Computational Design of Catalysts for Electrochemical Hydrogen Evolution and Nitrogen Reduction

Abstract

The development of efficient and carbon-neutral energy conversion technologies is a key issue for modern society's sustainability. The electrochemical hydrogen evolution reaction (HER) is regarded as a practical means for the production of high-purity hydrogen from abundant water. Such systems utilize renewable energy sources (wind or solar et al.) to power the electrocatalysis process with components of air (CO₂, N₂) and water as feedstocks to produce fuels or commodity chemicals, for example, electrochemical nitrogen reduction reaction (eNRR). Consequently, understanding the reaction mechanisms and the connection between catalytic performance and the intrinsic property of materials is of great importance. Density functional theory (DFT) based simulation methods proved powerful in exploring the in-depth insights of reaction mechanisms. This Thesis aims to apply DFT based computational methods to the study of electrocatalytic reactions like HER and eNRR and provide guidance for the design of catalysts for such processes. The first two chapters provide a systematic review of theoretical progress in exploring key electrocatalytic reactions' reaction mechanisms. The in-depth insights are presented based on the elucidation of reaction mechanisms, and the origins of different types of catalysts are discussed. The experimental aspects that could be combined with theoretical predictions are also addressed in this part. A significant emphasis is placed on comparing the different design principles for various types of materials. The first part of this thesis focuses on understanding alkaline HER mechanism using Pt/Ru dimer catalysts as the model catalysts. So far, the mechanistic understandings of alkaline HER are limited, especially for the debate on whether there is a singular descriptor for connecting the performance with the thermodynamic quantities. The dissociative chemisorption energy of water (ΔEdiss) as a singular activity descriptor following the analysis of several potential activity descriptors. ΔEdiss is proposed both because it has the capacity to identify the smallest theoretical thermodynamic overpotential and because it scales linearly with the kinetic barrier. These findings will be of immediate benefit to the guide rational development of electrocatalysts via electronic structural engineering to regulate ΔEdiss for the alkaline HER. The second part of the Thesis focuses on revealing the reaction mechanisms for eNRR on newly emerged materials. Firstly, a full picture (activity trends, electronic origins, and design strategies) of single-atom catalysts (SACs) supported on nitrogen-doped carbons as eNRR electrocatalysts are established. To construct such a picture, this work presents systematic studies of 60 types of transition metal SACs supported on nitrogencontaining carbon materials. The results show that the intrinsic activity trends could be established on the basis of the nitrogen adatom adsorption energy (ΔEN*). Furthermore, the influence of metal and support (ligands) on ΔEN* proved to be related to the bonding/antibonding orbital population and regulating the scaling relations for adsorption of intermediates, respectively. Accordingly, a two-step strategy is then proposed for improving the eNNR activity of TM-SACs. Also, the stability of N doped carbon supports and their selectivity in comparison to the competing hydrogen evolution need to be taken into consideration for screening the durable and efficient candidates. Finally, an effective strategy for designing active, stable, and elective SACs based on the mechanistic insights is elaborated to guide future eNRR studies. Secondly, the electrochemical nitrogen reduction reaction (eNRR) activity on binary metal boride is investigated as a model system of metal borides. To elaborate the mechanisms, molybdenum borides (Mo2B, α-MoB, and MoB₂) were first modeled; the results indicate that the crystal structures greatly impact the N₂ adsorption and therefore, the electrocatalytic activity. Our electronic structure investigation suggests that boron p orbital hybrids with dinitrogen π* orbital, and the population on p-π* orbital determine the N₂ adsorption strength. Therefore, the isolated boron site of Mo₂B with less filled pz orbital benefits the activation of N₂ and weaken the triple bond of dinitrogen. This isolated boron site concept was successfully extended to other metal borides in the form of M₂B (M stands for Ti, Cr, Mn, Fe, Co, Ni, Ta, W). Mo₂B, Fe₂B, and Co₂B were discovered as the most promising candidates with low limiting potentials due to the appropriate adsorption strength of reaction intermediates led by moderate pz filling. This part provides insights for designing metal borides as promising eNRR catalysts. The third part of this Thesis explores the reaction mechanisms for C-N bond formation during urea production of CO(₂) electrolysis with nitrogen containing feedstocks which is an emerging opportunity for the electrosynthesis of commodity chemicals. The feedstocks for such reactions are CO or CO₂ as carbon source, nitrate, nitrite, or NOx as the nitrogen source. The bottleneck for such systems still remains the lack of understanding of detailed reaction mechanisms and key reaction intermediates, which impedes the ration-al design of efficient catalyst materials for these reactions. To fill this knowledge gap, we applied ab initio molecular dynamics (AIMD) with explicit water molecules under different electrode potentials to explore the C-N bond formation mechanisms toward urea (CO(NH₂)₂) production on Cu (100) surface in neutral electrolyte. Based on the kinetic barrier values, we established reaction pathways towards urea and ammonia; the former involves C-N bond formation while the latter does not. We found a potential-dependent mechanism that accounts for the selective production of urea. For more positive electrode potentials, i.e., near -0.75 V vs standard hydrogen electrode (SHE), the coupling between NH* and CO* is kinetically viable which leads to the formation of urea. At more negative potentials, e.g., around -1.5 V vs. SHE, CO₂ reduction was accelerated and suppress urea and ammonia formations. These findings propose the potential-dependent mechanisms for C-N bond formation, which would benefit the design of new catalyst materials for electrosynthesis of more value-added products. At last, the challenges and perspectives of computational simulations based on DFT methods for electrocatalysis were discussed. These methods proved powerful in revealing the reaction mechanism as well as guiding the rational design of catalysts for electrocatalytic processes such as HER and NRR. Moreover, advanced simulation methods based on AIMD methods would be capable of capturing more realistic reaction environment and provide new comprehensive insights for the electrochemical process.