Home


Feature article Think Small! issue 1, Volume 2:

Solar Cell Manufacturers Bank on Thin-Film

by Olivier Nowak

The difficulty in securing a feedstock supply to produce conventional wafer-based solar cells has encouraged a frenzy of industry projects related to thin-film solar cells. Wtc has counted some 34 companies developing and in some cases building production lines for thin-film cells. Last year saw First Solar, Nanosolar and Ersol announce investments in the range of $100 million each to build new thin-film plants.
At this rate, thin-film cells could represent about 20% of the pv technology installed (on a year to year basis) worldwide in 2012. Wtc estimates that the market for thin-film cells is set to explode in the next few years, reaching $1.5 billion in 2012. While the usa has been steadily loosing ground to Japan and Europe in wafer-based cells, about half of the thin-film cells are made by us companies and this share is expected to remain high (although the production may not be located in the us).
Aside from the silicon shortage pushing manufacturers towards thinfilm, there are other good reasons to consider the technology. Thin-film cells have the potential to achieve mass production at low costs. The Target of some manufacturers is to achieve $1.3/W by 2012, i.e. a module that can deliver 1 W peak would cost $1.3 to produce. This would represent a marked improvement over today’s cost, estimated to be in the range of $2-5/W. New opportunities are thus arising for semiconductor equipment and materials suppliers to make these manufacturing targets a reality.
In use, thin-film cells provide advantages such as semi-transparency, flexibility and low weight.
The major disadvantage of thin-films cells is their low efficiency—up to 13%, depending on the technology, compared with 18% for conventional cells.
The efficiency of a thin-film cell initially degrades quickly, then reaches a plateau (which is the nominal efficiency in commercial products) where it slowly degrades. Long-term reliability still is an issue for certain modules.
Also, some thin-film cells require toxic elements such as cadmium which will have to be recycled when the cells reach the end of life.

Opportunities in manufacturing to reach industrialisation

While wafer cell production ressembles that of integrated circuits, thin-film cells are, in their manufacture, closer to displays.
Usually, the process starts with a conductive substrate that acts as the front or back contact. A few micrometers of materials are deposited on this substrate, e.g. by pecvd. Subsequently a second contact metal is deposited (up to around 0.5 ìm) and micro-structured. Finally, the cell is encapsulated, usually in a polymer film. Contrary to waferbased cells, the encapsulation is not a separate process, especially in the case of cigs (copper indium gallium diselenide) and CdTe (cadmium telluride).
In their favour, thin-films obviously require much less semiconductor material than wafers, i.e. 20–30 g/m² vs 1 kg/m² for wafer-based approaches (although this amount is dropping) and need less energy to produce; 150 kW/m² vs 550 kW/m² for waferbased approaches. Even accounting for the lower area efficiency, this makes thin-film more interesting per Watt.
Ultimately, manufacturers want to manufacture thin-film square-meter modules in a roll-to-roll process. But for the moment, the issue in the manufacturing of thin-film cells remains the uniform coating of large or rather wide areas or bands of material. Bands currently extend to a few tens of centimetres.
Since the manufacturing of cigs and other technologies like multi-junction cells is still evolving industrially, many variations exist on the basic approach and the processes used at each step. In this regard, thin-film cells offer excellent opportunities to develop new and advanced manufacturing approaches.

 

 

Thin-film facilitates buildingintegrated photovoltaic (BIPV)

Besides the general ramping up of photovoltaics, two phenomena sustain the quite optimistic market growth forecasted for thin film:


Bipv is the main target application for thin-film cells. Whereas wafer-based solar panels are add-ons to the building that require to be fixed on top of the construction elements, bipv components substitute themselves for construction elements with added electricity generation functionality. Thin-film modules also offer a decisive advantage over wafer-based in being able to make semi-transparent panels that substitute for window panes on facades, roofs, etc.
Thinfilm modules are light and easy to combine with steel plates for roofs or reservoirs, and also offer a varied range of appearances, some more aesthetically pleasing than deep blue poly-crystalline silicon wafers.
Mems oscillators fit between cmos clocks and quartz oscillators.
Another potential market for thin film cells is off-grid energy supply in impoverished, isolated areas such as in parts of Africa, or the electricity supply of isolated appliances with low power consumption. Here, thinfilm cells would provide cheaper systems than wafer-based modules.
Assuming flexible thin-film cells can be manufactured in volume, completely new applications include military solar tents, clothing and sails. However, the current applications are constrained by current battery chargers for portable devices like mobile phones and gps systems. These (often flexible, even rollable) chargers can deliver up to 15 W but measure up to for 30×120 cm² when unfolded.

CIGS fast emerging as competitor to silicon and CdTe

There are several thin-film cells technologies. Amorphous silicon (a-Si) is the most mature and since the mid 1990s available from a number of vendors. Up to now a-Si cells represented over 90% of the thin-film cell offer. The best a-Si cells reach 6.4% on glass and 4.2% on flexible substrates.
CdTe can also be considered a mature technology, currently available only from First Solar (usa) and Antec Solar (Germany) so far. Commercial modules achieve efficiency of 8 to 9%, with laboratory cells reaching 16.5%.
Copper indium–gallium selenide, sulfide (cigs/cis) include a range of compounds, the exact recipes depending on the manufacturer’s process. In use modules delivers 13% efficiency and 19.5% in the lab—the highest of all thin-film types. The technology has therefore aroused the interest of many newcomers. Wtc has counted at least 10 manufacturers with interest in cigs, but only one or two who can already deliver, e.g. one being Würth Solar. Several production (as opposed to pilot) lines are scheduled to start producing in 2007 and 2008.
Multi-junction cells combine amorphous silicon with micro-crystalline silicon (ìc-Si) or silicon germanium (SiGe) alloy to absorb a larger portion of the light spectrum. Despite the name, the typical grain size in ìc-Si actually does not exceed a few tens of nanometers. The largest wafer-based cell manufacturer worldwide, Sharp, has commercialised double junction (tandem) cells. It also announced it will start producing triple-junction cells in May this year.
  Although thin-film has yet to be fully industrialised, a new generation of solar cell technologies like organic/ dye-sensitized or nanocomposite cells are already on the starting blocks. And other new concepts—crystalline silicon on glass or even more radical concepts like microspheres—are also vying, quite literally, for their place under the sun.

Conclusion

With a booming photovoltaic market on the one side, and investors willing to fund their expansion on the other, thin-film solar cell manufacturers are rapidly developing their technologies and building out the production capacity that can help them tap the market. For all the equipment and material suppliers, and especially those that have no done so yet, this is a very good time to get involved in thinfilm solar cell manufacturing.