Smart growth of thermoelectric oxides by epitaxy assisted with operando monitoring and machine learning

CONTEXT AND MOTIVATION
This PhD project will be carried out at the Institut des Nanotechnologies de Lyon (INL) within the “Functional Materials and Nanostructures” team, within the framework of the funded PEPR DIADEM CINEMA project (2026-2030). The research focuses on the development of smart growth strategies for functional oxide thin films using molecular beam epitaxy (MBE) combined with in situ monitoring and machine learning approaches. Optimizing the relationship between growth conditions and material properties in functional materials still largely relies on empirical approaches such as trial-and-error or design of experiments. These methods are time and resource consuming and require highly stable processes that are rarely achieved in practice. This project proposes to develop a new methodology for smart growth control applied to p-type BaSnO₃, a wide-bandgap perovskite oxide of growing interest for transparent electronics and thermoelectric applications, based on the combined use of advanced in situ characterization tools and machine learning-driven data analysis. The oxide MBE chamber at INL is equipped with several real-time monitoring techniques including reflection high-energy electron diffraction (RHEED), spectroscopic ellipsometry, wafer curvature measurements, and optical flux monitoring. These tools provide complementary information about surface structure, growth dynamics, optical properties, strain, and stoichiometry during thin film deposition. The objective of the PhD thesis is to exploit these measurements, combined with machine learning approach, to develop data-driven models capable of correlating the growth parameter and functional properties of p-type BaSnO₃. BaSnO₃ is a perovskite oxide known for its exceptionally high electron mobility in its n-type form (La-doped), but its p-type counterpart remains largely unexplored and poorly controlled, making it a highly relevant and challenging target for this project.

OBJECTIVES
The PhD thesis combines data-driven analysis and materials research. The objective will be to develop two complementary approaches for growth optimization. The first approach is based on coupled operando measurements, combining real-time curvature monitoring, RHEED, and spectroscopic ellipsometry to directly probe the structural and optical state of the growing film in order a deep understanding of the growth physics and the truly impact of elaboration parameters on materials properties. The second approach relies on supervised machine learning (ML) and Baysian Optimisation (BO) algorithms to establish correlations between the key elaboration parameters (Ba flux, Sn flux, dopant flux, O₂ pressure, substrate temperature) and a functional material property such as the electrical resistivity. These methods are particularly well suited to small experimental datasets, as typically encountered in MBE. To train and validate these models, the database will first be built using n-type BaSnO₃ data extracted from the literature, a well-documented system that serves as an ideal benchmark. The PhD student will then construct their own experimental database from their own growth runs of p-type BaSnO₃, progressively enriching the model with original data. Once this methodology is validated on the n-type system, the final objective will be to apply both approaches — operando control and ML-guided optimization — to the growth of in situ measurements and establish correlations between growth conditions and material properties. By combining operando monitoring, machine learning and ex situ characterization, the project aims to establish predictive models linking growth parameters, defect chemistry, and transport properties, with the ultimate goal of achieving high-performance p-type BaSnO₃ with optimized resistivity and carrier concentration beyond the current state of the art.

The goal of INL is to encourage world-leading multidisciplinary research in the areas of micro and nanotechnologies and their applications. The pioneering research undertaken at the Institute ranges from materials and technology to devices and systems, thus enabling the emergence of dedicated technologies. The Institute is supported in its work by the Nanolyon Technology Platform.

The application areas cover major economic sectors: semiconductor industry, information technologies, healthcare and wellbeing, energy and the environment. The laboratory is located on two leading research campuses at Lyon Ouest-Ecully and LyonTech-La Doua. It has personnel of 200 people including 121 permanent staff. INL is one of the key laboratories of the “Université de Lyon” research and higher education centre. For further information, please consult our website https://inl.cnrs.fr

Location: The PhD thesis will be conducted at INL, on the site of Ecole Centrale de Lyon (ECL), in the team MFN (https://inl.cnrs.fr/en/equipe-materiaux/). The laboratory offers an interdisciplinary environment with strong expertise in epitaxial growth, nanomaterials, advanced characterization techniques and functional materials. The candidate will have access to state-of-the-art growth and characterization facilities.

Framework and Partnership: PEPR DIADEM – project CINEMA: https://www.pepr-diadem.fr/projet/cinema/
Dates / duration: the position is for 3 years, from Oct. 2026 to Sept. 2029.
Salary: Gross salary of approx. 2300€ brut/month; 1600 € net/month

Research profile & skills (required / highly desirable):
- Master degree (M2 or equivalent) in physics, materials science or a related field preferably with a master thesis related to IA data treatment
-  Interest in experimental materials science and thin-film growth
-  Knowledge in machine learning or data analysis applied to materials science
-  Experience in thin film growth and characterization
-  Motivation to work with advanced characterization and growth techniques
-  The candidate should be able to work independently while collaborating effectively with the research team.
-  Good communication skills and fluency in English are required.