Thesis defended on March 12, 2025, to obtain the degree of Doctor of the Université Grenoble Alpes - Specialty: Inorganic and Bio-inorganic Chemistry​

Abstract :
The discovery of CsPbBr3 perovskite nanocrystals (NCs) in 2015 and their unique light-emitting properties sparked interest in their colloidal synthesis. CsPbBr3 NCs with an ionic core differ from other semiconductors and are more prone to degradation caused by environmental factors such as light, heat, and humidity compared to their more covalent counterparts (e.g., CdSe, ZnO). To enhance the stability of NCs, researchers have employed a variety of ligands (stabilizing agents) that play a crucial role in controlling growth, shape, and ensuring colloidal stability. In this study, native ligands such as alkylammonium (e.g., dodecylammonium) and carboxylates (e.g., oleate), were used to neutralize surface charges and improve stability by forming a protective shell around the NCs. Investigations into surface chemistry of the NCs, using liquid-state and solid-state Nuclear Magnetic Resonance (ssNMR), have provided valuable insights into ligand binding thermodynamics and potential surface composition. These studies aim to enhance synthesis reproducibility and optimize stability for example via ligand exchange. Nevertheless, progress in understanding the surface termination and ligand stabilization of the NCs using ssNMR is clearly impeded by the current lack of sensitivity of this technique. A few studies made use of Dynamic Nuclear Polarization (DNP) to improve this sensitivity limitation but with limited efficiency compared to model systems (e.g. frozen solution). This thesis explores the main challenges encountered when using DNP to polarize perovskite NCs at 100 K and shows that x10-20 improvement in sensitivity can be demonstrated using state-of-the-art polarizing agents and a home-built closed-cycle cryogenic helium spinning system enabling DNP experiments down to 30 K. The work was conducted on CsPbBr3 NCs synthesized using a slightly modified version of the classical hot-injection method, aiming to improve the quality and stability of the monodisperse cuboidal structures. More precisely, we first discuss the sample preparation protocol used to conduct DNP experiments on NCs, considering ligand lability and potential surface modifications. We also discuss the effect of various experimental parameters such as dissolved oxygen (in the TCE solvent), sample storage, microwave absorption, sample heating, as well as glass quality and 1H T1 to explain why such systems are so difficult to polarize at 100 K. In order to improve this limitation, we discuss three strategies relying on the use of KBr to dilute the sample, state-of-the-art DNP polarizing agents and access to ultra-low temperature ~30 K. Despite these advances, correlating nuclei with low natural abundance (e.g., 13C and 15N) remains challenging. This work demonstrates the potential of combining DNP with NCs with isotopically labeled ligands (15N for ammonium and 13C for carboxylates) and outlines a method for recovering these costly ligands. Moreover, comprehensive characterization of the NCs using techniques such as UV-Vis and photoluminescence spectroscopies, XRD, SEM, and TEM confirms their structural and optical quality. TEM imaging complements atomic-level analyses of surface termination provided by NMR, offering a deeper understanding of the surface chemistry of CsPbBr3 NCs and highlighting the stability optimization pathways.​


Thesis supervisor : Gaël De Paëpe


Keywords :
DNP, NMR, Perovskite, Polarizing agents, Microwave, Labeled ligands