A group of scientists from the University of New South Wales (UNSW) in Australia and Russian heterojunction solar module producer Hevel Solar has developed a novel hydrogenation process that is claimed to have the potential to improve the stabilized efficiency of p-type heterojunction (SHJ) solar cell based on gallium-doped silicon wafers.
The solar industry usually applies n-type phosphorus–doped Czochralski–grown silicon (Cz–Si) wafers in the production of SHJ cells, as these ensure no susceptibility to the boron-oxygen light-induced degradation (B-O LID) that is typical for p-type boron-doped wafers and severely affects the performance of SHJ cells over time. N-type wafers ensure more stability, however, they are currently more expensive to produce than p-type wafers, which are the mainstream solution for the manufacturing of PERC cells. This means that using p-type wafers may potentially lead to a further cost reduction for the heterojunction technology, as wafer costs still represent 40% of a cell’s total cost.
In order to compete with n-type devices, however, p-type heterojunction cells will have to show improved performance. “The same advanced hydrogenation techniques (AHTs) we use in mass production for solving LID and LeTID in p-type PERC solar cells can be used in p-type SHJ solar cells to solve B-O LID when using boron-doped p-type Cz wafers,” research co-author, Brett Hallam, told pv magazine. “Even though the gallium-doped and n-type SHJ solar cells were stable in this work and didn’t need the process to improve stability, we have shown that these same processes can improve the efficiency of gallium-doped and n-type SHJ solar cells by 0.4-0.7% absolute.”
The research group explained that the expiration of Shin Etsu‘s gallium doping patent (US6815605B1) has encouraged the solar industry to adopt p-type gallium-doped Cz-Si wafers, which it describes as a potential mainstream solution for the SHJ segment for the next decade.
Two solar cells were developed with a new advanced hydrogenation process (AHP) at an existing SHJ line operated by Hevel using 156.75×156.75 mm p-type wafers doped with boron (B) and gallium (Ga), respectively. The first products were provided by Chinese manufacturer Longi and the second by Taiwanese wafer maker Sino-American Silicon (SAS). The scientists also built a standard n-type SHJ device as a reference.
All the wafers were initially treated with a potassium hydroxide (KOH) saw-damage etch and KOH anisotropic texturing. B-doped hydrogenated amorphous silicon (a-Si:H) layers were deposited in the rear side of the wafers and intrinsic and P-doped a-Si:H layers were then deposited in the front of the wafers using a plasma-enhanced chemical vapour deposition (PECVD). In the final step, indium-tin tin-oxide (ITO) transparent conductive oxide (TCO) layers were deposited on both sides through physical vapour deposition.
According to the scientists, the cell built with B-doped wafers exhibited an efficiency of 20.5% and an open-circuit voltage of 719.6 mV, while the Ga-doped device was found to have an efficiency of 22.6%, a fill factor of 78.2%, an open-circuit voltage of 730 mV and no degradation during light-soaking. “The conversion efficiency of the gallium-doped SHJ solar cells is still lower than the n-type reference cells, which was largely due to a reduced fill factor (FF),” the Russian-Australian group affirmed. “Further work is required to overcome this FF limitation to facilitate high-efficiency gallium-doped SHJ solar cells.”
“We’ve demonstrated this versatile process with a number of n-type SHJ solar cell manufacturers and there are commercial tools available for this, which are already being implemented in production for industrial n-type SHJ solar cells,” Hallam stated, referring to the immediate applicability of this process to industrial production. “We would likely need to do more work on understanding the exact requirements for different SHJ solar cell manufacturers with different toolsets and processing conditions.”
A cost assessment must still be carried out. “The costs would be the same as for methods solving LID/LeTID in p-type PERC cells and when advanced hydrogenation tools are implemented for industrial n-type SHJ solar cells and TOPCon solar cells,” the UNSW researcher stated.
According to him, a potential issue for p-type SHJ solar cells may be incoming wafer quality. “But this is true for any wafer used for SHJ solar cell manufacturing, regardless of whether it is p-type or n-type,” he emphasized. “Even for n-type wafers with multi-millisecond starting lifetimes, I feel that we should really be adding pre-treatments like gettering.”
The solar cell and the related hydrogenation process are presented in the paper Stability Study of Silicon Heterojunction Solar cells fabricated with Gallium- and Boron-doped Silicon Wafers, published in RRL Solar.
