Lead-halide perovskites, particularly CsPbBr3, are emerging as promising materials for X-ray detection in medical applications. This technology requires their deposition in thick layers (>100 µm), and close-space sublimation (CSS), initially developed by CEA-Liten, has shown highly encouraging results. However, this process remains poorly understood at the microscopic scale, and the relationship between microstructure and performance remains a major scientific and industrial challenge.

This thesis, in partnership with the SIMAP laboratory, aims to develop a comprehensive thermodynamic model of the CSS process. The candidate will (i) experimentally generate the essential thermodynamic data for modeling, (ii) simulate growth mechanisms, and (iii) validate them experimentally using dedicated instrumented growth furnaces and advanced characterization techniques. Machine learning tools will be implemented to establish predictive correlations between deposition parameters and layer properties.

The results will enable optimization of CsPbBr3 growth for more sensitive and stable X-ray detectors, with a strong impact on medical imaging. This work will also provide opportunities for high-impact publications and patents in a highly competitive field.