(The) architecture of Si-based photoelectrodes for highly efficient, low-cost, and durable solar-driven water splitting

Abstract

Photoelectrochemical (PEC) cells have attracted much great attention as an ideal solution for storing solar energy and producing value-added chemicals and fuels. In PEC cell, photo-excited carriers are generated in a semiconductor and drive a desired (photo)electrolysis at the semiconductor/electrolyte interface. Among the various semiconductor materials for PEC application, silicon has many advantages for constructing efficient PEC cells because it is earth-abundant, low-cost, and has a band gap (1.1 eV) suitable for absorbing most of the solar spectrum. However, Si has a surface condition that is unfavorable for water electrolysis. In addition, it is easily oxidized and damaged under water oxidation condition. Therefore, a multi-functional protection layer is indispensable for ensuring durability as well as high performance of a Si-based PEC cell. In conventional photoelectrode structures, most semiconductor materials have catalytic properties undesirable for photo-driven electrolysis, so cocatalyst is put on the PEC cell surface to reduce the overpotential. However, the competition between light absorption and electrocatalysis at a limited surface area on conventional photoelectrodes could limit the efficiency of a photoelectrode. In addition, semiconductors can be degraded by corrosion and/or oxidation, which cause significant degradation of PEC performance during water electrolysis. Therefore, an in-depth study about the best architecture for efficient and stable photoelectrode was required. In this work, we explain clearly the crucial coupling between the optical and electrochemical properties of a photoelectrode, which restricts its efficiency. Although many solutions for these problems have been proposed using Si-based photoelectrodes, a design that satisfies completely the need for high efficiency, cost-effectiveness, and robust stability does not yet exist.

Herein, we propose novel architectures for Si-based photoelectrodes able to break the critical coupling in conventional PEC cells. Our device modeling shows distinctly the possibility that our new photoelectrode architectures could achieve much improved performance compared to conventional photoelectrodes. In addition, in this work, we used a variety of tools (including such as atomic layer deposition (ALD), fabrication of nanostructures, and formation of alloy electrocatalyst) that could control precisely the PEC performance of photoelectrodes. First, we demonstrated new photoelectrode architecture to break this coupling using 3D-structured cocatalyst. Our device modeling and PEC measurements show clearly the effect of 3D-structured cocatalyst and provide guidelines for optimization of this new photoelectrode architecture. As a model system of the new photoelectrode architecture, we developed an oxide-passivated Si photoanode with locally defined Ni and NiFe inverse opal (IO) nanostructured cocatalyst. We confirmed that our oxide-passivated Si photoanode with micro-patterned Ni IO shows overpotential reduction of 128 mV for producing 25 mA/$cm^2$, compared to a Si photoanode with micro-patterned planar Ni in 1M KOH under 1 sun illumination. In addition, our oxide-passivated Si photoanode with micro-patterned NiFe IO produced a photocurrent density of 31.2 mA/$cm^2$ at 1.23 V vs. RHE, which is among the best of the state-of-the-art Si photoanodes.

Second, as another strategy for eliminating critical coupling of conventional photoelectrodes, we proposed a photoelectrode architecture which has the carrier-selective contacts using wide bandgap cocatalyst materials. If cocatalyst material allows penetration of most of the solar spectrum to the underlying light absorber, a photoelectrode can be fully covered with the cocatalyst layer. Therefore, this photoelectrode can provide strong catalytic properties without loss of light absorption. In this work, we developed ALD $CoO_x$/$SiO_x$/n-Si heterojunction photoanodes with has hole-selective contacts. Cobalt-oxide ($CoO_x$) is considered a promising multi-functional heterojunction material over Si because it has a wide bandgap, robust stability, and can even serve as an oxygen evolving catalyst. However, they exist in various phases and the relationship between a phase and its PEC performance has not yet been clearly defined. In this work, we controlled the phase of $CoO_x$ films intentionally by adjusting the ALD conditions. In addition, we demonstrated a crucial competitive relationship between the catalytic and energetic properties of the $CoO_x$/$SiO_x$/n-Si heterojunction. In addition, to overcome single-layered $CoO_x$ film, we developed double-layered $CoO_x$ film that provides the advantages of both CoO and $Co_3O_4$. Our DL ALD $CoO_x$/$SiO_x$/n-Si heterojunction photoanode showed a photocurrent density of 3.5 mA/$cm^2$ without a buried junction and maintained a saturating current density of 32.5 mA/$cm^2$ for 12 h in 1M KOH under 1 sun illumination.

Finally, for a highly efficient and cost-effective hydrogen evolution reaction (HER), we also developed a thin film $MoS_2$/p-Si photocathode by ALD. ALD is a technique suitable for PEC applications because it is capable of forming ultra-thin films with uniform thickness and stoichiometry at a high aspect ratio as well as planar substrates in a wafer scale. $MoS_2$ is an earth-abundant and low-cost material with a good HER catalytic property. In addition, it is considered a material that could possibly be used to substitute for Pt. In particular, if $MoS_2$ is formed on p-Si, it can serve as an electron-selective contact, which shows its excellent charge collection property. Our systematic study of ALD $MoS_2$ clearly demonstrates that the morphology, crystallinity, edge site density, stoichiometry, and thickness of ALD $MoS_2$ significantly affects its HER performance. In particular, the post-growth-sulfurization process can control the stoichiometry and crystallinity of $MoS_2$ film, which govern its electrochemical performance. Our optimized ALD $MoS_2$/Si photocathode exhibited 21.7 mA/$cm^2$ photocurrent density at 0 V vs. RHE, which is higher than that of Si photocathodes with Pt nano-particles and almost as high as the highest HER performance previously reported for $MoS_2$/Si photocathodes.

Our new photoelectrode strategies offer solutions to overcome the crucial constraints that obstruct the realization of highly efficient, low-cost, and durable PEC water splitting devices. In addition, since our strategies can be applied to various other PEC applications (HER, oxygen evolution reaction, and $CO_2$ conversions) due to their flexibility, we believe that these strategies promise solutions for commercialization of various PEC devices.