Thèse Couplage Électrochimique de la Réduction du Co2 et de la Dépolymérisation Oxidative de la Lignine H/F

École polytechnique

Établissement : École polytechnique
École doctorale : Ecole Doctorale de l'Institut Polytechnique de Paris
Laboratoire de recherche : LCM Laboratoire de Chimie Moléculaire
Direction de la thèse : Cédric TARD ORCID 0000000215151915
Début de la thèse : 2026-10-01
Date limite de candidature : 2026-03-31T23:59:59

Ce projet propose une approche innovante pour la valorisation électrochimique simultanée de la lignine et du CO2, intégrant des outils d'intelligence artificielle et de machine learning pour optimiser les rendements et automatiser l'analyse des produits. L'objectif est de transformer la lignine, un sous-produit abondant de la biomasse, en composés aromatiques fonctionnalisés (comme la vanilline ou l'acide syringique) tout en réduisant le CO2 en produits à valeur ajoutée (CO, formate, méthanol, méthane). La méthodologie repose sur l'utilisation d'électrodes à base de nickel et de palladium/argent optimisées pour la coupure sélective des liaisons C-O de la lignine et la réduction du CO2, respectivement.

Des algorithmes de clustering et des réseaux de neurones sont utilisés pour analyser automatiquement les données issues de la spectrométrie de masse, de la chromatographie et de la RMN, permettant de corréler les paramètres expérimentaux (potentiel, temps, type d'électrode) avec les rendements spécifiques en monomères. De plus, des modèles de classification supervisée permettent d'adapter les conditions expérimentales en fonction du type de lignine (bois dur, bois tendre, organosolv, kraft), optimisant ainsi le procédé pour chaque source de biomasse.

Ce projet offre une solution innovante et scalable pour la valorisation des déchets biomassiques et du CO2, tout en réduisant l'empreinte carbone par rapport aux méthodes traditionnelles.

The transition towards sustainable technologies is essential for the efficient utilization of renewable resources such as lignocellulose, a carbon-neutral and non-edible biomass. Lignin, the largest renewable source of aromatic building blocks, holds immense potential as a feedstock for bio-based chemicals. However, its complex and recalcitrant structure poses significant challenges for catalytic conversion. Electrochemical approaches have emerged as a promising, greener alternative to traditional thermo- and chemocatalytic processes, offering milder operating conditions and the possibility of using renewable electricity. Recent studies have demonstrated moderate yields of valuable aromatic compounds from lignin, but key challenges, such as poor solubility, electrode fouling, and limited selectivity, remain to be addressed.
Simultaneously, the electrochemical reduction of CO2 has gained attention as a means to convert this greenhouse gas into useful chemicals and fuels, such as CO, formate, methanol, or methane. By coupling lignin oxidation at the anode with CO2 reduction at the cathode, this project aims to develop a synergistic electrochemical platform that maximizes resource efficiency and sustainability. This integrated approach not only enhances lignin valorization but also contributes to CO2 utilization, addressing two major challenges in green chemistry.
This project will focus on overcoming critical barriers in lignin solubility, electrode material design, product analysis, and process optimization, with the ultimate goal of advancing electrochemical depolymerization and CO2 reduction as a viable and scalable industrial solution.

This project aims to develop an innovative and sustainable electrochemical platform that couples the oxidative depolymerization of lignin, a major component of lignocellulosic biomass, with the selective reduction of CO2. At the anode, lignin will be oxidized into valuable functionalized aromatic compounds, while at the cathode, CO2 will be reduced into high-value products such as carbon monoxide CO, formate, methanol, or methane. By integrating these two processes, we seek to create a synergistic system that maximizes resource efficiency and sustainability. While thermocatalytic approaches have been extensively explored, recent advances in electrocatalysis offer promising alternatives for both lignin valorization and CO2 conversion. Chemical modification strategies will be employed to enhance lignin's solubility and stability, and the resulting products will be characterized using advanced analytical techniques, including mass spectrometry and NMR spectroscopy. The ultimate goal is to establish a more efficient and sustainable route for lignin valorization, while simultaneously enabling the production of useful chemicals from CO2.

The methodology for this project will involve several interconnected steps to achieve the objectives, with a focus on automating data analysis and optimizing experimental conditions using AI and machine learning.
i.Chemical modification of lignin
Initially, we will focus on the chemical modification of raw lignin to stabilize it by forming 1,3-dioxane structures during extraction, thereby preserving its native structure and improving its solubility in both aqueous and organic media. Machine learning models will be employed to predict optimal modification conditions (pH, temperature, solvents) based on lignin type (softwood, hardwood, organosolv, kraft).
ii.Electrochemical experiments
The stabilized lignin will be used in electrochemical experiments under various conditions. We will investigate nickel-based materials for lignin electrooxidation, exploring different morphologies such as wires, plates, felts, and foams to optimize the depolymerization reaction. To automate the analysis of products, we will develop AI-driven analytical pipelines:
·Tandem mass spectrometry (MS/MS), chromatography (SEC, GC-MS), and NMR data will be processed using clustering algorithms to group products based on their spectroscopic profiles.
·Regression models will correlate experimental parameters (potential, time, electrode type) with specific monomer yields (e.g., vanillin, syringaldehyde).
This approach will enable real-time optimization of conditions to maximize target product formation.
iii.Classification and optimization based on lignin type
We will analyze lignin from different sources, including poplar, birch, spruce, and beech, to identify trends and mechanistic routes in the depolymerization process. Machine learning classification models will be used to:
·Classify lignin types based on FT-IR, NMR, and electrochemical profiles.
·Optimize experimental conditions (potential, time, electrode material) for each lignin type to maximize monomer yields.
This will allow tailored process optimization depending on the lignin feedstock.