A new approach to modeling TiO2−x-based memristors using molecular dynamics simulation

Abstract

This study aims to investigate the oxygen ion migration in defect rutile titanium dioxide (TiO2−x) in the presence of oxygen and titanium vacancies using a new approach based on molecular dynamic simulation. In this approach, the force field models along with Buckingham and Columbic potentials are used. For this purpose, the simulation is conducted in two different phases. In the first phase, the effect of temperature on the mean square displacements of oxygen ions is examined; besides, the diffusion and ionic conductivity of oxygen are studied in the absence of the electrical field. The results reveal that in the temperature range of 1100–1900 K, the oxygen vacancies tend to form clusters. These clusters consist of two oxygen vacancies. In the second phase, the memristor is applied to the model and hysteresis loops are studied in five cycles. The results of the second phase show that the new model correctly predicts the migration behavior of the oxygen ions in the memristor configuration. In conclusion, compared to the models using the Langevin equation method, the proposed model provides information which is closer to reality. Furthermore, the mean square displacements of the oxygen ions versus time are studied in the presence and absence of the electrical field at 300 K, and the results indicate greater diffusion in the presence of the electrical field.