Machine learning-assisted prediction of the electronic features of a Schottky diode interlaid with PVP:BaTiO3 composite

Abstract

This study employs two Machine Learning (ML) models to predict the electronic current and then analyze the main electronic variables of Schottky diodes (SDs), including leak current (I0), potential barrier height (ΦB0), ideality factor (n), series resistance (Rs), shunt resistance (Rsh), rectifying ratio (RR), and interface states density (Nss). The I-V characteristics are examined for both without and with an interlayer. The polyvinylpyrrolidone (PVP) polymer and BaTiO3 nanostructures are combined to form the nanocomposite interface. The ML algorithms that are employed include the Gaussian Process Regression (GPR) and Kernel Ridge Regression (KRR). The thermionic emission theory is used to gather training data for ML algorithms. Ultimately, the effectiveness of these ML methods in anticipating the electric characteristics of SDs is evaluated by contrasting the predicted and experimental findings in order to identify the optimal ML model. Whereas the GPR algorithm has given values that are closer to the actual values, the ML predictions of fundamental electric variables by practically both algorithms have the best level of agreement with the actual values. Also, the obtained findings indicate that when the nanocomposite interface is used, the amount of I0 and Nss for metal-semiconductor (MS) Schottky diodes reduces and φ B0 increases.