New processes for c-Si cells should improve efficiency, costs - Photovoltaics World
This article is from

New processes for c-Si cells should improve efficiency, costs

by Arnaud Duteil, Yole Développement

September 30, 2010 - While the venture-funded startups get most of the attention for their bets on disruptive solar technologies, a lineup of significant incremental developments now coming out of the lab are about to quietly boost the efficiency and decrease the cost of mainstream crystalline silicon (c-Si) cells.

Solar manufacturing equipment suppliers are currently introducing commercial plug-in processes to improve the emitter profile, metallization, passivation, and light-trapping surface texturing, often with innovative new uses of inkjet-type printing and laser treatments. Even more interesting is the potential for a relatively simple technology to move the contacts to the back of the wafer; this is now coming on the market. Of course, some of the leading solar companies have developed their own technologies to do these same things -- but now similar processes are becoming more widely available, and can be plugged more easily into many more production lines.

These technologies together appear to define a path to exceed 19% efficiency for many makers of c-Si cells. But making the right choices in the right combinations to get the best performance improvement for the lowest cost, and continuing to drive down costs with production experience, remains the opportunity for individual companies to gain competitive advantage.

Figure 1: Solar efficiency and production cost by technology, as of early 2010. Technologies with the best potential for economic mainstream energy generation in the near term are likely those already close to the sweetspot of reasonable cost and efficiency. (Source: Yole Développement)

Potential shortcut to back contact cells

Most significant of the new commercial processes hitting the market is the relatively simple back-contact process developed by the Energy Research Centre of the Netherlands (ECN), now licensed to Solland Solar and Canadian Solar, which moves the bus bars to the back and redesigns the silver fingers on the surface to a more efficient radial pattern to minimize shading and resistance. More importantly, it also simplifies module production by replacing stringing and soldering with pick-and-place on patterned foil with conductive adhesive, a high-speed process much like surface mounting ICs on a printed circuit board. Research cells have demonstrated ~17% efficiencies on multicrystalline silicon, and producers are now starting to ramp commercial production.

SunPower uses a more complex back-contact structure developed over many years to achieve its industry-leading efficiencies using monocrystalline substrates, and Kyocera gets results roughly similar to those from ECN with its proprietary process for multicrystalline substrate. But now the approach is likely to become much more widely used.

Figure 2: Key new processes for improving efficiency coming to market. (Source: Yole Développement)

Improvements coming in emitter profile, metallization

First of the new processes to be likely to most widely adopted, however, is control of the emitter profile across the cell surface. Doping more heavily under the contacts where concentration is most needed, and leaving the rest of the surface area more lightly doped to absorb more light, can potentially give about a 0.9% boost in efficiency, for perhaps $0.15 per wafer.

Several vendors are introducing solutions that are relatively easy to plug into existing process flows, though the technologies vary widely, using innovative approaches including ink jet-like printing, hot melt resists, and laser patterning.

  • Schmid is getting good traction, including apparently at SunPower, for its process that jets on an etch-masking pattern with a heated melt material that solidifies at room temperature to protect highly doped regions where the contacts will be, then removes dopant from the unprotected areas. The masking strips are then removed and the contacts printed.
  • RENA and Synova instead combine laser patterning with guided doping, guiding a laser through a phosphoric acid solution for precise ablation of the contact pattern followed or simultaneously with phosphor doping along the same line.
  • Centrotherm uses a laser to create openings in an oxide mask, then diffuses a heavy concentration of dopant into those lines. These same processes can also be used to make the dielectric on the back of the wafer for metal-wrap through back-contact cells.

The next new process in the pipeline that's likely to see wide adoption (after the improved emitter profile) is fine-line metallization, where some sort of jetting and electroplating will likely replace screen printing. Here, too, there are multiple competing solutions to make narrower, more conductive fingers and bus bars quickly and cheaply. Optomec offers an aerosol jet printing system with multiple heads that it reports can do collector lines in some 3sec per cell. RENA deposits a fine seed line (down to 50μm), then electroplates the conductive metal over that.

Other technologies soon coming out of the labs will improve passivation of both top and bottom surfaces to reduce recombination losses, and improve surface structure and texturing to trap more light -- moving from wet processing to laser, reactive ion etch, or nanoimprint lithography for more precise control.

All these technologies coming to fruition have of course been in development for some time -- labs and suppliers started to invest more aggressively in PV process technology once German feed-in tariffs jump-started the market and real demand started to develop for c-Si production tools. As the solar sector matures, cell makers are likely to get increasing technology development support from investment from their suppliers. And most of this effort will go toward c-Si since it will continue to account for some 70%+ share of market, and the smaller thin-film market also remains highly fragmented across different technologies. Progress on the thin-film side will thus likely be more incremental, as equipment suppliers work with specific customers to evolve the processes by experience, but are less eager to step up to invest heavily in independent development.

All these technology trends are discussed in detail in a report from Yole Développement: PV Technology Roadmap: Technology Trends in Crystalline and Thin-film Solar Cells.

Arnaud Duteil received a Master's Degree in Innovation and Technology Management from EM Lyon Business School and a Master's Degree in Engineering from ECAM. He is in charge of photovoltaic market analysis and reports at Yole Développement, Lyon, France.