Deciphering the optoelectronic evolution and interfacial kinetics of brilliant blue/p-Si photodetectors: synergistic experimental and machine learning approach

Engineering notes

Engineering notes will be added by the Power for Solar editorial team.

Chinese explanation / 中文解读

中文解读待补充：本站会优先为光伏效率、钙钛矿太阳能电池、储能技术、太阳能热利用、BIPV、并网技术等高价值论文补充中文说明。

Original abstract

Abstract This study presents an investigation into the interface-driven optoelectronic performance of Cu/BB/p-Si photodiode, combining experimental characterization with machine learning (ML) insights. The Brilliant Blue (BB) organic interlayer was found to play a dual role in passivating the Si surface and manipulating the interlayer carrier extraction process, which led to notable increase in the barrier height (Φ b ) from 0.541 to 0.571 eV and modulating the ideality factor (n) from 2.260 to 2.761. Under a simulated AM1.5 full-spectrum solar illumination of 100 mW/cm 2 , the Cu/BB/p-Si sensor demonstrated a high broadband detectivity of 1.467 × 10 12 Jones, a remarkable responsivity of 2.082 A/W and a linear dynamic range of 50.92 dB. To validate these physical observations, four ML algorithms were developed using Gaussian Process Regression (GPR), Kernel Ridge Regression (KRR), Random Forest Regression (RFR) and Support Vector Regression (SVR) based on experimental current-voltage (I-V) data under dark and varying illumination intensities (40–100 mW/cm 2 ). All ML models were evaluated, where the GPR model achieved the closest agreement with experimental data. The model’s robustness, highlighted by a perfect a10-index of 1.0, stems from its superior ability to capture the non-linear dependencies and interface-governed transport mechanisms facilitated by the BB layer. These results demonstrate that the synergy of interface engineering and data-driven modeling provides a powerful framework for optimizing next-generation surface-based optoelectronic technologies.

Links and sources

• PDF from original source

Comments