Crystalline Silicon Solar Cells Future Directions Stuart Bowden BAPVC January 2011 Stuart Bowden BAPVC January 12, 2011 1
Stuart Bowden Co-Director of Solar Power Labs at ASU Work relevant to BAPVC: Pilot solar cell production line on industry standard 6 wafers to advance processes, manufacturing science and metrology. Heterojunction solar cells for low temperature high efficiency cells Kerfless production of wafers at 10 to 100 µm using lasers. Nanostructured solar cells to achieve the 86.8% efficiency limit Tandems on InGaN. For QESST ERC: co-lead on Thrust 1 Terawatt PV and its implementation via the student pilot line testbed Solar Power Laboratory 2
Scale of the Problem - Motivation When I started in PV there was the hope that temperatures would return to normal Continually rising temperature makes our job more urgent Graph of from Today 1996 Just Is the kept world getting heating hotter up? Solar Power Laboratory 3
Scale of the solution At Historic 40% growth 1. All new US generation ~5years 2. All new World generation ~ 10 years 3. Total US Production ~ 15 years 4. Total World Production ~ 20 years elctricity generation, PV production (GW) 1.E+04 World Electricity production 1.E+03 US Electricity Production 1.E+02 New Generation (Wor 1.E+01 New Generation (US) Cumulative 1.E+00 PV production 1.E-01 PV production 1.E-02 1995 2000 2005 2010 2015 2020 2025 2030 Year
Learning Curves Plots from QESST and 1366
BOS and Efficiency With BOS prices equivalent to module prices higher efficiency modules reduce the cost of PV. Present Solar Power Laboratory 6
Current Silicon PV Market Silicon prices have declined dramatically recent years Crystalline silicon continues to dominate the PV industry at 87% of the market in 2010 Solar Power Laboratory 7
Current Silicon Device Technology Efficiency 15 22 % Diffused junction emitter SiN AR coating and passivation Screen printed contacts Aluminum back surface field Large interaction between component parts and processes Many Variations Solar Power Laboratory 8
IC Processes Solar Processes Process Past IC Heritage Present Solar cells Future Solar Cells Feed stock Siemens process Siemens process Fluidized bed reactor Ingot Cz Mono and/or multi Direct growth, ribbons Wafering blade sawn wire sawn Kerf-less Thickness 600 um ~200um 50-100 um Doping Passivation Contacts High temperature diffusions Thermally grown silicon dioxide Photolithography/ lift-off High temperature diffusions Silicon nitride Screen printed silver Copper Plating Low temperature depositions of a-si Al2O3, -ve SiN Solar Power Laboratory 9
Feedstock Electronic grade silicon fluctuates widely but at 40 $/kg : Better use of the Si feedstock is a key cost driver Current Future Efficiency 15% 20% Thickness 220 um 100 um Kerf 280 um Kerfless Si only cost* 0.3 $/W 0.05 $/W Wafer cost 0.7 $/W 0.1 $/W? * crystallization costs are in addition Solar Power Laboratory 10
Feedstock Even when electronic grade silicon is cheap, new production methods needed to meet scalability requirements. Photo: REC Solar Power Laboratory 11
Wafering Wafering wastes half the silicon in cutting and there is a limit to the device thickness Alternative wafering with ion implantation and peeling. At ASU we ve adopted laser cleaving. Direct growth of substrates for ribbons etc http://www.evergreensolar.com/ Stuart Bowden BAPVC January 12, 2011 12
Surface Passivation Silicon solar cell peak efficiency is 10-100 µm. Surface activity dominates device performance as we go thinner Many options for surface passivation such as SiN, Al2O3, organics Need to tailor the surface passivation to the doping concentration and type. future present Solar Power Laboratory 13
Light Trapping Essential for wafers Usually combined with the surface texturing Needs development alongside surface passivation as sharp tips are recombination sites. Photonics. Light trapping at the module level? Stuart Bowden BAPVC January 12, 2011 14
Metallization Silver price is being dragged up by the price of gold We can do roughly 10 times bigger industry than today. Fraunhofer-ISE Solar Power Laboratory 15
Material Abundance Stuart Bowden BAPVC January 12, 2011 16
Metallization option Cu plating Copper is 100 times cheaper than silver Plating gives a much denser metal finger Wet process Good for thin wafers Cu diffuses in Si at room temp Photo: Fraunhofer-ISE UNSW Stuart Bowden BAPVC January 12, 2011 17
Silicon Heterojunction Wide bandgaps such as asi give a junction as well as surface passivation. Heterojunction reduces recombination, enabling high Voc Low current due to absorption in the top a-si and transparent conducting oxide (TCO) Low fill factor Surface Inversion E V Depletion Region Stuart Bowden BAPVC January 12, 2011 18
Novel Device Design IBC solar cell has junction and all the contacts on the rear of the cell High J SC, low R series Ease of module manufacture Visually appealing http://www.sunpowercorp.com Multiple process steps Solar Power Laboratory 19
Manufacturability Critical to any new process technology is throughput. Many producers are above or near 1 GW p = 500 million wafers/year = 15 wafers/sec All inline processing vs batch processing Solar cells are more like CDs than ICs Module costs are increasingly important so need to think how cell are to be encapsulated. Rear contacts are very attractive. Solar Power Laboratory 20
Ideal Bulk Silicon Solar Cell 10 100 um thickness N-type to tolerate impurities and low cost feedstock Surfaces are critical recombination sites Low temperature processing for junctions and metallization Rear contact for ease of metallization and incorporation in module Efficiency ~ 25 % for reduced BOS Solar Power Laboratory 21
Acknowledgements Much of this material was taken from: www.pveducation.org QESST engineering research center kickoff meeting University PV courses especially Honsberg (ASU) and Buonassisi (MIT) Solar Power Laboratory 22
Advanced Silicon at Solar Power Lab. Nanostructures allow new physical mechanisms, which can be used to achieve solar cells with higher efficiency or new functionality Goal is to transition nanostructures to existing technologies
Conclusions Leverage points for lowering the cost of PV electricity 1) Increase efficiency to lower BOS 2) lower direct cost with earth abundant materials. 3) increase scalability. Decrease of waste in productions line. Solar Power Laboratory 24