In recent years, PERC cell efficiency has seen 0.4-0.5% of gain in mass production each year, despite the industry’s long-held belief in PERC’s limited rooms for raising efficiency. Some cell manufacturers claimed to have attained a cell efficiency of 23.5% in mass production, and that the figure will continue going up. PERC’s high cell efficiency and the large size wafer trend emerging in recent two years led to another round of massive production expansion. By the end of this year, new capacity additions for large format PERC cells will surpass 200 GW. The overexpansion of PERC seriously restrained the developments of other cell technologies.
PERC will remain the preferred option in the two to three years to come, or even five years. However, the fact that it is approaching to its theoretical limit indicates a new turn cell manufacturers ought to take. Presently, discussions on next-generation cell technologies succeeding PERC revolve around TOPCon and HJT processing. The two technologies are not at daggers drawn, but the one with faster pace in cell efficiency growth and cost reduction will secure larger share of capacity expansion after PERC.
TOPCon production capacity and efficiency
Source: InfoLink, New Technology Market Report. August 2021.
Conventional cell manufacturers favor TOPCon technology, as it is, like PERC, a high-temperature process, and thus more compatible with the latter. This year, many new PERC lines have been reserved for upgrading to TOPCon technology. Despite announcement of several manufacturers’ engagement in TOPCon, few expansion projects were materialized, due to difficulties in deposition processes. Only Trina, Suntech, and BYD planned to advance to mass production in the second half of this year, whilst Longi and Tongwei claimed to have GW-scale TOPCon capacity arranged respectively in Yinchuan and Meishan.
Presently, mainstream efficiency of TOPCon cells in mass production is around 23.7-23.8%, with some manufacturers claimed to have achieved 24.0%+. However, as efficiency of PERC markedly rose, TOPCon still fell behind by a certain margin.
TOPCon Major Challenges
Despite great similarities shared between boron diffusion and phosphorus diffusion, as well as equipment, the former saw more difficulties than the latter, as boron has lower solid solubility in silicon, which requires 900℃ to 1100℃. Moreover, the associated production of BBr3 diffusion severely damages quartz components. For now, more and more manufacturers turned to BCl3, of which the associated production hardly does harm to quartz components. However, subject to longer bond length of B-Cl, the diffusion evenness of BCl3 is slightly inferior to that if BBr3.
Currently, selective emitter (SE) is unable to apply to TOPCon front emitter, for such combination has rather limited rooms for minimizing sheet resistance, and thus greatly affects the improvement of short-circuit current and contact resistance.
Adoption of polysilicon passivation layer
The biggest difference between TOPCon and PERC lies in the fact that TOPCon introduced a tunnel oxide layer and an intrinsically polysilicon layer on the rear side. Presently, the mainstream process is to grow a 1.5-2cm tunnel oxide layer through thermal oxidation; meanwhile, sediment 150-200nm of intrinsically polysilicon layer through LPCVD, and then dope through phosphorus diffusion. However, production efficiency of this process is low. Additionally, given the issue of wraparound polysilicon layer, yield rates of the process are dissatisfying, sitting merely at 90-95%, far lower than the 98% of PERC. To break the bottleneck, more and more equipment suppliers started trying new deposition technologies, such as PECVD, PEALD, and PVD. Equipment for new techniques emerged successively, but without mass production data, disputes over TOPCon technology roadmap will persist.
Both TOPCon and PERC use high temperature sintering silver paste. However, TOPCon uses silver paste on both sides. Taking M6-sized cells for example, a TOPCon cell requires 130mg of silver paste, 60mg more than a PERC cell. In the meantime, lower yield rates restricted the reduction of production costs. Non-silicon costs of TOPCon and PERC still saw RMB 0.07-0.12/W of differences by far. Metallization costs of TOPCon can never go lower than that of PERC, owing to the application of silver pastes on both sides of a cell. Only by further raising cell efficiency and yield rates can TOPCon shorten the differences of cost per watt, as compared to that of PERC.
HJT production capacity and efficiency
Source: InfoLink, New Market Technology Report. August 2021.
With higher potential gain in efficiency and rather simple processes, HJT became an instant hit in the capital market. Recently, with Tongwei implementing 1 GW-HJT project, Huasun starting construction of Phase 2 2 GW-project, JA Solar announcing to introduce pilot HJT production lines, Akcome expanding production capacity, and the engagement in HJT of Golden Glass and China Resources Power, HJT technology saw heated discussions sparking. By far, more than 70 GW of HJT production capacity have been arranged. However, subject to higher equipment investment and production costs, capacities materialized will be lower, whilst production expansions come online slowly in 2021.
Without yield rate issues, the technology roadmap of HJT is clear. Among plasma-enhanced chemical vapor deposition (PECVD), radio-frequency plasma-enhanced chemical vapor deposition (RFCVD) has become the mainstream process for a-Si layer fabrication for its higher technology maturity and coating evenness. Some manufacturers are trying hot-filament CVD, which sees fast sedimentation and great wafer passivation performances. Magnetron sputtering became the mainstream process to form transparent conductive oxides (TCO) thin films, thanks to its easier evenness control, high target material utilization ratio, and low production costs. Reactive Plasma Deposition (RPD), with the sedimentation of reactive plasma, produces IWO thin films, which have better photoelectric performances, and is thus clang to by some manufacturers. Overall, PECVD+PVD is the most popular combination for HJT technology.
Presently, efficiency of HJT cells has officially passed the 24%+ mark. Methods of raising efficiency, such as SMBB and the replacement of nc-Si for a-Si, are all highly equipment-dependent. As equipment upgrade, suppliers and manufacturers both claimed to reach a cell efficiency of 25% beyond in mass production by 2022.
HJT Major Challenges
With broader localization of equipment manufacturing, the investment capex for HJT has declined to around RMB 450 million/GW and even RMB 400 million/GW, still higher than the RMB 150-200 million/GW of PERC and RMB 250/GW of TOPCon. Lofty costs of investments in equipment not only affect the activeness of upfront investments, but also indicate greater depreciated non-polysilicon costs later. Presently, low utilization rates of HJT manufacturers result in at least RMB 0.03/W higher depreciated costs than that of PERC.
Unlike PERC and TOPCon that uses high-temperature silver paste for sinter-bonding, HJT applies low-temperature cured silver paste, which is less matured, and thus higher selling prices. HJT requires higher silver paste per unit, because of its bifacial nature and higher resistivity of low-temperature cured silver paste. For instance, a M6 HJT cell costs approximately 200mg of silver paste, whilst a M6 PERC cell costs only 130mg. Metallization costs of HJT is RMB 0.12/W higher than that of PERC, as calculated with current unit consumption of low-temperature cured silver paste and its lofty prices.
To further reduce metallization costs, the industry focuses on new technology roadmaps, such as silver coated copper paste and copper electroplating. The development of copper electroplating is slow, given RMB 180 million/GW of additional investments, yield rates, and environmental concerns.
Some manufacturers completed tests and proceeded to reliability stage for silver coated copper paste, which is expected to enter small rate initial production by the fourth quarter. It is reportedly that 30% of copper allows cell efficiency to sustain but is less economically competitive, whereas 40% of copper may lead to 0.1-0.3% of losses in cell efficiency and most importantly, reliability concerns owing to exposed copper.
Indium containing target material
HJT saw RMB 0.05/W of additional non-silicon costs in the formation of TCO thin films, for its use of indium containing target material. Once HJT cell production explosively increase in the future, indium, a rare metal, will see prices skyrocket, despite the gradual replacement of foreign target material with those made in China. Driven by demand from solar cell manufacturers, indium prices in September have reportedly risen by 60% on levels in August.
Therefore, cost reduction of HJT relies greatly on minimizing the use of indium, which is presently proposed to be achieved through the substitution with AZO and indium recycling. The minimization of indium consumption sees no actual progress for the time being, since most HJT manufacturers are still in the midst of bringing capacities online.
Presently, new HJT production capacities all adopt the 182mm or 210mm format. Cutting process needs to be done prior to all HJT processes, for laser cutting damages the structure of HJT cells. In the meantime, HJT manufacturers introduced gettering process, which markedly raises the efficiency of HJT cells. Outcomes of gettering process and application of half-cut technique in large wafers will be the trending topics this and next year.
High efficiency n-type cell technology prospect
Despite more barriers, inherently high conversion efficiency, low degradation rates, and cheaper LCOE enables n-type cells to be the next-generation technology following PERC. Presently, both TOPCon and HJT have acquired efficiencies higher than that of PERC, with production cost being the pivoting factor determining their rapid developments. As more and more equipment manufacturers started conducting R&D on TOPCon and HJT, the former’s yield rate issue is likely to be solved through new process routes, such as PECVD and PVD. In recent years, HJT adopted SMBB and silver coated copper paste, and will thus narrow the difference between its non-polysilicon costs and that of PERC to RMB 0.2/W. By then, TOPCon and HJT will become more competitive and see capacity expansions accelerating.
Presently, production capacity and production volume of both TOPCon and HJT are expected to expand continually in recent one to three years. However, having better compatibility with existing PERC production lines, TOPCon enjoys more advantages than HJT in terms of expansion progress. In the short term, TOPCon will see both production capacity and volume growing faster than that of HJT. Cost is a crucial factor for the development HJT technology. More and more GW-scale HJT capacity expansions will take place, after silver coated copper paste and copper electroplating technology reach industrial maturity. HJT is expected to take another two to three years to achieve cost reductions (minimizing the use of polysilicon, indium, and silver) and see larger scale productions after 2023.