Exploring forward scattering mechanisms in TiO2 with carbon quantum dots: Insights into photovoltaic applications

The study investigates the forward scattering mechanisms in TiO2 with CQDs and their potential implications for photovoltaic applications. Thin films fabricated on microscopic glass were experimentally characterized using UV-Vis-IR spectrometer and SEM measurements. The experimental data was used to determine the effective scattering particle radius through Mie scattering modeling. At the same time, a convolutional neural networkbased algorithm and image post-processing were employed to study the thin film porosity. Machine learning models were then utilized to aggregate the data and identify the optimum CQD concentration for maximum forward scattering. The extinction spectra were found to be significantly dependent on the concentration of CQDs in the composite, with forward scattering dominating for wavelengths above 400 nm. The study also utilized a random-forest machine learning algorithm to predict the impact of varying CQD concentrations, providing insights beyond the dataset's limit of up to 15 %. Additionally, numerical simulations using finite difference time domain analysis and MEEP were employed to provide extinction spectra for different CQD concentrations, ultimately identifying an optimum CQD concentration of 6.9 % for maximum extinction performance. The study also analyzed the tradeoff between the effective scattering particle radius and porosity against CQDs concentration, shedding light on the impact of these parameters on forward scattering.