Computational Screening of New Dopants for Nife-Based Layered Double Hydroxide Catalysts for Seawater Splitting

Water electrolysis is a key technology to convert and store sustainable energy into high-energy-dense hydrogen. However, slow kinetics of oxygen evolution reaction (OER) at the anode hinders overall reaction rate of water splitting and delays realization of hydrogen energy society. OER consists of four electron steps and an ideal catalyst has 1.23 V potential barrier for each step, but a real catalyst has a step which requires higher potential than 1.23 V, i.e., overpotential. Therefore, exploring OER catalyst with minimized OER overpotential is an important challenge in increasing water electrolysis efficiency. On the other side, the direct use of abundant seawater as a splitting reactant has advantage in terms of resources and cost. It can also help commercialization of water electrolysis. Since seawater contains chloride which causes undesirable side reactions and accelerated corrosion of the anode material, appropriate OER catalyst is necessary to avoid these problems. Conventional trial-and-error method in catalyst design usually requires many years of R&D and high cost. Computational screening can preemptively narrow down the candidate group of high-efficient catalyst to reduce time and economic cost. NiFe-based layered double hydroxide (NiFe-LDH) is a promising OER catalyst due to the comparable activity to commercial IrO2catalyst and can provide easily tunable metal composition during synthesis process. Its unique layered structure and reversible oxidation state change in redox condition also have arisen interest of many researchers. In this study, starting from the mechanism study of OER, chloride evolution reaction (ClER), and chloride-induced corrosion on this material, we performed DFT calculations to predict which transition metal dopants can enhance the OER activity of NiFe-LDH without accelerating selectivity and corrosion problems in seawater. NiFe-LDH is known to experience its transition to NiFeOOH phase under OER condition. [001] and [110] facets were set as edge and terrace sides of NiFeOOH, respectively, after their surface energies were investigated. Surface Pourbaix diagram showed that the [001] facet has clean termination without any adsorbate while [110] facet has dissociated-H2O covered termination under alkaline OER condition. OER energetic profile showed that the Fe site on the [110] surface is an OER active site with a theoretical overpotential of 280 mV, and seawater conditions do not significantly affect the OER activity itself. ClER mechanism study revealed that it occurs via *ClOH intermediates on [110] metal sites or *Cl intermediates on deprotonated [001] oxygen sites. Energy calculation of surface chlorination and metal dissolution steps showed that gradual chlorination accelerates metal dissolution and [110] facet is more vulnerable to chloride-induced corrosion than [001] facet. As the third metal candidate of NiFe-LDH, 3d to 5d transition metals excluding heavy metals were screened. Gibbs free energy correction terms such as vibrational entropy, zero-point energy and solvation parameter were calculated only for NiFeOOH case, and applied them to other candidates to simplify the screening process. NiFeOOH [001] surface could not be better than NiFeOOH [110] surface due to the inherent low activity of the [001] facet, but interestingly seven dopants could reduce the overpotential of NiFeOOH [110] surface by affecting the Fe active site or being active sites themselves. Since these dopants increased the oxidation ability of the active site, all of them also lowered chloride oxidation potential. However, fortunately, none of their ClER operating potential was lower than their OER operating condition, which means that there still exist potential windows where 100% OER selectivity can be achieved. In the last screening step, dissolution energies of fully chlorinated metal sites were calculated for the seven candidates, and two cases of them showed negative dissolution energy, which indicates very unstable doping state. As a result, five candidates passed through all three screening criteria: OER activity, OER vs. ClER selectivity, and durability against corrosion. Further, the experimental validation has performed on three promising candidates considering the raw material price, and it revealed that their activity is higher activity than NiFe-LDH. In addition, for the two abandoned cases in the third screening step (about metal dissolution), the experimental data showed that they were not seem to be actually doped into NiFe-LDH, indicating that the computational screening of seawater compatibility really worked. This study supports the utility of theoretical screening in electrochemical catalyst design and presents computational approach method to consider seawater compatibility.