Metastability Triggered Reactivity in Clusters at Realistic Conditions: A Case Study of N-doped (TiO$_2$)$_n$ for Photocatalysis

Abstract: Here we report a strategy, by taking a prototypical model system for photocatalysis (viz. N-doped (TiO$2$)$_n$ clusters), to accurately determine low energy metastable structures that can play a major role with enhanced catalytic reactivity. Computational design of specific metastable photocatalyst with enhanced activity is never been easy due to plenty of isomers on potential energy surface. This requires fixing various parameters viz. (i) favorable formation energy, (ii) low fundamental gap, (iii) low excitation energy and (iv) high vertical electron affinity (VEA) and low vertical ionization potential (VIP). We validate here by integrating several first principles based methodologies that consideration of the global minimum structure alone can severely underestimate the activity. As a first step, we have used a suite of genetic algorithms [viz. searching clusters with conventional minimum total energy ((GA)$\textrm{E}$); searching clusters with specific property i.e. high VEA ((GA)$\textrm{P}^{\textrm{EA}}$), and low VIP ((GA)$\textrm{P}^{\textrm{IP}}$)] to model the N-doped (TiO$2$)$_n$ clusters. Following this, we have identified its free energy using ab initio thermodynamics to confirm that the metastable structures are not too far from the global minima. By analyzing a large dataset, we find that N-substitution ((N)$\textrm{O}$) prefers to reside at highly coordinated oxygen site to maximize its coordination, whereas N-interstitial ((NO)$\textrm{O}$) and split-interstitial ((N$_2)\textrm{O}$) favor the dangling oxygen site. Interestingly, we notice that each types of defect (viz. substitution, interstitials) reduce the fundamental gap and excitation energy substantially. However, (NO)$\textrm{O}$ and (N$_2)\textrm{O}$ doped clusters are the potential candidates for overall water splitting, whereas N$_\textrm{O}$ is congenial only for oxygen evolution reaction.