The silver price had been on the rise again (before it was suddenly interrupted by the Coronavirus). If things go back to the recent growth trend, there might be a new window of opportunity for plating, a technology rival for cell metallization based on screen printing and fire through pastes, which comes with the promise of saving on silver usage. TaiyangNews talked to Holger Kuehnlein, SVP Technology & Innovation at Rena Technologies, a global leader in high-quality wet etching technology solutions that has been also developing plating based solar cell metallization solutions for many years, about market technology trends. Founded in 1993, Rena provides wet chemical inline and batch processes around the world. The Gütenbach/Black Forest, Germany headquartered company is active in solar, semiconductor and medical technology and has installed more than 3,000 systems worldwide.
TaiyangNews: Before we come to the topic of plating, just a few general questions on technical trends & demand in the solar cell space. As Rena is a leading equipment supplier for solar cell manufacturing, what are the important technology trends you are seeing in the market?
Holger Kuehnlein: One really important trend is towards larger wafers, even up to 210 mm (called M12), which have been introduced to the market last year. Our lead customers show great interest as an increasing wafer size helps to bring down cost per watt even more. The discussions with our customers show that fabs for M12 wafers will come soon. Module fabs will be installed to run M12 cells as well to build larger modules using multi busbar interconnection and half- or one-third-cut cells.
The trend towards larger wafers brings up quite a variety of sizes, such as M2, G1, M4 and M6. But some manufacturers are introducing sizes beyond 166 mm or somewhere in-between. In other words: while we currently see several sizes coming up, we hope that the industry will agree on two or three different sizes that will become mainstream. These mainstream sizes will allow more standardized wafer making and quality on the one hand and on the other such standards would make it easier for the industry downstream, these are cell and module manufacturers, and even EPCs that prefer to work with standardized module sizes.
TaiyangNews: What does this trend to a larger variety of sizes and larger wafers mean to you as an equipment maker?
Holger Kuehnlein: Together with our key customers we discussed early about the equipment requirements. We defined the target of handling 210 mm wafers taking into consideration that the actual ingot diameter limit is 300 mm. The flexibility of our InOxSide and BatchTex equipment allows all wafers sizes up to 210 mm. We can adapt our RENA wet-chemical tools to the wafer size specified by our customer. However, different wafer sizes may impact the throughput of tools. Our customers have their total cost of operations in focus; therefore we believe that a few standard sizes will help the industry to supply the best equipment, materials and processes for these sizes.
TaiyangNews: Is the ability to process larger wafers limited to new machines or are you also offering a retrofit for the existing lines?
Holger Kuehnlein: Our devices are able to run 6-lanes M6 (166 mm) or 5 lanes for wafers up to M12 (210 mm). For our RENA Batch N platform, we offer our XL version, which enables the tool to handle up to 210 mm wafers.
For our large installed base, we will present at this year's SNEC exhibition the next generation RENA NIAK Inline platform, which is able to run 6-lanes M6 (166 mm) or 5 lanes for wafers up to M12 (210 mm). For our RENA Batch N platform, we offer our XL version, which can handle wafer sizes up to 210 mm, depending on carrier size.
For RENA's installed base, we will offer a retrofit for inline tools – whatever wafer size the customer is asking for. However, this may impact the throughput as larger wafers sizes mean less processing lanes in the machine.
For our BatchTex products, it is also simple: does the carrier fit into the tool? Can it be automated? If both questions are answered positively, we can offer retrofitting to the new size.
TaiyangNews: What is the largest size of wafers you can process with adaptation of existing equipment?
Holger Kuehnlein: Let ́s take a look at inline equipment first. It is rather a question of the number of lanes and throughput. For processing M12 wafers in existing standard inline equipment we can offer a retrofit to 4 lanes, reducing throughput, but enabling M12 technology. We have ramped the first inline production equipment for the use of M12 wafer size in Q4/2019 and were very pleased by process performance of this large format as well as the simplicity to transfer the process from M2 and M6. In mid 2020, we will offer our new machine platforms, which will attractively address the variety of wafer sizes we face today.
For batch processing, we also have ramped the first XL tools in the market in late 2019 and are satisfied with its performance on the M12 wafer material.
TaiyangNews: How do you see the market in general today?
Holger Kuehnlein: The market today is very bullish, at least from the big known players, which expanded continuously in the last years. We are also very positive based on the first quarter forecast regarding the general outlook of PV. While we see very strong further expansion of PERC, we also see that the belief in further optimization of PERC is getting weaker. Even though publications of 24% or higher PERC efficiency can be found all over, the big players seem to become a bit hesitant in believing that PERC production can reach this level of efficiency while maintaining its cost effectiveness.
What has essentially changed these days, it is not just technology guys that are talking about new technologies, even the top management is talking about advanced technologies in order to decide for the best economic solution.
TaiyangNews: The industry is talking about passivated contacts, SHJ (Silicon Hetero-Junction) and few are even talking about IBC (Interdigitated Back-Contact). What is your take on these cell technologies?
Holger Kuehnlein: We see excellent results for TOPCon (Tunnel Oxide Passivated Contact). The motivation for TOPCon is obvious – equipment wise, it is close to PERC processing technology and requires only limited changes. Also, existing module technology can be used for TOPCon cells and transfer to large wafer sizes seems feasible. Although TOPCon cells will require more processing steps in manufacturing and, to a certain extent, also more elaborate processing technology, our customers are convinced that efficiencies beyond 24% are possible in mass production using this cell architecture. However, there is still a lot of hesitation to roll out the technology in large volume manufacturing. Costs have to come down further and production yield needs to be stabilized on the same levels as for PERC. If TOPCon can reach 24% in production soon, we expect the competition with SHJ technology to increase even more.
Our customers are becoming much more aware of the importance of technology progress and its impact on cell efficiency: high efficiency calls for more elaborate technology. This is true for all technologies, PERC, TOPCon, SHJ or IBC. A new mainstream technology must be able to outrun the current working horse in efficiency and cost. That means, it is not easy to decide on something else because the cost proposition of PERC is excellent and unmatched.
What does it mean for Rena and our products? We offer texturing, etching and cleaning solutions. While for PERC and SHJ these processes are established and mature, we are developing with our customers cost-effective adapted solutions for single etching and cleaning for TOPCon technology. It's the same for our plating solution, which is a key enabler for TOPCON in several terms. Also, if PERC hits boundaries in efficiency levels sometime in the future, cost reduction solutions will get into focus – and plating is the answer to reduce metallization cost.
TaiyangNews: Plating has been around for quite some time but has somehow not been able to make it into the mainstream so far. What do you think is the reason?
Holger Kuehnlein: The main reason is a historically low silver price for more than 5 years. There was simply no motivation to look around for alternatives of silver. However, since May last year, the silver price has increased by 30%. During my last trip to China, cell makers have shown interest in our plating technology. While module prices are trending downwards, metallization costs are increasing in the total cost of ownership – even though the paste suppliers are doing a great job. PERC will reach its limit in terms of efficiency somewhere north of 23.5%. And TOPCon is looking interesting, but with increasing silver prices, the technology has even higher cost pressure as it uses silver paste on both sides.
Together with Jinko Solar we published several scientific papers about our plating solution for PERC along with our cost position. But with a low silver price prevailing in those times and consistently decreasing silver paste consumption, plating was not addressing the main pressure points. This is changing now with the recent rise in silver paste prices. With further price increases, plating technology will become more and more attractive again.
TaiyangNews: What is the determining factor – the volatility or the time period showing upward trend in price?
Holger Kuehnlein: More than volatility, it is about price charts compared to last year. As I mentioned earlier, we have seen a silver price recovery of 30% since May 2019. This is the key. Because even if you have a contract for silver, the next contracting will be more expensive. That means whatever measures you took to protect yourself from rising silver prices, they will not work any longer. So metallization costs would be eventually 30% higher than last year, which you can compensate by further reduction in silver paste consumption – and this is ongoing. Nevertheless, the margins for PERC products went down. Now we are seeing that the big players start looking for alternatives and are asking us if plating can help – and it actually does help.
TaiyangNews: At what level does silver price stop making sense for silver pastes?
Holger Kuehnlein: For sure this is one of the important company secrets of cell manufacturers in the world. However, if the prices go up to 750 USD per kg, then module price would definitely increase substantially. Actually, no one can compensate for silver price increases by 25 to 30% over today's level of about 570 USD per kg. It is not viable anymore to use silver paste based metallization at this level.
TaiyangNews: Could you please explain the process intricacies of your plating approach?
Holger Kuehnlein: For front side metallization we employ three layers of metal: nickel, copper and silver. The nickel layer has two functions: first and most important is contacting silicon in general, but also allowing to contact lowly doped emitters which enable emitters resulting in higher Voc, and secondly it is a barrier layer for copper, which is subsequently applied. But one cannot do full nickel plating as the bulk conductivity of the metal is not sufficient. Thus, a layer of copper is applied here, which serves as the conductive layer. About 50 mg of copper is typical for 5 BB cells. On top, a thin silver layer using 1.5 mg for a 100 nm layer is applied to protect copper from oxidation and ensure long term solderability. The silver coating also helps in making plating a drop-in replacement for silver paste based metallization for an easy integration by standard soldering techniques in module manufacturing.
TaiyangNews: Are Rena's solutions completely based on light induced plating and single sided?
Holger Kuehnlein: RENA plating technology is not limited to just one cell technology, but the performance benefit really depends on the cell architecture. In case of PERC, we propose only light induced plating to metallize the sunny side. Bifacial PERC plating doesn't make sense from an economical point as the silver content of the rear side is extremely low and local BSF formation is a sophisticated key process close to the optimum.
In case of PERT, double-sided plating subsequent for each side will require one step with light induced plating. The same approach is applied on TOPCon.
TaiyangNews: Then, what are the advantages of your plating solution for TOPCon?
Holger Kuehnlein: For TOPCon, plating is extremely interesting, because today silver paste is required on both sides. The same rule also applies to n-PERT and SHJ. Moreover, a polysilicon passivation layer for TOPCon has to be quite thick to enable a fire-through paste metallization. Due to this, current TOPCon structures still suffer from low bifaciality, expensive Poly-Si / Capping layer deposition. Here, we can bring a technology that can get rid of very high silver consumption, bring down costs and improve bifaciality.
TaiyangNews: There are ways to reduce silver consumption also with screen-printed cells. Does plating make sense compared to silver saving technologies such as multi busbars (MBB)?
Holger Kuehnlein: Even with silver saving technologies such as MBB, the PV industry sooner or later will reach a point where it will consume more than 40% of silver worldwide. That means, a part of the supply chain to the industry is prone to market volatility. So, concepts such as MBB do not offer immunity to the risk of rising silver prices.
TaiyangNews: Do you have any solution for SHJ?
Holger Kuehnlein: For plating on both sides of SHJ cells, we are using also the RENA inline platform InCellPlate. However, we are very cautious when recommending plating for SHJ. Even though we all know that Kaneka, for example, is producing world record efficiency SHJ cells using plating. However, this path faces some major challenges. The passivation film on SHJ cannot be used for masking like SiNx on PERC, PERT and TOPCon. R&D departments of companies and research centers worldwide have been looking at low cost masking technologies to avoid area plating on ITO top layers. The approach of copper sputtered seeds on the ITO and subsequent resist mask printing dominates. But this approach requires after plating a mask strip and seed metal etch back – 2 additional processes in a very cost sensitive industry.
TaiyangNews: What is the main difference of your approach to your main competitors?
Holger Kuehnlein: Every plating equipment has the challenge to contact brittle wafers to feed in the plating current. To achieve a high uptime and low maintenance on a 5,000 wafers-per-hour tool the contact clamp should be, in the best case, protected from the plating electrolyte. If not, one could see so-called contact marks after plating on the solar cell. But it is not just an optical issue, the deposited metal stack in this wafer area is incomplete which can cause corrosion and even copper diffusion can come up later on. RENA ́s InCellPlate completely avoids touching cells on the processes side and thus we have an excellent solution for these issues.
TaiyangNews: Can you provide some details about your machine platform?
Holger Kuehnlein: RENA InCEllPlate technology is an inline equipment running 5 cell lanes in parallel. Our machine is much more compact than others in the market. We recently improved the plating speed of copper deposition by 50% compared to the previous platform. This has enabled us to process 5,000 wafers per hour using a machine length in the range of 20 m.
TaiyangNews: What is the finger width you can attain with plating and what is the potential?
Holger Kuehnlein: We can easily demonstrate metal finger widths below 30 μm, but this is mainly depending on the laser spot size during SiNx patterning. While today lasers are limited to openings of 15 μm, institutes showed already patternings with 10 μm width which will result in finger widths below 25 μm.
TaiyangNews: How important is laser quality for realizing reliable rear metallization of TOPCON cells?
Holger Kuehnlein: It is important. It is definitely advantageous to have high performance lasers. Both European and Asian suppliers are offering such systems, with the European suppliers a bit in the lead regarding process experience for the necessary ablation processes.
TaiyangNews: One of the major concerns about plating solutions in China are environmental norms, which are strictly implemented in the major industrial regions. How do you intend to win the trust of investors and authorities for your solution?
Holger Kuehnlein: The effluent water from a metal plating line with 5,000 wph seems to be still a challenge for customers, but there are already established solutions in PCB industries running. RENA provides a technology that is based on vacuum distillation if requested. While our InCellPlate platform is already using smart rinsing systems to reduce rinse water to a level of 300 l/ h this subsequent system reduces the final volume to <2 l/h. The remaining volume of ~1,2 l/h has a high metal concentration and will be returned as a valuable liquid electrolyte back to suppliers. The great thing is that the full set of equipment installed has no physical drain from our plating machine out of the fab – and that is why it is called a ZERO wastewater solution.
TaiyangNews: Can you provide cost of ownership benefits with plating?
Holger Kuehnlein: The key driver of metallization costs is the consumable – the pastes. The metallization cost using fire through pastes for a bifacial TOPCon cell on M4 wafers is about 2.394 US cents per W. With plating, it is about 1.128 US cents per W. The benefit is 53% reduction in cost of ownership. However, technologies such as TOPCon benefit further as the above numbers do not factor in the benefit of potential increase in bifaciality plating brings along.
TaiyangNews: If you look 5 years into the future, what role do you think plating will play in solar cell manufacturing?
Holger Kuehnlein: I expect that RENA InCellPlate platform will address our customers metallization challenges in terms of cost reduction and efficiency increase. So if boundary conditions continue according to recent trends, plating will become mainstream for passivated contact cell technologies and even for PERC at the final end.
TaiyangNews: Thank you for the interview
This interview was published first in our TaiyangNews Market Survey on Metallization Pastes 2019-20, which can be downloaded for free here.