#1An Overview of Solar Cell Technology Mike McGehee Materials Science and Engineering Global Climate and Energy Project Center for Advanced Molecular Photovoltaics Precourt Institute Stanford University Nanosolar Konarka John Benner provided the slides with the NREL logo.#2Primary Photovoltaic (PV) Markets Residential Rooftop SUNGEVITY Ground- mounted (Usually utility scale) Commercial Rooftop 2#3How cheap does PV need to be to compete w/ coal? Average power price per households, $ per kWh¹ Denmark ⚫ California Tier 42 California Tier 5² Size of electricity market TWh¹ a year Italy Grid parity³ as of Today Netherlands 0.30 Norway Germany Hawaii Sweden United Kingdom Japan Australia 0.20 California Finland France New York Spain Texas 0.10 South Korea Greece China 0 500 1,000 J N India 1,500 2,000 Annual solar energy yield, kWh/kW₂1 Cost per watt at peak hours, $ per Wp1 2020 Source: CIA country files; European Photovoltaic Policy Group; Eurostat; Pacific Gas & Electric (PG&E); Public Policy June 2008 Institute of New York State; McKinsey Global Institute analysis#4Installed System Price per Watt, 2008-2011 6.00 $5.92 System GP $0.45/Wp 3Q10 Breakout Wafering GP $0.17/Wp Installed cost ($/Wp) 5.00 4.00 3.00 2.00 1.00 $3.72 p-Si + Wafer processing Systems Expenses $1.03 Cell $3.17 H $2.83 Module Assy Cell processing GP $0.12/Wp Module Assy GP ■System GP $0.25/Wp BOS Inverter 1Q 2Q 3Q 4Q 1Q 2008 ότ 3Q 4Q 2009 1Q ότ за 2010 4QP 1QP 2QP 3QP 4QP 1QP 2QP 3QP 2011 2012 4QP Original Source: Deutsche Bank, January 2011; Systems are global (i.e., blended across geographies) My source: R. Swanson, IEEE PV Specialists Conf., June 2011 National Renewable Energy Laboratory Labor ■Connection Permits Mod ASP Innovation for Our Energy Future#5PV is a booming industry, especially in China PV Production (MW) 25000 22500- 20000- 17500 15000- 12500- 10000- Rest-of-World China/Taiwan Row since 2007) (Broken out from 7500- Europe Japan 5000- North America 3803 Source: PV News, 2007-2011 2500- from May 2011 issue 2459 (revised: 07-2011) 1782 1199 0. 47 55 58 60 69 78 89 126 155 201 288 371 542 749 T T T T 11,315 7126 '90 '91 '92 '93 '94 '95 '96 '97 '98 '99 '00 '01 '02 '03 '04 '05 '06 '07 '08 '09 '10 5 Year 23,898#6But not everyone prospered in 2011 SOLYNDRA The new shape of solar Solyndra, SpectraWatt and Evergreen Solar went bankrupt. Jon Stewart, The Daily Show#7What makes the PV industry so • • • interesting? PV addresses the energy problem, which many passionately want to solve. By 2050 the world will need ~ 30 TW of power. Some think PV could provide 20 % of that. It takes a panel rated at 5 W, to average 1 W of power through the day and year, so we would need 30 TW of PV capacity. At $1/W, the industry would take in $30 trillion. • The industry is now well over $40 B/yr.#8There are many approaches to making PV cells and experts do not agree on which one is the best CPV MARKET Thin Films MATERIAL Crystalline Silicon STRUCTURE Ground Mounted Rooftop Mounted Amorphous Silicon CdTe / CIGS Organic Multi- Crystalline Mono- Crystalline 20x-100x National Renewable Energy Laboratory 500x Cu(In,Ga)Se, ~ 1-2 um 2 2 c-Si 180 um Innovation for Our Energy Future#9Efficiency (%) 50 50 48 Best Research-Cell Efficiencies Multijunction Concentrators ▼Three-junction (2-terminal, monolithic) ▲ Two-junction (2-terminal, monolithic) Single-Junction GaAs 44 ▲ Single crystal 40 40 36 46 32 A Concentrator Thin film crystal Crystalline Si Cells ■Single crystal Multicrystalline Thick Si film Silicon Heterostructures (HIT) Thin-Film Technologies Cu(In, Ga)Se2 o CdTe O Amorphous Si:H (stabilized) Nano-, micro-, poly-Si Multijunction polycrystalline Emerging PV o Dye-sensitized cells Organic cells (various types) ▲ Organic tandem cells Inorganic cells Quantum dot cells Spectrolab (metamorphic, 299x) Fraunhofer ISE Boeing- (metamorphic, 454x) Spectrolab NREL NATIONAL RENEWABLE ENERGY LABORATORY (lattice matched, 364x) Solar Junction (lattice matched, 418x) Boeing-Spectrolab (metamorphic, 179x) Boeing-Spectrolab (metamorphic, 240x) Spire Semiconductor (metamorphic, I 406x) 43.5% NREL (inverted, metamorphic) NREL (inverted, NREL metamorphic, Boeing- Boeing- 325.7x) Sharp Spectrolab Spectrolab (IMM, 1-sun) NREL (inverted, 35.8% Spectrolab metamorphic, 1-sun) NREL/ NREL Japan Energy Spectrolab FhG-ISE Spectrolab 32.6% (117x) Varian NREL Varian (216x) 28 (205x) NREL SunPower (96x) Spectrolab (4.0 cm2, 1-sun) Radboud Univ. IES-UPM (1026x) Alta Stanford (140x) Amonix (92x) FhG-ISEAA Devices 29.1% 28.2% (232x) 27.6% 26.4% Kopin 224 24 20 20 IBM (T. J. Watson Research Center) Varian Radboud FhG- UNSW NREL Univ. A- Stanford Spire UNSW Radboud Univ. Alta ISE Devices 25.0% UNSW UNSW Cu(In, Ga)Se2 UNSW UNSW/ (14x) Sanyo Sanyo Sanyo 23.0% Sanyo Sanyo Spire Georgia Eurosolare ZSW Sandia ARCO National Lab Westing- house UNSW Varian Georgia Tech Georgia Tech Tech FhG-ISE 20.4% 20.3% UNSW NREL NREL NREL NREL NREL NREL ZSW First Solar NREL NREL 17.3% 16 RCA No. Carolina University So. Florida Univ. Mobil State Univ. ARCO Boeing AstroPower NREL (small-area) NREL Stuttgart Sharp (large-area) (45 μm thin- NREL Solar Kodak Solarex Euro-CIS 12 Boeing NREL Boeing United Solar film transfer) (CdTe/CIS) United Solar (asi/ncSi/ncSi) 12.5% Photon Energy AMETEK Matsushita Kodak Boeing ARCO 8 Monosolar United Solar EPFLO EPFL Kaneka United Solar Sharp IBM (CTZSSe) IBM (CTZSSe) 11.1% 10.1% Konarka 8.6% (2 μm on glass) NREL/Konarka Univ. Linz Solarmer Konarka 8.3% UCLA Boeing RCA Solarex EPFL Heliatek University EPFL Groningen Heliatek NREL 4 of Maine RCA RCA RCA RCA University Linz Plextronics (ZnO/ PbS-QD) 4.4% Univ. RCA O RCA 0 University Linz L Siemens Dresden (ZnO/PbS-QD) NREL I 1975 1980 1985 1990 1995 2000 2005 2010 (Rev. 9-2011) Lots of records in 2011!#10• • • More factors that make the plot interesting The overall global economy has been turbulent for a few years. Government policies are constantly changing. When an industry based on manufacturing grows faster than 40 %/year in spurts, it is hard for the supply chain to always provide what is needed.#11Conventional p-n junction photovoltaic (solar) cell p-type n-type (1) hole depletion layer + (2) (3) (4) V 0 + electric field electron diffusion potential (Vi) Ec EF Ev light#12Efficiency limits Sources of energy loss Thermalization of excess energy M CB Below band gap photons not absorbed -VB Efficiency (%) 40 (b) 30 50 Black-body limit is GaAs InP 20 20 Cu(In, Ga)(S,Se) 2 CulnSe2 CdTel AM1.5 a-Si:H Ge Cu(In,Ga)(S,Se)₂ 10 CulnS2 CdS Cu₂S AMO CuGaSe₂ 0.5 1.0 2.0 2.5 1.5 Bandgap (eV) Increasing Voc and decreasing Jsc#13Harris Group Multijunctions: The Road to Higher Efficiencies LELAND STAN JUNGE ORSANMALD Higher-efficiency MJ cells require new materials that divide the solar spectrum equally to provide current match Ge provides lattice match but the bandgap is too small GalnP 1.8 eV GaAs 1.4 eV Ge GalnP 1.8 eV GaAs 1.4 eV GalnNAs -10 7 8 9 1 Energy (eV) ISC2009 Tokyo University, March 3, 2009 2 3 0.7 eV Conventional MJ Cell 1.0 eV New Solar Junction MJ Cell JSH 13#14Solar Junction World Record ONREL NATIONAL RENEWABLE ENERGY LABORATORY PV Cell & Module Performance Group Efficiency vs. Concentration Ratio Solar Junction HIPSS Data Temperature: 25°C April 1, 2011 Solar Junction Wed, Mar 16, 2011 11:28 AM ASTM G173 Direct Device Temperature = 25.0°C Area used 03124 cm² Irradiance: 417.9 kW/m² NREL HIPSS Confidential PV Performance Characterization Team T T T 20 World's most efficient solar cell ever produced: 43.5% at Efficiency (%) 46 44 42 40 38 36 34 HIPSS 32 418-655 suns X25 multi-source 30 10 100 1000 10000 Concentration Ratio Solar Junction Area: 0.3124 cm² Spectrum: ASTM G173 Direct HIPSS PFN trigger settings: 250.3 & 376.0 HIPSS error bars are +5% of value Concentration = HIPSS L / X25 1-sun Isc Current (A) 1.8 1.6 1.4 12 1.0 0.8 0.6 04 0.2 ידידיוידידי 0.0 00 10 05 1.0 1.5 20 25 3.0 35 Voltage (V) Voc = 3.412 V L = 1.869 A 89.17% Fill Factor Efficiency 43.5±22% V₁ = 3.066 V Imax = 1.854 A P = 5.685 W HEX PROPRIETARY AND CONFIDENTIAL. PROPERTY OF SOLAR JUNCTION. 14#152010 Production by Cell Type 6% 2% 5% Source: PV News, May 2011 National Renewable Energy Laboratory Module Sales 87% ■ Silicon ■ CdTe ■CIGS ■Thin-film silicon Innovation for Our Energy Future#16Silicon PV Silicon Feedstock Ingot Growth Slicing Wafers Photovoltaic System Module Encapsulation Cell Fabrication National Renewable Energy Laboratory Innovation for Our Energy Future#17Simple Cell Technologies continue to improve - 19.6% efficient planar cells on CZ silicon Source: J-H Lai, IEEE PVSC, June 2011 National Renewable Energy Laboratory Ag (screen-printed, fired) SIN (PECVD) SiO2 (thin, thermally- grown) n* (P)-Si (ion implantation) p(B)-Si (starting wafer) SiO2 (thin, thermally- grown) SIN (PECVD) X Al-Si Eutectic (screen- printed, alloyed) p* (AI)-Si (LPE during alloying) Innovation for Our Energy Future#18In September 2011 there were protests at a Chinese PV factory over pollution in the river Chemical and Engineering News, September 26, 2011#19Loan Transactions involving Chinese Banks to Chinese Solar Companies since Jan 2010* Company Amount ($M) Banks China Sunergy 160 China Development Bank Dago New Energy 154 Bank of China Hanwa SolarOne 1,000 Hanwa SolarOne 885 JA Solar 4,400 Jinko Solar 7,600 Bank of China Bank of Shanghai China Development Bank Bank of China LDK Solar 8,900 China Development Bank Suntech 7,330 China Development Bank Trina Solar 4,400 Yingli Green Energy 179 China Development Bank China Citic Bank, Bank of China Yingli Green Energy 5,300 China Development Bank Yingli Green Energy 144 Bank of Communications Yingli Green Energy 257 Bank of Communications Total 40,709 Source: Mercom Capital Group, llc All amounts in millions of dollars. *As of Sept. 26, 2011 http://www.mercomcapital.com/news.php#CHINA After this presentation was given, a article with a response from Suntech was published at http://www.greentechmedia.com/articles/read/The-Reality-of-Chinas-Billions-in-Solar-Loans/.#20Conclusions on Silicon PV • Progress has been better than many expected. • • • Modules are being sold at $1/W, but not for profit. $1/W w/ profit seems inevitable. It is not yet clear that $0.5/W Si cells can be made sustainably.#21Thin Film Solar Cells • A thin film of semiconductor is deposited by low cost methods. • Less material is used. • Cells can be flexible and integrated directly into roofing material. Metal P-type CdTe 3~8 um N-type CdS Transparent Conducting Oxide 0.1 um 0.05 um Glass Superstrate ~1000 um#22CdTe Solar Cell with CdS window layer Back Contact: Cathode Absorber layer Window Layer Front Contact: Anode Metal P-type CdTe 3~8 um N-type CdS Transparent Conducting Oxide Glass Superstrate Incident Light CdS: tends to be n-type, large bandgap(2.42eV) 0.1 um 0.05 um ~1000 um 22#23Cadmium Telluride Solar Cells glass SnO2 CdS CdTe ● Direct bandgap, E₁=1.45eV Good efficiency (Record: 17.3%) High module production speed • Long term stability (20 years) • 1 μm CdS/CdTe Image from Rommel Noufi Schematic from Bulent Basol ZnTe:Cu Ti high P P = 10-100 Torr T - 800°C CdTe crystals in perforated ampoule 600°C •slit emits raper substrate translation#24CdTe: Industrial Status First Solar is the leader. It takes them 2.5 hours to make a 11 % module. Manufacturing Capacity 308MW 100MW 25MW 716MW 2,236MW 2,732MW Average Manufacturing Cost 1,502MW 1,228MW 2006: $1.40/watt 2007: $1.23/watt 2008: $1.08/watt 2009: $0.87/watt 2010: $0.77/watt 2010 2011* 2012* 2005 2006 2007 2008 2009 The energy payback time is 0.8 years. www.firstsolar.com 24#25One reason cells on the roof don't have 17.3 % efficiency QE 1.0 10 nm CdS 0.8 30 nm 0.6 35 nm CdS/CdTe solar cell QE (8 Devices) 0.4 95 nm 150 nm. 0.2 0.0 400 240 nm 500 600 700 800 900 λ [nm] The challenge in industry is to implement thin CdS layers without having a pinhole. From Reuben Collins 25 25#26How much of a problem is the toxicity of Cd? It is probably manageable. First Solar will recycle the panels when the customer is done with them. 26 26#27Cu(In Ga₁x)Se₂ • World record efficiency = 20.4 %. • Many companies are evaporating, printing, sputtering and electrodepositing it. • Some are manufacturing ~30-50 MW/yr. • Handling a 4-element compound is tough. 1-x ZnO, ITO - 2500Å CdS - 700Å CIGS 1-2.5μm Mo -0.5-1μm Glass, Metal Foil, Plastics Shell Solar, CA Global Solar Energy, AZ Energy Photovoltaics, NJ ISET, CA ITN/ES, CO NanoSolar Inc., CA DayStar Technologies, NY/CA MiaSole, CA HelioVolt, Tx Solyndra, CA SoloPower, CA Wurth Solar, Germany SULFURCELL, Germany CIS Solartechnik, Germany Solarion, Germany Solibro, Sweden CISEL, France Showa Shell, Japan Honda, Japan#28Solyndra's CIGS modules Direct Sunlight Diffuse Sunlight Reflected Light Outer Tube Metal Cap Hermetic Seal www.solyndra.com#29A comparison of Solyndra's modules to their competitors SOLYNDRA CONVENTIONAL www.solyndra.com#30Wind Performance SOLYNDRA CONVENTIONAL www.solyndra.com#31Air Flow SOLYNDRA Ability to Avoid Heating CONVENTIONAL CONVENTIONAL Please view the videos on their website to see the manufacturing and installation processes. www.solyndra.com#32• • • What went wrong? There are significant disadvantages to using cylinders (e.g. more area, more dark current). Just about everything in the factory had to be custom built to enable the use of cylinders. The glass cylinders are not as cheap as those used in fluorescent light tubes. The price of crystalline silicon dropped faster than Solyndra expected. • Building a second factory wasn't a great idea. Martin Roscheisen explained the disadvantages of the Solyndra approach back in 2009. (http://www.nanosolar.com/company/blog/tubular-pv)#33PRINTED SEMICONDUCTOR Nanoparticle Dispersion Printed Semiconductor + Rapid Thermal Processing (RTP) Precursor Layers PV Film Stack Electrode Nanoparticle Layer Transparent Conductors Thin-film Device www.nanosolar.com