A coupled quantum-molecular mechanics approach for performance analysis of defective Silicon based photovoltaic solar cells

Abstract

Although, molecular mechanics (MM) based approaches are capable of simulating the dynamic charge motion of atoms over time scales up to femto-seconds, the accuracy is an issue. As a result, quantum-mechanics (QM) based approaches are a choice to predict accurate dynamic motion of atomic charges. However, computational cost of QM simulations is significantly higher than that of MM simulations. In this study, a computationally efficient coupled QM/MM model is developed by combining the QM and MM approaches, enabling simulation of larger domains with accurate estimates. The proposed methodology is implemented with the help of QMMM package available in large-scale atomic/molecular massively parallel simulator (LAMMPS), to investigate the dynamic charge motion in the presence of cracks in Silicon. A Silicon domain of dimensions 48.87 Å × 48.87 Å × 5.43 Å is considered in the simulations. Furthermore, a small domain around the crack tip, with dimensions 4 Å × 4 Å is identified for carrying out QM analysis and denoted as embedded region (ER). Simulations are performed considering four different cases: (i) pristine Silicon, (ii) Silicon with an initial edge crack, (iii) pristine Silicon with Graphene deposition, and (iv) Graphene deposited Silicon containing an initial edge crack. In the coupled model, first, for the given load step, considering the minimum energy criteria molecular dynamics simulations are performed over the entire domain. This is followed by QM simulations over an identified ER based on first principle studies using the plane wave density functional theory. The latest atom positions from the QM simulations are updated in the MM domain before proceeding to the next load step. The electrical performance of Silicon solar cells is studied by estimating the effective Bader charge and total electric power. The effective Bader charge for atoms in QM domain is observed to be significantly higher indicating charge accumulation around the crack tip. This is further evidenced through the total electric power estimations, where pristine Silicon with Graphene deposition is observed to possess the highest power followed by cases iv, ii and i.