# Unlocking the Potential of Mg-Doped Rare Earth Manganites: Machine Learning-Guided Synthesis and Insights into Structural and Optical Properties

## TL;DR

This paper uses machine learning and a sol-gel method to create new nanoceramics with promising optical and structural properties for advanced devices.

## Contribution

Novel synthesis of Mg-doped rare earth manganites using machine learning and sol-gel methods, revealing their structural and optical properties.

## Key findings

- XRD confirmed orthorhombic Pnma crystal structure for both M0.5Ca0.25Mg0.25MnO3 compounds.
- LCMMO and PCMMO showed band gaps of 3.25 eV and 3.71 eV, respectively, with low Urbach energies.
- SEM, UV-Vis, and FTIR analyses confirmed chemical homogeneity and optical properties of the nanoceramics.

## Abstract

By leveraging machine learning insights from prior perovskite studies and employing the sol–gel method, we successfully synthesized two novel perovskite nanoceramics—M0.5 Ca0.25Mg0.25MnO3 (M = La, Pr)—as multifunctional nanomaterials. X-ray diffraction (XRD) confirmed their orthorhombic Pnma crystal structure. The Williamson–Hall method estimated average particle sizes of 59.5 nm for PCMMO and 21.8 nm for LCMMO, while the Scherrer method provided corresponding values of 32.59 nm and 20.43 nm. SEM, UV-Vis, and FTIR analyses validated the chemical composition, homogeneity, and optical properties of the synthesized compounds, revealing band gaps of 3.25 eV (LCMMO) and 3.71 eV (PCMMO) with Urbach energies of 0.29 eV and 0.26 eV, respectively. These findings provide valuable insights into the structural and optical properties of LCMMO and PCMMO, highlighting their potential as multifunctional materials for advanced device applications.