Insight into the PEC and interfacial charge transfer kinetics at the Mo doped BiVO4 photoanodes

BiVO4 is a promising photoanode material for the photoelectrochemical (PEC) oxidation of water; however, its poor charge transfer, transport, and slow surface catalytic activity limit the expected theoretical efficiency. Herein, we have investigated the effect of Mo doping on SnO2 buffer layer coated BiVO4 for PEC water splitting. SnO2 and Mo doped BiVO4 layers are coated with layer by layer deposition through a precursor solution based spin coating technique followed by annealing. At 5% doping of Mo, the sample (SBM5) shows a maximum current density of 1.65 mA cm−2 at 1.64 V vs. RHEl in 0.1 M phosphate buffer solution under AM 1.5 G solar simulator, which is about 154% improvement over the sample without Mo (SBM0). The significant improvement in the photocurrent upon Mo doping is due to the improvement of various bulk and interfacial properties in the materials as measured by UV-vis spectroscopy, electrochemical impedance spectroscopy (EIS), Mott–Schottky analysis, and open-circuit photovoltage (OCPV). The charge transfer kinetics at the BiVO4/electrolyte interface are investigated to simulate the oxygen evolution process in photoelectrochemical water oxidation in the feedback mode of scanning electrochemical microscopy (SECM) using 2 mM [Fe(CN)6]3− as the redox couple. SECM investigation reveals a significant improvement in effective hole transfer rate constant from 2.18 cm s−1 to 7.56 cm s−1 for the hole transfer reaction from the valence band of BiVO4 to [Fe(CN)6]4− to oxidize into [Fe(CN)6]3− with the Mo doping in BiVO4. Results suggest that Mo6+ doping facilitates the hole transfer and suppresses the back reaction. The synergistic effect of fast forward and backward conversion of Mo6+ to Mo5+ expected to facilitate the V5+ to V4+ which has an important step to improve the photocurrent.


Introduction
Photoelectrochemical (PEC) splitting of water is one of the most promising methods for simultaneous conversion of hydrogen from water using solar energy as a sustainable and clean energy source and zero carbon footprint; the process has inherently high power and energy densities. 1-5 PEC water splitting consists of a photoanode and photocathode on which an oxygen evolution reaction (OER) and a hydrogen evolution reaction (HER) respectively are taking place. The overall water splitting reaction, however, is limited by the sluggish OER kinetics; therefore, it is important to develop efficient photoanodes for the improvement in the overall water splitting process. Some of the important materials under continued investigation as photoanode materials for the OER are TiO 2 , a-Fe 2 O 3 , WO 3 and BiVO 4 . 6-14 Among these materials, BiVO 4 is the most researched photoanode due to its suitable band position, bandgap and high theoretical efficiency ($7.5 mA cm À2 ) and high solar to hydrogen (STH) conversion efficiency ($9%). [15][16][17][18] However, the slow surface catalytic activity, short hole diffusion length, fast electron-hole recombination are major challenges with the BiVO 4 . 6,[19][20][21][22] To overcome these challenges, a number of the strategies have been widely investigated such as; nanostructure control, [23][24][25][26] band engineering, 17,27-34 heteroatom doping, [35][36][37][38][39][40][41] generation of oxygen vacancy 22,[42][43][44] and the oxygen evolution catalyst (OEC) incorporation. 18,45-53 SnO 2 has been used for the heterojunction formation in the BiVO 4 system which suppresses the back electron-hole recombination process. 17 Additionally, the SnO 2 underneath of BiVO 4 blocks the surface state of the ITO/FTO. These modications improve the injection of the photogenerated holes to the electrode-electrolyte interface. During OER at the photoanode surface, photogenerated holes are expected to react with the OH À to form OHc radical intermediate, which is converted to O 2 and very small part of them dimerize to H 2 O 2 or OH À . Some of the OHc intermediate diffuse out into the electrolyte. The charge transfer kinetics at the electrode-electrolyte interface is in the range of nanosecond, which is very difficult to investigate by the conventional electrochemical method. Electrochemical impedance spectroscopy (EIS) 54 and the transient absorption spectroscopy (TAS) 21 are two important techniques implied to investigate the PEC processes and generate important parameters like photoinduced carrier lifetime and diffusion length.
Scanning electrochemical microscopy (SECM) is a powerful technique to investigate the charge transfer kinetics for in situ measurements at the solid-liquid and liquid-liquid interfaces. [55][56][57] SECM is decisively applied for the investigation of the mechanism of interfacial charge transfer processes in OER, HER, oxygen reduction reaction (ORR) and hydrogen oxidation reaction (HOR) on the Pt, Pd, Au Hg, and the other electrodes. 56,58,59 Interfacial charge transfer kinetics at the semiconductor-electrolyte interface has been investigated for the photo-induced charge mediated reactions. 60 Bard group has demonstrated the detection, quantication, and evolution of decay kinetics of the photogenerated hydroxyl radicals in the PEC on the semiconductor interface. 61,62 Surface interrogation SECM (SI-SECM) technique has been utilized to quantify photogenerated hydroxyl radicals (ads) and dimerization of the photogenerated radicals at the photoanode.
Considering the shortcoming of BiVO 4 , present investigation is aimed to improve the catalytic efficiency by developing the BiVO 4 photoanodes with SnO 2 as interlayer, BiVO 4 was doped with Mo to further improvement of the catalytic activity. The SnO 2 coating over the ITO plate was carried out for suppressing the charge recombination process through the generation of heterojunction of SnO 2 and BiVO 4 . The optical, chemical, and electronic properties of the materials have been investigated to understand the improvement in the PEC efficiency on Mo doping. The improvements in the photocurrent on Mo doping are analyzed based on the relative improvements in the bulk and surface properties as measured by the EIS and Mott-Schottky analysis. The decrease in the charge transfer resistance (R ct ) shows the improvements in the bulk property of the BiVO 4 . The increase in the capacitance upon the Mo doping suggests better activity at the electrode-electrolyte interface due to the enhancement of the active surface sites, which leads to the enhancement in the PEC efficiency. 20 Strong correlation among the optical property of the material, the open circuit photovoltage (OCPV), and onset potential was discussed in relation with the improvement in the PEC efficiency on Mo doping. The increase in the at band potential and OCPV suggests the improvements in the charge separation upon the Mo doping which resulted in the enhancement in PEC efficiency. 63-66 SECM has been applied to investigate the photo-induced interfacial charge transfer kinetics in situ at the electrode-electrolyte interface, the interfacial photo-generated hole transfer kinetics was correlated with the efficiency of PEC process across different catalysts investigated.

Fabrication of photoanode
SnO 2 /BiVO 4 heterojunction was prepared by the spin coating technique. In this typical synthesis procedure, SnCl 4 (98%, 0.24 mL, 0.2 M) was dissolved in 10 mL of ethylene glycol and sonicated for 20 min and kept for stirring overnight prior to the spin coating. An aliquot of 100 mL as prepared precursor solution of SnO 2 was spin-coated on the ITO substrate at 2000 rpm for 1 min, followed by annealing at 250 C on the hot plate for 5 min. This process was repeated 8 times to get the optimum thickness of SnO 2 . 67 Aer spin coating; the modied electrodes were annealed at 450 C in a tube furnace for 2 h at 5 C per minute heating rate to form crystalline SnO 2 . Aer having the SnO 2 layer over the ITO substrate, the BiVO 4 lm was formed on the ITO/SnO 2 substrate by metal-organic decomposition method. In this typical synthesis method, Bi(NO 3 ) 3 $5H 2 O (0.2 mmol) was dissolved in 5 mL of ethylene glycol-water mixture (8 : 2, volume ratio) subsequently NH 4 VO 3 (0.2 mmol) was added slowly, the mixture was sonicated for 30 min and kept on stirring for overnight at room temperature. The above solution was spin-coated over ITO (for control experiments) and ITO/ SnO 2 substrate at 2000 rpm for 1 min and then annealed at 350 C for 5 min on the hot plate. The coating was carried out for repeated 8 times to achieve the appropriate thickness for maximum efficiency, 67 and aer completion of the coating, the samples were heated at 450 C for 3 h at the heating rate of 5 C per minute in a tube furnace. Mo doping was achieved by adding ((NH 4 ) 2 $MoS 4 , 99.95%) at 1, 3, 5 and 7 atom percentage, resulting in a mixture containing Bi/(V + Mo) ¼ 1 : 1 to replace V position in the crystal lattice of BiVO 4 . The photoanodes thus fabricated using without Mo and at 1, 3, 5 and 7 atom percentage of Mo are designated as SBM0, SBM1, SBM3, SBM5, and SBM7 respectively.

Photoelectrochemical measurements
Photoelectrochemical measurements were performed using the CH Instrument (920 D model) using a three-electrode cell with an Ag/AgCl (3.0 M KCl) reference electrode, glassy carbon rod as counter and modied ITO coated with the catalyst material as the working electrode. 0.5 M Na 2 SO 4 solution in 0.1 M potassium phosphate buffer solution (PBS, pH ¼ 7) was used as electrolyte. All photoelectrochemical studies were carried out using Ag/AgCl (3 M KCl) reference electrode and potentials were converted and reported to reference hydrogen electrode (RHE) using the following eqn (1).
where, E Ag/AgCl is working potential and E 0 Ag/AgCl is standard potential (i.e. 0.2243 V).
To measure the charge transfer efficiency and the charge transport efficiency, 0.1 M Na 2 SO 3 as hole scavenger was added in the electrolyte. All samples were front-illuminated because of signicantly higher photocurrent than that of back illumination. Solar simulator having 1 sunlight tted with AM 1.5 G lter was used as a light source. The xenon arc lamp is used as a monochromatic light source; power of the monochromatic light is measured by digital power meter from Newport. The photocurrent was measured by linear sweep voltammetry (LSV) technique with 5 mV s À1 scan rate and chopped light voltammetry was recorded. Chronoamperometry was used for the stability test of the photoanode materials. Electrochemical Impedance Spectroscopy (EIS) was used to measure interfacial charge transfer resistance (R ct ) at 1.44 V by applying sinusoidal wave of amplitude 10 mV in the frequency range from 10 5 to 10 À1 Hz under 1 sun illumination. Further, the relaxation frequency and time constant of the electrochemical process were measured from EIS data to quantify the efficiency of the photoanode materials. Mott-Schottky experiments were carried out at 100 Hz frequency for the measurements of donor density and at band potential of photoanode material which inherently affects the PEC activity. Incident photon to current efficiency (IPCE) was measured with the setup similar to that of PEC measurement with monochromatic light from 350 to 650 nm with a 10 nm step. The incident light power was measurement at each wavelength with a calibrated photodiode. SECM study was performed on 920 D bi-potentiostat (CH Instrument) using four electrodes system. SECM Teon cell was in-house fabricated for holding the substrate at the base with Oring, reference and counter electrodes. Photoanode materials were used as the substrate, Ag/AgCl (3 M KCl) as a reference and glassy carbon rod as a counter electrode and results are reported in terms of RHE potential. Commercial Pt ultra microelectrode (UME) having RG value of 5 and a diameter 9 mm was used as the probe. Pt microelectrode was polished with a micro polishing cloth with 0.05 mm alumina powder successively and then cleaned in 0.5 M H 2 SO 4 solution for 20 cycles of cyclic voltammetry scans in the potential window of 1.44 V to 0.29 V vs. RHE at the scan rate of 50 mV s À1 . Ferricyanide solution of 2 mM concentration was used as a redox couple in 0.1 M PBS of pH 7. The potential of the probe was chosen at 0.64 V in the region of steady diffusion current aer recording the CV in 2 mM ferricyanide solution with a scan rate of 50 mV s À1 . Samples were illuminated from the front side using the solar simulator as the light source. Probe Approach Curve (PAC) technique was used to record the approach curve to the substrate in dark and also under the illumination of light to measure the kinetic parameter using four electrodes system at different polarization potentials, from the tting of the probe approach plots the interfacial charge transfer kinetics were obtained. The mapping of photoanodes was carried out by the SECM technique (constant height mode) under the illumination condition to map the catalytic activity of the catalyst substrate.
Characterization of the materials XRD analysis of the prepared samples was performed by using a Rigaku powder diffractometer (9 kW Rotating Anode) with Cu K a radiation (l ¼ 1.5406 A). Raman spectra of photoanodes were recorded by using Lab RAM HR 800 Microlaser Raman system with an Ar + laser of 516 nm. The morphology of the photoanodes was examined by eld emission scanning electron microscopy (FE-SEM, JEOL model JSM-7600F). X-ray photoelectron spectroscopy (XPS, MULTILAB, VG Scientic, Al K a radiation as monochromator) was used to investigate the binding energy of the components of Mo doped BiVO 4 photoanodes.

Result and discussion
Structure analysis by XRD and Raman spectroscopy hedron from zero to non zero dipole moment in the crystal lattice, thus making the crystal more polar. This enhanced dipole moment due to the distorted polyhedron are reported to promote the charge separation on photo-excitation, which would enhance the overall photocatalytic activity. 69 Raman spectroscopy is used to investigate the crystallization, local structure and electronic properties of the materials.
where n is the Raman stretching frequency for V-O in cm À1 and R is the bond length inÅ. It can be seen that the V-O bond varies from 1.

SEM and EDS analysis
Morphology of photoanode materials was characterized by scanning electron microscopy (SEM). BiVO 4 was uniformly coated on ITO/SnO 2 as shown in Fig. 3. The materials have shown the granular type of morphology, and the grain size is observed to be increased marginally with Mo doping. Similar observation has been reported in the literature. [82][83][84][85] On 7% Mo doping, the sample has shown chains of grains with vacant spaces in the matrix. Elemental analysis was carried out by energy dispersive spectroscopy (EDS) as shown in Fig. S1 of the ESI. † In SBM0, the atomic ratio of Bi and V is 1 : 1 as shown in Table S1 (ESI †). The atomic percentage of Bi, V, O, and Mo was found as these materials were taken during the synthesis  procedure. Sn content corresponding to inner layer SnO 2 was observed. Further, the thickness of the coating was measured and shown in Fig. S2 (ESI †). The thickness of the combined layers of SnO 2 and BiVO 4 was found to be around 440 nm.

UV-vis spectral measurements
The optical absorption property of a semiconductor reecting the electronic property of the material is a key factor for determining the photoelectrocatalytic activity. All photoanodes materials are characterized by UV-vis diffuse reectance spectra of as shown in Fig. 4A Where a is optical absorption coefficient, hn is photon energy, E g is the bandgap and A is a probability constant. The numerical values of n are 1/2 and 2 for indirect and direct transition, respectively. Thus the nature of the transition can be determined from the linearity of plots of (ahn) 1/2 and (ahn) 2 vs. hn and bandgap can be determined from the x-axis intercept. Fig. 4B shows the Tauc Fig. 4D. For IPCE measurements, LSV was recorded for the wavelength ranging from 370 nm to 550 nm, and photocurrents were sampled at 1.64 V, and then IPCE was measured by using the following equation 89 where J is the photocurrent density, P in is the power of incident photon (monochromatic light), l is wavelength in nm. The onset wavelength of IPCE is 500 nm, which corresponds to the  For further insight into the charge transfer and separation process, Mott-Schottky analysis was carried out; the scans were recorded in the potential window from 0.64 V to 1.64 V (vs. RHE) with an increment of 0.025 V at 100 Hz frequency as shown in Fig. 5A.
where C (F) is space charge capacitance, q is elementary charge, A is electrode surface area, 3 is relative permittivity of BiVO 4 (68), 90 3 0 vacuum permittivity (8.854 Â 10 À12 Fm À1 ), N D (cm À3 ) is donor density, V is applied potential, V FB is at band potential and k B is Boltzmann constant (1.38 Â 10 À23 J K À1 ) and T is absolute temperature. From the slope of a plot of 1/C 2 vs. V, donor density is measured, which is the inherent property of photoanode materials. Donor density was calculated from the slope of Fig. 5A and tabulated in Table 3. N D of SBM0 is found as 5.23 Â 10 19 cm À3 , the measured value is in accordance with the literature reports. 68 Further, the at band potential, which is also an important property of photoanode materials and qualitative measurement of the degree of band bending at the electrode-electrolyte interface, 63,64,91 was calculated from Mott-Schottky plot as shown in Fig. 5A. At higher band bending, the electron-hole recombination will be difficult at the interface, which will result in the improvements of PEC efficiency and stabilizes the photoanodes. [63][64][65]91 The at band potential of SBM0 is found as 0.94 V, on Mo doping, the at band potential is observed as,  bending is a qualitative measure of in-built potential and charge recombination. 65,66 The OCPV was calculated from the difference in open circuit potential in dark and illumination as shown in Fig. 5B and tabulated in Table 3  The interfacial charge transfer efficiency and charge transport efficiency are two limiting parameters on which the overall efficiency of PEC depends. Charge transfer and transport efficiency measurements were performed using a hole scavenger method. 92,93 LSV was recorded in 0.1 M PBS for water splitting. Hole scavenger Na 2 SO 3 (0.1 M solution) was used by assuming complete and fast oxidation of sulte; the LSV plots are shown in Fig. S4 (ESI †). Then, current densities were sampled at 1.64 V and h tranfer and h trasport were calculated using following equations.
J max: ¼ integration of light absoption integration of solar light spectrum (8) where, J max. was calculated from the photocurrent obtained from the silicon diode detector. J H 2 O is the current density for water oxidation; J Na 2 SO 3 is the current density for the oxidation of sulte. Since the water oxidation process is sluggish, sodium sulte was used as a hole scavenger. It is supposed that oxidation of sulte is 100%; thus the charge transfer efficiency was calculated with respect to 100% faradaic efficient, i.e. how fast charge gets transferred from the electrode-electrolyte interface to water molecule for oxidation. The charge transfer efficiency was calculated and tabulated in

Electrochemical impedance measurements
The conductivity and charge transfer resistance are measured from the analysis of Nyquist plot; it also provides the qualitative insight of the charge transfer processes in bulk as well as at the interface of the photoanode, corresponding results are shown in Fig. 5C. Impedance results are tted with the equivalent circuit as shown in Fig. S5 (in ESI †) and the tting parameters are tabulated in Table 4. R ct value of SBM0 is found as 16.8 kU. When Mo is doped, R ct values decreased considerably. The decrease in R ct value suggests faster charge transfer at the interface on Mo doping. This variation in R ct values supports the PEC activity measurements. The photoanodes were further characterized for their charge relaxation processes from Bode plot analysis as shown in Fig. 5D. Frequencies of phase maxima were sampled for different catalysts, which correspond to the relaxation frequency of photogenerated charge and the results are summarized in Table 4. The relaxation frequency of the SBM0 sample is 6.88 Hz, and the relaxation frequency is increased linearly with the Mo doping from SBM0 to SBM7. Further, the relaxation time constant (s) of the electrochemical process was calculated using s ¼ 1/2pf where f is relaxation frequency. The decrease in s indicates the faster electrochemical process on Mo doping, which supports the improvements in the PEC efficiency. Diffusion length is an important parameter in characterizing the interfacial processes, and the overall PEC efficiency depends heavily on the hole diffusion process. The measurement of the diffusion length from impedance measurements, however, includes the assumption that the relaxation time constant is the time taken by the hole to oxidize the water molecule, which contains the diffusion of the hole inside the lms and also its diffusion at the electrolyte interface to oxide water molecule. Diffusion length of the photo-generated holes is calculated by using the equation, L D ¼ (D Â s) 2 where L D is diffusion length, D is the diffusion coefficient of the photogenerated hole, taken as 0.05 cm 2 s À1 , 94 and s is the relaxation time, the values as obtained are tabulated in Table 4. L D of SBM0 is found to be 340 mm. The diffusion length thus obtained from the impedance measurements is decreased with the Mo doping in BiVO 4 . The measured diffusion process includes the diffusion inside the solid catalysts and also at the interface since the observed diffusion length is signicantly higher compared to the thickness of the lms, the holes are expected to be transported outside the electrochemical interface. The thickness of the lms has a negligible contribution to the overall measured diffusion length of the material. The lower diffusion length at the electrochemical interface is associated with the fast charge transfer process at the interface, which is expected to have a higher PEC current.
As seen from Table 4 the capacitance of the SBM0 is significantly low (8.33 Â 10 À6 F) compared to the materials containing Mo. The observed capacitance improvement on the addition of Mo indicates the signicant improvement in the surface charge density on the incorporation of Mo in BiVO 4 . These results show that there are surface improvements as well with the improvements in the bulk property of the photoanode upon Mo doping in the BiVO 4 , which enhances the overall PEC efficiency.

Testing of stability of the photoanodes
Stability of photoanode materials is important for prolonged application of the catalyst; it was performed using chronoamperometry technique at 1.44 V vs. RHE with chopped light voltammetry method, as shown in Fig. 6, the photocurrent of SBM0 was decreased by 37% aer initial excitation. On Mo doping, the recombination step has been reduced drastically, and in place of decay in current, growth in the photocurrent response is observed. The stability test was further performed for continuous illumination of light for 900 s, as shown in Fig. S6 (in (ESI †)) for two catalyst samples. In both, the samples SBM1 and SBM5, the photocurrent is improved initially and then stabilized. The improvement of photocurrent is observed previously and explained on the basis of the charging effect of the photoanode. 95,96 The photocharging effect has been discussed based on both the surface and bulk modications in the materials. The redox reaction through the transformation of V(V) to V(IV) is discussed as one of the important reasons behind the enhanced photocurrent due to photocharging. The photocharging effect is observed to be enhanced in the present case on the incorporation of Mo in the catalyst. The reduction potential of V(V) to V(IV) is more positive than the Mo(VI) to Mo(V) reduction; however, the redox kinetics in the later case is signicantly faster. 97 In view of this, during photocharging process, Mo(VI) will get reduced to Mo(V) rst due to the kinetic effect; aerward Mo(V) would transfer the electron to V(V) and facilitate the reduction of V(V) to V(IV). As reported previously, this facilitated reduction of V(V) to V(IV) due to the presence of Mo would enhance the catalytic activity on photocharging. 95 The photocharging effect is also discussed to be due to the interfacial factor through the modication of band structure upon photocharging. The OCPV is measured in all the materials; its value is increased with the Mo doping. The increase in the OCPV can be correlated with the relatively more signicant bending of the conduction band than the valence band under photo-illumination. The increase in the OCPV indicates the lesser possibility of recombination and enhancement of the hole transfer property through the interface, which eventually would increase the PEC catalytic process in the material. 65,66 Interfacial charge transfer kinetics using SECM The charger transfer at the semiconductor-electrolyte interface and quantication of the active sites in in situ measurements have been performed by the SECM technique. [98][99][100][101][102][103] Investigation of the interface has been carried out by the tip feedback mode. When a tip is far from any surface, tip current, i T,N depends on the number of electrons transferred, concentration of electroactive species, diffusion coefficient and radius of ultra-microelectrode. 98 When the surface is an insulator, the tip current is i T i T,N which results negative feedback. When surface is conducting, redox active species is generated at the surface, tip current, i T i T,N as a result, positive feedback is observed.  6 ] 3À can be used as the acceptor of the photogenerated holes and electrons to characterize the redox kinetics of the photo-catalyst for the PEC water-splitting reaction. 105 Fig. 7A 6 ] 4À is kinetically more favorable by DG $ À0.4 eV than the water oxidation. 106 The cathodic potential at the probe is tuned to the reduction potential of [Fe(CN) 6 ] 3À to [Fe(CN) 6 ] 4À so that there is no side reaction such as oxygen reduction take place. Therefore, feedback current can be assumed mainly from the reduction of [Fe(CN) 6 ] 3À . The hole transfer kinetics at BiVO 4 /interface is measured by the feedback current of the reduction of [Fe(CN) 6 ] 3À in dark and illumination conditions. In the dark, photoanode behaves as an insulator because of the non-availability of the free electron or hole at the interface and hence negative feedback is observed, however on illumination, it behaves as conducting substrate because of photogenerated holes and electrons. The photogenerated holes at the interface oxidize [Fe(CN) 6 ] 4À to [Fe(CN) 6 ] 3À , [Fe(CN) 6 ] 3À species diffuses to the probe and increases the mass transfer process, resulting in the positive feedback response. The normalized tip current (I T ) is calculated by using I T ¼ i T /i T,N where, i T is the real time tip current during the approach to the substrate electrode and i T,N is steady current of the tip when the tip is far from the substrate. Prior to the approach of the probe to the substrate, the CV of ultra microelectrode was performed in 2 mM ferricyanide solution as shown in Fig. S7A † shows a good response from the probe. The negative feedback response was tted using eqn (S1) † for the measurement of RG value. The positive feedback response under illumination was used for the hole kinetics measurements. The positive feedback response of all photoanodes are shown in Fig. 7, the observation of positive feedback indicates the transfer of a hole from the illuminated electrode surface. The normalized apparent heterogeneous charge transfer rate constant (k) and effective heterogeneous charge transfer rate constant k eff (in cm s À1 ) at BiVO 4 /electrolyte interface is obtained from the tting of experiment approach curve to the theoretical SECM kinetics model using eqn (S1)-(S6). † 107 All approach curves have been numerically tted using the above equation and then kinetics parameters are calculated. Some of the ttings of the approach plots are shown in Fig. S8 of the (ESI). † The effective heterogeneous charge transfer rate constant k eff is calculated using the relation k eff ¼ kD diffusion /r T where D diffusion is the diffusion coefficient of the redox probe [Fe(CN) 6 ] 3À and tabulated in Table 5. The low value of k eff for SBM0 shows sluggish hole transfer process at the interface. When Mo was doped in to the BiVO 4 hole transfer rate constant is found to improve signicantly. This improvement suggests that interfacial charge transfer is facilitated upon Mo doping. Further, the effect of applied bias on the hole transfer rate constant is investigated by approaching the probe at different applied potentials. The rate constant for all the materials remained unchanged with the applied potentials indicating its limiting value even at lowest applied potential.
The localized PEC activity of the BiVO 4 photoanodes was analyzed by imaging the surface in the constant height mode at the applied potentials of 0.94 V and 1.14 V to the substrate. A cathodic potential of 0.64 V was applied to the probe for the reduction of the [Fe(CN) 6 ] 3À to [Fe(CN) 6 ] 4À . Therefore, during imaging of the substrate, feedback current at the probe was monitored, and the images are shown in Fig. S9 to S11. † The catalyst substrate is mapped with various current regions, and the current response is increased marginally with more anodic applied potential to the substrate. The overall current response of the tip scanned over the substrate is increased with Mo doping in BiVO 4 up to 5% doping level. The images are further characterized with bigger current region at doping level up to 5%, originated from the overlapping diffusion layer across the catalyst substrate. At 7% doping, in place of overlapping region, the catalytic currents are characterized by small patches separated from each other. The observation can be correlated from  Paper the separated grains at 7% doping compared to the overlapped diffusion layer structure in all other samples. Catalytic current obtained from the overlapping diffusion layer is benecial in catalyst design, as it provides a similar catalytic activity to that of the continuous lms requiring fewer amounts of catalysts. The SECM imaging thus further indicates the betterment in the catalytic activity on Mo doping and also reveals the regional distribution of the catalytic current over the catalyst substrates.
Results thus indicate signicant improvement in the catalytic activity of BiVO 4 through the inclusion of SnO 2 heterojunction and doping of Mo, which has been explained using some of the important interfacial measurements.

Conclusion
Present investigation was aimed to improve the photoelectrocatalytic efficiency of BiVO 4 through the incorporation of SnO 2 interlayer and doping of Mo. The signicant improvement ($154%) in the photocurrent was observed upon 5% Mo doping in SnO 2 /BiVO 4 . Strong correlation among the optical property of the material, the open circuit photovoltage (OCPV), and onset potential was observed in relation to the improvement in the PEC efficiency on Mo doping. The increase in the at band potential and OCPV suggests the improvements in the charge separation upon the Mo doping which resulted in the enhancement in PEC efficiency. SECM investigation reveals signicant improvement in effective hole transfer rate constant from 2.18 cm s À1 to 7.56 cm s À1 with the Mo doping in BiVO 4 . The electrochemical impedance investigation supports the improvements in the charge transfer and transport efficiency by improvements in the bulk and surface properties of BiVO 4 . The facilitated reduction of V(V) to V(IV) on Mo doping is also responsible for the improvement in the catalytic activity. The improvement in the catalytic activity has been evaluated from the improvement in the physicochemical at the bulk of the catalysts, its surface, and most importantly due to the improvement in its interfacial charge transfer processes. Mild expansion in the crystal lattice is also observed on replacement of V by Mo, the improvement in the catalytic activity however, related primarily to the electronic nature compared to any morphological changes.

SEM
Scanning electron microscope EDS Energy dispersive spectroscopy XPS X-ray photoelectron spectroscopy CV Cyclic voltammetry LSV Linear sweep voltammetry PEC Photo-electrocatalysis EIS Electrochemical impedance spectroscopy OCPV Open-circuit photovoltage UME Ultra-microelectrode PAC Probe approach curve SECM Scanning electrochemical microscopy

Author contributions
The manuscript was written through the contributions of all authors. All authors have given approval to the nal version of the manuscript.

Funding sources
This project is fully funded by our institute Bhabha Atomic Research Centre, Government of India.

Conflicts of interest
All the authors of this manuscript declare that there is no conicts of interest exists to disclose.