Photoelectrocatalytic H2 evolution in water with molecular catalysts immobilised on p-Si via a stabilising mesoporous TiO2 interlayer

A versatile platform for the immobilisation of molecular catalysts on a readily-prepared Si photocathode with a mesoporous TiO2 layer is reported.

surface to be analyzed (S ≈ 0.2 cm 2 ) exposed. The cells were then allowed to dry thoroughly for 18 h in air before use.

Characterisation of Photoelectrodes
SEM images were recorded on a FEI Philips XL30 FEG ESEM instrument at 5 kV acceleration voltage. ATR FT-IR spectra of the compounds or the functionalised TiO 2 were recorded on a Nicolet iS50 spectrometer. XPS was performed on an ESCALAB 250Xi spectrometer (Thermo Fisher Scientific, East Grindstead, UK) utilising a monochromatic Al-Kα source (50-300 W, 0.2-1 mm spot size). The quantification of the amount of immobilised NiP or CoP 3 (mole per geometrical area) on the Si|mesoTiO 2 |catalyst electrodes was evaluated by UV-visible spectroscopy after desorption of the catalyst from the corresponding electrode. Typically, the Si|mesoTiO 2 |catalyst electrode (S ≈ 0.5 cm 2 ) was immersed for 1.5 h in a MeOH bath containing tetrabutylammonium hydroxide (0.1 M). The NiP and CoP 3 solutions' absorptions were then measured (l = 1 cm) at 350 and 400 nm, respectively, and the concentration was estimated using the molar absorption curves in Figure S7. UV-vis spectra were collected using a Varian Cary 50 Bio UV-vis spectrometer.

Photoelectrochemical Studies
LSVs and CPP were performed with an Ivium CompactStat potentiostat. A Newport Oriel Xenon 150 W solar light simulator (100 mW cm −2 , AM1.5G and IR water filters, λ > 400 nm) was used as the light source. A three-electrode configuration was employed in a custom-made airtight two-compartment PEC cell with a Nafion membrane separating the compartments. A platinum mesh was used as counter electrode and an Ag/AgCl/KCl (sat.) electrode as reference electrode. All electrochemical measurements were performed at room temperature in aqueous acetic acid solutions (0.1 M, pH 3.0 or 4.5), except for Si|mesoTiO 2 |H 2 ase and related control experiments, where a MES buffer (2-(Nmorpholino)ethanesulfonic acid, 50 mM, pH 6.0) was used. LSVs were conducted at a scan rate of 5 mV s −1 with chopped light alternating between dark and light every 5 s. The onset potential was defined as the potential at which a photocurrent density of |J| = 10 µA cm −2 was achieved by the respective electrode. The applied potential during CPP was 0.0 V vs. RHE, and continuous illumination was maintained, apart from hourly dark chops lasting for 2 min each. CPP of CoP 3 and H 2 ase was ceased after 4 h and 5 h, respectively; all others were continued for 24 h.
Prior to the CPP experiments, the electrolyte solution in both compartments of the PEC cell was purged with N 2 containing 2 % CH 4 as an internal standard for gas chromatography (GC) measurements. The amount of gaseous H 2 was analyzed by headspace gas analysis using an Agilent 7890A Series GC equipped with a 5 Å molecular sieve column (N 2 carrier gas at a flow rate of approximately 3 mL min -1 ). The GC oven holding the columns was kept isothermal at 45 °C, and a thermal conductivity detector was employed. Aliquots (75 µL) of the headspace gas were removed for GC analysis at regular time intervals. The FE of the photocathodes was calculated by comparing the expected amount of H 2 produced as indicated by the total charge passed through the electrode and the actual amount produced. Analytical measurements were performed in triplicate and the standard deviation of each data point is denoted by error bars.

IPCE Measurements
IPCE measurements were conducted in the same electrochemical cell set-up as used for PEC performance experiments, with the solar light simulator coupled to a monochromator (MSH300, LOT Quantum design). The sequence carried out at each wavelength was 1 min of illumination, followed by 5 min in the dark. The current was collected at two points per second, with the initial 10 and final 10 points of each light cycle averaged; the electrode's dark current was subtracted from this average to give the final photocurrent. Sample photocurrent data were normalised to the output of a power meter (Thorlabs PM100D Compact Power and Energy Meter Console). Measurements were performed in triplicate and the standard deviation at each wavelength is denoted by error bars.

Analysis of TiO 2 Charging Current
The charging and discharging of TiO 2 's CB were studied by two successive chronoamperometric experiments, conducted on Si|mesoTiO 2 and Si|mesoTiO 2 |NiP electrodes each. These were conducted at room temperature in a one-compartment PEC cell in a three-electrode configuration with an acetic acid solution (0.1 M, pH 4.5). In the first chronoamperometry phase, a potential of 0.0 V vs. RHE was applied for two min under solar light illumination (AM1.5G, 100 mW cm −2 ,  > 400 nm), corresponding to the charging of the CB of TiO 2 . After this first phase, 20 sec were allowed to pass where the electrode was left in the dark with no applied potential. In the following second chronoamperometry phase, corresponding to the discharging step, 0.0 V vs. RHE was applied in the dark. In some cases, a solution of MV in the electrolyte solution was injected partway through the second chronoamperometry (final concentration in PEC cell = 10 mM). The recorded current is normalised and given as a percentage. A similar experiment was also conducted on Si|mesoTiO 2 |NiP (without the addition of MV).