Multi-Electron Transfer at H-Terminated p-Si Electrolyte Interfaces: Large Photovoltages under Inversion Conditions

Abstract

Photovoltages for hydrogen-terminated p-Si(111) in an acetonitrile electrolyte were quantified with methyl viologen [1,1′-(CH3)2-4,4′-bipyridinium](PF6)2, abbreviated MV2+, and [Ru(bpy)3](PF6)2, where bpy is 2,2′-bipyridine, that respectively undergo two and three one-electron transfer reductions. The reduction potentials, E°, of the two MV2+ reductions occurred at energies within the forbidden bandgap, while the three [Ru(bpy)3]2+ reductions occurred within the continuum of conduction band states. Bandgap illumination resulted in reduction that was more positive than that measured with a degenerately doped n+-Si demonstrative of a photovoltage, V ph, that increased in the order MV2+/+ (260 mV) < MV+/0 (400 mV) < Ru2+/+ (530 mV) ∼ Ru+/0 (540 mV) ∼ Ru0/– (550 mV). Pulsed 532 nm excitation generated electron–hole pairs whose dynamics were nearly constant under depletion conditions and increased markedly as the potential was raised or lowered. A long wavelength absorption feature assigned to conduction band electrons provided additional evidence for the presence of an inversion layer. Collectively, the data reveal that the most optimal photovoltage, as well as the longest electron–hole pair lifetime and the highest surface electron concentration, occurs when E° lies energetically within the unfilled conduction band states where an inversion layer is present. The bell-shaped dependence for electron–hole pair recombination with the surface potential was predicted by the time-honored SRH model, providing a clear indication that this interface provides access to all four bias conditions, i.e., accumulation, flat band, depletion, and inversion. The implications of these findings for photocatalysis applications and solar energy conversion are discussed.

Methyl Termination of p-Type Silicon Enables Selective Photoelectrochemical CO2 Reduction by a Molecular Ruthenium Catalyst

“…Relative to the metallic electrodes, the illuminated p-type CH 3 -Si photoelectrode demonstrates a significantly more anodic onset potential of −0.80 V vs Fc +/0 , consistent with a 460 mV photovoltage under these conditions. Our observed photovoltage is similar to the V ph reported by Keller et al for [CoCp 2 ] +/0 at p-type H-Si photoelectrodes (440 mV), 25 and we quantified consistent photovoltages from cyclic voltammograms of a 1 mM [CoCp 2 ][PF 6 ] solution (Figure S5). Cyclic voltammograms of C-trans-[Ru(tpy)(Mebim-py)-(NCCH 3 )][PF 6 ] 2 in solution were obtained at a p-type CH 3terminated Si (111) photoelectrode under white light illumination with an irradiance of 339 mW/cm 2 (Figure 2).…”

“…Voltammetry experiments were conducted in a single-compartment cell with minimal distance between electrodes to mitigate uncompensated resistance; high analyte concentration and stirring conditions were used to minimize mass transport effects that could limit photovoltage. The [CoCp 2 ] +/0 redox couple was selected for this purpose because it has a negative reduction potential that falls within the Si bandgap (but close to the conduction band edge), 25 which is a good model of the Si surface when in contact with the Ru catalyst molecule that has similar reduction potentials. 25 The onset potential for reduction of [CoCp 2 ] + was determined by extrapolating a linear fit of the voltammograms to the xintercept.…”

“…The [CoCp 2 ] +/0 redox couple was selected for this purpose because it has a negative reduction potential that falls within the Si bandgap (but close to the conduction band edge), 25 which is a good model of the Si surface when in contact with the Ru catalyst molecule that has similar reduction potentials. 25 The onset potential for reduction of [CoCp 2 ] + was determined by extrapolating a linear fit of the voltammograms to the xintercept. The difference in onset potential for the photoelectrode relative to the dark electrodes is equal to the photovoltage.…”

“…However, this dependence can reflect photon flux limitations, or it can arise from an increase in the number of conduction band electrons. It is well established that when a large number of conduction band electrons are photogenerated, the formation of an inversion layer is favored, leading to longer carrier lifetimes. ,− To distinguish between these possibilities, we sought to quantify the photon flux and compare it to the rate of product formation. As the molecular catalyst is a strong visible light absorber, many spectral photons are screened by the solution.…”

Photoelectrochemical CO2 Reduction to CO Enabled by a Molecular Catalyst Attached to High-Surface-Area Porous Silicon

“…No redox features associated with 1 could be discerned for 1 @porSi (under CO 2 or Ar) using cyclic voltammetry or linear sweep voltammetry (Figures S22 and S23). This is likely due to 1 @porSi being far from an ideal electrode for cyclic voltammetry, with a range of microenvironments (due to both diffusion gradients down the pores and roughness of the pore walls), tortuous diffusion pathways for electrolyte and substrates, and complex behavior of the Si photoelectrode under these conditions . As discussed further in Section SVI.…”