In HJT solar cell processing, the first part in deposition is about core layer deposition, which involves multi-layer stacking of intrinsic-amorphous and doped-amorphous silicon layers of opposite polarities on each side of the wafer and controlling them at the nanometer scale (see The Core Story Of HJT). The second half of deposition in HJT processing deals with the application of the transparent conductive oxide layer, which acts as both an antireflective coating and a conductive electrode to extract and laterally conduct the electrical current. The TCO films are usually applied on both sides. While the majority of HJT aspirants aim at gaining from the bifacial aspect, one can also apply silver on the rear side to make it monofacial.
Applying TCO is also a very important step in HJT processing. Here, enough care must be taken to retain the passivation properties of the underlying amorphous silicon layers. The quality of the TCO influences lateral charge collection. In addition, the transparency and resistivity of the TCO are also very important. The TCO application is typically accomplished in Physical Vapor Deposition (PVD) tools by means of sputtering, and indium tin oxide (ITO) is the most commonly used material for it.
Reactive Plasma Deposition (RPD) is an alternative method. Where RPD differs from PVD is that it utilizes an evaporating mechanism for coating the target material on the wafer surface without damaging the existing films. This process typically uses indium tungsten oxide (IWO) targets. Japan's Sumitomo holds the patents for RPD along with the target, while S.C New Energy has also acquired a license to use and develop the technology further. Several companies evaluated the technology during the process development phase and are satisfied with the results. The TCO films deposited by RPD still have good mobility and offer a good balance between the electrical and optical parameters, according to Yang from Risen speaking from experience of having an RPD tool at the company's first process development line. However, the story beyond R&D is not as rosy; a few issues have come to light in the production environment, such as high performance loss that still cannot be explained or solved at the moment, Yang emphasized. Risen is cooperating with the equipment supplier and working on finding a solution for such issues, "No matter what, we will not give up on RPD," said Yang.
This probably is the reason why RPD was not part of any of the Chinese HJT aspirants' plans for the second round of capacity additions. Consequently, PVD is currently the most widely followed approach.
The topic that's being discussed the most in the PVD segment is how to reduce or even eliminate the use of indium, the key ingredient of the typically used TCO in HJT, which is ITO. The fact that indium is only second to silver in terms of non-silicon costs of HJT cell production is what puts it in focus. Its limited availability is another issue. Kaining Ding, head of R&D department at Germany's Research Center Juelich, based on a simple calculation estimates that the indium consumption is about 3.6 to 4.8 mg/W in a best-case scenario of 25.5% efficiency and target utilization rate of 100%. At this level, at least more than 3,000 tons of indium is required for a TW of HJT cell production, while the annual supply is about 2,000 tons including the primary supply and secondary supply from recycling. The supply cannot be increased either, given indium is a byproduct of lead and zinc mining. PV cannot be part of recycling for at least the next 25 years. Indium, on the other hand, is also used in other industries, for example in flat panels. Once HJT starts to hit big volumes, the supply-demand imbalance can result in indium prices shooting up. To make the usage of indium sustainable, the consumption has to be reduced by 90%, basically to 0.38 mg/W, according to Ding.
Still some way to go: While RPD is known for supporting high efficiency levels, the technology is not yet ready for high volume production. (Source: TaiyangNews).
One prospective candidate to provide an alternative is aluminum doped zinc oxide (AZO), which is low cost and non-toxic but is a relatively less inert material chemically. While that's on the reliability side, from a performance point of view, substitution of ITO with AZO is not a problem on the rear side at all, as optics matter less on the backside of the cells, according to Ding. He also lists several options to balance the opto-electronics in the back; increase doping, reduce oxygen content that is introduced during sputtering and decrease the finger pitch on the back. "By this, 50% reduction in indium consumption can already be achieved without causing any efficiency losses," said Ding, while reliability is a topic that needs to be addressed separately. However, the main issue pertains to the front side where opto-electronic characteristics are very critical. Several structures are under evaluation: a thin layer of ITO is capped with silicon oxide as capping and antireflective coating; ITO buffer layer under the AZO; ITO as the capping layer and AZO sandwiched by two ITO layers, while research related to replacing ITO completely with AZO is at its primitive stage. These structures have different performance losses and savings in indium. A point to be noted is that ITO capping helps to sustain the damp heat testing condition, given the non-inert nature of the AZO. The good news, according to Ding, even when ITO is needed to sandwich the AZO, a reduction of up to 85% in indium usage can be achieved.
The topic has certainly attracted the attention of the equipment makers. VON ARDENNE has also shown good results from employing stacks of ITO and AZO. The German company compared the important solar cell characteristics with different TCO configurations. Taking a cell with 100 nm of ITO applied on both sides as reference, cells with ITO on the front side and a stack of AZO and ITO and a reverse configuration have been evaluated. The overall efficiencies for these three cells with different TCO configurations are more or less the same. When looking at other parameters, the cells with AZO as part of the TCO stack have lower Jsc due to losses incurred as a result of the absorption of AZO.
However, the losses are somewhat compensated with the slight gain in Voc and fill factor. Overall, the results are promising, but the long-term stability and module performance are still under evaluation, according to Sebastian Gatz, Vice President Photovoltaics at VON ARDENNE.
Replacing ITO with AZO is still a subject of evaluation at the manufacturers' level. Huasun's CTO Wenjing Wang said that his company also evaluated the material as part of its R&D, but the process is not yet ready for mass production. While not discussing the details, Wang would only say that there are some issues that need to be resolved before the process is implemented in high volume production. The same status quo is maintained on the subject by nearly all HJT makers (see The Status Quo Of HJT).
The Text is an excerpt from 3rd edition of TaiyangNews' Heterojunction Technology 2022 report, which provides an overview on the most recent HJT developments as the technology is entering the GW scale production level and can be accessed free of charge here.