Domaine, spécialité : CHIMIE
Mots-Clés : Intelligence Artificielle, HTE
Unité d’accueil : NIMBE/LCMCE
Summary
We are offering two internships as part of a research project aimed at optimizing chemical reactions used to synthesize therapeutic compounds, in particular via photo-induced cross-coupling reactions. By using CO2, these methods become more ecological and useful for pharmaceutical production. The project combines High Throughput Experimentation (HTE), which enables reactions to be miniaturized and parallelized, with Artificial Intelligence (AI) models to optimize chemical reactions, saving time and resources in the process.
Full description
The synthesis of compounds for therapeutic purposes requires continuous optimization of methods, particularly in photoinduced cross-coupling reactions, an essential chemical tool for the pharmaceutical industry. Photocatalyzed carboxylating cross-coupling reactions, using CO2, present a more environmentally-friendly approach, useful for the synthesis of pharmaceuticals and the formation of radiolabeled products. High throughput experimentation (HTE) has transformed the discovery and optimization of reactions, through their miniaturization and parallelization, saving considerable resources and time. SCBM’s recent acquisition of HTE technology, combined with LCMCE’s expertise in AI and chemical reaction modeling, aims to merge the two technologies for efficient exploration of synthetic methodologies, with the potential to rationalize the optimization of reactions in the pharmaceutical industry.
The project aims to develop AI models for optimizing chemical reactions, using catalytic tests carried out by HTE as training data. Three reactions are envisaged, with increasing levels of difficulty, and common features:
- being a carboxylating cross-coupling involving an electrophile (RX), a nucleophile (R’E) and CO2
- using a complex photocatalytic system (a metal, a ligand, a photocatalyst, a base)
- requiring wide variations in experimental conditions (pressure, irradiation, time, solvent)
In the project, we will start by determining a set of essential reaction conditions to be carried out (order 102, WorkPackage 1 (WP1)). Then, chemical reactions will be carried out using HTE campaigns (WP2, 96 reactions per plate per day), considering the previously selected reaction conditions. The data will be used to develop predictive algorithms (WP3). Finally, the validity of the model will be tested on more chemically complex structures (WP4).
Internship #1, focused on high-throughput synthesis, will involve working on WP1 and WP2, in collaboration with another research intern who will be in charge of WP3 (supervision predominantly by Eugénie Romero-Laboureur).
Internship #2, focused on algorithm development, will involve working on WP3 in collaboration with research intern #1 (with Emmanuel Nicolas as the main supervisor).
Location
CEA-Saclay, (91) Essonne, France
Internship conditions
- Durée du stage : 6 mois
- Niveau d’étude requis : Bac+5
- Formation : Master 2
- Poursuite possible en thèse : Oui
- Date limite de candidature : 3 février 2025
Required skills
Méthodes, techniques :
Plateforme HTE (High Throughput Experimentation)
Modèle IA/ML
Synthèse organique
Langages informatiques et logiciels :
Python
Langue : Anglais
Supervisor of the internship
Emmanuel Nicolas
Tél. : 01 69 08 26 38
Email :