Catalytic chemical vapour deposition will improve the efficiency of solar cells: Study

The past few decades have seen great progress in solar energy technology in response to the ever-growing threat of climate change.

By Ground report 07 Mar 2024

New Update

Catalytic chemical vapour deposition will improve the efficiency of solar cells: Study

The past few decades have seen great progress in solar energy technology in response to the ever-growing threat of climate change. Specifically, people have increasingly focused on silicon heterojunction (SHJ) solar cells in this field because they can achieve very high energy conversion efficiencies.

JAIST study: passivation layer's importance

The study conducted by researchers from the Japan Advanced Institute of Science and Technology (JAIST) highlighted the significance of the passivation layer in silicon heterojunction (SHJ) solar cells. This layer, made of intrinsic hydrogenated amorphous silicon (i-a-Si:H), is pivotal in preventing the recombination of light-excited charge carriers.

The team focused on optimizing the catalytic chemical vapor deposition (Cat-CVD) process to create high-quality thin passivation films without damaging the substrate, a notable advantage over plasma-based methods.

publive-image — Flowchart of the BO process with practical constraints. Image credit: Keisuke Ohdaira from JAIST

Finding the ideal Cat-CVD parameters for the best i-a-Si:H film quality has been challenging due to the numerous adjustable factors, such as substrate temperature, gas flow rates, and deposition time. The study aimed to streamline this process and reduce the reliance on trial and error to determine the most effective combination of conditions.

BO optimizes Cat-CVD process

In a study that Professor Keisuke Ohdaira led and published in ACS Applied Materials and Interfaces on 21 February 2024, a research team from JAIST developed an innovative and efficient strategy to find the optimal deposition conditions for Cat-CVD.

Co-authors from JAIST included Ryota Ohashi, a Master’s student, Research Assistant Professor Huynh Thi Cam Tu, and Senior Technical Specialist Koichi Higashimine. Dr. Kentaro Kutsukake, a researcher at RIKEN, also joined them in the effort.

The study said, "We base the proposed strategy on a practical implementation of an optimization technique known as “Bayesian optimization” (BO). This popular methodology, which draws on machine learning, can determine the maxima of a function of an unknown shape."

The study shows that in the context of Cat-CVD, a sequential process in Bayesian Optimization (BO) inputs known deposition conditions and their corresponding outputs into the algorithm. The algorithm then makes predictions for the output of any unevaluated deposition conditions. This method continually feeds the results of tested predicted deposition conditions back into the algorithm until the optimum values.

However, applying BO to the Cat-CVD problem isn't straightforward, as Prof. Ohdaira points out. He states, "Simple Bayesian optimization, which aims solely to maximize carrier lifetime, does not incorporate functions that regulate film thickness. As a result, this lack of regulation could potentially propose thick films for future experimental conditions and this could result in operational problems."

Study introduces method for film deposition

The study addressed the challenge of unfeasible experimental conditions in film deposition by introducing a method called “constrained Bayesian Optimization” (BO). This approach utilized a trio of predictive models: the first assessed the film’s charge carrier lifetime, the second gauged deposition pressure from gas flow rates to avoid impractical scenarios, and the third regulated film thickness considering deposition time, crucial for Cat-CVD processes.

With an initial dataset of 14 samples detailing deposition conditions and film performance, the researchers needed only eight optimization cycles to enhance carrier lifetimes, achieving full optimization within twenty cycles. This method also uncovered new parameter combinations that significantly increased carrier lifetimes.

Professor Ohdaira emphasized the importance of substrate temperature and the SiH4 gas flow rate during deposition, which were key to minimizing carrier recombination. Thus, constrained BO not only identified optimal conditions but also expanded scientific understanding.

As per the study, constrained BO’s potential extends beyond solar cells to a broad spectrum of devices. Professor Ohdaira suggests that this technique could broadly optimize complex material processes like film deposition, a cornerstone in producing electronic devices.