamine patients (Table 9). The system presented high selectivity since each working electrode was modified with the corresponding antibody; namely, BZD antibody was immobilized on the first, amphetamine antibody on the second and cocaine antibody on the third working electrode [103].

2.12.2. Pharmaceutical Formulations

As in most sample preparations, tablets are generally weighed and homogenized, followed by dissolving a certain amount to make a stock solution. For DZP and OZP, the tablet portion was dissolved in water, sonicated, filtered, then diluted with buffer, and analyzed by the standard addition method [96]. NZP in tablet form was crushed and dissolved in 0.01 mol/L HCl/NaCl, sonicated, filtered, and further diluted with the same solution to bring the concentration within linearity [97]. BZP and ALP tablet [42,43] portions were dissolved in methanol, filtered, and diluted with B-R buffer (1:9) prior to analysis. BDDE [42] and CPE [43] were applied to the cyclic voltammetric analysis of ALP and BZD and their individual quantification by DPV (Table 9). Before using, the BDDE was pretreated electrochemically through an anodic (2.00 V, 60 s) and a cathodic (−2.00 V, 30 s) step, to clean the electrode surface and to generate an H-terminated electrode surface, respectively [42].

Finally, for DZP, CLZP, and ALP [35], a 4–5 mg portion of crushed tablets was dissolved in B-R buffer with ethanol and applied directly onto the electrode for analysis.

2.12.3. Beverages

Alcoholic beverages were studied by spiking DZP, CLZP, or ALP, diluting tenfold with the B-R buffer solution (pH 2.00 or 7.00), and then analyzed in standard conditions. Based on the correlation of their voltammetric peaks with the electroactive groups existing in their molecule, BZDs were classified into three classes, which were detected after spiking beer, gin, cognac, white wine, and red wine. A dual test, at pH 2.00 and pH 7.00, was necessary for red wine to be able to discriminate between the three BZDs classes [35].

Similarly, DZP and MDZ were analyzed with an LSG-PEI disposable sensor after spiking into B-R buffer pH 4.00, tenfold diluted alcoholic and non-alcoholic samples (Table 9). Before the actual voltammetric recordings, the LSG-PEI was pretreated electrostatically at −2.0 V for 120 s, to partially reduce the oxygen-containing groups from the graphene surface and to improve its conductivity. 3,4-MDMA and substances commonly found in beverages, like glucose, ethanol, ascorbic, and citric acids, did not interfere with the BDZ determination. However, the method could not discriminate between structurally related BDZs, which present reduction signals in the same potential region [104].

ITV at an eLLI-based sensor was also applied to evaluate physicochemical parameters of four “designer” BZDs, among them being the diffusion coefficients of the cationic forms of the analytes, which were 1.6 × 10⁻⁵ cm²/s, 9.7 × 10⁻⁵ cm²/s, 7.3 × 10⁻⁵ cm²/s, and 8.4 × 10⁻⁵ cm²/s for FPZM, FBZM, CLNZ, and DCZP. This technique allowed their quantification (Table 9) and the discrimination between the two very similar species CLNZ and FPZM [105].

A voltammetric ET consisting of six working electrodes (graphite epoxy composite, Pt, and graphite modified with one of the following NPs: Cu, CuO, Pt, WO₃), an auxiliary, and a reference electrode was designed for the simultaneous determination of DZP, FNZP, and LZP. To obtain reproducible results every time, the 4th cycled voltammogram was considered, and before each measurement, the electrode’s surface was cleaned by applying a potential of 1.20 V for 60 s in a pH 10.00 solution. Principal Component Analysis was applied to select the sensors. For quantitative analysis (Table 9), the Discrete Wavelet Transform was used to reduce data complexity and to build an Artificial Neuronal Network model [106].

Table 9. The performance characteristics of electrodes developed for multidrug analysis, including BZDs.

Analyte (BZD)	Electrode	Technique/Conditions/Ep (V)	Linear Range (mol/L)	LoD (mol/L)	Sample	Ref.
ALP BZP DZP NDZP	m-AgSAE	DPV/ PBS pH 11/~−1.06 B-R buffer pH 2.00/~−0.35 0.10 mol/L NaOH/~−1.22 B-R buffer pH 10.20/~−1.10	6.00 × 10−7–1.00 × 10−4 6.00 × 10−7–1.00 × 10−4 4.00 × 10−7–1.00 × 10−4 6.00 × 10−7–1.00 × 10−4	3.54 × 10−7 3.73 × 10−7 9.00 × 10−8 1.28 × 10−7	Synthetic urine	[95]
DZP OZP	MIP-MMWCNTs-COOH/PGE	DPV/ABS pH 6.00/ DZP: −0.89 OZP: −1.10	9.83 × 10−8–2.00 × 10−4 4.88 × 10−8–1.50 × 10−4	5.97 × 10−8 2.10 × 10−8	Spiked urine, tablets	[96]
NZP	eLLI/macroITIES eLLI/microITIES	ITV/0.01 mol/L NaCl, 0.01 mol/L HCl, pH 2.00	1.00 × 10−5–3.0 × 10−5 2.00 × 10−6–3.00 × 10−5	2.15 × 10−6 (+) 1.86 × 10−6 (−) 4.20 × 10−7 (+) 3.80 × 10−7 (−)	Spiked urine, blood, pharmaceuticals	[97]
7-amino-NZP	eLLI/macroITIES eLLI/microITIES	ITV/0.01 mol/L NaCl, 0.01 mol/L HCl, pH 2.00	1.00 × 10−5–3.0 × 10−5 2.00 × 10−6–3.00 × 10−5	7.10 × 10−7 (+) 1.23 × 10−6 (−) 2.10 × 10−7 (+) 4.20 × 10−7 (−)	Spiked urine, blood
ALP CDZO CLZP DZP OZP	p(DA-CS)-AuNPs/GCE	DPV PBS pH 7.00/−0.33 PBS pH 6.00/−0.42 PBS pH 10.00/−0.58 PBS pH 8.00/−0.52 PBS pH 7.00/−0.32	5.00 × 10−8–4.20 × 10−7 7.50 × 10−8–4.00 × 10−7 2.80 × 10−7–6.50 × 10−7 2.00 × 10−8–1.40 × 10−7 6.30 × 10−8–1.39 × 10−6	5.00 × 10−8 7.50 × 10−8 2.80 × 10−7 2.00 × 10−8 6.30 × 10−8	Human serum	[99]
ALP CDZO CLZP DZP OZP	p(DA-FA)/GCE	DPV/0.10 mol/L NaOH ~−0.40 ~−0.25 ~−0.16 ~−0.40 ~−0.40	3.10 × 10−8–5.20 × 10−8 3.00 × 10−8–7.10 × 10−8 2.50 × 10−8–9.00 × 10−7 2.70 × 10−8–4.10 × 10−8 2.50 × 10−8–4.70 × 10−8	3.10 × 10−8 3.00 × 10−8 2.50 × 10−8 2.70 × 10−8 2.50 × 10−8	Human plasma	[100]
ALP CDZO CLZP DZP OZP ALP CDZO CLZP DZP OZP ALP CDZO CLZP DZP OZP	AgNPs/N-GQD/Cs/Au	DPV/0.10 mol/L NaOH Oxidation ALP, CDZO, DZP, OZP: 0.50; CLZP: 0.60 DPV SWV	5.60 × 10−5–1.56 × 10−4 5.20 × 10−5–2.50 × 10−4 8.40 × 10−5–6.25 × 10−4 5.40 × 10−5–1.42 × 10−4 5.40 × 10−5–4.54 × 10−4 3.80 × 10−6–2.47 × 10−4 6.19 × 10−5–9.90 × 10−4 3.80 × 10−6–4.95 × 10−4 7.70 × 10−6–4.95 × 10−4 6.30 × 10−8–1.39 × 10−6 7.30 × 10−5–1.92 × 10−4 7.04 × 10−5–2.50 × 10−4 5.40 × 10−5–1.00 × 10−3 5.60 × 10−5–4.54 × 10−4 5.60 × 10−5–2.94 × 10−4	5.60 × 10−5 5.20 × 10−5 8.40 × 10−5 5.40 × 10−5 5.40 × 10−5 3.80 × 10−6 6.19 × 10−5 3.80 × 10−6 7.70 × 10−6 6.30 × 10−8 7.30 × 10−6 7.04 × 10−5 5.40 × 10−5 5.60 × 10−5 5.60 × 10−5	Electrolyte Human plasma Electrolyte	[102]
BZDs	Ab-LSG	DPV/5.00 mmol/L [Fe(CN)6]3−/4− in 0.10 mol/L PBS + 0.10 mol/L KCl/~−0.05	1 pg/mL–500 ng/mL	9.7 ng/mL	Saliva	[103]
ALP BZP	BDDE	DPV/ALP: B-R buffer pH 5.00/−0.84; BZP: B-R buffer pH 11.00/−1.10	8.00 × 10−7–1.00 × 10−4 1.00 × 10−6–1.00 × 10−4	6.40 × 10−7 3.10 × 10−7	Pharmaceuticals	[42]
ALP BZP	CPE	DPV/B-R buffer:methanol (9:1) pH 3.00/ALP: −0.78; BZP: ~−0.50	8.00 × 10−7–1.00 × 10−4 8.00 × 10−7–1.00 × 10−4	LoQ: 4.20 × 10−7 3.80 × 10−7	Pharmaceuticals	[43]
ALP CLZP DZP	Graphite SPE	SWV/B-R buffer pH 7.00 Standard solution ALP: ~−0.6 to ~−0.7 CLZP: ~−0.55, ~−0.75, ~−0.9 DZP: ~−0.75 to − ~0.9	2.50 × 10−8–1.00 × 10−6 7.50 × 10−7–6.00 × 10−4 2.50 × 10−8–1.00 × 10−6	1.40 × 10−8 7.50 × 10−7 3.60 × 10−8	Pharmaceuticals	[35]
ALP CLZP DZP		Deoxygenated solution ALP: ~−0.75, ~−0.97 CLZP: ~−0.55, ~−0.95 DZP: ~−0.85	1.50 × 10−5–2.00 × 10−4 1.00 × 10−6–2.00 × 10−4 5.00 × 10−6–3.00 × 10−4	1.10 × 10−5 7.50 × 10−7 3.11 × 10−6	Beverages
DZP MDZ	LSG-PEI	SWV/B-R buffer pH 4.00/DZP:~−1.15; MDZ:~−1.05	2.50 × 10−6–2.50 × 10−5 2.50 × 10−5–1.00 × 10−4	6.60 × 10−7 6.10 × 10−7	Beverages	[104]
CLNZ DCZP FBZM FPZM	eLLI	ITV/B-R buffer pH 2.00	2.00 × 10−6–5.00 × 10−5 2.00 × 10−6–5.00 × 10−5 2.00 × 10−6–5.00 × 10−5 2.00 × 10−6–5.00 × 10−5	1.40 × 10−7 1.60 × 10−7 2.40 × 10−7 1.90 × 10−7	Beverages	[105]
DZP FNZP LZP	ET	CV/B-R buffer pH 10.00/PCA, DWT	3.50 × 10−5–1.10 × 10−4 3.20 × 10−5–9.60 × 10−4 3.10 × 10−5–9.30 × 10−5	6.00 × 10−6 5.60 × 10−6 4.60 × 10−6	Electrolyte	[106]

(+)—positive signals defined as the cation transfer from the aqueous to the organic phase; (−)—negative signals defined as the cation transfer from the organic to the aqueous phase.

Table 9. The performance characteristics of electrodes developed for multidrug analysis, including BZDs.

Analyte (BZD)	Electrode	Technique/Conditions/Ep (V)	Linear Range (mol/L)	LoD (mol/L)	Sample	Ref.
ALP BZP DZP NDZP	m-AgSAE	DPV/ PBS pH 11/~−1.06 B-R buffer pH 2.00/~−0.35 0.10 mol/L NaOH/~−1.22 B-R buffer pH 10.20/~−1.10	6.00 × 10−7–1.00 × 10−4 6.00 × 10−7–1.00 × 10−4 4.00 × 10−7–1.00 × 10−4 6.00 × 10−7–1.00 × 10−4	3.54 × 10−7 3.73 × 10−7 9.00 × 10−8 1.28 × 10−7	Synthetic urine	[95]
DZP OZP	MIP-MMWCNTs-COOH/PGE	DPV/ABS pH 6.00/ DZP: −0.89 OZP: −1.10	9.83 × 10−8–2.00 × 10−4 4.88 × 10−8–1.50 × 10−4	5.97 × 10−8 2.10 × 10−8	Spiked urine, tablets	[96]
NZP	eLLI/macroITIES eLLI/microITIES	ITV/0.01 mol/L NaCl, 0.01 mol/L HCl, pH 2.00	1.00 × 10−5–3.0 × 10−5 2.00 × 10−6–3.00 × 10−5	2.15 × 10−6 (+) 1.86 × 10−6 (−) 4.20 × 10−7 (+) 3.80 × 10−7 (−)	Spiked urine, blood, pharmaceuticals	[97]
7-amino-NZP	eLLI/macroITIES eLLI/microITIES	ITV/0.01 mol/L NaCl, 0.01 mol/L HCl, pH 2.00	1.00 × 10−5–3.0 × 10−5 2.00 × 10−6–3.00 × 10−5	7.10 × 10−7 (+) 1.23 × 10−6 (−) 2.10 × 10−7 (+) 4.20 × 10−7 (−)	Spiked urine, blood
ALP CDZO CLZP DZP OZP	p(DA-CS)-AuNPs/GCE	DPV PBS pH 7.00/−0.33 PBS pH 6.00/−0.42 PBS pH 10.00/−0.58 PBS pH 8.00/−0.52 PBS pH 7.00/−0.32	5.00 × 10−8–4.20 × 10−7 7.50 × 10−8–4.00 × 10−7 2.80 × 10−7–6.50 × 10−7 2.00 × 10−8–1.40 × 10−7 6.30 × 10−8–1.39 × 10−6	5.00 × 10−8 7.50 × 10−8 2.80 × 10−7 2.00 × 10−8 6.30 × 10−8	Human serum	[99]
ALP CDZO CLZP DZP OZP	p(DA-FA)/GCE	DPV/0.10 mol/L NaOH ~−0.40 ~−0.25 ~−0.16 ~−0.40 ~−0.40	3.10 × 10−8–5.20 × 10−8 3.00 × 10−8–7.10 × 10−8 2.50 × 10−8–9.00 × 10−7 2.70 × 10−8–4.10 × 10−8 2.50 × 10−8–4.70 × 10−8	3.10 × 10−8 3.00 × 10−8 2.50 × 10−8 2.70 × 10−8 2.50 × 10−8	Human plasma	[100]
ALP CDZO CLZP DZP OZP ALP CDZO CLZP DZP OZP ALP CDZO CLZP DZP OZP	AgNPs/N-GQD/Cs/Au	DPV/0.10 mol/L NaOH Oxidation ALP, CDZO, DZP, OZP: 0.50; CLZP: 0.60 DPV SWV	5.60 × 10−5–1.56 × 10−4 5.20 × 10−5–2.50 × 10−4 8.40 × 10−5–6.25 × 10−4 5.40 × 10−5–1.42 × 10−4 5.40 × 10−5–4.54 × 10−4 3.80 × 10−6–2.47 × 10−4 6.19 × 10−5–9.90 × 10−4 3.80 × 10−6–4.95 × 10−4 7.70 × 10−6–4.95 × 10−4 6.30 × 10−8–1.39 × 10−6 7.30 × 10−5–1.92 × 10−4 7.04 × 10−5–2.50 × 10−4 5.40 × 10−5–1.00 × 10−3 5.60 × 10−5–4.54 × 10−4 5.60 × 10−5–2.94 × 10−4	5.60 × 10−5 5.20 × 10−5 8.40 × 10−5 5.40 × 10−5 5.40 × 10−5 3.80 × 10−6 6.19 × 10−5 3.80 × 10−6 7.70 × 10−6 6.30 × 10−8 7.30 × 10−6 7.04 × 10−5 5.40 × 10−5 5.60 × 10−5 5.60 × 10−5	Electrolyte Human plasma Electrolyte	[102]
BZDs	Ab-LSG	DPV/5.00 mmol/L [Fe(CN)6]3−/4− in 0.10 mol/L PBS + 0.10 mol/L KCl/~−0.05	1 pg/mL–500 ng/mL	9.7 ng/mL	Saliva	[103]
ALP BZP	BDDE	DPV/ALP: B-R buffer pH 5.00/−0.84; BZP: B-R buffer pH 11.00/−1.10	8.00 × 10−7–1.00 × 10−4 1.00 × 10−6–1.00 × 10−4	6.40 × 10−7 3.10 × 10−7	Pharmaceuticals	[42]
ALP BZP	CPE	DPV/B-R buffer:methanol (9:1) pH 3.00/ALP: −0.78; BZP: ~−0.50	8.00 × 10−7–1.00 × 10−4 8.00 × 10−7–1.00 × 10−4	LoQ: 4.20 × 10−7 3.80 × 10−7	Pharmaceuticals	[43]
ALP CLZP DZP	Graphite SPE	SWV/B-R buffer pH 7.00 Standard solution ALP: ~−0.6 to ~−0.7 CLZP: ~−0.55, ~−0.75, ~−0.9 DZP: ~−0.75 to − ~0.9	2.50 × 10−8–1.00 × 10−6 7.50 × 10−7–6.00 × 10−4 2.50 × 10−8–1.00 × 10−6	1.40 × 10−8 7.50 × 10−7 3.60 × 10−8	Pharmaceuticals	[35]
ALP CLZP DZP		Deoxygenated solution ALP: ~−0.75, ~−0.97 CLZP: ~−0.55, ~−0.95 DZP: ~−0.85	1.50 × 10−5–2.00 × 10−4 1.00 × 10−6–2.00 × 10−4 5.00 × 10−6–3.00 × 10−4	1.10 × 10−5 7.50 × 10−7 3.11 × 10−6	Beverages
DZP MDZ	LSG-PEI	SWV/B-R buffer pH 4.00/DZP:~−1.15; MDZ:~−1.05	2.50 × 10−6–2.50 × 10−5 2.50 × 10−5–1.00 × 10−4	6.60 × 10−7 6.10 × 10−7	Beverages	[104]
CLNZ DCZP FBZM FPZM	eLLI	ITV/B-R buffer pH 2.00	2.00 × 10−6–5.00 × 10−5 2.00 × 10−6–5.00 × 10−5 2.00 × 10−6–5.00 × 10−5 2.00 × 10−6–5.00 × 10−5	1.40 × 10−7 1.60 × 10−7 2.40 × 10−7 1.90 × 10−7	Beverages	[105]
DZP FNZP LZP	ET	CV/B-R buffer pH 10.00/PCA, DWT	3.50 × 10−5–1.10 × 10−4 3.20 × 10−5–9.60 × 10−4 3.10 × 10−5–9.30 × 10−5	6.00 × 10−6 5.60 × 10−6 4.60 × 10−6	Electrolyte	[106]

(+)—positive signals defined as the cation transfer from the aqueous to the organic phase; (−)—negative signals defined as the cation transfer from the organic to the aqueous phase.

3. Conclusions

One can conclude that there is an increasing need for the development of cost-effective, sensitive platforms, like e-fingers [72], for the non-invasive simultaneous detection and monitoring of drugs. In this respect, electrodes modified with various (nano)materials (CNTs, QDs, NPs, polymeric films, etc., or a combination of them) have been reported for BZDs quantification in different matrices. However, it should be emphasized that most electrodes presented were conventional solid electrodes, bare or modified, used in batch electrochemical systems that require at least 5 mL of analyzed solutions, thus consuming more reagents than drop-based analytical systems, such as SPE and wearable sensors. Therefore, finding proper upscaling-compatible sensor fabrication methods may be a challenge for further research directions. On the other hand, in terms of time consumption, many of the presented procedures for electrode modification involved multiple steps, and therefore, future research should be directed towards finding rapid, single-step modification methods. One can consider that the first steps in these directions (miniaturization and reducing preparation time and costs) were made by the development of PADs.

It is essential that these devices can be miniaturized in order to analyze small sample volumes and to be used with portable devices so that the BZDs can be detected on-site, at the place of potential use, as well as in POC diagnosis. The present literature emphasized that there are increasing concerns in the development of various disposable (e.g., PADs) and even wearable sensors. To develop more eco-friendly analytical methods, it is important that the sample pretreatment step is minimized to avoid reagent consumption and to shorten the analysis time, and electrochemical analysis usually fulfills these requirements.

There is also a trend towards using various advanced mathematical models (chemometric tools) for data processing to enable selective multicomponent determination even with electrodes that are not very selective. ET also falls into this category.

Finally, it can be noted that the number of electrochemical methods reported in the last decade for the determination of individual BZDs is not very large, while for some components of this class, as well as for the “designer” BZDs, there is very little or even no data in this regard.

The current information contained in this review may be a useful starting point for further research in the field which may be directed towards (i) electroanalysis of more BZDs including “designer” BZDs; (ii) improving selectivity for structurally related molecules detection (e.g., by proper combination of molecularly imprinted polymers with nanomaterials with electrocatalytic activity or by using chemometric approaches); (iii) miniaturization; (iv) the further design and development of disposable sensors for POC testing and for the on-the-spot illicit drug detection; and (v) the use of more eco-friendly preparation methods and sensors.