Machine learning Accelerated Identification and NMR Signatures of Sugar adsorption on Sn-zeolites H/F

IFPEN; Solaize; On-site; EUR 31,000 - 40,000; Full time

Job summary

IFP Energies nouvelles, à Solaize, lance un poste post-doctoral pour développer des MLIPs afin de modéliser l adsorption de sucres sur des zéolites Sn et d’étudier les signatures NMR sous solvants.

Le travail s’effectue en collaboration avec l’University Charles (Prague) et CRMN (Villeurbanne) dans le cadre du projet MAINS financé par l’ANR; le candidat doit être titulaire d’un doctorat en chimie computationnelle ou science des matériaux.

Qualifications

• Solide connaissance en chimie et science des matériaux, avec expérience en DFT et/ou ML potentials.
• Connaissance des zéolites ou de la catalyse serait un plus.
• Anglais technique requis (B2).

Responsibilities

• Construire des jeux de données ab initio de référence pour des systèmes Sn-zeolite/sugar/water.
• Ajuster des modèles ML foundation et lancer des MD à ns pour explorer sites Sn et modes d’adsorption.
• Former des prédicteurs ML pour les décalages 13C et 119Sn NMR et comparer avec les données expérimentales.
• Contribuer à des relations structure–activité pour des réactions de glucose.
• Réaliser une revue bibliographique régulière et rendre compte au sein du projet.
• Présenter des talks et rédiger des articles en collaboration avec les partenaires CRMN et Charles University.

Skills

Chemistry knowledge

Materials science knowledge

DFT experience

ML potentials experience

Zeolites knowledge

Catalysis knowledge

English (B2)

Education

PhD in computational chemistry or computational materials science

Tools

DFT

ML potentials

Job description

Machine learning Accelerated Identification and NMR Signatures of Sugar adsorption on Sn-zeolites H/F

Machine learning Accelerated Identification and NMR Signatures of Sugar adsorption on Sn-zeolites H/F

Statut

Post-Doc

Contrat

12 (possibly extended to 24 months)

69360 Solaize, France

Contexte

Biomass-based sugars are a key feedstock for the sustainable manufacture of platform chemicals. Zeolites having Sn cations in their framework have proven efficient for the catalysis of several important sugar transformation reactions. Available static density functional theory (DFT) calculations suggest several modes of interactions between sugars and the potential active sites. However, they cannot capture the dynamics, confinement and explicit solvent effects that govern sugar/Sn-site interactions in zeolite pores. To address those limitations of the static approach, one can resort to molecular dynamics. To this end, the postdoctoral fellow will develop machine-learning interatomic potentials (MLIPs) to model sugar adsorption and reactivity in Sn-zeolites under hydrated conditions. Solid-state nuclear magnetic resonance (NMR) stands out as one of the most powerful tools to investigate molecular uptake and interactions at pore surfaces in zeolites, thus a focus will be put on the computation of relevant NMR features to be compared with experiments.

The work will be carried out at IFP Energies nouvelles (Solaize) in close collaboration with Charles University (Prague) and the CRMN (Villeurbanne) within the project MAINS funded by Agence Nationale de la Recherche (ANR). The PhD degree must have been awarded no more than three years before the start of the postdoctoral contract.

The post-doctoral fellow will (i) build reference ab initio datasets (periodic calculations) for {Sn-zeolite/sugar/water} systems; (ii) fine-tune foundation MACE models and run ns-scale MLIP molecular dynamics to screen open/closed/hydrated Sn sites, sugar tautomers, adsorption modes and hydration levels; (iii) train ML predictors of 13C and 119Sn NMR chemical shifts to compute dynamically averaged spectra, to be compared with experimental DNP NMR data obtained at our partner CRMN in Lyon; and (iv) contribute to structure–activity relationships for glucose isomerization and retro-aldol reactions. He/She will be in charge of regular literature survey on the topic, of regular reporting to the project team, of presenting talks at international conferences, and writing drafts of papers. He/She will participate to the computational/experimental synergy within the project, by discussing possible interpretation of experimental results acquired at CRMN.

Short stays at our collaborators at Charles University, Prague, Czech RepublicFrequent visits at our collaborators at CRMN, Villeurbanne, France (close to Lyon) Possible home-office (from France) part of the time

Compétences techniques et aptitudes Knowledge in chemistry and materials science, with experience in DFT and/or ML potentials; knowledge of zeolites or catalysis is a plus.Ability to work with several research teams, with autonomy and high scientific rigor and ethics.

Langues English: at least B2 level

Candidates should hold a PhD in computational chemistry or computational materials science