Kinetics of Photoelectrochemical Oxidation of Methanol on Hematite Photoanodes

The kinetics of photoelectrochemical (PEC) oxidation of methanol, as a model organic substrate, on α-Fe2O3 photoanodes are studied using photoinduced absorption spectroscopy and transient photocurrent measurements. Methanol is oxidized on α-Fe2O3 to formaldehyde with near unity Faradaic efficiency. A rate law analysis under quasi-steady-state conditions of PEC methanol oxidation indicates that rate of reaction is second order in the density of surface holes on hematite and independent of the applied potential. Analogous data on anatase TiO2 photoanodes indicate similar second-order kinetics for methanol oxidation with a second-order rate constant 2 orders of magnitude higher than that on α-Fe2O3. Kinetic isotope effect studies determine that the rate constant for methanol oxidation on α-Fe2O3 is retarded ∼20-fold by H/D substitution. Employing these data, we propose a mechanism for methanol oxidation under 1 sun irradiation on these metal oxide surfaces and discuss the implications for the efficient PEC methanol oxidation to formaldehyde and concomitant hydrogen evolution.


Photoelectrochemistry
The effect of the concentration of methanol is shown in Figures S1.A and S1.B. Figure S1.A shows the increasing plateau photocurrent response for methanol oxidation on a α-Fe 2 O 3 photoanode as the concentration of methanol is increased. This constant increasing of the photocurrent may be due to diffusion limitations that are overcome at high concentrations of methanol in the electrolyte. Figure S1.A shows that from 90 to 98% of methanol the plateau photocurrents obtained were constant, this indicates that above 90 % methanol, diffusion is not limiting the reaction. Therefore, a 95% methanol in 0.1 M NaOH electrolyte was selected for our studies on α-Fe 2 O 3 . Figure S1.B shows that the plateau photocurrent when methanol is oxidized on TiO 2 reaches a saturation point at a concentration of 4% of methanol in 0.1 M NaOH and it remains constant as the concentration of methanol increases up to 50%. Figure S1. Current/potential response of the measured photoanodes under approximately 1 sun front illumination conditions through electrode/electrolyte (EE) interface, a) α-Fe 2 O 3 photoanode under dark conditions (black dashed line) and EE illumination, measured in 0.1M NaOH aqueous solution (blue) and 25% methanol (pink), 85% methanol (green), 90% methanol (yellow), 95% methanol (red) and 98% methanol (dark red) in 0.1M NaOH and b) TiO 2 photoanode under EE illumination, measured in 0.1M NaOH aqueous solution (blue) and 4% methanol (red), 25% methanol (green), 50% methanol (purple) in 0.1M NaOH.

Formaldehyde detection
Formaldehyde generation was performed setting a bulk photoelectrolysis in a 100 mL three electrode photoelectrochemical cell under approximately 1 sun of 365 nm LED light and 0.55V vs Ag/AgCl. In order to quantify formaldehyde, 3 aliquots were extracted at the following times of reaction 13, 31.6 and 47.6 min. The extracted volume (10 mL in the first aliquote, and 5 mL for the two remaining) was mixed with 10 mL of 1M NaOH and 0.1 g of 4-Amino-5hydrazino-1,2,4-triazole-3-thiol and diluted to 25 mL in water, following the reported procedure. 1 An individual UV-Visible absorption spectrum of the final mixtures was then measured. Using the calibration curve ( Figure S2) the amount of produced formaldehyde was quantified. Finally, form the bulk photoelectrolysis ( Figure S3) the theoretical amount of formaldehyde was estimated taking into account the total charge measured over the testing period, the Faradaic constant and the number of electrons generated in the reaction, which in this case is 2. The Faradaic efficiency (FE) was estimated as:

Photo-Induced Absorption Spectroscopy and Transient Photocurrent measurements on hematite at 0.00 V Ag/AgCl
Photo-induced absorption spectroscopy and transient photocurrents were measured at different applied potentials in the PEC cell for the oxidation of methanol. Figure S4.A shows the PIA signal corresponding to the photogeneration and accumulation, steady-state and reaction of holes as a function of excitation light intensity for the oxidation of methanol on the α-Fe 2 O 3 photoanode at 0.00 V. The corresponding TPC signals are shown in Figure S4.B which gives us information about the electron extraction

Effect of Recombination on the Decays
The photo-induced absorption signal probed at 650nm, when the LED light is turned off corresponds to the decay of the photogenerated holes. In the case of the oxidation of methanol at 0.55 V Ag/AgCl , where no back electron/hole, or 'surface', recombination 2 takes place, the decay corresponds to the reaction with methanol of the holes accumulated at the surface of hematite. At lower applied potentials, such 0.00 V Ag/AgCl and matched surface hole density compared to 0.55 V Ag/AgCl (approximately 0.3 holes.nm -2 ), the PIA signal decays with faster kinetics. In these conditions of low applied potential where no water oxidation occurs, we assign this faster decay to the presence of back electron/hole recombination taking place, with kinetics competitive with methanol oxidation. Figure S5. Normalised surface hole absorption, probed at 650nm when the PEC cell was held at 0.00 V (light red) and 0.55 V (dark red) Figure S6.A shows the PIA signal corresponding to the photogeneration and accumulation, steady-state, and reaction of holes as a function of excitation light intensity for the oxidation of methanol on the TiO 2 photoanode at -0.80 V. The corresponding TPC signals are shown in Figure S6.B.

Initial Rates Analysis of methanol oxidation on hematite
The rate law analysis is based on the kinetic model described by Le Formal et. al.,3 in which the photogeneration, accumulation and reaction of the holes within the time ( ௗ ೞ ௗ௧ ) in any photoanode (followed by PIAS) in absence of back electron/hole recombination can be described by eq 1.
Where ‫ܬ‬ ௦ ௦௨ is the flux of photogenerated holes to the surface, ݇ ு మ ை ை௦ and ݇ ெைு ை௦ are the rate constants for water and methanol oxidation respectively. ߚ and ∝ are the orders of water and methanol oxidation reactions, with respect to the density of accumulated holes. The term ఉ accounting for water oxidation as side or competitive reaction 4 is taken into account in this model. At steady-state conditions, where the rate law analysis is reported, is zero and ‫ܬ‬ ௦ ௦௨ is equivalent to the photocurrent, ݆ ሺሻ . Therefore, eq 1 can be rewritten as At potentials where no water oxidation can occur i.e., 0.00 V Ag/AgCl for hematite and -0.80 V Ag/AgCl in the case of anatase (shown in Figure 4), according to the J-V curves shown in Figure  1.a in the main text and Figure S1.b, respectively, the rate law analysis from eq 2 can be rewritten in logarithmic base as Figure S7. Initial rates analysis, gradient of PIA signal within the first 30 ms, ௗ ೞ ௗ௧ versus surface hole density, ‫݀‬ ௦ , of the oxidation of methanol at 0.55 V Ag/AgCl applied potential on α-Fe 2 O 3 .
An initial rates analysis can also be made taking into account the decay of the photogenerated holes accumulated at the surface of the photoanodes. When the LED illumination is turned off, the flux of photogenerated holes towards the surface of the hematite ‫ܬ(‬ ௦ ௦௨ ) is zero and where no water oxidation takes place, the rate law analysis from eq 1 can be rewritten in logarithmic base as (4) Figure S7 shows the initial rates analysis, within the first 30ms of decay, for the oxidation of methanol at 0.55 V Ag/AgCl , where a second order of reaction (α) is extracted from the gradient of the log(p s )-logቀ ௗ ೞ ௗ௧ ቁ plot which is in agreement with the analysis done in quasi steady-state conditions and shown in Figure 4.
We now turn to the observed rate constants shown in Figure S7, where ݇ ெைு ை௦ calculated from the data collected at 0.55 V Ag/AgCl , where no back electron / hole recombination takes place, is ~ 20 holes -1 nm 2 s -1 , which is in accordance with the observed rate constant calculated from the quasi steady-state analysis (33 holes -1 nm 2 s -1 ).  Figure S8 shows the photo-induced absorption (PIA) signals ( Figure S8.A) and their corresponding transient photocurrents ( Figure S8.B) for the oxidation of d 4 -methanol in 0.1 M NaOD in deuterated water.

Rate Law Analysis of water compared to methanol oxidation on hematite
Turning to potentials where water oxidation might compete with methanol oxidation reaction, i.e. 0.55 V Ag/AgCl , the rate law analysis would be described by eq 2 as ݇ ெைு ≫ ݇ ு మ ை , observed graphically in Figure S9, validating the rate law analysis at high applied potentials with no competition of water oxidation reaction. Figure S9. Rate law analysis: photocurrent density, ݆ ሺሻ versus surface hole density, ‫݀‬ ௦ , for the oxidation of methanol at 0.55 V (dark red) and 0.00 V (light red), and water at 0.55 V (black) on α-Fe 2 O 3 . The PIA signal was converted to surface hole density using the reported molar extinction coefficient of 640 M -1 cm -1 for α-Fe 2 O 3 . 3 Water oxidation data has already been published by Le Formal et. al. 3