From Synthesis to Machine Learning-Assisted Analysis: Investigating Silver Nanowires Effects on Conductive Films

In this work, we synthesize silver nanowires (AgNWs) and investigate how their morphology (e.g., aspect ratio) and concentration can affect the electrical performance of conductive films. To model and predict these effects, we build machine learning (ML) models, including random forest, extreme gradient boosting and kernel ridge regressor, and optimize their hyperparameters using Bayesian optimization. To improve model interpretability and quantify the individual and combined effects of morphology and concentration, we integrate these models with permutation importance analysis. This explainable ML approach provides both accurate prediction and advances understanding of how nanomaterials’ properties affect film performance. Our results lay the foundation for optimizing AgNWs-based film design for electronic devices and provide practical guidelines for tailoring nanomaterial properties.