Constrained Bayesian Optimization for the Deposition Conditions of Doped a-Si:H with High Passivation Performance for...

Area of Science:

• Materials Science
• Photovoltaics
• Chemical Engineering

Background:

• High conversion efficiency in silicon heterojunction (SHJ) solar cells depends critically on passivation quality from both intrinsic and doped amorphous silicon layers.
• Optimizing deposition conditions for doped amorphous silicon (a-Si:H) is essential for achieving superior field-effect passivation.

Purpose of the Study:

• To effectively determine deposition conditions for doped a-Si:H using Bayesian optimization (BO) to maximize passivation performance.
• To enhance the effective minority carrier lifetime (τ_eff) while maintaining film thickness and conductivity within specified ranges.

Main Methods:

• Utilized Bayesian optimization (BO) with multiple prediction models to optimize deposition parameters for doped a-Si:H.
• Conducted BO cycles for n-type a-Si:H (21 cycles) and p-type a-Si:H (7 cycles) starting with initial random samples.
• Fabricated SHJ solar cells and symmetric passivation stacks using optimized a-Si:H films for performance evaluation.

Main Results:

• Achieved a τ_eff of approximately 5.4 ms for n-a-Si:H and 1.2 ms for p-a-Si:H via BO.
• Fabricated SHJ solar cells demonstrated a τ_eff of 2.4 ms and an open-circuit voltage (V_OC) of 0.701 V.
• Symmetric passivation stacks showed an implied V_OC exceeding 0.7 V, confirming sufficient passivation.

Conclusions:

• Bayesian optimization is an effective strategy for optimizing deposition conditions of doped a-Si:H for enhanced solar cell performance.
• The optimized a-Si:H films significantly improve passivation, leading to higher effective minority carrier lifetimes and open-circuit voltages in SHJ solar cells.