Bioproductions
Projets financés
Méthodologies et outils transversaux
MAMABIO

MAMABIO

Méthodologie d'apprentissage machine pour la simulation accélérée et prédictive à l'échelle atomique de la transformation de molécules biosourcées.

Les réactions de transformation de molécules biosourcées jouent un rôle essentiel dans la transition énergétique actuelle, de même que les outils numériques requis pour le développement de l’Industrie 4.0. Le projet MAMABIO est au carrefour de ces deux enjeux, dans la mesure où il vise à proposer des méthodologies numériques accélérées, afin de construire des modèles cinétiques à haut potentiel prédictif, dans l’objectif final de développer des procédés de transformation de la biomasse efficients.

Les verrous actuels à ces développements portent sur : 

  • La dynamique moléculaire ab initio (AIMD), requise dans de nombreux cas, représente :
    • Un coût de calcul élevé avec un niveau de théorie (donc de précision) accessible restreint ;
    • des difficultés méthodologiques, a fortiori pour des réactions mal décrites telles que la transformation de molécules biosourcées
  • La nécessité de données expérimentales de référence pour valider les méthodes développées et apporter des données cinétiques complémentaires

Les objectifs du projet :

Méthodologies numériques accélérées

Développement d'outils de Machine Learning (ML) avancés pour accélérer le calcul de constantes de vitesse précises à partir de calculs ab initio.

Modèles cinétiques à haut potentiel prédictif,

Obtention de données cinétiques transitoires à partir de données spectroscopiques operando et de chimiométrie.

Procédés efficients de transformation de molécules biosourcées

 

Organisation du projet: 

Mamabio - Organisation
 
 
 

Durée du projet :
 

Juin 2023 - Mai 2028

 

Responsable scientifique :
 

Céline Chizallet (IFPEN)

 

Le consortium :
 

Établissements d'enseignement supérieur
SUP.png
Instituts de recherche
ONR.png

 

Titre-image Publications

HAL : Dernières publications

  • [hal-04504311] Reference‐Quality Free Energy Barriers in Catalysis from Machine Learning Thermodynamic Perturbation Theory

    For the first time, we report calculations of the free energies of activation of cracking and isomerization reactions of alkenes that combine several different electronic structure methods with molecular dynamics simulations. We demonstrate that the use of a high level of theory (here Random Phase Approximation—RPA) is necessary to bridge the gap between experimental and computed values. These transformations, catalyzed by zeolites and proceeding via cationic intermediates and transition states, are building blocks of many chemical transformations for valorization of long chain paraffins originating, e.g., from plastic waste, vegetable oils, Fischer–Tropsch waxes or crude oils. Compared with the free energy barriers computed at the PBE+D2 production level of theory via constrained ab initio molecular dynamics, the barriers computed at the RPA level by the application of Machine Learning thermodynamic Perturbation Theory (MLPT) show a significant decrease for isomerization reaction and an increase of a similar magnitude for cracking, yielding an unprecedented agreement with the results obtained by experiments and kinetic modeling.

    ano.nymous@ccsd.cnrs.fr.invalid (Jérôme Rey) 14 Mar 2024

    https://ifp.hal.science/hal-04504311v1
  • [hal-04878044] 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.

    ano.nymous@ccsd.cnrs.fr.invalid (Jérôme Rey) 09 Jan 2025

    https://ifp.hal.science/hal-04878044v1
  • [hal-04592530] Importance of Dynamic Effects in Isobutanol to Linear Butenes Conversion Catalyzed by Acid Zeolites Assessed by AIMD

    Dehydration of alcohols into alkenes is a key reaction for the production of fuels and chemicals from biomass. However, the mechanism of these reactions is highly questionable, hindering the rational optimization of efficient catalysts. In the present work, the formation of linear butenes starting from isobutanol catalyzed by proton-exchanged zeolites is unraveled by ab initio molecular dynamics (AIMD). Comparison with static calculations done for a gas phase reaction catalyzed by a proton and for the prototypical chabazite zeolite framework shows that AIMD estimations of the free energy barriers are significantly different from the static ones. Moreover, a common transition state (TS) is found for two competing reactions, namely, the isomerization of isobutanol into butan-2-ol (the dehydration of the latter yielding linear butenes) and the synchronous dehydration and isomerization of isobutanol into products related to linear butenes in a single step. The existence of a post-TS bifurcation prevents a traditional estimation of rates by transition state theory. To circumvent this problem, we quantify relative transmission coefficients using the Bennett–Chandler theory, which shows a clear tendency for decrease of relative frequency for isobutanol isomerization and increase of that for synchronous dehydration and isomerization when switching from 100 to 500 K. This work represents a step forward for the accurate determination of rates for key reactions in alcohol dehydration reactions.

    ano.nymous@ccsd.cnrs.fr.invalid (Monika Gešvandtnerová) 29 May 2024

    https://ifp.hal.science/hal-04592530v1

Date de modification : 08 janvier 2025 | Date de création : 08 novembre 2023 | Rédaction : B-BEST