Two for the price of one: reducing solar silver costs using a dual-print process.
Recently our company partnered with DuPont Microcircuit Materials, a leading manufacturer of photovoltaic metallization pastes, to analyze potential cost-savings and cell conversion efficiency improvements in the manufacture of c-Si solar cells. The cost savings centered on silver paste consumption reductions, a central focus in the photovoltaic market for some time. With silver prices at $1300/kg at the time of our study, cutting cost was more important than ever. The drive to reducing silver use has already led to more efficient grid architectures, including the design of narrower finger widths, moving from the 80 to 90[micro]m range to widths of 50 to 60[micro]m in production. Printing these finer dimension fingers has been achieved through advances in paste technology, print platform capability and improvements in screen production with ultra-fine meshes. However, paste laydown reductions have resulted in thinner busbar paste volumes, so one must be careful to balance this so that subsequent solder adhesion (during the attachment of tabbing ribbons) is not adversely affected.
The work with DuPont MCM utilized several components to analyze single- and dual-print processes and their relative effectiveness in relation to paste reduction, cell reliability and cell efficiency. These elements included DEK-manufactured ultra-fine mesh screens, both traditional and optimized DuPont MCM silver metallization pastes and DEK electroformed nickel stencils. With the single-print process, standard mesh screens were used to print the cell rear side, and ultra-fine mesh screens were used for the front-side grid in total (fingers and busbars) with a latest-generation front-side paste. Using ultra-fine meshes in single-print mode permitted the desired reduction in paste consumption--as much as 30% in this evaluation--as well as slight increases in electrical efficiency but, as I cautioned, did result in reduced solder adhesion forces. Because a finer mesh was used, the busbar area was printed thinner than the minimum required for good adhesion. So, with a single print scenario and ultra-fine screens, we achieved what I would call a win-win-loss: reduced silver paste consumption (win), slight improvements in electrical efficiency (win), and reduced solder adhesion (loss).
Taking the analysis further, the next phase leveraged dual-print techniques to find out if paste laydown reductions, cell efficiency improvements and good solder adhesion could be achieved. For this evaluation, busbars were printed with an optimized--or floating--busbar paste through a 400 mesh (400 wires per inch) screen. The advanced busbar paste is. non-contacting and enables the busbar to do its job (conduct current away from the fine lines) and mechanically adhere to the wafer surface without making electrical contact with the wafer itself. Therefore, with the optimized busbar paste--as opposed to the traditional front-side grid paste--less paste can be used without adversely affecting adhesion. In fact, in our evaluation, the optimized busbar paste actually increased adhesion force beyond the normal standard. Next, specialized dual-print electroformed stencils, 30[micro]m thick with 30pm apertures, were used to print the fine finger widths with a standard front-side paste. Because the stencil design permits a 100% open area in the apertures (as opposed to mesh screens), extremely fine lines can be achieved and, therefore, paste consumption reduced further. So, here we have a win-win-win scenario: a relative increase in cell efficiency of >0.5%, reduced paste consumption of up to 40% and higher reliability modules due to the improvement in solder adhesion forces.
With the dual-print process, the rear side is printed, followed by the screen Print of the front-side busbars and then the stencil print of the front-side grid/fingers. With the extra print step, an additional printer is required. But, a detailed cost of ownership analysis reveals that even with the extra equipment costs, the dual-print processes as outlined above yield significant cost savings. Our calculations show that with the dual-print scenario described here and a consistent cell efficiency of 17.25% (note that this assumes no increases in efficiency, although our work showed a greater than 0.5% relative efficiency gain), a cell manufacturer would save $120,000 per year, even taking into account the cost of the additional printer and stencils.
With all the pressure to reduce manufacturing costs and lower the overall price tag of solar power, the results of this work are a tremendous step forward. This is arguably the best current solution for silver paste consumption reduction--one that also improves, efficiency and enhances reliability. And work is already underway to advance stencil technologies for, perhaps, a single print solution with similar results. I'll keep you posted.
For more information on the study conducted by DEK and DuPont, please contact the author.
TOM FALCON is a senior process development specialist at DEK Solar (dek.com): email@example.com. His column runs bimonthly.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||SOLAR TECHNOLOGIES|
|Publication:||Printed Circuit Design & Fab|
|Date:||Apr 1, 2013|
|Previous Article:||Negative Z to adjust for board warp: if you want to compensate for bow, you've got to get low.|
|Next Article:||Testing new grounds in automotive electronics: manufacturers are increasingly designing products for ease of test.|