Introduction

Heterogeneous photocatalysis is considered as the potential strategy for water decontamination due to its high efficiency, environmental benignity, cost effectiveness, and potential utilization of solar energy1,2. The photocatalytic reaction mechanisms can be mainly divided into radical and non-radical pathways. Radical-type photocatalysts can generate hydroxyl radical (•OH) under light irradiation and form a gradient distribution of radical at the solid-liquid interface3. Meanwhile, the reactivity of •OH is independent of its production method, leading to its high reactivity and the universality of organic degradation1. In contrast, the holes generated by non-radical-type photocatalysts are only localized on the surface to undergo redox reactions. Compared with radical systems, the reactivity of non-radical-type photocatalysts towards organic molecules is more complicated. In particular, BiVO4, as a typical non-radical-type photocatalyst, exhibits inhibited surface kinetics for the degradation of organic molecules including phenol, ciprofloxacin, and sulfonamides4,5,6,7. However, although the photogenerated hole of BiVO4 has higher redox potential (ca. +2.5 V vs. NHE)8, the active species [•OH (+1.8 ~ 2.7 V vs. NHE)9, high-valent metal-oxo10,11,12 and singlet oxygen (1O2, +1.88 V vs. NHE)13 with lower redox potentials generated in advanced oxidation processes exhibit higher rate constant for organics3,14. It suggests a contradiction between the reactivity of non-radical-type photocatalytic system for organics and the ideal photocatalytic model controlled by thermodynamics. Generally, similar to the overpotential theory, the redox potential of photogenerated holes should be positively correlated with the oxidation reaction trend of organic molecule. Therefore, under the satisfied thermodynamic conditions, the reason why organic molecules are difficult to be degraded by non-radical-type heterogeneous photocatalytic system may be the limited interface reaction kinetics. According to kinetic theory, solvation is a key influencing factor in aqueous reactions. Among them, the stabilization of the reactants by the solvation effect will significantly increase the activation energy of the reaction. In particular, the holes of non-radical photocatalysts belong to short-range active species, which can only oxidize the molecules on the surface. The results of first-principles and ab initio molecular dynamics studies show that water molecules mainly exist in two interface interaction modes on the most stable (010) plane of BiVO4: water molecules coordinating with Bi3+ sites through lone pair electrons and interacting with oxygen terminal of VO4 tetrahedron through hydrogen bonds15,16,17. Strong BiVO4-water interface interaction may significantly reduce the reaction rate of photogenerated holes with organic molecules. However, the influence of solvation effects on the degradation efficiency of organics in non-radical-type heterogeneous photocatalytic systems is seriously neglected. Therefore, it is necessary to further explore the effect of solid–liquid interface on the organic degradation kinetics in the non-radical-type photocatalytic system and decrease the interfacial electron transfer barrier from organic molecule to photocatalyst.

In various strategies to improve interfacial electron transfer, sulfite can effectively trap the photogenerated holes of the photocatalyst to generate active radicals (e.g., SO3•–, SO5•–, SO4•‒, •OH), which can promote the gradient distribution of active species at the solid-liquid interface and the degradation of organic molecules18,19. For example, Deng et al. found that BiOBr combined with 20 mM sulfite could completely degrade 20 mg L−1 methyl orange within 30 min under visible light20. Wei et al. combined g-C3N4 with sulfite under visible light to significantly improve the removal performance of organic dyes and phenol21. However, in previous reports, the solid-liquid interface structure, interaction mode between sulfite and photocatalysts, and the photocatalytic activation law of sulfite have not been thoroughly elucidated.

Sulfamethoxazole (SMX) are the oldest group of antibiotics used in human and veterinary medicines for bacterial infections treatment. Considering its polarity, amphotericity, photo- and thermal stability, SMX has a high migration ability in the environment, which has become an emerging issue and potential threat to marine life and human health. Searching for suitable methods with high-efficiency degradation of SMX is one of the most challenging missions. Herein, we achieved sulfite-mediated indirect interfacial electron transfer between BiVO4 and organic molecules. The differences in interfacial interaction and electron transfer of sulfite and SMX on the surface of BiVO4 were explored by electrochemistry and theoretical calculations, while their existence mode at the solid-liquid interface was revealed. Furthermore, the chain transformation of sulfite at the solid-liquid interface was confirmed by quenching experiments and electron spin resonance (ESR). Finally, combined with semiconductor physics, interfacial chemistry, and double-layer theory, the role of sulfite in the interfacial electron transfer of the non-radical BiVO4 photocatalytic system is elaborated.

Results

Oxidation kinetics of SMX in the Na2SO3/BiVO4 system

According to the structural characterization (Supplementary Figs. 1–5), BiVO4 is a monoclinic decahedral structure with exposed (010)/(110) facets, which has an excellent visible light response. However, no change in SMX concentration is recorded in the presence of BiVO4 under visible light (Fig. 1), which indicates the limited kinetics of SMX degradation on the surface of BiVO4. Furthermore, the photolysis of SMX in the absence of photocatalysts can be neglected, due to its lack of visible light absorption22. Since the chain transformation of sulfite (Na2SO3) is difficult to trigger under visible light, SMX is not degraded in the presence of sulfite alone. In contrast, once 0.5 mM sulfite is added to the photocatalytic system of BiVO4, SMX exhibits a rapid degradation with a removal efficiency of 85.6% within 15 min. Therefore, it is necessary to further explore the sulfite-mediated mechanism and predominant active species in the Na2SO3/BiVO4 system.

Fig. 1: Kinetics of interfacial reaction of BiVO4 activated by sulfite.
figure 1

Degradation of SMX by BiVO4, Na2SO3, Na2SO3/BiVO4, and blank systems under visible light. Experimental conditions: [SMX] = 4 μM, [Na2SO3] = 0.50 mM, [BiVO4] = 1.0 g L−1, initial pH = 7.0, λ ≥ 420 nm, and T = 25 °C. The center of the data point is the mean of the two sets of data, and error bars represent the standard deviation of the two sets of data.

Full size image

Mechanism on sulfite activation in the system

According to the activation mechanism, sulfite is converted to a SO3•– via one-electron oxidation (Eq. 1)20. Since the ground-state O2 is a double-radical with triplet state, it will rapidly add to the SO3•– to form SO5•– (Eq. 2)18,23,24. Subsequently, SO5•– is reduced by SO3•– into peroxymonosulfate (PMS) and SO4•– (Eqs. 3, 4)23. Among them, SO4•– with the high redox potential can further oxidize H2O/OH- to generate •OH (Eqs. 5, 6)23. Therefore, SO3•–, SO5•–, SO4•– and •OH may be formed in the Na2SO3/BiVO4 system.

$${{SO}}_{3}^{2-}+{h}^{+}({{BiVO}}_{4})\to {{SO}}_{3}^{{{\bullet }}-}$$
(1)
$${{SO}}_{3}^{{{\bullet }}-}+{O}_{2}\to {{SO}}_{5}^{{{\bullet }}-}$$
(2)
$${{SO}}_{5}^{{{\bullet }}-}+{{SO}}_{3}^{2-}\to {{SO}}_{4}^{{{\bullet }}-}+{{SO}}_{3}^{2-}$$
(3)
$${{SO}}_{5}^{{{\bullet }}-}+{{SO}}_{3}^{2-}\to {{SO}}_{5}^{2-}+{{SO}}_{3}^{{{\bullet }}-}$$
(4)
$${{SO}}_{4}^{{{\bullet }}-}+{H}_{2}O\to {{SO}}_{4}^{2-}+{{\bullet }}{\rm{OH}}+{H}^{+}$$
(5)
$${{SO}}_{4}^{{{\bullet }}-}+{{OH}}^{-}\to {{SO}}_{4}^{2-}+{{\bullet }}{\rm{OH}}$$
(6)
$${{SO}}_{4}^{{{\bullet }}-}+{{SO}}_{3}^{2-}\to {{SO}}_{4}^{2-}+{{SO}}_{3}^{{{\bullet }}-}$$
(7)
$${{\bullet }}{\rm{OH}}+{{SO}}_{3}^{2-}\to {{SO}}_{3}^{{{\bullet }}-}+{{OH}}^{-}$$
(8)

Initially, the photocatalytic degradation of SMX by BiVO4 in the presence of sulfite under different atmospheres is carried out. From Fig. 2a, SMX has the highest degradation performance under the condition of air purging, while the dissolved oxygen (DO) concentration is rapidly consumed to the lowest at the initial reaction stage (Supplementary Fig. 6). After that, due to stirring and air purging, the DO concentration gradually increases. According to the sulfite activation mechanism, O2 only can be consumed by SO3•–, implying that sulfite is rapidly oxidized at the initial reaction stage. Furthermore, the degradation rates of SMX slightly decrease without air purging, indicating that higher DO concentration is conducive to the production of active species. In addition, under anaerobic conditions (Ar purging), the degradation of SMX can be neglected, demonstrating that O2 acts as the initiator of the active species generation and SMX is difficult to be oxidized by SO3•–.

Fig. 2: Dominated active species for SMX degradation in the Na2SO3/BiVO4 system.
figure 2

a Degradation of SMX in different atmosphere conditions. Effects of b TBA and c EtOH on SMX degradation in the Na2SO3/BiVO4 system; d ESR signals for the detection of reactive species generated in the Na2SO3/BiVO4 system. Experimental conditions: a [SMX] = 4 μM, [Na2SO3] = 0.50 mM, [BiVO4] = 1.0 g L−1, [TBA] = 1–100 mM, [EtOH] = 1–100 mM, [DMPO] = 100 mM, initial pH = 7.0, λ ≥ 420 nm, and T = 25 °C. The center of the data point is the mean of the two sets of data, and error bars represent the standard deviation of the two sets of data.

Full size image

To further investigate the active species generation, quenching experiments and ESR were performed. Since SO4•– are more sensitive to α-H, the rate constant (k) of tert-butyl alcohol (TBA) and ethanol (EtOH) with SO4•– [\({k}_{{{\rm{TBA}},{\rm{SO}}}_{4}^{{{\bullet }}{{\mbox{-}}}}}=(4.0-9.1)\times {10}^{5}{M}^{-1}{s}^{-1}\), \({k}_{{{\rm{EtOH}},{\rm{SO}}}_{4}^{{{\bullet }}{{\mbox{--}}}}}=(1.6-7.7)\times {10}^{7}{M}^{-1}{s}^{-1}\)] are quite different, while •OH have less effect [\({k}_{{\rm{TBA}},{{\bullet }}{\rm{OH}}}=6.0\times {10}^{8}{M}^{-1}{s}^{-1}\), \({k}_{{\rm{EtOH}},{{\bullet }}{\rm{OH}}}=1.9\times {10}^{9}{M}^{-1}{s}^{-1}\)]11,25. As observed in Fig. 2b, the addition of TBA (1 mM and 10 mM) just slightly retarded the oxidation of SMX. Significant inhibitory action can be observed only when TBA concentration is increased to 100 mM. In contrast, there is a strong inhibition with the addition of 1 mM EtOH, and the inhibitory effect gradually enhance with the increase of EtOH concentration (Fig. 2c). Therefore, SO4•– may be the major active species in the Na2SO3/BiVO4 system (\({k}_{{{\rm{SMX}},{\rm{SO}}}_{4}^{{{\bullet }}{{\mbox{-}}}}}=0.93\times {10}^{9}{M}^{-1}{s}^{-1}\))26, while •OH play an auxiliary role (\({k}_{{\rm{SMX}},{\rm{\bullet }}{\rm{OH}}}=6.78\times {10}^{9}{M}^{-1}{s}^{-1}\))12. In addition, ESR was carried out to probe the production of active species (Fig. 2d). Strangely, there is a distinct signal of DMPO-•OH in the BiVO4 system under visible light, which opposite to the degradation of SMX in the BiVO4 system (\({k}_{{\rm{SMX}},{{\bullet }}{\rm{OH}}}=6.78\times {10}^{9}{M}^{-1}{s}^{-1}\)). This may originate from the hole-assisted H2O nucleophilic attack for DMPO (Supplementary Fig. 7a), but does not actually generate •OH25. When sulfite is added to BiVO4 suspension, the signal of DMPO-•OH was further enhanced, indicating that the introduction of sulfite significantly promoted the generation of strong oxidative radicals. However, the DMPO-SO4•– adduct signal is not clearly detected, which may be attributed to the poor stability of DMPO-SO4•– and easy decomposition to DMPO-•OH (Supplementary Fig. 7b)25. In addition, SO5•– are also produced during the chain transformation of sulfite. However, according to previous reports, SO5•– has weak reactivity, indicating that SO5•– almost did not take part in the degradation process of SMX23. Therefore, we can conclude that SO4•– is the predominant active species for SMX degradation in the Na2SO3/BiVO4 system, while the contribution from •OH, SO5•–, SO3•– is driven to a lesser extent.

The mechanism of sulfite activation by BiVO4 under visible light was further explored by control experiments. It could be seen from Fig. 3a that with the increase of BiVO4 concentration, the degradation performance of SMX is gradually improved, implying that the higher reactive surface area of BiVO4 is conducive to the increase of SO4•– generation proportion. Meanwhile, there is an observe obvious wavelength dependence of light irradiation in the Na2SO3/BiVO4 system (Fig. 3b). This due to the blue-shift of the cutoff wavelength increasing the carrier concentration of BiVO4. Moreover, the degradation performance of SMX display a volcanic relationship with the change of sulfite concentration (Fig. 3c). This may originate from that excessive sulfite will quench the active species (e.g., SO4•– and •OH) (Eqs. 7, 8). In summary, under the consistent of sulfite concentration, the increase of BiVO4 concentration and the blue-shift of the cutoff wavelength promotes the higher proportion of SO4•– production. Meanwhile, there is an optimal dosage of sulfite under the consistent of the BiVO4 concentration and light irradiation conditions. Wherein, insufficient sulfite reduced the yield of active species, and excessive sulfite produced a stronger self-quenching effect. According to DO (Supplementary Fig. 6) and pH (Supplementary Fig. 8) change of solution, it can be seen that a large amount of sulfite is consumed at the initial reaction stage. This indicates that the higher proportion of sulfite oxidation at the initial stage will be conducive to the generation of strong oxidative radicals (SO4•– and •OH). In addition, the Na2SO3/BiVO4 system presents obvious pH dependence (Fig. 3d), and displays highest degradation performance for SMX under neutral conditions. Wherein, under acidic conditions, the protonation of sulfite enhances its ionization potential and inhibits its activation (Supplementary Table 1). It could be seen from Supplementary Figs. 9, 10 that under alkaline conditions, BiVO427, SO32- and SMX are all negatively charged, which increases the electrostatic repulsion among BiVO4, SMX, and SO4•–, resulting in inhibition of SMX degradation.

Fig. 3: Controlled experiment.
figure 3

Effect of a BiVO4 concentration, b light irradiation range, c Na2SO2 concentration, and d pH on the degradation of SMX. Experimental conditions: [SMX] = 4 μM, [Na2SO3] = 0.50 mM, [BiVO4] = 1.0 g L−1, initial pH = 7.0, λ ≥ 420 nm, and T = 25 °C. The center of the data point is the mean of the two sets of data, and error bars represent the standard deviation of the two sets of data.

Full size image

Investigation of the interface reaction

The interfacial reaction of sulfite on the surface of BiVO4 was further investigated by chopping LSV. As revealed in Fig. 4, the current density of BiVO4 gradually enhances with the increase of positive bias voltage under visible light, which displays the excellent photoelectrochemical properties of BiVO4. However, when SMX was added to the solution, the current density of BiVO4 is obviously suppressed, indicating that SMX is difficult to undergo oxidation by direct electron transfer on the surface of BiVO4. Interestingly, the current curve of BiVO4 is drastically red-shifted and its current density is essentially increased with the addition of sulfite, which indicates that the holes of BiVO4 can be effectively trapped by sulfite. When SMX was further added to the electrolyte, the current density is further enhanced, indicating that SMX, as a receptor of SO4•–, significantly promoted the interface electron transfer of BiVO4.

Fig. 4: Photoelectrochemical experiment.
figure 4

Photocurrent linear sweep voltammetry (LSV) scan of BiVO4 photoanodes with Na2SO3 and SMX (0.1 M Na2SO4, [Na2SO3] = 0.5 mM, [SMX] = 4 μM, pH = 7.0) under chopped illumination (λ = 510 ± 60 nm).

Full size image

Mechanism of interface reaction in the BiVO4 photocatalytic system

The photoexcitation of BiVO4 and the interfacial interaction between sulfite and BiVO4 were investigated by theoretical calculations. Figure 5a and Supplementary Fig. 11 illustrates the valence band (VB) of BiVO4 is mainly composed of hybrid orbitals of Bi6s and O2p, while its conduction band (CB) is mainly composed of V3d, indicating that the Bi-O structure acts as the region where holes principally accumulate28. Meanwhile, the valence band top of BiVO4 composed of Bi6s orbitals reduces the effective mass of holes and increases the mobility, which effectively promotes the surface oxidation reaction. Fortunately, the (110) plane of BiVO4 is mainly composed of Bi-O layers, forming the hole gathering plane (Fig. 5c). As shown in Fig. 5c and Supplementary Table 2, sulfite has higher adsorption energy and charge transfer amount on the surface of BiVO4 compared with SMX (Supplementary Fig. 12). In addition, sulfite has a lower adiabatic ionization potential (AIP) than that SMX (Supplementary Table 1), indicating that sulfite is more prone to one-electron oxidation. In summary, there is a strong coupling effect between adsorption and single-electron oxidation of sulfite on the surface of BiVO4.

Fig. 5: Electronic structure and solid–liquid interface reaction mechanism of BiVO4.
figure 5

a DOS calculation for BiVO4; b the proposed photocatalytic decomposition process of SO32− on the BiVO4; c adsorption of monodentate SO32− on the BiVO4 with (110) facet; d quasistatic energy profile and charge transfer pathways of BiVO4 under continuous illumination in contact with the aqueous solution. Jredox is the target charge transfer from the valence band to the redox reagent, SCL is space charge layer, Vph is the open-circuit photovoltage. EF,n and EF,p are the quasi-Fermi levels of electrons and holes under illumination, respectively; e model of the double-layer structure of BiVO4 in contact with the aqueous solution under equilibrium conditions; f transformation pathways of major species in Na2SO3/BiVO4.

Full size image

According to the above experimental and theoretical analysis results, the interaction and reaction of sulfite and SMX on the surface of BiVO4 can be summarized. The main active species for the degradation of organic molecules in the photocatalytic system include superoxide radical (•O2−), hydroxyl radical (•OH), and hole (h+). The contribution of •O2− to the degradation of organic molecules is excluded due to its weak oxidation, while •OH is not generated in the BiVO4 photocatalytic system. Therefore, photogenerated holes are the only active species in the BiVO4 photocatalytic system. In particular, h+ belonging to short-range active species can only oxidize molecules in the compact layer3. According to the ideal photocatalysis model, the non-radical pathway of the photocatalysis system includes diffusion of organic molecules to the photocatalyst surface, reaction with photogenerated h+, and desorption into bulk solution. Simultaneously, the thermodynamics of surface reaction is determined by the potential difference between the photogenerated h+ and organic molecules. The higher the overpotential, the stronger the reaction trend. According to previous reports, the top of the valence band of BiVO4 is ~2.5 V vs. NHE8. Compared with BiVO4, the active species [•OH (+1.8 ~ 2.7 V vs. NHE)9, high-valent metal-oxo10,11,12 and 1O2 (+1.88 V vs. NHE)13] with lower redox potential exhibit higher reactivity to organic molecules. And graphitic carbon nitride (g-C3N4), which is also a photocatalyst, has only the position of valence band of 1.5–2.0 V vs. NHE29,30, showing excellent SMX degradation performance31. Therefore, the limited SMX degradation reaction of BiVO4 might be controlled by kinetics. Especially, the identification of the difference between g-C3N4 and BiVO4 can better reflect the kinetic control effect in the heterogeneous photocatalytic system. From the structural point of view, the surface of BiVO4 is composed of oxygen terminal or hydroxyl group, because both Bi3+ and V5+ are extremely oxyphilic. However, g-C3N4 formed from thermal polymerization is relatively weak in hydrophilicity, and its structure is a heptazine-ring conjugated organic polymer linked by amino nitrogen32. Therefore, g-C3N4 lacking surface hydrophilic groups are less bound to surface water molecules. Meanwhile, according to our previous studies, pi-pi interactions can be formed between g-C3N4 and organic molecules with aromatic rings33,34. Among them, the pi-pi interaction can cause the organic molecules with aromatic rings to dislodge the water molecules in the compact layer to adsorb on the g-C3N4 surface, while the partial overlap of the wave function in space between organic molecule and g-C3N4 is conducive to their charge transfer. Therefore, g-C3N4 has general organic molecular degradation performance31. In contrast, the surface of BiVO4 composed of terminal oxygen or hydroxyl groups is more hydrophilic, leading to stronger interactions between BiVO4 and water molecules in the compact layer. Compared with water molecules, organic molecules have weaker polarity and larger volume. This results in a weaker interaction between organic molecules and BiVO4, and is not conducive to the diffusion of organic molecules from the bulk solution to the solid-liquid interface. Therefore, combined with the non-radical mechanism and surface structural features of BiVO4, it is difficult for organic molecules with low polarity and high ionization potential to enter the compact layer of BiVO4 and perform non-radical electron transfer. The steric segregation between organic molecules and holes triggers the deactivation of the photocatalytic system. This phenomenon is named surface solvation-induced inactivation (SSII). Compared with organic molecules, the wave function of sulfite in the form of anion is more diffuse in space, which is conducive to the formation of electrostatic interaction-dominated hydrogen bonds with the hydroxyl on the surface of BiVO4 and into the compact layer. Meanwhile, sulfites with low ionization energy are favorably oxidized by the holes of BiVO4 (Supplementary Tables 1 and 2), thereby promoting the generation of active species. Therefore, the introduction of sulfite realizes the indirect degradation of SMX by BiVO4, which expands the steric reaction ability of BiVO4 to organic molecules.

Based on above results, an overall model of Na2SO3/BiVO4 system from photoexcitation to surface reaction is proposed. As depicted in Fig. 5d, the contact of BiVO4 as an n-type semiconductor with water will lead to the formation of a space charge layer (SCL) and an upward energy band bend. When BiVO4 is excited under visible light, the quasi-Fermi level of holes (EF,P) drops significantly and generates photovoltage. Under the action of photovoltage and band bending, holes will grab electrons from substrate molecules with suitable redox potentials. Meanwhile, the BiVO4 will be solvated and form double layer (including compact layer and diffuse layer) mainly composed of solvent molecules (Fig. 5e). This severely hinders the interfacial interaction and electron transfer between substrate molecules and BiVO4. According to the above results, SMX can only be in the diffuse layer, while sulfite can enter the compact layer of BiVO4. Once sulfite is oxidized by the hole of BiVO4 with one electron, the generated SO3•− will then undergo chain transformation to generate active species and enter into the diffuse layer to participate in the oxidation of SMX (Fig. 5e, f).

Quantitative structure-activity relationships

In addition, the degradation of various pollutants by Na2SO3/BiVO4 system are further investigated. According to the above results, SO4•– is the predominant active species in the Na2SO3/BiVO4 system, while •OH play an auxiliary role. Among them, •OH are non-selective active species, while SO4•– have certain selectivity. It could be seen from Supplementary Fig. 13 that no significant degradation of phenol, aniline, and benzoic acid is observed in the BiVO4 photocatalytic system. However, the concentration of nitrobenzene exhibits a significant decrease, which might originate from the reduction of the nitro group by solvated electrons diffusing out of the tight layer. In contrast, the Na2SO3/BiVO4 system exhibited obvious photocatalytic degradation performance for various pollutants (Fig. 6a). Wherein, aromatic compounds with electron-donating substituents (phenol and aniline) present better degradation rate, while aromatic compounds with electron-withdrawing groups (benzoic acid and nitrobenzene) are more difficult to be degraded. To further explore the reasons for the differences in the degradation performance of various pollutants, a correlation between the adiabatic ionization potential (AIP) of organic molecules and the rate constant (k) is constructed. Generally, AIP can be used to describe the tendency of organic molecules to lose electrons, which is similar to the electron transfer in the oxidation of organic compounds by active species. As delineated in Fig. 6b, there is a good linear correlation between k and the AIP of various pollutants. In conclusion, Na2SO3/BiVO4 system is more inclined to the degradation of electron-rich organics, indicating the structure-dependent degradation of organic molecules.

Fig. 6: Quantitative structure-activity relationship.
figure 6

Degradation performance for a monosubstituted benzenes; b relationship between the k and AIP value. Experimental condition: [organic] = 4 μM, [Na2SO3] = 0.50 mM, [BiVO4] = 1.0 g L−1, initial pH = 7.0, λ ≥ 420 nm, and T = 25 °C. The center of the data point is the mean of the two sets of data, and error bars represent the standard deviation of the two sets of data.

Full size image

Discussion

In this work, the oxidation kinetics of organic molecular controlled by the solid-liquid interface in the BiVO4 photocatalytic system is revealed. Initially, the holes belong to short-range active species and are localized on the photocatalyst surface. Meanwhile, the generation of solid–liquid interface in the actual system triggers the spatial isolation between the photocatalyst and the organic molecules, which leads to the deactivation of the degradation of organic molecules on the surface of BiVO4. However, sulfites with lower AIP and higher adsorption energy on the surface of BiVO4 can accumulate in the compact layer of BiVO4 and trap the generated photogenerated hole for single-electron oxidation. The generated sulfite radical undergoes chain transformation under the induction of oxygen to generate active species (SO4•−) to degrade SMX. This results in a sulfite-mediated indirect electron transfer from SMX to BiVO4. A model of Na2SO3/BiVO4 system from photocatalyst excitation to interface reaction is established. In addition, there is a good linear correlation between the rate constant (k) and AIP of various pollutants in the Na2SO3/BiVO4 system, indicating the structure-dependence of organic molecules degradation. This study provides insight for further understanding the effect of solid-liquid interface on reaction kinetics in heterogeneous photocatalytic system.

Methods

Chemicals and materials

Sources of chemicals and materials are provided in Supplementary Method 1.

Preparation of BiVO4

Bismuth vanadate (BiVO4) was synthesized by a hydrothermal method as reported in the literature35. Typically, 12 mmol Bi(NO3)3 ∙ 5H2O was dissolved into 64 mL of 1 M HNO3. After stirring for 0.5 h, 12 mmol NH4VO3, and 0.1 mol urea were quickly poured into the above solutions, and the color of the precipitate gradually changed from blood red to bright yellow. Subsequently, the beaker was placed in an 80 °C oil bath for 24 h with magnetic stirring. After natural cooling, the yellow precipitation was collected by centrifugation, washed with ethanol (EtOH) and water repeatedly, and then dried at 60 °C overnight.

Experimental procedure

Batch experiments were carried out in a 100 mL glass beaker at 25 °C under magnetic stirring. In a typical test, 50 mg of BiVO4 was ultrasonically dispersed into the beaker containing 50 mL of 4 μM SMX, and then a certain mass of Na2SO3 was introduced under magnetic stirring. Subsequently, the initial pH was adjusted to 7.0 by using 0.05 M H2SO4 or NaOH. Meanwhile, the Xenon lamp (CEL-HXF300, Beijing China Education Au-light Co., Ltd) equipped with 420 nm cutoff filter was preheated at least 30 min to ensure the stability of the light source, and then the light intensity was adjusted to 400 mW cm−2 using an optical power meter (CEL-NP2000-10A, Beijing China Education Au-light Co., Ltd). At certain time intervals, 800 μL of the reacting solution was withdrawn, mixed with 800 μL of methanol to quench reactive species, filtered through a 0.22 μm PTFE filter, and then for further analysis. All experiments were conducted in duplicates at least, and the averaged values with standard deviations were reported.

Analytical methods

The concentration of organics was quantified by on a Waters e2695 equipped with a C18 column and a UV-vis detector with details shown in Supplementary Table 3. The pH values were monitored by a Shanghai Leici pH meter with daily calibration. The concentration of dissolved oxygen (DO) was measured by a JPB-607A portable meter (Leici, Shanghai, China). ESR signals of radical spin-trapped with 5,5-dimethyl-1-pyrroline N-oxide (DMPO) were obtained using a JES-FA200 (JEOL Co., USA) spectrometer equipped with a 300 W Xe lamp (420 nm filter) as visible light source. The material characterization techniques, trapping experiments, and photoelectrochemical measurements are given in the Supplementary Information (Supplementary Method 2–4).

Density functional theory (DFT) calculation analysis

The theoretical calculations of isolated and periodic systems were performed by Quantum Espresso (QE 6.5), Gaussian 16 C.01, and CP2K 9.1, respectively36,37,38,39. The CP2K 9.1 input file was generated through Multiwfn 3.8_dev40. Details on DFT calculation can be found in Supplementary Method 5.