Roman Fanta

All Publications

• Why Nondynamic Correlation Matters forpipiStacking? Lessons from the Benzene Dimer. The journal of physical chemistry letters Fanta, R., Jurecka, P., Dubecky, M. 2025: 10982-10988

  Abstract

  Two leading methods for benchmarking pipi interactions are fixed-node diffusion Monte Carlo (DMC) and coupled cluster with singles, doubles, and perturbative triples [CCSD(T)]. The parallel-displaced benzene dimer (BZPD) is a key model for assessing the performance of theoretical approaches in describing these interactions. Reference calculations, symmetry-adapted perturbation theory, and correlation energy decompositions highlight the subtle but critical role of weak nondynamic correlation effects in achieving benchmark interaction energies for BZPD. While single-determinant DMC (SDDMC), using nodes from DFT or Hartree-Fock, performs well for many noncovalent systems, our analysis shows that neglecting weak nondynamic correlations in mean-field trial wave functions leads to systematic underbinding in SDDMC energy differences, in contrast to CCSD(T)/CBS results. These findings underscore the limitations of SDDMC for describing pipi interactions and emphasize the need for improved nodal descriptions to establish DMC as a reliable benchmark method for larger complexes.

  View details for DOI 10.1021/acs.jpclett.5c02576

  View details for PubMedID 41084851

• Spin State Modulation in M-N-C Single-Atom Catalysts for Oxygen Electrocatalysis ACS CATALYSIS Jung, H., Fanta, R., Hossain, M., Bajdich, M. 2025

  View details for DOI 10.1021/acscatal.5c04591

  View details for Web of Science ID 001570163000001

• Resolution of Selectivity Steps of CO Reduction Reaction on Copper by Quantum Monte Carlo. The journal of physical chemistry letters Fanta, R., Bajdich, M. 2025: 1494-1500

  Abstract

  Electrochemical reduction of carbon monoxide to valuable fuels and chemicals on copper surfaces remains a challenging area in catalysis due to a limited understanding of adsorption mechanisms and reaction pathways. Although density functional theory (DFT)-based studies have investigated these processes, their accuracy varies across different functionals. Here, we present the application of fixed-node diffusion Monte Carlo (FNDMC) to benchmark the adsorption energies of CO*, H*, and key CO reduction reaction (CORR) intermediates, COH* and CHO* on the Cu(111) surface. Our results for CO* and H* adsorption energies closely align with experimentally measured chemisorption reactions, highlighting the limitations of DFT and providing site-specific energy comparisons that are often not available experimentally. Additionally, we explore the effect of explicit solvation, demonstrating how water stabilizes the COH* over CHO*, thus suggesting a critical role of COH* in CORR. Finally, we release our high-accuracy FNDMC benchmarks for testing and developing new DFT functionals for electrocatalysis. Overall, this study underscores the potential of FNDMC for detailed surface chemistry studies and offers new insights into catalytic processes.

  View details for DOI 10.1021/acs.jpclett.4c03409

  View details for PubMedID 39898589