Recent Advances in Photoelectrochemical Sensors for Analysis of Toxins and Abused Drugs in the Environment

Abstract

Toxic pollutants in the environment, such as toxins and abused drugs, have posed a major threat to human health and ecosystem security. It is extremely desirable to develop simple, low-cost, sensitive, and reliable techniques for the detection of these pollutants in the environment. As a booming analytical method, photoelectrochemical (PEC) sensors possess low background noise and high sensitivity. The performances of PEC sensors are fundamentally related to the photoelectric conversion efficiency, which mainly depends on the properties of photoactive materials. This review aims to summarize the engineered photoactive materials, i.e., semiconductors and semiconductor-based heterojunctions, as well as their actual applications, with emphasis on sensing mechanisms in PEC sensors for the analysis of toxins and abused drugs in the environment. Finally, the future research perspectives in this field are also discussed.

1. Introduction

Increasingly serious environmental pollution is a worldwide problem which is caused by the rapid industrialization and urbanization of the past few decades. There are many toxic pollutants in the environment, such as toxins and abused drugs, which pose a major threat to human health and ecosystem security [1,2]. Microcystins are a class of toxins mainly produced by cyanobacteria, and microcystin-leucine arginine (MC–LR) is the kind of freshwater toxin with the strongest toxicity and greatest potential to do harm; it brings many accidents and threats to human health because it is a powerful inhibitor of two key enzymes, protein phosphatases type 2A and type 1, in cellular processes [3,4,5]. The World Health Organization (WHO) proposed that the maximum allowable amount in drinking water be stipulated as 1 μg L⁻¹ for MC–LR [6].

Abused drugs are a new kind of environmental pollutant. With their continuous release into the environment, the global environmental burden and ecological risk are increasing [2,7,8]. Non-steroidal anti-inflammatory drugs (NSAIDs) are a category of drugs. Although they possess analgesic, antipyretic, and anti-inflammatory effects, they are also considered harmful to human health and ecosystem security [9,10,11]. Therefore, NSAIDs are regarded as a new pollutant by the WHO. As reported, NSAIDs and other abused drugs such as antibiotics and antioxidants are frequently detected in the aquatic systems [12,13]. Consequently, it is extremely desirable and urgent to develop sensitive and reliable techniques for the detection of toxins and abused drug pollutants in the environment.

Various analytical techniques, including a variety of chromatographic techniques [14,15,16], enzyme-linked immunosorbent assays [17,18,19], and spectrophotometric techniques [20,21], are frequently used for sensitive determination of toxins and abused drugs. However, the requirements of tedious sample pretreatment, skilled operators, and expensive and precise instruments greatly limit their application scope and in situ applications. Therefore, it is of great significance to develop a low-cost and convenient routine analysis method for toxins and abused drugs in the environment.

New emerging photoelectrochemical sensors (PEC sensors) are sensors based on the principle of photoelectric chemistry for the analysis and detection of biological and chemical molecules; they refer to the process of generating current by causing valence electron transfer inside the photoexcited material and triggering the chemical reaction under the irradiation of a light source [22,23,24,25,26,27]. Since light and photocurrents are used as excitation sources and recognition signals, respectively, PEC sensors can reduce background noise and achieve higher sensitivity than other conventional technologies. In addition, the use of an electronic reader makes PEC instruments simple, cost-effective, and miniaturized [28]. These obvious characteristics make the PEC sensors possible candidate sensors for the rapid and accurate monitoring of toxins and abused drugs in the environment.

Photoelectric conversion efficiency generally has a great influence on the sensitivity of the PEC sensors, which mainly depends on the properties of photoactive materials [29,30]. The photoactive materials under light irradiation can cause charge separation and transfer, and subsequently generate photocurrent as the detection signal, which usually acts as the conversion layer in the conversion of photons to electric energy [31]. Up to now, reviews have focused on the processes of PEC-sensor technology and its application in various environmental pollutants [28,30,32,33,34,35,36]. However, as far as we know, there is no integral overview of PEC sensors for monitoring toxins and abused drugs in the environment. Herein, this paper summarizes the applications of photoactive materials in PEC sensors in the analysis of toxins and abused drugs in the environment in recent years. In addition, the challenges and future prospects in this field are also discussed. We hope to provide some insights into the development of more mature PEC sensor strategies in order to meet the actual needs of toxins and abused drugs monitoring in the environment.

2. Principle of PEC Sensors

Typical PEC sensors mainly consist of three parts: an excitation light source, a detection system containing an electrolyte and electrode, and a signal acquisition system. Photoactive materials are applied as a photoelectric converter at the interface of a working electrode to produce photocurrent signals under light irradiation. The properties of photoactive materials or the electrolyte environment can be directly or indirectly changed by the target analytes, resulting in photocurrent variation and quantitative analysis [32].

As shown in Figure 1, a semiconductor photoactive materials-modified conductive substrate induces electron–hole (e⁻-h⁺) pairs under light irradiation in a typical PEC sensing process. The holes on the valence band (VB) transfers to the conduction band (CB). Correspondingly, an electrochemical workstation could measure the photocurrent signal that originated from the separation and migration of photogenerated charge carriers. In accordance with the electron migration direction, the photocurrent signal can be divided into anode sensing, corresponding to the photoinduced electrons transfer from the electrode to an external circuit, triggering a redox reaction with electron donors (D) by photoinduced holes; meanwhile, the inverse leads to a cathodic sensing via electron acceptors (A), trapping the excited electrons from the CB. Based on the different formation process of photocurrents at the interface, PEC sensing involves three main principles: one is that the analyte is directly used as an electron donor or acceptor, and feasible PEC detection is achieved through a simple redox strategy [37,38]; the other concerns observing the quantitative response change of the photocurrent via the indirect physical and chemical interaction between the targets and the photoactive materials modified by the recognition elements (such as aptamers [6,39,40,41], antibodies [42,43], and molecularly imprinted polymers [44,45]) to effectively indicate the concentration of the targets; the last one is the space effect-based sensing, i.e., that the analyte molecules interact with the photoactive materials to form a complex, contributing to a steric hindrance effect on the charge transform between interfaces and an enhanced recombination of e⁻-h⁺ pairs [46,47].

3. Photoactive Materials in PEC Sensors for Analysis of Toxins and Abused Drugs in the Environment

Photoactive materials act as a transducer converting the interactions between the analytes and PEC active materials, which play a crucial role in the PEC sensing system and demonstrate superior light capture capabilities, satisfactory charge separation efficiency, and appropriate energy level [37]. Excellent photoactive nanomaterials with high photoelectric conversion efficiency can greatly contribute to the excellent analytical performance of PEC sensors [48]. According to previous reports, photoactive materials can be divided into semiconductor and semiconductor-based heterojunctions. The semiconductor-based heterojunctions can effectively increase the photoelectric conversion efficiency of photoactive materials by means of heterojunctions, which usually have the following four categories, semiconductor–semiconductor heterojunctions (S–S heterojunctions), semiconductor–carbon heterojunctions (S–C heterojunctions), semiconductor–metal heterojunctions (S–M heterojunctions) and multi-component heterojunctions (MC heterojunctions), as shown in Figure 1. In addition, PEC sensing is based on anodic sensing as an example.

3.1. Semiconductors

Based on their good band gap alignment, controllable size, and excellent biological compatibility, a wide range of semiconductor materials, such as metal oxides [49,50,51,52], metal chalcogenides [53,54], Bi-based semiconductor materials [55,56,57], graphite-like carbon nitride (g-C₃N₄) [58,59], and organic polymers [60,61], have attracted enormous interest in the construction of PEC sensors for the analysis of toxins and abused drugs in the environment. However, owing to the short absorption threshold, high e⁻-h⁺ recombination rate, and mismatch of band structure, the photoelectron production efficiency is low, which seriously hinders the feasibility of a single semiconductor. For example, Li et al. developed a photoelectrochemical detector for the determination of organic compounds based on the excellent oxidation power of nanostructured TiO₂ photoanodes, which stemmed from the photo-hole generated under UV illumination [62]. Chen et al. fabricated a novel, simple, high-sensitivity, and high-selectivity MC–LR photoelectrochemical sensor on highly ordered and vertically aligned TiO₂ nanotubes with molecularly imprinted polypyrrole (PPy) as the recognition element under a UV lamp [49]. Highly ordered and vertically aligned TiO₂ nanotubes exhibit a sensitive photo-oxidation current due to their outstanding photocatalytic ability. PPy can recognize the MC–LR targets via the multiple hydrogen bond interactions. Then, the π bond delocalized electron system of PPy facilitates the separation of photogenerated electron–hole pairs and improves the photoelectric conversion efficiency that would enhance the PEC current response. Özcan et al. used commercial TiO₂-modified FTO electrodes for the photoelectrochemical determination of paracetamol under three UV fluorescent lamps [63]. As UV light only accounts for about 3–4% of solar energy on earth [64], it is necessary to reduce the band gap of single semiconductor, such as TiO₂ (3.2 eV) and ZnO (3.4 eV) [65], to make the photoactive materials exhibit high visible light responsive activity for the useful utilization of solar energy.

Element doping can be regarded as a simple and effective approach to adjusting the PEC performances of semiconductors [66,67], which narrows the bandgap energy of semiconductors and shifts the threshold wavelength to the visible light region because of changes in the nanocrystal electronic structure [68]. Wang et al. successfully grew novel Ti³⁺ self-doped branched rutile TiO₂ nanorod arrays (NRAs) on FTO transparent conductive glass [69]. The Ti³⁺ defects in the TiO₂ nanoparticles, the related oxygen vacancies, as well as the nanobranch structure, expended the optical absorption of the TiO₂ NRAs from 400 nm to 510 nm. The doped and branched TiO₂ NRAs with larger specific surface areas and stronger light absorption could significantly improve their photoelectrochemical properties. Li et al. constructed co-doped ZnO diluted magnetic semiconductor (DMS) for fabricating PEC aptasensors of acetamiprid (Figure 2a) [70]. On account of the sp-d exchange interactions between the band electron and the localized d electrons of Co²⁺, Co²⁺ doping is beneficial in expanding the optical absorption bandwidth of ZnO to the visible region and promoting the separation of photogenerated carriers. The linear range of the fabricated aptasensor for acetamiprid was 0.5–800 nmol L⁻¹, and the detection limit was 0.18 nmol L⁻¹. Zheng et al. fabricated a boron and graphene quantum dots-co-doped g-C₃N₄ (named GBCN) as a dopamine PEC sensor, which showed a low detection limit of 0.96 nM, a wide detection range of 0.001–800 μM, and good stability (no obvious photocurrent change was found both after 20 light source on/off cycles and 90-day storage), as shown in Figure 2b [71]. The excellent properties of the fabricated PEC-active material might be due to the following reasons: (1) GBCN has more dopamine active adsorption sites with the obviously increased specific surface area of g-C₃N₄ after co-doping; (2) co-doping makes the adsorption red shift from 470 to 540 nm and increases the absorption intensity; (3) the synergistic effect of boron and graphene quantum dots effectively promotes the migration of photogenerated electrons from the conduction band of g-C₃N₄ to graphene quantum dots, as well as facilitating charge separation.

Oxygen vacancies [48,72,73,74,75] and nitrogen vacancies [76] have also been reported to increase the charge separation efficiency and broaden the light adsorption range. Naldoni et al. synthesized novel crystalline core/disordered shell black TiO₂ nanoparticles (NPs) with a reduced bandgap of only 1.85 eV, which could fit for visible light absorption, as shown in Figure 3 [72]. The experimental data analysis shows that oxygen vacancies are present in the bulk anatase crystalline phase, while the disordered NP surface appears to be nearly stoichiometric. Whereas, the experimental data analysis also proves that the narrowing of the bandgap is caused by the synergistic existence of oxygen vacancies and surface disorder. Cui et al. utilized the synergistic effect between the oxygen vacancies concentration and lattice strain of BiOBr nanosheets in order to improve the photoelectric response [74]. Yan et al. developed a ciprofloxacin (CIP) PEC sensor by using nitrogen-deficient graphitic carbon nitride (ND-g-CN) as a PEC-active material [76]. Nitrogen vacancies can be served as charge traps to effectively suppress the charge recombination, which works in conjunction with the 2D thin-sheet structures of the available charge separation and transfer to increase light capture and enhance PEC performance. The detection range of the designed PEC sensor for CIP is 60 to 19,090 ng L⁻¹, and the detection limit is 20 ng L⁻¹.

3.2. Semiconductor-Based Heterojunctions

Semiconductor-mediated photocatalysis suffers from the drawbacks of short photogenerated electron–hole pair lifetime and the limited visible light absorption. Fabricating semiconductor heterojunctions has been one of the most widely used strategies over the past decade [77]. In general, the sensitivity range of PEC sensors is strongly dependent on the charge transfer of photoactive materials, which benefits from the high carrier density [78]. The formation of semiconductor-based heterojunction materials allows for the incorporation of materials that can adjust the bandgap energy, increase the charge carrier density, and then further suppress the recombination of photogenerated electron–hole pairs and attain their purpose of improving optoelectronic conversion efficiency.

3.2.1. Semiconductor–Semiconductor Heterojunctions

The Bi-based semiconductors, such as BiPO₄ [79,80,81,82], BiVO₄ [83,84,85], Bi₂WO₆ [86,87,88], and BiFeO₃ [89,90]_, are linked with the Bi 6s and O 2p hybrid orbitals of the narrowed bandgap as well as the deep valance band, which have received extensive attention in the study of PEC because of their efficient visible light photocatalytic activity [91]. Bi₂WO₆, as the simplest member of the Aurivillius family of layered perovskites, possesses good photon absorptivity under visible light irradiation [92]. However, unfortunately, it suffers from rapid photorecombination of carriers in a single component system. Adhikari et al. designed Bi₂S₃-sensitized Bi₂WO₆ microstructures topotactically via the hydrothermal technique under a controlled hydrothermal atmosphere and synthesized different weight percentage Bi₂S₃/Bi₂WO₆ microstructures for the evaluation of ofloxacin photo-oxidation under visible light [93]. In Figure 4b, after Bi₂S₃ is introduced into the Bi₂WO₆, we observe that a wide absorption band has been formed in the visible light range. The calculated energy bandgap is estimated to be 2.12 eV for 10 wt% Bi₂S₃/Bi₂WO₆. As shown in Figure 4a, both Bi₂S₃ and Bi₂WO₆ are excited to form a photogenerated electron–hole pair under the light illumination. The photogenerated electron in Bi₂S₃ flows to the conduction band of Bi₂WO₆, associated with the migration of photogenerated holes in the valence band of Bi₂WO₆ to the valence band of Bi₂S₃ simultaneously. Due to the favorable band positions of Bi₂S₃ and Bi₂WO₆, the electron–hole recombination could be effectively separated, which is helpful to the high photocatalytic degradation of ofloxacin. Therefore, when ofloxacin is added in the system, competition between photogenerated electrons to flow through the system and oxidize ofloxacin occurs, causing the decreasing of the photocurrent. On the basis of the surface plasmon resonance effect of metallic Bi, Li et al. coupled Bi and CuS, a narrow bandgap semiconducting material, with Bi₂WO₆ to prepare the Bi-CuS/Bi₂WO₆ heterojunction with enhanced separation of photo-generated charge carriers, charge transfer, and absorption wavelengths under visible light [87]. Then, a sensitive label-free PEC sensor found on Bi-CuS/Bi₂WO₆ nanocomposite was fabricated for the determination of paracetamol (PA), which is a ubiquitous NSAID. When PA is put into the electrolyte, the isolated holes on Bi₂WO₆ VB enter CuS to oxidize PA and acquire electrons, thus continuously producing photoelectrons to enhance the photocurrent signals.

TiO₂ is the most widely used photoactive material, which has the advantages of strong photoactivity, good biocompatibility, low cost, and strong chemical stability in various environments [94,95]. Nevertheless, the light absorption in the visible region of TiO₂ is limited to a great extent because of the wide bandgap. TiO₂ can only absorb UV light (λ < 400 nm), which can kill biomolecules [96]. Therefore, the modification of TiO₂ with narrow-bandgap semiconductors is crucial for the PEC biosensing applications to not only enhance its absorption in the visible range but also to improve the lifetime of charge carriers. CdS, CuS, MoS₂, and Ag₃VO₄ are sensitive to the visible light excitation due to their narrow bandgap, about 2.27 eV, 2.1 eV, 1.9 eV, and 2.2 eV, respectively [73,97,98,99]. Wei et al. fabricated an ultrasensitive PEC immunoassay for the detection of MC–LR based on a sensitive PEC material of CdS/TiO₂ nanorod arrays (CdS/TiO₂ NRAs) to immobilize antigens [100]. The photogenerated holes of CdS NPs help to weaken the self-oxidation, limit the recombination of electron–hole pairs, as well as broaden the light absorption of TiO₂ NRAs then immobilize antigens as a visible light-driven material. The photocurrent values of the proposed immunosensor are linear from 0.005 to 500 μg L⁻¹ concentration of MC–LR, and the limit of detection is 0.001 μg L⁻¹. Tang et al. prepared a CuS–TiO₂ heterojunction composite via dispersedly depositing CuS NPs on TiO₂ nanospheres’ surface, and they then constructed a PEC aptasensor for sensitive detection of MC–LR in an aquatic environment with a detection limit of 2.0 × 10⁻⁵ nM and good stability, of whose initial value the photocurrent response at the PEC aptasensor retained ~91.8% after 10 days’ storage in 4℃ refrigerator (Figure 4c) [6]. As shown in Figure 4d, the absorption edge of CuS–TiO₂ heterojunction composite (~504 nm) is much longer than that of pure TiO₂ (~388.5 nm). The intensified visible light absorption works together with the improved separation of photogenerated charges and the reduced self-aggregation of CuS nanoparticles to obtain the superior analytical performance for MC–LR. Fan et al. developed a highly sensitive atrazine PEC aptasensor by exploring TiO₂ nanotubes decorated with MoS₂ quantum dots (MoS₂ QDs/TiO₂ NTs) as a photoelectrode [47]. Because the visible light absorption range of TiO₂ was effectively extended from UV to visible light by MoS₂ QDs, the bioactivity of the photoelectrode was well maintained. Therefore, the photoelectrode exhibits high binding capacity of aptamer molecules to targets, which helped the designed atrazine PEC aptasensor achieve high sensitivity and specificity. Zhang et al. obtained a heterojunction of Ag₃VO₄ nanoparticles-deposited TiO₂ nanotube arrays (TNTAs) on a Ti foil electrode, which cooperated with oxygen vacancies of TNTAs caused by the electrochemical reduction treatment to greatly improve the visible light absorption of TNTAs, as well as the charge separation of TNTAs [73].

In addition, other new heterojunctions have also been reported as photosensitive material for the construction of PEC sensors for the analysis of toxins and abused drugs in the environment, such as the CuCo₂O₄@CoO core–shell hybrid rod for selectively sensing diclofenac [37], ZnPc/CN (zinc phthalocyanine/graphitic nitride) nanocomposite for preparing sulfadimethoxine aptasensor [101], hollow ZnS–CdS nanocage-based molecularly imprinted (ZnS–CdS/rMIP) PEC sensor for sensitive measurement of oxytetracycline [54], CdS nanorod arrays (CdS NRs)/Ag₂S on FTO for PEC immunoassay of ochratoxins, and so on. The report results show that the semiconductor–semiconductor heterojunctions usually exhibit significantly improved photocurrent response compared with the pure semiconductor, which could help to increase the analytical performance of the PEC sensors remarkably.

3.2.2. Semiconductor–Carbon Heterojunctions

Carbon materials, such as graphene, reduced graphene oxide (rGO), graphene quantum dots (GQDs), graphite-phase carbon nitride (g-C₃N₄), carbon nanotubes (CNTs), and so on, all show the inherent advantages of strong chemical stability, good biocompatibility, low toxicity, superior electron mobility, and high specific surface area [30,102]. Therefore, they are usually combined with other semiconductor materials by forming nanohybrids to fabricate high performance PEC sensors [103,104]. The dramatic analytical performances of the PEC sensors are achieved in the following ways: the adsorption of more electroactive species, shortening charge transport time and distance, acceleration of the transfer of photogenerated electrons, and suppression the recombination of charge carriers.

Graphene [105], N-doped graphene [3,48,78], and reduced graphene oxide (rGO) [106,107,108] are the most common carbon materials used to design semiconductor–carbon heterojunctions in PEC sensors. Because of the non-photoactivity and zero bandgap, graphene was usually employed as a carrier or electron transfer promoter for fabricating PEC sensors [109,110,111,112]. Yan et al. proposed a novel cathodic “signal-off” strategy for oxytetracycline PEC aptasensing based on BiOI–graphene nanocomposite [113]. In this sensor, p-type semiconductor BiOI was used as photoactive species, while doping BiOI with a certain amount of graphene can not only improve the absorption in the visible light region but also facilitate the electron transfer of the photoanode. Except pure graphene, N-doped graphene is also used in the PEC field according to the increased bandgap of graphene with the doped N element [114,115]. Zhang et al. designed a novel visible light-driven self-powered PEC aptasensor for MC–LR detection based on 3D N-doped graphene hydrogel/hematite nanocomposites (NGH/Fe₂O₃), as shown in Figure 5a [78]. Three-dimensional NGH could trap the holes to suppress the recombination, thus promoting the transmission of photo-induced electrons and increasing the conductivity. Therefore, the coupling NGH with Fe₂O₃ could generate a Schottky junction that could promote the separation of charges and increase the performance of the aptasensor with a wide linear range of 1 pM to 5 nM. Ge et al. synthesized Bi₂MoO₆ nanoparticles-anchored boron and nitrogen co-doped graphene (BNG) nanosheets nanocomposites with oxygen vacancy [48]. The incorporation of BNG nanosheets increased the oxygen vacancies, broadened the light adsorption range, and accelerated the charge transfer. So, a highly efficient PEC aptasensor for sensitive analysis of lincomycin was constructed based on the prepared Bi₂MoO₆/BNG photoactive nanocomposite. In addition to graphene and doped graphene, rGO has also been applied to PEC sensing for its improved water dispersibility and chemical reactivity. However, rGO is a p-type semiconductor with a narrow bandgap and relatively poor conductivity, so rGO is frequently coupled with other semiconductor materials in PEC sensing field [116,117]. Lu et al. developed a visible light-driven PEC aptasensor for MC–LR based on rGO/Ti–Fe–O nanotubes (NTs) [107]. It benefited from the large π-conjugated structure and unique electrical conductivity of rGO which endows a superhigh carrier migration rate. Therefore, the photo generated e^- can transfer to the surface of rGO rapidly, which effectively promotes the separation of photo-generated e⁻-h⁺ (as shown in Figure 5b). The fabricated aptasensor exhibits outstanding determination performance for MC–LR with a detection limitation of 5 fM.

Besides graphene-class materials, other carbon materials [38,39,40,118,119] are also used to synthesize semiconductor–carbon heterojunctions as photo-responsive materials for constructing PEC sensors. Liu et al. developed a MC–LR PEC sensor on a surface molecular-imprinted TiO₂-coated multiwalled carbon nanotubes (MI–TiO₂@CNTs) hybrid heterojunction nanostructure [119]. The sensor displayed a superior photocurrent response sensitivity and a wide linear range from 1.0 pM to 3.0 nM, which was mainly attributed to the high specific surface area and excellent photoelectric activity of the TiO₂@CNTs heterojunction nanostructure. As shown in Figure 5c, TiO₂ incurred charge separation under illumination, resulting in e⁻-h⁺. When TiO₂ NPs chemically bonded to the surface of CNTs, the electrons transferred from the CB of TiO₂ to the surface of CNTs due to their unique electrical conductivity and high electron storage performance. Therefore, the residual holes in TiO₂ VB could react with MC–LR that was absorbed on the electrode with MI recognition to produce an increased oxidation current. Gao et al. employed an electrophoretic deposition strategy to successfully fabricate a thin-film PEC sensor of functionalized N-doped graphene quantum dots embedded in BiOBr nanosheets that were loaded on ITO substrate (F-NGQDs/BiOBr/ITO) for paracetamol (PA) [38]. It was found that the thin film with NGQDs/BiOBr heterojunction showed enhanced charge transfer and absorption wavelengths under visible light, thus yielding an increase in the observed photocurrent. As shown in Figure 5d, the VB and CB positions of NGQDs are higher than those of BiOBr. The NGQDs are excited and undergo charge separation to generate e⁻-h⁺. Since BiOBr has different CB energy levels, photogenerated electrons are rapidly transferred to BiOBr, and then the separated holes on the VB of BiOBr can enter NCQDs and further oxidize PA to obtain electrons, thereby continuing to produce photoelectrons. Li et al. fabricated a novel bismuth vanadate/two-dimensional carbon nitride/deoxyribonucleic acid (BiVO₄/2D-C₃N₄/DNA) MC–LR aptamer PEC sensor with a detection sensitivity range of 5 × 10⁻⁷ to 10 μg L⁻¹ [120]. The photogenerated charge transport process schematic diagram is shown in Figure 5e. 2D-C₃N₄ plays a key role in multifunctional interface reconciliation. To be specific, 2D-C₃N₄ not only serves as a photogenerated hole-oriented transfer medium from the BiVO₄ photoanode to the detective target; it also fixes the DNA aptamer parallelly via the π–π bond to shorten the hole transfer distance from the semiconductor to the target.

Figure 5. (a) and (b) Mechanism diagram of the MC–LR aptasensors based on NGH/Fe₂O₃ Schottky junctions [78] and rGO/Ti–Fe–O NTs [107], respectively. (c) Photoexcitation and electron-charge transfer process of the MI–TiO₂@CNTs/FTO [119]. (d) Electron-transfer mechanism of PA PEC sensor based on F-NGQDs/BiOBr/ITO electrode [38]. (e) Schematic of photogenerated charge transport process on BiVO₄/2D-C₃N₄/DNA aptamer PEC sensor [120].

Figure 5. (a) and (b) Mechanism diagram of the MC–LR aptasensors based on NGH/Fe₂O₃ Schottky junctions [78] and rGO/Ti–Fe–O NTs [107], respectively. (c) Photoexcitation and electron-charge transfer process of the MI–TiO₂@CNTs/FTO [119]. (d) Electron-transfer mechanism of PA PEC sensor based on F-NGQDs/BiOBr/ITO electrode [38]. (e) Schematic of photogenerated charge transport process on BiVO₄/2D-C₃N₄/DNA aptamer PEC sensor [120].

3.2.3. Semiconductor–Metal Heterojunctions

The surface plasmon resonance (SPR) effect is an optical near-field effect on the localized surface of noble metal nanoparticles (e.g., Au, Ag, and Pd), resulting in enhanced local electromagnetic fields [121]. Thanks to this effect, the light absorption and photoelectric conversion efficiency of semiconductors can be greatly improved according to the excellent electron mediators of the noble metal nanoparticles through the rapid separation and transportation of photogenerated e⁻-h⁺ pairs at the semiconductor/metal interface [122]. Au NPs are the most commonly used noble metal [123,124]. Xu et al. constructed a facile and sensitive PEC sensing platform based on plasmonic Au/graphitic carbon nitride composites (Au/g-C₃N₄) for the detection of 4-chlorophenol (4-CP) [125]. As shown in Figure 6a, upon the illumination on the surface of Au/g-C₃N₄ composites, the photogenerated electrons of g-C₃N₄ can be quickly injected into Au. Due to the SPR effect of Au NPs, the interaction between Au NPs and incident light can generate a strong local electromagnetic field that induces plasmon oscillation on the local surface of Au NPs. Therefore, the photogenerated electrons are retained on the surface of Au NPs, which contributes to the effective separation of photogenerated e⁻-h⁺. At last, large number of electrons accumulating on the CB of g-C₃N₄ and on the surface of Au NPs can flow to the external circuit via the ITO substrate, resulting a significant enhancement in the photocurrent density. Upon 4-CP introduction, holes in the g-C₃N₄ VB can be sacrificed by 4-CP, which suppresses the e⁻-h⁺ recombination then further enhances the photocurrent density. Employing ZnCuInSe (Zn-doped CuInSe₂ quantum dots)/Au/TiO₂ nanowires sandwich structure, Geng et al. developed a facile PEC aptasensor for kanamycin detection with a linear response range from 0.2 to 250 nM [126]. The Au NPs, which were sandwiched between the TiO₂ nanowires and the ZnCuInSe QDs layer, play a dual role in enhancing the photocurrent, as shown in Figure 6b. One is to serve as a photosensitizer that extends the light absorption of the electrode to the visible range; the other is to act as an electron relay that promotes charge transfer between the TiO₂ nanowires and ZnCuInSe QDs. Zhang et al. fabricated an atrazine aptasensor based on a Au NPs-modified 3D-ordered microporous (3DOM) TiO₂ nanostructure frame with high sensitivity, selectivity, and stability [127]. The fabricated PEC aptasensor exhibits the detection limit of 0.167 ng/L, outstanding anti-interference ability in 100-fold concentration of other endocrine disrupting compounds, and stable photocurrent response after over 2000 s visible light illumination. As shown in Figure 6c, under the visible light illumination, the photogenerated electrons are excited to the CB of 3DOM and the holes left. Au NPs can also produce hot electrons and holes via the SPR effect. According to the energy level of hot electrons and hot holes of Au NPs, as well as the CB and VB of TiO₂, the hot electrons generated by Au NPs are easily transferred to the CB of TiO₂, and then both are transferred to the external circuit through the FTO to generate photocurrent. As a result, visible light absorption is effectively enhanced and the recombination of photogenerated carriers is efficiently decreased.

In addition to Au, Ag [128,129,130], Cu [131,132], Pd [133,134], Bi [135,136], Pt [137] are also applied to construct composites with semiconductors, which are used as the substrates for the PEC sensor. Wang et al. developed a 4-chlorophenol (4-CP) PEC sensor based on ternary composites of Ag NPs, graphitic carbon nitride (GCN), and carbon spheres (CS) (Ag/GCN/CS), which shows that an enhanced PEC response might be attributed to the synergistic effect between the SPR of Ag NPs and the electron-transfer behavior of CS [128]. The plausible PEC mechanism of the Ag/GCN/CS/ITO is discussed in two incident light spectral regions of 420–550 nm and 550–800 nm, considering the absorption edge of pure GCN (about 550 nm), as shown in Figure 6d. Under the illumination of 420–550 nm, Ag NPs, as electron sinks, can capture the electrons generated by the excited GCN, thereby inhibiting e⁻-h⁺ recombination. Meanwhile, in the wavelength range of 550–800 nm, a strong oscillation of electrons occurs on the surface of Ag NPs according to the SPR effect, resulting in the transfer of hot electrons from the surface of Ag NPs to the CB of GCN, which can effectively reduce the recombination rate of electrons and holes. Xu et al. prepared a nanocomposite of Bi-doped ultrathin polymeric carbon nitride (Bi/CV-PCN) with carbon vacancies (CVs), displaying high PEC activity and stability, to utilize as the photoactive material for fabricating an enrofloxacin (ENR) aptasensor. As shown in Figure 6e, the electrons in the VB of PCN are excited to the CVs under visible light irradiation, following further transition to the CB of PCN. Since the CB of PCN is more negative than the Fermi level of Bi, the electrons are subsequently transferred from PCN to metal Bi. The addition of Bi metal plays two roles in improving the light trapping ability of PCN, besides accelerating the separation of photogenerated charge carries according to the built-in electric field originating from the SPR effect of Bi. At last, the holes left in the VB of PCN directly oxidize the ENR to produce photocurrent. The synergistic effect of the CVs that form an intermediate energy level, and the SPR effect of Bi, effectively boost the visible light harvesting ability and improve charge separation efficiency, finally resulting in a remarkable improvement on the PEC performance.

Figure 6. (a) Photocurrent generation mechanism schematic diagram of Au/g-C₃N₄/ITO electrode [125]. (b) Dual role of the Au nanoparticles in the ZnCuInSe/Au/TiO₂ sandwich structure serving as a plasmonic photosensitizer and an electron relay [126]. (c) The mechanism of atrazine PEC sensor based on Au NPs/3DOM TiO₂/FTO [127]. (d) The plausible PEC mechanism of the Ag/GCN/CS/ITO in different spectral regions of (i) 420–550 nm and (ii) 550–800 nm [128]. (e) Electron-transfer mechanism of ENR sensing based on the Bi/CV-PCN composite [136].

Figure 6. (a) Photocurrent generation mechanism schematic diagram of Au/g-C₃N₄/ITO electrode [125]. (b) Dual role of the Au nanoparticles in the ZnCuInSe/Au/TiO₂ sandwich structure serving as a plasmonic photosensitizer and an electron relay [126]. (c) The mechanism of atrazine PEC sensor based on Au NPs/3DOM TiO₂/FTO [127]. (d) The plausible PEC mechanism of the Ag/GCN/CS/ITO in different spectral regions of (i) 420–550 nm and (ii) 550–800 nm [128]. (e) Electron-transfer mechanism of ENR sensing based on the Bi/CV-PCN composite [136].

3.2.4. Multi-Component Heterojunctions

Considering the synergistic effect, multi-component heterojunctions composed of visible light active components and electron transfer systems are spatially integrated for the purpose of simultaneously achieving an enhanced visible light response and transfer of charge carriers. Semiconductor, carbon material, and noble metal nanoparticles amount to a good combination, which effectively combines the light absorption and electron transfer ability and SPR effect [128]. Ouyang et al. designed a heterojunction based on coupling CeO₂ quantum dots (QDs) with graphitic carbon nitride (g-CN) and Au nanoparticles (ACG) for fabricating a self-powered PEC MC–LR aptasensor with a detection limit of 0.01 pM [40]. As shown in Figure 7a, because of the combination of CeO₂ QDs and the g-CN nanosheet into a highly dispersed nanocomposite, the matched band potentials and closely contacted interface between them can efficiently suppress the recombination of photo-generated electron–hole pairs and promote charge separation, which benefits the as-obtained PEC aptasensor to exhibit excellent analytical performances. The Au nanoparticles not only provide many binding sites for MC–LR aptamers through Au-S bond; they also greatly increase the light absorption capacity because of the SPR effect. Similarly, the CoO/Au/g-C₃N₄ Z-scheme heterojunction [138], Au nanoparticles, and graphene-doped CdS composite [139] were designed to fabricate the PEC sensor.

Semiconductorsand dye molecules-co-sensitized TiO₂ is also a good combination. The dye molecules can be adsorbed on the surface of semiconductors, increasing their photo response owing to the greater absorption of the light in the visible region [140]. Lima et al. used TiO₂ to co-sensitize with N-methylphenazonium methyl sulfate (PMS) and CdTe QDs; the composite was then immobilized in a Nafion film that severed as the photoelectroactive element for the development of a PEC sensor for the detection of tannic acid [141]. PMS was exploited as a co-modifier of the TiO₂, aimed to produce a stable materialthat photosensitive to visible LED light in this study. Sousa et al. fabricated a naringin PEC sensor on a zero-biased fluorine-doped tin oxide (FTO) electrode modified with CdS and chloroprotoporphyrin IX iron(III) (CPPI)-sensitized TiO₂, designated CPPI–TiO₂/CdS/FTO [142]. The proposed mechanism for the naringin PEC detection is shown in Figure 7b. The CPPI adsorbed on TiO₂ particles can absorb light to produce photoexcited states (CPPI^*) then inject electrons into the conduction band of TiO₂, resulting, simultaneously, in an oxidized form of CPPI⁺. On the other hand, the CdS can also be excited by visible light to produce electrons and holes in the CB and VB, respectively. Therefore, the photoelectron at the CB of TiO₂ can react with the holes at the VB of CdS via a mean z-scheme. Eventually, the photoelectrons on the CB of CdS can be collected from the FTO electrode to generate an anodic photocurrent. In addition to the above-mentioned multi-component heterojunctions, other combinations have also been reported for the analysis of toxins and abused drugs in the environment, such as CdS/Fe₂O₃ co-sensitized TiO₂ nanorod arrays [143], TiO₂/CNTs/CdTeQDs [144], In₂O₃−In₂S₃−Ti₃C₂ composite [145], and so on.

Figure 7. (a) The mechanism of MC–LR PEC aptasensor based on ACG heterostructure [40]. (b) Proposed mechanism for the naringin photoelectrochemical detection based on the CPPI–TiO₂/CdS platform. CPPI* means the photoexcited states of CPPI [142].

Figure 7. (a) The mechanism of MC–LR PEC aptasensor based on ACG heterostructure [40]. (b) Proposed mechanism for the naringin photoelectrochemical detection based on the CPPI–TiO₂/CdS platform. CPPI* means the photoexcited states of CPPI [142].

4. Summary and Outlook

Due to the high sensitivity and low cost of PEC sensors, numerous PEC sensors have been fabricated for the rapid and accurate monitoring of toxins and abused drugs in the environment. It was found that the photoelectric conversion efficiency was obviously improved via the innovative design and controllable preparation of photoactive materials. This review mainly summarizes the composition of the engineered photoactive materials, the sensing mechanisms, and actual applications of the fabricated PEC sensors.

Although great progress in the PEC sensing of toxins and abused drugs in the environment has been made by far, there is still plenty of room for development and further applications in the future. Firstly, multi-signal amplification strategies and multi-component heterojunctions are usually used to construct PEC sensors, which might increase the fabrication difficulty and cost, as well as reduce the stability of the PEC sensors. Therefore, it is of great importance to develop highly efficient signal amplification strategies and to prepare high-quality and environmentally friendly photoactive heterojunctions because employment of the often-used, traditional QDs (e.g., CdS) would lead to a second pollution. In addition, most of the reported PEC sensors are focused on the quantitative single-target analysis. A broad–specific PEC immunosensor was constructed for the simultaneous detection of ochratoxin A, ochratoxin B, and ochratoxin C because of the electronegativity of the chlorine atom that increases the antibody binding ability. Considering the difficulty of broad-spectrum antibodies, research can focus on developing an aptamer with multiple recognition sites for multiple-target detection simultaneously. Moreover, integrating PEC sensors with microfluidics and chips is also a good strategy by which to increase detection efficiency via multi-channel simultaneous detection. A microfluidic system with a paper working electrode combined with the PEC sensing technique is also an intelligent design that could facilitate miniaturization and assist integration into portable devices. Furthermore, qualitative detection of unknown toxins and abused drugs in the environment is also urgently needed. The designing of photoactive materials with specific sites and coupling PEC sensing with other methods, such as cyclic voltammetry (CV) and differential pulse voltammetry (DPV), have been reported as effective ways. However, a highly stable light source is usually required to excite the PEC sensors, which also greatly limits their practicality. Consequently, solar-powered and self-powered PEC sensors combined with energy conversion devices are being developed.