Tomáš Bučko
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Tomáš Bučko has a total of 174 co-authors. Here are the top co-authors by number of shared publications:
J. Hafner
Department of Chemistry and Biochemistry, University of Regina, Regina, Saskatchewan S4S 0A2, Canada, and Faculty of Physics and Center for Computational Materials Science Materialphysik, Universität Wien, Sensengasse 8, A-1090 Wien, Austria
15 shared publicationsJürgen Hafner
Fakultät für Physik and Center for Computational Materials Science, Universität Wien, Sensengasse, Wien 1090, Austria
15 shared publications- Michael Badawi
Université de Lorraine, CNRS, Laboratoire Lorrain de Chimie Moléculaire, F-54000 Nancy, France
12 shared publications L. Benco
Fakultät für Physik and Center for Computational Materials Science, Universität Wien, Sensengasse 8, A-1090 Wien, Austria, Institute of Inorganic Chemistry, Slovak Academy of Sciences, Dubravska cesta 9, SK-84536 Bratislava, Slovak Republic, and J. Heyrovsky Institute of Physical Chemistry, Academy of Sciences of the Czech Republic, Dolejskova 3, CZ-18223 Prague 8, Czech Republic
11 shared publications- Céline Chizallet
IFP Energies nouvelles
10 shared publications Lubomir Benco
Institut für Materialphysik and Center for Computational Materials Science, Universität Wien, Sensengasse 8, A−1090 Vienna, Austria
10 shared publications
Latest Publications
- Open Access
- 30 March 2026
- Published by Elsevier in SSRN Electronic Journal
Isobutanol Dehydration Catalyzed by (100) Surface of Ferrierite: From Static DFT to Machine Learning Accelerated Molecular Dynamics
Ferrierite is industrially used as a catalyst for the dehydration of isobutanol into butenes and the critical part of this transformation is presumably catalyzed by Bronsted acid (BA) sites located on its external surface. In this work, we present a DFT investigation of a complete reaction network over the T1O3 BA site at the (100) surface of this zeolite. Exploration of reaction mechanisms by means of the static approach revealed that the transformation towards all products proceed in two steps - dehydration and deprotonation. Free energy static calculations showed that the dehydration step is rate determining for all competing reaction channels. Free energetics of the dehydration reaction was therefore investigated using ab initio molecular dynamics accelerated by machine learned forcefields (MLFFs) with accuracy controlled via a correction scheme based on machine learning perturbation theory. The activation free energies for transformations towards branched and linear products computed using the PBE+D2 functional (111.3 and 125.7 kJ/mol, respectively) suggest a preference of the former reaction channel. Separating dispersion correction from MLFFs allowed us to efficiently recompute free energies using other dispersion corrections (no correction, D3, D3(BJ) and D4) and confirm that this factor does not affect prediction on the preferred reaction channel.Our results are in agreement with our previous investigation of analogous reactions catalyzed by bulk chabazite, but not with experimental selectivities observed for ferrierite, which suggests that the considered BA sites are not at the origin of the specific selectivity of ferrierite for linear alkenes.
- 30 March 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessIsobutanol Dehydration Catalyzed by (100) Surface of Ferrierite: From Static DFT to Machine Learning Accelerated Molecular Dynamics
Ferrierite is industrially used as a catalyst for the dehydration of isobutanol into butenes and the critical part of this transformation is presumably catalyzed by Bronsted acid (BA) sites located on its external surface. In this work, we present a DFT investigation of a complete reaction network over the T1O3 BA site at the (100) surface of this zeolite. Exploration of reaction mechanisms by means of the static approach revealed that the transformation towards all products proceed in two steps - dehydration and deprotonation. Free energy static calculations showed that the dehydration step is rate determining for all competing reaction channels. Free energetics of the dehydration reaction was therefore investigated using ab initio molecular dynamics accelerated by machine learned forcefields (MLFFs) with accuracy controlled via a correction scheme based on machine learning perturbation theory. The activation free energies for transformations towards branched and linear products computed using the PBE+D2 functional (111.3 and 125.7 kJ/mol, respectively) suggest a preference of the former reaction channel. Separating dispersion correction from MLFFs allowed us to efficiently recompute free energies using other dispersion corrections (no correction, D3, D3(BJ) and D4) and confirm that this factor does not affect prediction on the preferred reaction channel.Our results are in agreement with our previous investigation of analogous reactions catalyzed by bulk chabazite, but not with experimental selectivities observed for ferrierite, which suggests that the considered BA sites are not at the origin of the specific selectivity of ferrierite for linear alkenes.
- 29 October 2025
- Published by Elsevier in Journal of Catalysis
- Vol. 453
Linear versus branched products: How dynamical effects influence the transformations of isobutanol catalyzed by acidic zeolites
- 29 October 2025
- Published by Elsevier in Journal of Catalysis
- Vol. 453
Linear versus branched products: How dynamical effects influence the transformations of isobutanol catalyzed by acidic zeolites
- Open Access
- 28 September 2025
- Published by AIP Publishing in The Journal of Chemical Physics
- Vol. 163 (12)
Machine learning accelerates Raman computations from molecular dynamics for materials science
Raman spectroscopy is a powerful experimental technique for characterizing molecules and materials that is used in many laboratories. First-principles theoretical calculations of Raman spectra are important because they elucidate the microscopic effects underlying Raman activity in these systems. These calculations are often performed using the canonical harmonic approximation, which cannot capture certain thermal changes in the Raman response. Anharmonic vibrational effects were recently found to play crucial roles in several materials, which motivates theoretical treatments of the Raman effect beyond harmonic phonons. While Raman spectroscopy from molecular dynamics (MD-Raman) is a well-established approach that includes anharmonic vibrations and further relevant thermal effects, MD-Raman computations were long considered to be computationally too expensive for practical materials computations. In this perspective article, we highlight that recent advances in the context of machine learning have now dramatically accelerated the involved computational tasks without sacrificing accuracy or predictive power. These recent developments highlight the increasing importance of MD-Raman and related methods as versatile tools for theoretical prediction and characterization of molecules and materials.
- 28 September 2025
- Published by AIP Publishing in The Journal of Chemical Physics
- Vol. 163 (12)
Open AccessMachine learning accelerates Raman computations from molecular dynamics for materials science
Raman spectroscopy is a powerful experimental technique for characterizing molecules and materials that is used in many laboratories. First-principles theoretical calculations of Raman spectra are important because they elucidate the microscopic effects underlying Raman activity in these systems. These calculations are often performed using the canonical harmonic approximation, which cannot capture certain thermal changes in the Raman response. Anharmonic vibrational effects were recently found to play crucial roles in several materials, which motivates theoretical treatments of the Raman effect beyond harmonic phonons. While Raman spectroscopy from molecular dynamics (MD-Raman) is a well-established approach that includes anharmonic vibrations and further relevant thermal effects, MD-Raman computations were long considered to be computationally too expensive for practical materials computations. In this perspective article, we highlight that recent advances in the context of machine learning have now dramatically accelerated the involved computational tasks without sacrificing accuracy or predictive power. These recent developments highlight the increasing importance of MD-Raman and related methods as versatile tools for theoretical prediction and characterization of molecules and materials.
- 23 September 2025
- Published by AIP Publishing in The Journal of Chemical Physics
- Vol. 163 (12)
Adsorption of benzene on graphene studied by ML-accelerated ab initio molecular dynamics simulations
Periodic ab initio molecular dynamics simulations accelerated by machine learning were conducted to investigate the adsorption of benzene on graphene at a coverage of 0.11 ML and a temperature of 150 K. For this purpose, seven density functional approximations (DFAs) were used, differing in the exchange–correlation functional (PBE, SCAN, HSE06, and vdW-optB86b) and the long-range dispersion correction (D2, D3, D4, and MBD). Thermal effects on the structure, enthalpy (ΔadsH), and Gibbs free energy (ΔadsG) of adsorption were analyzed. The relatively fast surface diffusion of the adsorbate was identified as the dominant anharmonic effect causing an increase in ΔadsG of up to 50% compared to static harmonic calculations. In contrast, the anharmonic effect on ΔadsH was shown to be nearly negligible. The computed thermodynamic data were used to predict the desorption kinetics parameters, i.e., activation energy (Ea), Arrhenius pre-factor (ν), and position of desorption maximum (Tmax). Among the DFAs tested, the PBE + D4 method was found to provide the best overall agreement with the experimental data. The SCAN + MBD, PBE + D3, and PBE + D2 methods predict Ea values that agree with the experimentally determined value within the reported uncertainty but significantly underestimate Tmax.
- 23 September 2025
- Published by AIP Publishing in The Journal of Chemical Physics
- Vol. 163 (12)
Adsorption of benzene on graphene studied by ML-accelerated ab initio molecular dynamics simulations
Periodic ab initio molecular dynamics simulations accelerated by machine learning were conducted to investigate the adsorption of benzene on graphene at a coverage of 0.11 ML and a temperature of 150 K. For this purpose, seven density functional approximations (DFAs) were used, differing in the exchange–correlation functional (PBE, SCAN, HSE06, and vdW-optB86b) and the long-range dispersion correction (D2, D3, D4, and MBD). Thermal effects on the structure, enthalpy (ΔadsH), and Gibbs free energy (ΔadsG) of adsorption were analyzed. The relatively fast surface diffusion of the adsorbate was identified as the dominant anharmonic effect causing an increase in ΔadsG of up to 50% compared to static harmonic calculations. In contrast, the anharmonic effect on ΔadsH was shown to be nearly negligible. The computed thermodynamic data were used to predict the desorption kinetics parameters, i.e., activation energy (Ea), Arrhenius pre-factor (ν), and position of desorption maximum (Tmax). Among the DFAs tested, the PBE + D4 method was found to provide the best overall agreement with the experimental data. The SCAN + MBD, PBE + D3, and PBE + D2 methods predict Ea values that agree with the experimentally determined value within the reported uncertainty but significantly underestimate Tmax.
- Open Access
- 4 September 2025
- Published by American Chemical Society (ACS) in The Journal of Physical Chemistry C
- Vol. 129 (37) , 16742-16755
Anharmonic Correction to the Adsorption Free Energy of Oxygen-Containing Intermediates on Pt(111) by Machine-Learned Force Field-Based Thermodynamic Integrations
- 4 September 2025
- Published by American Chemical Society (ACS) in The Journal of Physical Chemistry C
- Vol. 129 (37) , 16742-16755
Open AccessAnharmonic Correction to the Adsorption Free Energy of Oxygen-Containing Intermediates on Pt(111) by Machine-Learned Force Field-Based Thermodynamic Integrations
- Open Access
Machine Learning Accelerates Raman Computations from Molecular Dynamics for Materials Science
Raman spectroscopy is a powerful experimental technique for characterizing molecules and materials that is used in many laboratories. First-principles theoretical calculations of Raman spectra are important because they elucidate the microscopic effects underlying Raman activity in these systems. These calculations are often performed using the canonical harmonic approximation which cannot capture certain thermal changes in the Raman response. Anharmonic vibrational effects were recently found to play crucial roles in several materials, which motivates theoretical treatments of the Raman effect beyond harmonic phonons. While Raman spectroscopy from molecular dynamics (MD-Raman) is a well-established approach that includes anharmonic vibrations and further relevant thermal effects, MD-Raman computations were long considered to be computationally too expensive for practical materials computations. In this perspective article, we highlight that recent advances in the context of machine learning have now dramatically accelerated the involved computational tasks without sacrificing accuracy or predictive power. These recent developments highlight the increasing importance of MD-Raman and related methods as versatile tools for theoretical prediction and characterization of molecules and materials.
Open AccessMachine Learning Accelerates Raman Computations from Molecular Dynamics for Materials Science
Raman spectroscopy is a powerful experimental technique for characterizing molecules and materials that is used in many laboratories. First-principles theoretical calculations of Raman spectra are important because they elucidate the microscopic effects underlying Raman activity in these systems. These calculations are often performed using the canonical harmonic approximation which cannot capture certain thermal changes in the Raman response. Anharmonic vibrational effects were recently found to play crucial roles in several materials, which motivates theoretical treatments of the Raman effect beyond harmonic phonons. While Raman spectroscopy from molecular dynamics (MD-Raman) is a well-established approach that includes anharmonic vibrations and further relevant thermal effects, MD-Raman computations were long considered to be computationally too expensive for practical materials computations. In this perspective article, we highlight that recent advances in the context of machine learning have now dramatically accelerated the involved computational tasks without sacrificing accuracy or predictive power. These recent developments highlight the increasing importance of MD-Raman and related methods as versatile tools for theoretical prediction and characterization of molecules and materials.
- Open Access
- 18 September 2024
- Published by American Chemical Society (ACS) in Journal of the American Chemical Society
- Vol. 146 (39) , 26863-26876
Rapid Characterization of Point Defects in Solid-State Ion Conductors Using Raman Spectroscopy, Machine-Learning Force Fields, and Atomic Raman Tensors
The successful design of solid-state photo- and electrochemical devices depends on the careful engineering of point defects in solid-state ion conductors. Characterization of point defects is critical to these efforts, but the best-developed techniques are difficult and time-consuming. Raman spectroscopy─with its exceptional speed, flexibility, and accessibility─is a promising alternative. Raman signatures arise from point defects due to local symmetry breaking and structural distortions. Unfortunately, the assignment of these signatures is often hampered by a shortage of reference compounds and corresponding reference spectra. This issue can be circumvented by calculation of defect-induced Raman signatures from first principles, but this is computationally demanding. Here, we introduce an efficient computational procedure for the prediction of point defect Raman signatures in solid-state ion conductors. Our method leverages machine-learning force fields and "atomic Raman tensors", i.e., polarizability fluctuations due to motions of individual atoms. We find that our procedure reduces computational cost by up to 80% compared to existing first-principles frozen-phonon approaches. These efficiency gains enable synergistic computational-experimental investigations, in our case allowing us to precisely interpret the Raman spectra of Sr(Ti0.94Ni0.06)O3-δ, a model oxygen ion conductor. By predicting Raman signatures of specific point defects, we determine the nature of dominant defects and unravel impacts of temperature and quenching on in situ and ex situ Raman spectra. Specifically, our findings reveal the temperature-dependent distribution and association behavior of oxygen vacancies and nickel substitutional defects. Overall, our approach enables rapid Raman-based characterization of point defects to support defect engineering in novel solid-state ion conductors.
- 18 September 2024
- Published by American Chemical Society (ACS) in Journal of the American Chemical Society
- Vol. 146 (39) , 26863-26876
Open AccessRapid Characterization of Point Defects in Solid-State Ion Conductors Using Raman Spectroscopy, Machine-Learning Force Fields, and Atomic Raman Tensors
The successful design of solid-state photo- and electrochemical devices depends on the careful engineering of point defects in solid-state ion conductors. Characterization of point defects is critical to these efforts, but the best-developed techniques are difficult and time-consuming. Raman spectroscopy─with its exceptional speed, flexibility, and accessibility─is a promising alternative. Raman signatures arise from point defects due to local symmetry breaking and structural distortions. Unfortunately, the assignment of these signatures is often hampered by a shortage of reference compounds and corresponding reference spectra. This issue can be circumvented by calculation of defect-induced Raman signatures from first principles, but this is computationally demanding. Here, we introduce an efficient computational procedure for the prediction of point defect Raman signatures in solid-state ion conductors. Our method leverages machine-learning force fields and "atomic Raman tensors", i.e., polarizability fluctuations due to motions of individual atoms. We find that our procedure reduces computational cost by up to 80% compared to existing first-principles frozen-phonon approaches. These efficiency gains enable synergistic computational-experimental investigations, in our case allowing us to precisely interpret the Raman spectra of Sr(Ti0.94Ni0.06)O3-δ, a model oxygen ion conductor. By predicting Raman signatures of specific point defects, we determine the nature of dominant defects and unravel impacts of temperature and quenching on in situ and ex situ Raman spectra. Specifically, our findings reveal the temperature-dependent distribution and association behavior of oxygen vacancies and nickel substitutional defects. Overall, our approach enables rapid Raman-based characterization of point defects to support defect engineering in novel solid-state ion conductors.
- Open Access
- 23 July 2024
- Published by Royal Society of Chemistry (RSC) in Catalysis Science & Technology
- Vol. 14 (18) , 5314-5323
Machine learning thermodynamic perturbation theory offers accurate activation free energies at the RPA level for alkene isomerization in zeolites
The determination of accurate free energy barriers for reactions catalyzed by proton-exchanged zeolites by quantum chemistry approaches is a challenge. While ab initio molecular dynamics is often required to sample correctly the various states described by the system, the level of theory also has a crucial impact. In the present work, we report the determination of accurate barriers for a type B isomerization of a monobranched C7 alkene (4-methyl-hex-1-ene) into a dibranched tertiary cation inside a protonated chabazite zeolite. This is done by using the Machine Learning Thermodynamic Perturbation Theory (MLPT) at the Random Phase Approximation (RPA) level, on the basis of blue-moon sampling dynamic data obtained at the Generalized Gradient Approximation (GGA) level (PBE+D2). The comparison of PBE+D2 and RPA profiles shows that the former overstabilizes cationic intermediates with respect to neutral ones. The transition state of the isomerization is a non-classical edge protonated cyclopropane, the stabilization of which is lower than that of the π-complex when PBE+D2 is replaced by RPA, but higher than that of the classical tertiary carbenium. Consequently, the backward isomerization barrier is decreased. Applying the MLPT approach to recompute the free energy barriers with various dispersion correction schemes to the PBE energies shows that none of the schemes is sufficient to improve both the forward and backward barriers with respect to the RPA reference. These data complement previously determined alkene cracking barriers [Rey et al., Angew. Chem., Int. Ed., 2024, 63, e202312392], thanks to which it is possible to compare the presently determined barriers with reference experimental data [Schweitzer et al., ACS Catal., 2022, 12, 1068–1081]. The agreement with experiments is significantly improved at the RPA with respect to GGA. Chemical accuracy is approached (maximum deviation of 6.4 kJ mol−1), opening the door to predictive kinetic modelling starting from first principles approaches.
- 23 July 2024
- Published by Royal Society of Chemistry (RSC) in Catalysis Science & Technology
- Vol. 14 (18) , 5314-5323
Open AccessMachine learning thermodynamic perturbation theory offers accurate activation free energies at the RPA level for alkene isomerization in zeolites
The determination of accurate free energy barriers for reactions catalyzed by proton-exchanged zeolites by quantum chemistry approaches is a challenge. While ab initio molecular dynamics is often required to sample correctly the various states described by the system, the level of theory also has a crucial impact. In the present work, we report the determination of accurate barriers for a type B isomerization of a monobranched C7 alkene (4-methyl-hex-1-ene) into a dibranched tertiary cation inside a protonated chabazite zeolite. This is done by using the Machine Learning Thermodynamic Perturbation Theory (MLPT) at the Random Phase Approximation (RPA) level, on the basis of blue-moon sampling dynamic data obtained at the Generalized Gradient Approximation (GGA) level (PBE+D2). The comparison of PBE+D2 and RPA profiles shows that the former overstabilizes cationic intermediates with respect to neutral ones. The transition state of the isomerization is a non-classical edge protonated cyclopropane, the stabilization of which is lower than that of the π-complex when PBE+D2 is replaced by RPA, but higher than that of the classical tertiary carbenium. Consequently, the backward isomerization barrier is decreased. Applying the MLPT approach to recompute the free energy barriers with various dispersion correction schemes to the PBE energies shows that none of the schemes is sufficient to improve both the forward and backward barriers with respect to the RPA reference. These data complement previously determined alkene cracking barriers [Rey et al., Angew. Chem., Int. Ed., 2024, 63, e202312392], thanks to which it is possible to compare the presently determined barriers with reference experimental data [Schweitzer et al., ACS Catal., 2022, 12, 1068–1081]. The agreement with experiments is significantly improved at the RPA with respect to GGA. Chemical accuracy is approached (maximum deviation of 6.4 kJ mol−1), opening the door to predictive kinetic modelling starting from first principles approaches.
- 1 May 2024
- Published by AIP Publishing in The Journal of Chemical Physics
- Vol. 160 (17)
Thermodynamics of the gas-phase dimerization of formic acid: Fully anharmonic finite temperature calculations at the CCSD(T) and many DFT levels
A proof-of-concept study is undertaken to demonstrate the utility of the machine learning combined with the thermodynamic perturbation theory (MLPT) to test the accuracy of electronic structure methods in finite-temperature thermodynamic calculations. As a test example, formic acid dimer is chosen, which is one of the systems included in the popular benchmark set S22 [Jurečka et al., Phys. Chem. Chem. Phys. 8, 1985–1993 (2006)]. Starting from the explicit molecular dynamics and thermodynamic integration performed at the PBE + D2 level, the MLPT is used to obtain fully anharmonic dimerization free and internal energies at the reference quality CCSD(T) level and 19 different density functional approximations, including GGA, meta-GGA, non-local, and hybrid functionals with and without dispersion corrections. Our finite-temperature results are shown to be both qualitatively and quantitatively different from those obtained using the conventional benchmarking strategy based on fixed structures. The hybrid functional HSE06 is identified as the best performing approximate method tested, with the errors in free and internal energies of dimerization being 36 and 41 meV, respectively.
- 1 May 2024
- Published by AIP Publishing in The Journal of Chemical Physics
- Vol. 160 (17)
Thermodynamics of the gas-phase dimerization of formic acid: Fully anharmonic finite temperature calculations at the CCSD(T) and many DFT levels
A proof-of-concept study is undertaken to demonstrate the utility of the machine learning combined with the thermodynamic perturbation theory (MLPT) to test the accuracy of electronic structure methods in finite-temperature thermodynamic calculations. As a test example, formic acid dimer is chosen, which is one of the systems included in the popular benchmark set S22 [Jurečka et al., Phys. Chem. Chem. Phys. 8, 1985–1993 (2006)]. Starting from the explicit molecular dynamics and thermodynamic integration performed at the PBE + D2 level, the MLPT is used to obtain fully anharmonic dimerization free and internal energies at the reference quality CCSD(T) level and 19 different density functional approximations, including GGA, meta-GGA, non-local, and hybrid functionals with and without dispersion corrections. Our finite-temperature results are shown to be both qualitatively and quantitatively different from those obtained using the conventional benchmarking strategy based on fixed structures. The hybrid functional HSE06 is identified as the best performing approximate method tested, with the errors in free and internal energies of dimerization being 36 and 41 meV, respectively.