APS Logo

Mechanism of the Metal-Insulator Transition And Its Length Scales in Complex Oxides: Theory, Experiment and Machine Learning

Invited

Abstract

Metal-insulator transition compounds, namely materials that depending on the environment present either metallic or insulating characteristics - and phase transition correlated materials, more generally - are of wide scientific and technological interest, yet many open questions still remain in our understanding: is the transition from a metallic to an insulating state driven by the electronic or the lattice degrees of freedom? What sets the relevant length scales of the transition? What are the relevant features we can use to design new MIT materials? In this talk, I will provide a theoretical and experimental discussion of new results that have helped us address this question. I will show that the transition in the rare-earth nickelates RNiO3 with R a rare-earth, and Ca2RuO4 materials and their thin films and heterostructures is driven by the electron-lattice coupling, rather than the electronic energy. We will use a simple model to understand the way this is affected by dimensionality and strain[1,2]. Then, I will discuss how a heterostructure can be used as a 1D model system to understand the length scales involved in the transition [3]. Finally, I will show how machine learning tools can be used to guide new understanding of this class of systems, and introduce our materials library, machine learning model and new relevant physical features identified using this model [4].

[1] ‘Disentangling lattice and electronic contributions to the metal–insulator transition from bulk vs. layer confined RNiO3’, PNAS, 2019, 116 (29)
[2] 'The Electron-Lattice Coupling Driving the Metal-Insulator Transition: Energy Landscapes of of Ca2RuO4, RNiO3 and their Heterostructures', In Preparation, A.B. Georgescu and A.J. Millis
[3] ‘Length scales of interfacial coupling between metal and insulator phases in oxides’, Nature Materials, Vol,19, 2020
[4] 'A Database and Machine Learning Model to Identify Thermally Driven Metal-Insulator Transition Compounds', https://arxiv.org/abs/2010.13306

Presenters

  • Alexandru Bogdan Georgescu

    McCormick School of Engineering, Department of Materials Science and Engineering, Northwestern University, Simons Foundation, Center for Computational Quantum Physics, Flatiron Institute, McCormick School of Engineering, Department of Materials Science & Engineering, Northwestern University

Authors

  • Alexandru Bogdan Georgescu

    McCormick School of Engineering, Department of Materials Science and Engineering, Northwestern University, Simons Foundation, Center for Computational Quantum Physics, Flatiron Institute, McCormick School of Engineering, Department of Materials Science & Engineering, Northwestern University