Building Blocks to Boost Solar Productivity (Part 2)
Today’s blog post is part two of the three part series running this week looking at the interrelated building blocks that are key contributors to producing solar modules at a cost of less than US$1 per watt.
Thinner Wafers and Higher Throughputs
Silicon wafers for PV are much thinner and more fragile than those used in typical integrated circuit (IC) fabs. While the IC industry starts with 700-750 micron thick wafers, the solar industry has been driving to thinner wafers to save on silicon consumption, the largest single materials cost in solar cell production.
Today, a solar wafer is about 170-190 microns thick, about half the thickness of a decade ago; and solar industry roadmaps are targeting wafer thicknesses of 100 microns or less within the decade. The optimal thickness of a wafer is a balance between silicon cost, production yields, cell efficiency, and processing costs. Silicon costs decrease with wafer thickness, but given the interplay of wafer thickness and sawing kerf losses, the decrease in net silicon cost with final PV wafer thickness is sub-linear (figure 1). Ongoing improvements in wafer sawing technology are reducing kerf losses, and experimental “kerfless” direct wafering processes in which wafers are formed from ribbon growth, ingot cleaving, epitaxial lift-off techniques, etc. offer the possibility of much thinner wafers with minimal kerf losses. In the coming decade, it is reasonable to expect 80 micron thick wafers by a variety of sawn and kerfless methods, and eventually 40 microns or less by kerfless methods.
Figure 1 Polyslicon price and wire diameters
The limit to how thin solar wafers will become is likely to be determined by the need to absorb the majority of the sunlight striking the wafer and by the degree to which the industry implements low-cost production methods for passivating wafer surfaces and interfaces to minimize recombination losses. Various light trapping schemes have been demonstrated in laboratory devices, and new generations of passivation methods for p-type surfaces are starting to be commercialized.
Compared to semiconductor fabs, silicon wafers move much more quickly through solar production lines, requiring sophisticated handling to minimize breakage. Solar cell fabs use far fewer process steps than IC fabs, but the volume of material is much higher. Solar tools are increasingly targeted at through-put rates above 2500 wafers per hour as typical solar fabs continue to grow. While the newest generation of gigawatt fabs will have many parallel processing lines, the throughput of individual tools will continue to increase, in large part through continued innovation in tool design and operation.
Trends in increasing solar tool productivity without sacrificing yields are evident in the tools used to section crystalline silicon ingots in wafers. New crimped wire technologies can significantly increase cropping, bricking and sawing speeds. For example, crimped wire on Applied HCT squarers increase tool productivity by 70% and decrease overall cost of ownership by up to 30%. These increases in processing tool capacity are being paralleled by a continuing increase in the use of tool and materials diagnostic and characterization techniques. Going forward, step-wise increases in tool capacity through both tool size and productivity and implementation of more sophisticated on-line metrology tools will further increase solar fab throughput, while maintaining the tight process and materials controls needed for commercial cost competitiveness.
Come back tomorrow for the last blog post, "Advances in Automation" of this three part series.