Though generally reliable, silicon solar modules can be subject to unforeseen degradation, leading to a duty life shorter than the expected 25-year life cycle. Potential-induced degradation (PID) has proven difficult to characterize and study. This dissertation is dedicated to developing a physical model to understand the kinetics of PID of the shunting type, and explain the factors that may lead to the design of PID-robust modules.
A bias-temperature stress (BTS) methodology to study ion migration in dielectric films is presented, which accounts for the contribution of bulk traps in the dielectric. Using this method, an Arrhenius relationship for the diffusivity of Na+ in SiNx is determined, for which the prefactor is D0=1.4E-14 cm^2/s, and the activation energy is Ea = 0.14 eV, with a 95% confidence interval of [0.07, 0.21] eV. Based on this result, we bound the transit time of sodium ions through highly resistive SiNx anti-reflective coatings, within 1 h and 2 days, under temperature and electric fields relevant to PV operation.
A numerical solution to the coupled Poisson-Nernst-Planck system of equations is presented, based on the finite element method (FEM), that can accurately simulate ionic transport in dielectrics and stacks of materials. The FEM implementation adequately describes the accumulation of charge in the semiconductor interface of metal-insulator-semiconductor capacitors (MIS). Using this model, we evaluate diffusion coefficients of Na+ in SiO2 under BTS conditions.
A methodology to simulate PID degradation in PV modules is derived, which uses the result from the ion transport model to simulate the characteristic J-V of the devices. PID is adequately described by the presence of metallic shunt at the p-n junction of the cell, for which, the metal conductivity depends on the sodium concentration. An upper bound for the diffusivity of Na in stacking faults that result in PID is estimated to be 1E-14 cm^2/s, based on comparison of the simulated PID time series with experimental reports of PID-s.