Efficiency enhancement approaches look set to further fragment the PV equipment supply-chain, providing increased options for c-Si cell producers — who must arm themselves with a wider knowledge of the different options presented. Coherent’s Finlay Colville explains the different c-Si cells technology approaches and equipment types to support them.
by Finlay Colville, Coherent, Inc.
As high-efficiency cell concepts are ushered in by most c-Si cell manufacturers, the equipment supply-chain has been forced to reposition itself. From custom-line build to drop-in turnkey supply, decisions are being made on technology approaches. The result is increased choice for cell manufacturers, which demands a wider knowledge of the different options presented.
December 3, 2009 – As the 2009 solar exhibition season draws to a close, it is a good time to reflect on the key changes within the crystalline-silicon (c-Si) tooling on offer to cell producers. What’s behind the new equipment being promoted to them? And why is there such a wide choice of c-Si cell concepts being pursued now through the supply-chain? While many had hoped for some kind of equipment tooling standardization — at least within the c-Si segment of the market — it now seems that equipment on offer to c-Si producers is becoming as varied as that previously targeted at thin-film manufacturers.
Before reviewing the trends, let’s examine what’s behind ” target=”_blank”>the new landscape for equipment on offer to c-Si and thin-film fabs over the past 12 months. The first issue to consider is not so much the change in capacity/production or supply/demand — ” target=”_blank”>well documented as the year has progressed — but the fundamental difference in utilization rates between the c-Si and thin-film producers. During 2005-2008, utilization rates were less important to the equipment supply chain: growth was strong, demand was high, and capacity expansions were being funded regularly. Largely as a reflection, equipment suppliers were able to pick-and-choose the market segments (c-Si or thin-film) or even sub-segments (thin-film amorphous or CIGS) to sell their equipment within. After the events of 2009, it is not a huge surprise that equipment revenues have decreased dramatically, nor that equipment suppliers have been forced to fast-track new production tooling ahead of projected capacity increases from 2010/2011 onwards. What is interesting is the difference here between c-Si and thin-film, and at the heart of this are the different utilization rates within these two segments.
|Figure 1. A comparison of utilization rates across the c-Si and thin-film segments. Utilization rates are expected to pick up considerably from 2010 onwards, in particular within thin-film fabs currently being commissioned.|
Crystalline-silicon utilization rates have been traditionally higher than thin-film fabs, and here it is just as important to consider committed fab-expansions (where lines have been ordered) than true nameplate capacities reported in the press. Figure 1 illustrates this, showing the difference between c-Si and thin-film up to 2009, and with a forecast out to 2012. For tool suppliers to thin-film, caution is suggested in the near-term, as the order-of-the-day for thin-film suppliers is certainly to increase existing production output within installed (or committed) fabs. As such, it is fully expected that equipment revenues into thin-film will remain hard hit until utilization rates there improve. Additionally, for c-Si tooling, changes in equipment technologies on offer are equally impacted: not due to utilization rates, but on account of the inflection point brought on by higher-efficiency demands which calls out for new c-Si tooling on the market. Gone are the days of production tooling for c-Si lines catering to ‘standard’ c-Si cells with 13%-15% average efficiencies. Tooling for high-efficiency cell lines is based on new equipment, new technologies, and new cell concepts pursued. Therefore, near-term equipment focus from the supply-chain has been diverted to meet the immediate demands of c-Si producers seeking to increase cell efficiencies, where the goal now is to obtain average cell efficiencies in volume production at the 18%-20% level.
Different c-Si cell types
To understand how the equipment supply for c-Si fabs is changing, it is necessary to capture the different c-Si cell concepts being considered. For a long-time the exclusive focus of the research community, such high-efficiency — or ‘advanced’ — cell types are well understood and have been reviewed frequently in recent years. Further, roadmaps suggesting industry adoption here have been well publicized, no more so than in Europe with a collective call for prescribed cell thickness/efficiency/cost improvements based on new c-Si cell types out to 2020.
There are four basic c-Si cell types that receive most attention. First, there is the ‘standard’ cell type, characterized by discrete front and back contacts deposited via screen-printing and firing. Standard cells typically use lower-quality (and lower-cost) boron-diffused p-type silicon as the bulk material. Such cell types are by far the most common in the market today, and consequently have the widest choice when it comes to proven equipment suppliers.
The next cell type relies on modifications to the ‘standard’ cell where either the front surface bus-bars — or both fingers and bus-bars — are located at the rear of the cells. Here the phosphorous diffused emitter remains on the front surface, but conductive pathways (‘vias’) transfer (‘wrap’) the current from front to back through the bulk. These cells, called metal or emitter wrap-through (MWT or EWT) concepts, allow for efficiency gains not only from increased sunlight capture on the front surface, but more significantly due to reduced losses introduced during the module interconnection packaging.
Then there are the full back-junction cells, where the emitter is located also at the rear as part of an interdigitated structure of alternating n+ and p+ doped regions. While back-junction cells are dependent on higher-grade/higher-cost n-type silicon (with higher minority carrier lifetimes), they have an excellent track-record already within the industry for cell efficiencies in excess of 20%, albeit from a single source.
The final cell type also relies upon n-type silicon, but with amorphous-silicon films deposited on both surfaces for low surface recombination. The so-called heterojunction-with-intrinsic-thin-film (HIT) cell has been championed by another top-tier c-Si manufacturer with high-efficiency in volume production.
|Figure 2. Standard c-Si cell types with discrete front and back contacts are projected to account for around 90% of installed c-Si capacity by the end of 2012. Within this share however, several classes of ‘standard’ cells will emerge, driven by the emergence of selective emitter based concepts starting to enter the market.|
In Figure 2, comparison is made between these different cell types. A deeper dive into the trends seen so far within overall c-Si cell production reveals an even more interesting statistic: the percentage of cell output from the top-tier c-Si producers today. Indeed, almost all of their production output comes from the use of non-standard (or customized) production lines, and not through the use of complete turnkey systems. More on this later, when reviewing the expected pull on new c-Si cell equipment and which technologies are being pursued throughout the supply-chain.
The new equipment choices for c-Si cell producers
In the past, the choice for c-Si cell producers was more polarized. New entrants often sided with a quick route to market, based upon standard turnkey lines. While not pushing the boundaries in terms of cell efficiencies, such an approach offered a low-risk, quick ramp-up strategy. The other advantage here was the availability of an extensive network of module integration suppliers, where connecting individual cells with standard front and back contacts was routine and guaranteed. Selling c-Si cells therefore had the widest possible module-supplier outlet for pure-play cell makers, or simply made equipment selection for downstream-integrated players (both cell and module builders) easier.
For those manufacturers pursuing non-standard cell types, full ownership of cell production technology and manufacturing tooling was a perquisite to success. And viewing the case-studies on offer here is perhaps one of the most telling observations ahead of widespread high-efficiency cell adoption throughout the industry. In comparing the relative capacities and productions of the three non-standard cell types shown in Fig. 2, there is a strong correlation. Back-junction and HIT cells were enabled by downstream-integrated players, with full-ownership on equipment (and module) tooling: conversely, wrap-through cells have struggled to gain adoption, with manufacturers often critically reliant on module tool suppliers providing customized tooling for all-back-contact interconnection.
Today, equipment tooling for c-Si fabs is governed by several factors:
- The different approaches taken by established c-Si manufacturers in high-volume production vs. new entrants looking to ramp-up quickly;
- The scope of full turnkey lines offered to the market, and how much they are split into front- or back-end-of-line specialist tooling; and
- The range of qualified high-efficiency processes available in the market, and whether these are adopted as complete lines from single (or ‘partnered’) tool suppliers or as stand-alone customized inline tools.
The segregation between established players and new-entrants is likely to continue to be the main factor driving the adoption of turnkey lines; it remains unlikely that established cell producers are going to ‘standardize’ production, or lose control of in-house differentiation within a competitive market landscape. However, new entrants will now be able to fast-track technology by adopting one of the high-efficiency turnkey lines primed for full market release during 2010. Partial turnkey lines may emerge, offering value-added features either at the front- or back-end, depending on where the high-efficiency gains are sought from cell makers. Indeed, it may be that many of the current next-generation turnkey lines morph into optimized front- or back-end solutions. Possibly, the most likely route for new equipment adoption throughout the industry will come from customized tooling across the various process stages within the complete cell manufacturing line. In fact, for those seeking the largest efficiency enhancement, multiple new-tool/technology adoption is required — something that further decreases the pull on turnkey lines. Customized equipment adoption is not new to the leading cell makers — only now the choice is more varied with different high-efficiency cell concepts to pursue. The wildcard here however is the level to which established market-leaders for specific parts of the standard cell production flows adapt their tooling to modify for high-efficiency or integrate directly up/downstream with new equipment and process steps needed for the cell concepts being proposed.
While each high-efficiency cell concept has specialized process steps and associated tooling necessary to enable these, ultimately the selection of equipment from the market may well owe more to safeguarding and differentiating patents licensed, with intellectual property being protected. Arguments are already being given as to the validity of one high-efficiency concept over another from a technical standpoint, but in truth, technology exclusivity may form the underlying basis for subsequent sales and marketing strategies. While potentially confusing those starting out in high-efficiency cell decision making, this may divert the attention of cell producers to taking more ownership in customized line improvements, or simply performing customized retrofits by strategic partnering further down the equipment supply-chain. Customized retrofitting plays far more into the hands — and comfort zone — of established cell manufacturers; those whose initial capacity expansions were in fact differentiated in the first place.
Selective emitters to the fore — not for the first time
There are several approaches to improving the efficiency of the standard c-Si cell concepts:
- Increased light trapping by improving surface texturing;
- Optimizing passivation layers (both front and rear surfaces) for reduced recombination losses;
- Redistributing the phosphorous diffusion levels below and between the front surface contact grid; and
- Changing metallization techniques for improved aspect ratios and increased current collection.
While each is subject to its own body of research, the advanced cell concepts based upon the standard cell type typically draw on a combination of the above efficiency-enhancement steps. No more so is this evident in what’s known as the selective emitter cell types, which form the basis of the third route highlighted above. Selective emitter is rapidly becoming a buzzword within the industry, and down through the equipment supply-chain, for almost the first time. Selective emitters provide probably the most immediate route to increased cell efficiency within c-Si production today. Typically however, redesigning the front finger process comes hand-in-hand with addressing improvements to front metallization techniques compared to the historical wide (and high) lines deposited by screen-printers.
When reviewing the range of selective emitter schemes to choose from, even here there are a number of options:
- Screen-printed phosphorus-containing paste;
- Diffusion-masking; and
Within each also, there are different approaches. For example, laser-doping can be introduced via diffusion of phosphorous within the PSG layer prior to removal, or by the introduction of a phosphorous-containing layer after or during the SiNx deposition stage. Variants exist combining some of the steps above.
Selective emitters are not new to the industry. The research community has been acutely aware for some time that c-Si cells based upon selective emitter formation were fundamental to roadmap evolution. And efforts to commercialize them go back over ten years, before many within the existing equipment supply chain were actively involved in the solar industry. Indeed, with the research labs having had a number of years head-start compared to in-house company-located R&D efforts, intellectual property and licensing becomes a key issue. The full ramifications of this will only be played out in time.
Of more interest though are historical perspectives on the selective emitter concept: in particular, the research activity at the University of New South Wales (UNSW) during the mid-1980’s coupled with the licensing and commercial implementation by BP-Solar shortly after, within their Saturn lines. It wasn’t branded selective emitter then, but more appropriately high-efficiency. The concept in question was the well-known laser-grooved buried-contact (or LGBC) cell. Revolutionary in its day out of UNSW, herculean in proportions the success of BP-Solar to produce over 150MW of these cells, LGBC technology laid the foundations for many of the selective emitter schemes today, some of which can simply be regarded as high-finesse next-generation versions of the LGBC cell concept. What’s different now, compared to the time period when BP-Solar were implementing these cells in production, is the market size and the participation of an equipment supply-chain co-developing the necessary production tooling to fast-track industrial qualification. For now, the equipment supply-chain is actively seeking to drive the adoption of selective emitter cell types: albeit implemented by necessity and no longer viewing them with curiosity as an esoteric, technology-led exercise.
Other factors affecting adoption timelines
While equipment manufacturers are hard at work finalizing tool supply for the new c-Si cell concepts, the timing is of course dictated by a number of external factors, including:
- Silicon material cost and supply, and the overall competiveness of c-Si vs. thin-film panels at the levelized cost of energy stage;
- Average wafer thickness decreases for c-Si cells, and their impact on handling equipment used within production lines;
- The timing for rear-passivation layers (stacks) being implemented to increase efficiency there;
- Increased thin-film utilization rates, and how quickly this adds to the overall competitive landscape relative to the demand available; and
- Everything next-generation related, including radically new approaches to c-Si design, organic PV, and other potential wildcard entrants.
So, who knows, maybe reviewing the equipment landscape after the trade-show season at the end of 2010 will feature more twists and turns. Tracking market trends with so many diverse approaches to technology on offer could simply not be more important in deciding strategy and supply of equipment within the solar industry today.
Finlay Colville received his BSc in physics at the U. of Glasgow in 1990 and Ph.D in laser physics at the U. of St. Andrews in 1995, and is director of marketing for solar at Coherent Inc., 5100 Patrick Henry Drive, Santa Clara, CA 95054 USA; ph +44-7802-238-775; email Finlay.Colville@coherent.com; www.Coherent.com/Solar.