Printing is not a new concept for the photovoltaics (PV) industry; in fact, it has relied on screen printing to create the top electrodes for crystalline silicon (c-Si) PV for some time. This is a somewhat unexciting application, but one that accounts for almost $200 million in sales of silver inks and pastes. In addition, this means that the PV industry is at least familiar with functional printing and what it can deliver.
An obvious next step is to use functional printing to create the core PV layer itself. The reason why a solar cell firm would want to do this is primarily one of cost. Although PV producers might be loath to admit it, at any given efficiency level, their products are something of a commodity. There really isn't that much to choose between one brand and another other than price. This is one reason why panel makers build the factories close to key geographical markets; solar panels are heavy and transportation can add considerably to costs. So making the right production location decision can be critical.
So can choosing the right manufacturing technology and--in theory at least--printing seems to fit in quite well in this regard. Not only is it an approach that is well understood in the PV industry, but printing is generally considered a low cost approach compared with other deposition technologies that are taken from the semiconductor industry. This is true at the level of capital expenditure and operational expenditure. As far as the latter is concerned, printing supposedly reduces the amount of energy and material consumed in the manufacture of solar cells compared with other more traditional deposition approaches. This is obviously important intrinsically, but also helps preserve the image of PV as a "green technology." In addition, printing is uniquely a deposition and a patterning technology. This may be important for texturing electrodes (or sometimes even the absorber layer), which translates directly into higher performance cells.
Finally, printing has the advantage that it is very well suited to flexible substrates and this in turn means that it is a good choice for building integrated PV (BIPV) products; these value-added products are yet another way that PV companies can distinguish their products in the marketplace.
This is the theory anyway. The reality is a little different. For a start, using printing, as opposed to a high-temperature manufacturing processes, usually leads to a substantial reduction in performance in the final cell, and this is only partially offset by the fact that one process can be used for deposition and texturing. Since performance (i.e, energy conversion efficiency) is the key measure of the usefulness of a cell, this is a big limitation. In addition, casual descriptions of printing PV make it sound like all one can merely buy a screen printing machine or industrial inkjet and be up and running printing PV in a month or so. The reality is that printing PV can be very difficult technically, with lots of operational problems standing between the inception of an idea and it realization. This is a fact that many in the PV industry seem blissfully unaware of.
With all that said, printed PV definitely seems like an opportunity for the printed electronics (PE)/functional printing community, which seems to have floundered somewhat in the past year. First this community has never been able to keep up with some of its more extravagant promises from its early years. Second, it has been hit hard by the worldwide economic downturn. What PV represents is a remaining addressable market where printing can be applied and which is still in growth mode, although at lower rates than were expected a year or so back. This has caused a repositioning in the PE industry. For example, several conductive silver ink makers are now specifically rebranding their inks and pastes for the PV market and at least one PE start up that was focusing on printed transistors just a few months back is now a PV company!
Where Printing Fits in the PV Marketplace
Until recently, almost all commercially produced solar cells relied on crystalline silicon (c-Si) technology. This technology has been able to deliver adequate conversion efficiency at acceptable costs, which makes it suitable for a wide range of applications, from providing electricity in remote locations that are not served by the electrical power grid to large-scale power plants. However, c-Si is inherently unprintable in the usual sense. The only way that printing can be applied to semiconductors such as c-Si is in the form of transfer printing where high-performance semiconductor devices are created using classic semiconductor processes and then transferred to flexible substrates using a printing-like process. Semprius is doing something like this to produce concentrated PV devices using GaAs technology.
Thin-film and organic PV (and the related area of dye sensitive cells, DSCs) are a much better prospect for printing and both have grown considerably in the past few years, in absolute and relative terms. With regard to absolute growth, the reason consists in all the factors that have promoted PV as a whole: renewability, low carbon emissions, price stability, tax incentives, feed-in tariffs and so on. With regard to relative growth, the initial reason for the interest in TFPV (and perhaps OPV/DSCs) was the shortage of c-Si. But this shortage has dissipated and the main drivers for TFPV, OPV and DSCs now are (at least potentially) the advantages of low-cost, easier manufacturing and the ability to do that manufacturing on flexible substrates, which is a plus compared to the dominant crystalline silicon PV. We have already noted that printing of these newer forms of PV is easier said than done and an additional issue with many of the materials that are used for TFPV and OPV/DSC is that they cannot easily be turned into inks. This often yields some competitive advantage to skilled ink makers and proprietary printing processes.
Although silicon-based PV of one kind or another is by far the most common kind of PV, it is the one to which printing has been applied the least, if one (once again) excludes the electrodes/contacts. Where printing has been applied is mostly in niches, with perhaps the most important being Innovalight. This company uses silicon nanoparticle ink for its a-Si PV cells. The ink is in turn created with "radiofrequency plasma" technology. However, although it uses silicon, the expected performance of Innovalight's products, which are due to appear in 2009, are closer to that of OPV than any other kind of silicon PV and it is not really appropriate to think of this technology as being grouped with c-Si or a-Si PV from an addressable market standpoint. The other kind of printing that might have some general appropriateness for TFPV is transfer printing. The new direction that Semprius has taken toward PV seems to illustrate how printed silicon could be used in the PV market. We note, however, that Semprius' current activity involves GaAs, not silicon.
Printing As Way to Make TFPV Competitive with c-Si PV
The TFPV/OPV technologies--other than the ultra-expensive GaAs solution--may never quite catch up to c-Si in terms of efficiency. However, they might be able to demonstrate the right combination of efficiency, cost, flexibility and low weight to make them competitive with c-Si PV in a number of different important addressable markets. In fact, we are already beginning to see this happen to some extent with the growing number of on-grid applications for a-Si PV and more dramatically CdTe PV.
NanoMarkets believes that printing will be one enabling technology to reach these goals. As discussed at the beginning of this chapter, there are plenty of reasons to believe that printing is a good way to lower costs in the PV industry when compared to more traditional vapor deposition methods, although as we also discussed there are major challenges to this goal. There are few examples, if any, of firms using printing to create the absorber layer for CdTe PV at the present time, although the literature does talk about screen printed CdTe. Part of the issue here, however, is that the CdTe sector is overwhelmingly dominated by one firm--First Solar. Thus the lack of a presence for printing in the CdTe sector says more about the fact that this one specific firm does not use printing than it does about the potential for using printing in CdTe. We note that even now, screen printing is being evaluated as a way to reduce costs and improve material consumption for the CdS and CdTe layers in CdTe-based PV.
Printing is being used by a few firms in the budding CIGS sector. One such company, Nanosolar, has developed a CIGS-based ink and recently started shipping these solar panels. Nanosolar claims that printing solves an important uniformity issue for CIGS, which is keeping the particles uniform in the proper atomic ratios, especially over large areas. In the CIGS sector, there is a need for high throughput low-cost processes because other common production methods (evaporation of the elements in vacuum; sputtering of the metals followed by selenization with H2Se) suffer from relatively slow throughput, poor material utilization, and relatively high vacuum. Apart from Nanosolar, several companies and institutes are evaluating ink formulations, some based on nanomaterials, for the CIGS absorber layer.
CIGS is usually billed as the TFPV technology of the future, so perhaps the fact that printing seems reasonably well accepted in the CIGS sector speaks well for the future of printed PV. On the other hand, nothing is certain in the CIGS sector at the present time, including how successful it will be in the long run. Nanosolar is known to have built significant capacity for printed CIGS PV, but it is far from clear how much of that capacity is being used.
Meanwhile, printing has already become very closely associated with OPV in part because this is the chosen fabrication method for Konarka, the firm with the largest mindshare and dominant IP portfolio in this space. This company uses a roll-to-roll printing process to produce its OPV cells, which it sells under the brand Power Plastic. Its process technology is similar to the printing used to make photographic film, and thus Konarka was able to get up and running fairly quickly by purchasing Polaroid's manufacturing facility in New Bedford, Mass. Konarka's take on the virtues of printing is that its use instead of traditional vapor deposition techniques increases throughput, significantly reduces energy consumption, and thus lowers the cost per watt of its product--a must for PV technologies to compete with traditional energy forms.
Printing and Electrodes for TFPV
But with all the interest in printing PV, the fact remains that the most prominent role of printing in the PV industry today is in c-Si PV, which use thick-film printed silver electrodes on the front, and often on the back as well. Although this is a reasonably large market, it is one in which suppliers are mostly well established so it is not really an opportunity in terms of new market entry potential. The one caveat here is that there may be some room for printing of nanosilver contacts in c-Si, or indeed any kind of PV cell, if higher conductivity in electrodes are shown to lead directly to higher performance of the cells.
TFPV and OPV/DSC technologies are at a stage where the materials and fabrication methods used for metallization are not completely established, so printing could find some opportunities here. However, in many cases transparent conducting oxides (TCOs) are being used for top contacts and TCOs have not shown themselves to be well-suited to printing, although both ZnO and ITO inks exist. Transparent conductive polymers are sometimes easier to print, but would mostly play a role in the OPV sector.
The interesting, longer term potential for printing is in the nanomaterials sector. Theoretically, nanomaterials (including carbon nanotube preparations) could be highly conductive, offer good transparency and be relatively inexpensive. And current research directions seem to indicate that printing will be an important part of the deposition and patterning of nanomaterial-based electrodes.
Finally, while much of the attention in the area of electrode development is currently focused on the front/top electrode, this is because demanding performance requirements are necessary for this electrode. For precisely this reason, printing, which as we noted earlier has its own performance issues, could prove to be well matched to the back/bottom electrode and to the related reflective layer where this exists.