Effects of PCPDTBT:PCBM Ratio on the Electrical Analysis and the Prediction Of I-V Data Using Machine Learning Algorithms for Au/PCPDTBT:PCBM/n-Si MPS SBDs

Abstract

In this study, Au/Poly[2,6-(4,4-bis-(2-ethylhexyl)-4H-cyclopenta[2,1-b;3,4-b′]dithiophene)-alt-4,7(2,1,3-benzothiadiazole)] (PCPDTBT) : [6,6]-phenyl C61 butyric acid methyl ester (PCBM) /n-Si heterojunction Schottky barrier diodes (SBDs) with 1:1 and 2:1 PCPDTBT:PCBM doping ratios were produced, and the electrical analysis of metal-polimer-semiconductor (MPS) SBDs with different concentrations was investigated. Ideality factor (n), saturation current values (I0) and barrier heights (F0) of the materials were obtained based on the current-voltage (I-V) measurements performed. According to the results obtained, the PCBM concentration has significant effects on the electrical properties of the Au/PCPDTBT:PCBM/n-Si MPS SBD. To predict the electrical characterization of a system in detail, based on its doping concentration, the I-V data set consisting of 2 samples is typically split into a 70% training set and a 30% test set, which is used to train machine learning algorithms. Various methods, including Fine Tree, Cubic SVM, Fine KNN, Boosted Trees, Bagged Trees, Subspace KNN, RUSBoosted Trees, Wide Neural Network, Trilayered Neural Network, and Logistic Regression Kernel, have been analyzed. The obtained results indicate that certain algorithms can predict the I-V data of Au/PCPDTBT:PCBM/n-Si MPS SBD with full accuracy, i.e., 100%.