Theodorus de Bruin
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
Theodorus de Bruin has a total of 167 co-authors. Here are the top co-authors by number of shared publications:
- Hervé Toulhoat
Sorbonne Universite
12 shared publications - Carlos Nieto-Draghi
IFP Energies nouvelles
11 shared publications - Pascal Raybaud
IFP Energies Nouvelles, Rond-Point de l’échangeur de Solaize, BP3, 69360 Solaize, France
8 shared publications - Céline Chizallet
IFP Energies nouvelles
6 shared publications - Carine Michel
ENS de Lyon CNRS Laboratoire De Chimie Lyon France
5 shared publications - Stephan Niklaus Steinmann
CNRS, ENS de Lyon, LCH, UMR 5182
5 shared publications
Latest Publications
- Open Access
- 27 September 2026
- Published by American Chemical Society (ACS) in ChemRxiv
A Critical Review of Theoretical Approaches to Compute Ionic Conductivity in Solid-State Electrolytes: The Case Study of Li3YCl6 and Li3YBr6
The development of next-generation batteries strongly depends on the design of electrolytes that ensure efficient ionic transport. In this context, computational in silico approaches have become increasingly important. Among them, molecular dynamics simulations are particularly powerful for investigating constitutive elements of batteries and providing a deeper understanding of structure-property relationships in solid-state electrolytes. Yet, for these systems, obtaining accurate ionic conductivity and activation energy is not straightforward as finding a force field to accurately describe the system and a robust workflow is challenging. In this work, a classical force field (CFF), based on the SoftBV bond-valence approach, and a machine learning potential (MLP) trained on density functional theory (DFT) energies and forces using the MACE neural network architecture, were optimized and compared to investigate the two solid electrolytes, Li3YCl6 (LYC) and Li3YBr6 (LYB). Our study proposes an integrated workflow (structure generation, property analysis with relevant statistical sampling) using standard methodologies based on the Einstein equations and a new method called STACIE[Toraman et al., J. Chem. Inf. Model., 65, 19, 2025, DOI:10.1021/acs.jcim.5c01475] (STable AutoCorrelation Integral Estimator) deriving these properties through the analysis of power spectral densities for transport property evaluation. Based on the evaluation and analysis of the contribution of the crosscorrelations between mobile ions, we provide clear rules to validate the use of Nernst-Einstein’s approximation to estimate the ionic conductivity. Results show trends in agreement with the literature. Notably the correct order of the conductivity activation energies Ea(LYC) > Ea(LYB), observed experimentally, is maintained. Overall, while CFF can provide an adequate representation when guided by experimental data, the strength of the MACE potential lies in its predictive capability. Using configurations generated by DFT calculations, the MLP generates trajectories with accurate crystallographic features from which realistic transport properties are extracted. Compared to their CFF counterparts, MLP results also show better mechanical stability at higher temperature for LYB, a softer system based on the analysis of solid-phase mechanical properties.
- 27 September 2026
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessA Critical Review of Theoretical Approaches to Compute Ionic Conductivity in Solid-State Electrolytes: The Case Study of Li3YCl6 and Li3YBr6
The development of next-generation batteries strongly depends on the design of electrolytes that ensure efficient ionic transport. In this context, computational in silico approaches have become increasingly important. Among them, molecular dynamics simulations are particularly powerful for investigating constitutive elements of batteries and providing a deeper understanding of structure-property relationships in solid-state electrolytes. Yet, for these systems, obtaining accurate ionic conductivity and activation energy is not straightforward as finding a force field to accurately describe the system and a robust workflow is challenging. In this work, a classical force field (CFF), based on the SoftBV bond-valence approach, and a machine learning potential (MLP) trained on density functional theory (DFT) energies and forces using the MACE neural network architecture, were optimized and compared to investigate the two solid electrolytes, Li3YCl6 (LYC) and Li3YBr6 (LYB). Our study proposes an integrated workflow (structure generation, property analysis with relevant statistical sampling) using standard methodologies based on the Einstein equations and a new method called STACIE[Toraman et al., J. Chem. Inf. Model., 65, 19, 2025, DOI:10.1021/acs.jcim.5c01475] (STable AutoCorrelation Integral Estimator) deriving these properties through the analysis of power spectral densities for transport property evaluation. Based on the evaluation and analysis of the contribution of the crosscorrelations between mobile ions, we provide clear rules to validate the use of Nernst-Einstein’s approximation to estimate the ionic conductivity. Results show trends in agreement with the literature. Notably the correct order of the conductivity activation energies Ea(LYC) > Ea(LYB), observed experimentally, is maintained. Overall, while CFF can provide an adequate representation when guided by experimental data, the strength of the MACE potential lies in its predictive capability. Using configurations generated by DFT calculations, the MLP generates trajectories with accurate crystallographic features from which realistic transport properties are extracted. Compared to their CFF counterparts, MLP results also show better mechanical stability at higher temperature for LYB, a softer system based on the analysis of solid-phase mechanical properties.
- Open Access
- 18 July 2025
- Published by American Chemical Society (ACS) in ChemRxiv
Molecular simulation of CO2 and H2O adsorption/deformation on model cement materials
Carbon capture and storage (CCS) is a critical technology for mitigating atmospheric CO2 levels and achieving carbon neutrality, with the potential to capture billions of tons of CO2 annually. To ensure secure geological storage, it is essential to deeply understand well cement degradation, car-bonation processes, and wellbore integrity under CO2 injection conditions. This study employs Monte Carlo simulations to investigate the interactions between CO2 and H2O in cement materials (portlandite and tobermorite) at 66°C and 100 bar, focusing on thermodynamic mechanisms pre-ceding chemical degradation. Model validation was conducted by comparing the heat of water ad-sorption (Qst) with literature values, which confirmed the accuracy of the simulations and empha-size the role of surface water in CO2 adsorption. The analyses reveal CO2 enrichment at water in-terfaces and demonstrate the significant impact of water content on material interactions. These findings provide valuable insights for refining predictive models and guiding future research on cement integrity under CO2 exposure.
- 18 July 2025
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessMolecular simulation of CO2 and H2O adsorption/deformation on model cement materials
Carbon capture and storage (CCS) is a critical technology for mitigating atmospheric CO2 levels and achieving carbon neutrality, with the potential to capture billions of tons of CO2 annually. To ensure secure geological storage, it is essential to deeply understand well cement degradation, car-bonation processes, and wellbore integrity under CO2 injection conditions. This study employs Monte Carlo simulations to investigate the interactions between CO2 and H2O in cement materials (portlandite and tobermorite) at 66°C and 100 bar, focusing on thermodynamic mechanisms pre-ceding chemical degradation. Model validation was conducted by comparing the heat of water ad-sorption (Qst) with literature values, which confirmed the accuracy of the simulations and empha-size the role of surface water in CO2 adsorption. The analyses reveal CO2 enrichment at water in-terfaces and demonstrate the significant impact of water content on material interactions. These findings provide valuable insights for refining predictive models and guiding future research on cement integrity under CO2 exposure.
- Open Access
- 19 June 2025
- Published by Royal Society of Chemistry (RSC) in EES Batteries
- Vol. 1 (4) , 788-802
Challenges and opportunities in using Kinetic Monte Carlo for battery research and innovation
Kinetic Monte Carlo is already exploited to understand and improve batteries via molecular-level insights, but would be even more reliable with improved algorithms for time-scale disparities, more flexible lattices and more accurate time-constants.
- 19 June 2025
- Published by Royal Society of Chemistry (RSC) in EES Batteries
- Vol. 1 (4) , 788-802
Open AccessChallenges and opportunities in using Kinetic Monte Carlo for battery research and innovation
Kinetic Monte Carlo is already exploited to understand and improve batteries via molecular-level insights, but would be even more reliable with improved algorithms for time-scale disparities, more flexible lattices and more accurate time-constants.
- 14 March 2025
- Published by American Chemical Society (ACS) in Langmuir
- Vol. 41 (11) , 7272-7282
A Coarse-Grained Model Describing the Critical Micelle Concentration of Perfluoroalkyl Surfactants in Ionic Aqueous Phase
In this study, dissipative particle dynamics (DPD) simulations were employed to determine the critical micelle concentration (CMC) of perfluoroalkyl and polyfluoroalkyl substances (PFAS) in ionic aqueous solutions. This approach provides precise CMC data for PFAS surfactants in the presence of various ionic species, thereby addressing a gap in the current literature. Additionally, this study contributes to the development of open-source molecular force fields for charged perfluorinated compounds, which are currently limited. These models incorporate hydration free energy values obtained from density functional theory (DFT) and account for ionic interactions through a well-established linear relationship. Hydrophobic interactions between the surfactant tail and water were fine-tuned to match the CMC of chosen surfactants. Then, the DPD models successfully predicted CMC values for a diverse range of surfactants, including those based on hydrocarbons and PFAS, demonstrating the ability to represent realistic salinities encountered in natural waters. Experimental validation of the methodology was conducted using sodium n-nonyl sulfate (SNS) and sodium n-dodecyl sulfate (SDS) via interfacial tension measurements, confirming the accurate representation of the CMC changes with salinity. This study enhances our understanding of the behavior of PFAS surfactants in ionic aqueous solutions and provides a valuable tool for predicting CMC values in complex environmental systems.
- 14 March 2025
- Published by American Chemical Society (ACS) in Langmuir
- Vol. 41 (11) , 7272-7282
A Coarse-Grained Model Describing the Critical Micelle Concentration of Perfluoroalkyl Surfactants in Ionic Aqueous Phase
In this study, dissipative particle dynamics (DPD) simulations were employed to determine the critical micelle concentration (CMC) of perfluoroalkyl and polyfluoroalkyl substances (PFAS) in ionic aqueous solutions. This approach provides precise CMC data for PFAS surfactants in the presence of various ionic species, thereby addressing a gap in the current literature. Additionally, this study contributes to the development of open-source molecular force fields for charged perfluorinated compounds, which are currently limited. These models incorporate hydration free energy values obtained from density functional theory (DFT) and account for ionic interactions through a well-established linear relationship. Hydrophobic interactions between the surfactant tail and water were fine-tuned to match the CMC of chosen surfactants. Then, the DPD models successfully predicted CMC values for a diverse range of surfactants, including those based on hydrocarbons and PFAS, demonstrating the ability to represent realistic salinities encountered in natural waters. Experimental validation of the methodology was conducted using sodium n-nonyl sulfate (SNS) and sodium n-dodecyl sulfate (SDS) via interfacial tension measurements, confirming the accurate representation of the CMC changes with salinity. This study enhances our understanding of the behavior of PFAS surfactants in ionic aqueous solutions and provides a valuable tool for predicting CMC values in complex environmental systems.
- 3 December 2024
- Published by American Chemical Society (ACS) in ACS Catalysis
- Vol. 14 (24) , 18442-18456
Deciphering Faujasite Zeolite Dealumination at the Atomic Scale
- 3 December 2024
- Published by American Chemical Society (ACS) in ACS Catalysis
- Vol. 14 (24) , 18442-18456
Deciphering Faujasite Zeolite Dealumination at the Atomic Scale
- 19 April 2024
- Published by American Chemical Society (ACS) in Journal of Chemical Theory and Computation
- Vol. 20 (9) , 3779-3797
Managing Expectations and Imbalanced Training Data in Reactive Force Field Development: An Application to Water Adsorption on Alumina
ReaxFF is a computationally efficient model for reactive molecular dynamics simulations that has been applied to a wide variety of chemical systems. When ReaxFF parameters are not yet available for a chemistry of interest, they must be (re)optimized, for which one defines a set of training data that the new ReaxFF parameters should reproduce. ReaxFF training sets typically contain diverse properties with different units, some of which are more abundant (by orders of magnitude) than others. To find the best parameters, one conventionally minimizes a weighted sum of squared errors over all of the data in the training set. One of the challenges in such numerical optimizations is to assign weights so that the optimized parameters represent a good compromise among all the requirements defined in the training set. This work introduces a new loss function, called Balanced Loss, and a workflow that replaces weight assignment with a more manageable procedure. The training data are divided into categories with corresponding "tolerances", i.e., acceptable root-mean-square errors for the categories, which define the expectations for the optimized ReaxFF parameters. Through the Log-Sum-Exp form of Balanced Loss, the parameter optimization is also a validation of one's expectations, providing meaningful feedback that can be used to reconfigure the tolerances if needed. The new methodology is demonstrated with a nontrivial parametrization of ReaxFF for water adsorption on alumina. This results in a new force field that reproduces both the rare and frequent properties of a validation set not used for training. We also demonstrate the robustness of the new force field with a molecular dynamics simulation of water desorption from a γ-Al2O3 slab model.
- 19 April 2024
- Published by American Chemical Society (ACS) in Journal of Chemical Theory and Computation
- Vol. 20 (9) , 3779-3797
Managing Expectations and Imbalanced Training Data in Reactive Force Field Development: An Application to Water Adsorption on Alumina
ReaxFF is a computationally efficient model for reactive molecular dynamics simulations that has been applied to a wide variety of chemical systems. When ReaxFF parameters are not yet available for a chemistry of interest, they must be (re)optimized, for which one defines a set of training data that the new ReaxFF parameters should reproduce. ReaxFF training sets typically contain diverse properties with different units, some of which are more abundant (by orders of magnitude) than others. To find the best parameters, one conventionally minimizes a weighted sum of squared errors over all of the data in the training set. One of the challenges in such numerical optimizations is to assign weights so that the optimized parameters represent a good compromise among all the requirements defined in the training set. This work introduces a new loss function, called Balanced Loss, and a workflow that replaces weight assignment with a more manageable procedure. The training data are divided into categories with corresponding "tolerances", i.e., acceptable root-mean-square errors for the categories, which define the expectations for the optimized ReaxFF parameters. Through the Log-Sum-Exp form of Balanced Loss, the parameter optimization is also a validation of one's expectations, providing meaningful feedback that can be used to reconfigure the tolerances if needed. The new methodology is demonstrated with a nontrivial parametrization of ReaxFF for water adsorption on alumina. This results in a new force field that reproduces both the rare and frequent properties of a validation set not used for training. We also demonstrate the robustness of the new force field with a molecular dynamics simulation of water desorption from a γ-Al2O3 slab model.
- Open Access
- 10 March 2024
- Published by Wiley in Advanced Functional Materials
- Vol. 34 (30)
A Perspective on the Molecular Modeling of Electrolyte Decomposition Reactions for Solid Electrolyte Interphase Growth in Lithium‐Ion Batteries
The solid electrolyte interphase (SEI) is a thin heterogeneous layer formed at the anode/electrolyte interface in lithium‐ion batteries as a consequence of the reduction of the electrolyte. The initial formation of the SEI inhibits the direct contact between the electrode and the electrolyte and thus protects the battery. However, the composition, structure, and size of the SEI evolve over time and the growth of the SEI is considered the primary mechanism leading to the gradual deterioration of the battery performance. Despite the importance of the SEI and its growth, the atomistic understanding of the underlying elementary reaction steps remains partial. Molecular modeling of the electrolyte decomposition is key to gain detailed insights that are complementary to experiments for the reactions occurring in this heterogenous interphase. In this perspective, the electron transport mechanisms are first described from the anode to the electrolyte through the SEI and highlight the importance of the inorganic/organic interface within the heterogeneous SEI: it is where the electrolyte decomposition reactions are likely to occur. Finally, a view is provided on the current progress on molecular modeling techniques (e.g., Density Functional Theory, force fields, machine learning potentials) of the SEI and the challenges each method faces.
- 10 March 2024
- Published by Wiley in Advanced Functional Materials
- Vol. 34 (30)
Open AccessA Perspective on the Molecular Modeling of Electrolyte Decomposition Reactions for Solid Electrolyte Interphase Growth in Lithium‐Ion Batteries
The solid electrolyte interphase (SEI) is a thin heterogeneous layer formed at the anode/electrolyte interface in lithium‐ion batteries as a consequence of the reduction of the electrolyte. The initial formation of the SEI inhibits the direct contact between the electrode and the electrolyte and thus protects the battery. However, the composition, structure, and size of the SEI evolve over time and the growth of the SEI is considered the primary mechanism leading to the gradual deterioration of the battery performance. Despite the importance of the SEI and its growth, the atomistic understanding of the underlying elementary reaction steps remains partial. Molecular modeling of the electrolyte decomposition is key to gain detailed insights that are complementary to experiments for the reactions occurring in this heterogenous interphase. In this perspective, the electron transport mechanisms are first described from the anode to the electrolyte through the SEI and highlight the importance of the inorganic/organic interface within the heterogeneous SEI: it is where the electrolyte decomposition reactions are likely to occur. Finally, a view is provided on the current progress on molecular modeling techniques (e.g., Density Functional Theory, force fields, machine learning potentials) of the SEI and the challenges each method faces.
- 21 February 2024
- Published by American Chemical Society (ACS) in The Journal of Physical Chemistry C
- Vol. 128 (8) , 3269-3280
Lessons Learned from Semiempirical Methods for the Li-Ion Battery Solid Electrolyte Interphase
- 21 February 2024
- Published by American Chemical Society (ACS) in The Journal of Physical Chemistry C
- Vol. 128 (8) , 3269-3280
Lessons Learned from Semiempirical Methods for the Li-Ion Battery Solid Electrolyte Interphase
- 28 January 2024
- Published by Elsevier in Materials Today Communications
- Vol. 38
Mechanistic insights to CO2 adsorption and activation on hydroxylated chromia (0001) surface
- 28 January 2024
- Published by Elsevier in Materials Today Communications
- Vol. 38
Mechanistic insights to CO2 adsorption and activation on hydroxylated chromia (0001) surface
- 17 January 2024
- Published by American Chemical Society (ACS) in ACS Catalysis
- Vol. 14 (3) , 1639-1652
Impact of Point Defects, External Surfaces Sites, and Si/Al Ratio on Al–O Bonds Hydrolysis Kinetics at the Initiation of Faujasite Dealumination
- 17 January 2024
- Published by American Chemical Society (ACS) in ACS Catalysis
- Vol. 14 (3) , 1639-1652
Impact of Point Defects, External Surfaces Sites, and Si/Al Ratio on Al–O Bonds Hydrolysis Kinetics at the Initiation of Faujasite Dealumination