Understanding photophysical and electrocatalytic processes during photocatalysis in a powder suspension system is crucial for developing efficient solar energy conversion systems. We report a substantial enhancement by a factor of 3 in photocatalytic effi-ciency for the oxygen evolution reaction (OER) by adding trace amounts (~0.05 wt%) of noble metals (Rh or Ru) to a 2 wt% cobalt oxide-modified Ta3N5 photocatalyst particulate. The optimized system exhibited high quantum efficiencies (QEs) of up to 28 and 8.4% at 500 and 600 nm in 0.1 M Na2S2O8 at pH 14. By isolating the electrochemical components to generate doped cobalt oxide electrodes, the electrocatalytic activity of cobalt oxide when doped with Ru or Rh was improved compared with cobalt oxide, as evidenced by the onset shift for electrochemical OER. Density functional theory (DFT) calculation shows that the ef-fects of a second metal addition perturbs the electronic structure and redox properties in such a way that both hole transfer kinetics and electrocatalytic rates improve. Time resolved terahertz spectroscopy (TRTS) measurement provides evidence of long-lived electron populations (>1 ns; with mobilities μe ~0.1-3 cm2 V-1 s-1), which are not perturbed by the addition of CoOx-related phases. Furthermore, we find that Ta3N5 phases alone suffer ultrafast hole trapping (within 10 ps); the CoOx and M-CoOx decorations most likely induce a kinetic competition between hole transfer toward the CoOx-related phases and trapping in the Ta3N5 phase, which is consistent with the improved OER rates. The present work not only provides a novel way to improve electrocatalytic and photocatalytic performance but also gives additional tools and insight to understand the characteristics of photocatalysts that can be used in a suspension system.
|Original language||English (US)|
|Number of pages||10|
|State||Published - May 31 2016|
Bibliographical noteKAUST Repository Item: Exported on 2020-10-01
Acknowledgements: The research reported in this work was supported by the King Abdullah
University of Science and Technology. The authors also
acknowledge the PSN and IDRIS computation centers for providing