# Composition-based machine learning for predicting and designing Mn4+-doped phosphors

## TL;DR

This paper uses machine learning to predict and design Mn4+-doped phosphors based on their composition, avoiding complex structural data.

## Contribution

The study introduces the largest experimental dataset of Mn4+-activated phosphors and uses composition-based machine learning for accurate property prediction.

## Key findings

- K-Nearest Neighbors and Extra Trees Regressors achieved high accuracy in predicting excitation and emission wavelengths.
- The models generalized well to Eu3+-doped systems and showed strong agreement with experimental data.
- An inverse design model was developed to suggest new phosphor compositions for desired optical properties.

## Abstract

We present a data-driven approach to predict the excitation wavelength, emission wavelength, and crystal field energy levels (4T1, 4T2) in Mn4+-doped phosphors based solely on elemental composition. For the first time, we construct the largest and most compherensive experimental dataset of Mn4+-activated phosphors to train and accurately predict the properties without relying on complex structural descriptors. Among several evaluated models, the K-Nearest Neighbors and Extra Trees Regressors achieved the highest accuracy for predicting excitation and emission wavelengths, respectively. Importantly, to evaluate generalization, we test these models on Eu3+-doped systems and achieve high predictive accuracy. An inverse design model is further developed to suggest candidate phosphor compositions for target optical outputs. By avoiding complex descriptors while preserving accuracy and interpretability, this work provides a foundation for theory-informed discovery of luminescent materials.

We collect the largest experimental dataset and apply machine learning to study composition–property relations in Mn4+-doped phosphors. The resulting models outperform prior studies and show good agreement with experimental data.