Best opportunity to make cost per watt more competitive with other electrical generation technologies is naturally to start with those technologies that are already achieving best efficiencies at reasonable costs, in the central region of the efficiency vs cost tradeoff matrix.

 

Though there’s healthy headroom for development in CIGS and tandem thin film silicon cells, we would argue that crystalline silicon technology, with its leading efficiency and largest market share, is likely to see the sharpest cost reductions. Though thin film technologies are continuing to capture an increasing share of the solar market, crystalline silicon modules still dominate 70% of the market. And compared to the fragmented variety of thin film technologies, c-Si remains the most significant market for equipment suppliers, and is getting most of the development investment. That’s producing a new generation of production equipment and materials coming on the market this year and next that should enable widespread improvement in the efficiency and cost of solar modules. 

Crystalline silicon cells could increase efficiency by 50% in the next few years if R&D results can be brought into general production.

 

The sweet spot is likely finding ways to upgrade the efficiency of today’s multi crystalline cells to somewhere close to mono multicrystalline levels, as the mono cells ~20% lower price advantage gives some headroom to add new processes.         

Easiest new process for crystalline silicon cell makers to plug in —and thus likely to be the first widely implemented—is to redistribute the emitter profile on the front surface, by doping a little more heavily under the contacts to improve conductivity, but leaving the surface less heavily doped elsewhere to absorb more light. Equipment makers are offering a range of highly innovative new solutions, using lasers and ink jet-like deposition of masks to control dopant density, the best solutions reporting a gain of 0.9% efficiency for a cost of around $0.15 per wafer. The next step will likely be fine line metallization, making thinner contacts with better electrical characteristics that shade the top surface less, by replacing screen printing with inkjet or aerosol jetting and electroplating. The industry will next likely look to optimize the passivation layers on top and bottom to reduce recombination losses, then  move to improving surface structure and texturing to trap more light, moving from wet processing to laser, reactive ion etch, or nanoimprint lithography. 

 

A more radical transformation is to move the contacts to back of the cell, leaving more of the front surface open to collect light, and allowing optimization of doping and surface treatments as well, once the contacts are out of the way. Some leading producers like SunPower and Kyocera are of course already doing this with their own proprietary technology, but commercial processes are now coming on the market to allow other producers to do so as well. The Energy Research Centre of the Netherlands (ECN) has developed a relatively simple metal wrap through back contact technology, now licensed for production to Solland Solar and Canadian Solar that allows simpler module assembly with a process much like mounting ICs on printed circuit boards, and has boosted multicrystalline efficiencies to 17% at the research level. Widespread use of this technology could potentially significantly improve the efficiencies of much of the c-Si cell market in the near term.

 

There’s plenty of potential as well of course for low cost thin film PV technologies to improve efficiencies enough to really challenge wafer-based modules. But the best producers of most of the thin film technologies have been holding around 11%-12% efficiency for the last couple of years, suggesting development may have reached a plateau.

CIGS looks like the most promising of the thin film technologies, showing impressive 19% efficiency in R&D on small cells, suggesting big potential for improvement at the production level. There’s a clear path to reducing cost as well, by moving away from vacuum deposition processes. But depositing consistent layers at high throughput is a challenging task, and one that will likely depend for solution largely on serious investment by equipment suppliers. Few solar companies –or few solar investors—want to spend the money or the time to develop their own deposition tools as well as their own cells and modules. So development is largely being done in cooperation with equipment suppliers on their standard turnkey lines, by tweaking and testing the process while using it in production. Improvements are thus likely to be steady but gradual, not radical new processes. Major progress may primarily depend on the willingness of equipment makers to invest in this still small market or rather, more likely, on one of the tool suppliers deciding to step up an invest seriously in development.

CdTe of course leads the industry in cost per Watt, and accordingly dominates the solar farm market and will continue to be a significant alternative. But the technology seems unlikely to be the solution to global solar power generation, simply because it is limited by single company capacity, and by potential questions about material availability in the long term.

Tandem cells that add a second layer to amorphous silicon to capture more wavelengths of light are already beating straight amorphous silicon on cost per Watt as well as efficiency, and are likely to completely edge out the single junction a-Si cells in the market over the next five years. But tandem cells themselves haven’t seen much efficiency improvement over the last several years, so they too may be approaching a practical limit of efficiency.