Heterojunction is one of the two advanced cell architectures the solar industry has been banking upon to improve the performance of today's PV device. The current cell technology incumbent, PERC has hit its efficiency threshold, and even the large wafer trick that allowed it to generate more power is not exclusive to PERC anymore. Consequently, cell/module manufacturers have once again started focusing on high efficiency, and this is where HJT makes its presence felt. HJT, now on the radar of many PV manufacturers, has made significant progress in recent times, especially since our previous HJT report published in Dec. 2020. This 3rd edition of our TaiyangNews Report on HJT Technology once again provides an overview of the most important developments associated with the key stages of process, supply and value chain.
HJT brings a lot to the table – high efficiency and fewer process steps to name a couple – but it also demands a whole different manufacturing method and setup. In fact, the modification required for HJT begins at the wafer that HJT is based on – the n-type wafer platform. While meeting the wafer quality spec for HJT was a major challenge, there is now an industry-wide practice where the wafers are annealed to mimic the gettering effect that finally lifts the quality barrier. One of the most significant wafer related developments is adapting to larger wafers. With this development, though, half cell processing is inevitable. So instead of cutting a fully processed cell, which increases the edge losses, HJT players are now seriously evaluating the prospects of processing pre-cut wafers. Now the question is whether to cut the wafer at the cell lines or right at the brick level to produce half-cut wafers.
In terms of HJT processing, surface preparation through the wet-chemical station is where it all begins. While the core steps of wet-chemical treatment for HJT would not change from SDE, texturing and cleaning, each step needs to be individually optimized. For example, HJT requires a high degree of cleaning. RCA vs. ozone based used to be a bone of contention in the past, with the former being the most preferred. Now, the industry has moved to ozone-based cleaning methods, keeping lower operational costs in view.
At the core of HJT processing are the application of doped and intrinsic amorphous silicon layers. As for the deposition of the core layer, while there were, and also are, some alternatives, it is PECVD that more or less enjoys a monopoly. As for PECVD suppliers, previously well-known companies from China GS-Solar, Ideal Energy, S.C New Energy and Maxwell still remain the big names. Given its high market penetration, Maxwell holds the lead over others. H2GEMINI, a German-Chinese collaboration, is a new entrant among PECVD suppliers. Tool platforms today can support high throughputs of up to 8,000 wafers per hour. As for depositing the core layer itself, replacing the doped amorphous silicon layers with microcrystalline silicon films is a direction that is being contemplated, initially on the front and then eventually on the rear as well.
TCO deposition is the next process step in sequence, and also in importance. While PVD sputtering is the predominant technology to apply TCO, RPD – even with its inherent advantages – is yet to make an entry into the mainstream. The leading PVD tool suppliers include Von Ardenne, Maxwell, Singulus and GS-Solar. H2GEMINI, having developed its own reactor design, is willing to offer a PVD tool through an OEM. Throughput has never been a problem for PVD; tools putting out 10,000 wafers per hour are available in the market.
S.C New Energy is currently the lone supplier of RPD, which is also proprietary. While ITO is typically the TCO deposited with PVD, IWO is deposited with RPD. In either case, indium – a scarce commodity – is a key constituent of both these TCOs. Reducing ITO consumption, if not eliminate, is an important topic of discussion in terms of cost reduction and sustainability. AZO is an alternative that has now been evaluated by several equipment vendors as well as cell makers – and with encouraging first results.
Screen printing is the only commercial technology for metallization also for HJT. The low temperature cured silver paste is the printing media here. To compensate for the inherent limitation of low conductivity, paste consumption in screen printing is high compared to other cell technologies. Several paste suppliers are offering advanced paste formulations to reduce the consumption without affecting the performance attributes of the paste. The ability to support high printing speeds and fine-line printing are other matters of focus.
Increased number of busbars has proved to be an effective means of reducing paste consumption. While the 12-busbar layout had already been implemented, the industry has just started increasing the busbar count to 15 busbars with a few even planning to go beyond 18 or even 24.
In order to further reduce the silver content of pastes, thereby costs, the industry is evaluating silver-coated copper pastes. With proper optimization of paste content – while the performances can be safeguarded, reliability is something that needs to be evaluated further. Plating has also been an alternative to screen printing, but is yet to get a stamp of approval for HJT by the industry.
Interconnection is another key process that requires optimization at the module level. Against conventional wisdom, soldering is still the mainstream technology used for HJT modules in China, while the Western world often relies on ECA based interconnection that supports low temperatures. A couple of European equipment vendors are offering tools to support ECA based interconnection, and nearly every leading stringing tool supplier claims to support the HJT compatible low temperature soldering process. A few changes in BOM, especially related to the encapsulation materials, can enhance the reliability aspect of the HJT module.
PV manufacturers and equipment vendors have started hitting record-high efficiency levels, which bodes well for the industrialization of the technology.
Several PV manufacturers have attained average cell efficiencies around 24.5%, for modules today's best has reached up to 22.5 %. Unlike in the past, when HJT modules had only rather low power ratings due to smaller wafers and focus on the residential segment, now there is the first commercial HJT panel based on very large wafers that has power ratings of close to 700 W. As a few cell/module HJT makers have started selling their HJT cells, module makers with no cell capacities are also venturing into building HJT modules. Due to advantages such as a low temperature coefficient, better low light performance and highest bifaciality, the first companies are exploring expansion into utility-scale projects with HJT, while residential is a key application for the technology due to its higher costs per watt.
Almost all leading PV cell/module manufacturers are involved with HJT in R&D, process development or pilot lines, while a few companies have started high-volume production and are already expanding. While installed production capacities are low, at around 6 GW end of last year, HJT continues to attract more and more industry participants, with announcements totaling about 80 GW as of the end of last year.
This is the executive summary of TaiyangNews latest report on Heterojunction Technology 2022, which can be downloaded for free here.