Our paper published in Nature Catalysis (2018) can be read from https://www.nature.com/articles/s41929-018-0134-1
To solve the energy crisis and environmental problems, utilizing the inexhaustible solar energy to release clean and storable hydrogen from water has been intensively investigated. Techno-economical analyses have predicted that direct decomposition of water into hydrogen and oxygen on particulate photocatalysts via one-step excitation is one of the simplest and most cost-effective approaches towards applicable hydrogen production on a plant scale. The core issue is the development of particulate photocatalysts enabling visible-light-driven overall water splitting.
We have developed various particulate (oxy)nitride photocatalysts with d0 or d10 electronic configurations (J. Phys. Chem. C 2007, 111, 7851). In particular, tantalum nitride (Ta3N5) is an attractive material for the solar hydrogen production because of the narrow band gap of 2.1 eV and the simple composition. In 2002, our group reported Ta3N5 photocatalytically active for hydrogen or oxygen evolution from aqueous solution in the presence of sacrificial reagents for the first time (Chem. Lett. 2002, 31, 736). Since then, its performance in the sacrificial hydrogen or oxygen evolution reactions, Z-scheme water splitting and photoelectrochemical water oxidation has been dramatically improved through detailed examinations of ammonia-nitridation conditions, morphological control, and surface modifications (J. Am. Chem. Soc. 2012, 134, 19993; Chem. Eur. J. 2013, 19, 7480; Angew. Chem. Int. Ed. 2017, 56, 4739). However, despite the tremendous effort over the last one and a half decades, one-step-excitation overall water splitting had not been achieved on Ta3N5 photocatalysts until this work.
The conventional Ta3N5, prepared from tantalum oxide (Ta2O5) by long-term nitridation at high temperature, consists of aggregated polycrystalline particles including grain boundaries and defects, which are detrimental to the utilization of photoexcited carriers in up-hill reactions such as overall water splitting. We expected that high-quality Ta3N5 particles could be obtained through the nitridation of other Ta-based oxide precursors. Alkali tantalate ATaO3 (A= Li, Na and K) with the volatile elements Li, Na and K were candidates, because the evaporation of alkali metal species during high-temperature nitridation would facilitate the generation of the Ta3N5 structure. We found that rod-like Ta3N5 single crystals were evolved on potassium tantalate (KTaO3) (110) planes selectively through a short-time nitridation, whereas polycrystalline Ta3N5 was formed from lithium tantalate (LiTaO3) or sodium tantalate (NaTaO3). The lattice spacing of KTaO3 is close to that of Ta3N5 while those of LiTaO3 and NaTaO3 are not. The rapid volatilization of K species from KTaO3 (110) planes may have resulted in the generation of Ta3N5 on the (110) facets of KTaO3 rather than on the (100) facets. Such Ta3N5 nanorods were demonstrated to be single crystals free from grain boundaries. When loaded with the core-shell-structured rhodium and chromium oxide cocatalyst by photodeposition process, the single-crystal Ta3N5 nanorods show activity in one-step-excitation overall water splitting under visible light and simulated sunlight.
Although the efficiency is still low at present, it is possible that the photocatalyst performance could be improved by increasing the proportion of such high-quality Ta3N5 nanorod single crystals on KTaO3. Our findings demonstrate the importance of suppressing defect generation through the refinement of photocatalyst synthesis in achieving one-step-excitation overall water splitting under visible light. The nanoscale single-crystal growth by designing appropriate precursors with matching lattice parameters and volatile components is expected to contribute to the development of photocatalysts for solar energy conversion.