PROGRESS IN PHOTOVOLTAICS: RESEARCH AND APPLICATIONS, VOL 2, 231-234 (1994) Research Short Communication: Solar Cell Eficiency Tables (Version 4) Martin A. Green’ and Keith Emery’ Centre for Photovoltaic Devices and Systems, University of New South Wales, Kensinyton 2033, Australia National Renewable Eneryy Laboratory, 161 7 Cole Boulevard, Golden, CO 80401, USA Updated tables showing the highest independently confirmed eficiencies for solar cells and modules are presented. Guidelines for inclusion of results into these tables are outlined and several new entries since January 1994 are briefly described. INTRODUCTION regular feature of ‘Progress in Photovoltaics’ is the publication of solar cell and module efficiency tables summarizing the highest independently confirmed results for a range of More details of guidelines for the inclusion of results in these tables are described in the corresponding article in the most recent January issue of the journal. Briefly, the main criterion for the inclusion of results in these tables is that they must have been measured independently by at least one of the designated test centres described in the January issue,3 under standard test conditions. Special attention is also paid to the definition of cell area, with three area definitions currently being used: total area, aperture area and designated illumination area. ‘Active area’ efficiency measurements are not included. There are also constraints on minimum allowable cell and module areas, although some flexibility here is desirable (0.05 cm2 for a concentrator cell, 0.25 cm2 for a one-sun cell and 800 cm2 for a module). for a tandem cell, 1 cmZ Table 1-111 show the updated results. Changes in the tables from those previously published are set in bold type. Apart from corrections, there are several new entries. for a 4-cm2 passivated emitter and In Table I, an improved crystalline silicon solar efficiency of 23.5% rear locally diffused (PERL) cell fabricated at the University of New South Wales (UNSW) has been confirmed independently at Sandia National Laboratorie~.~ In the same table, the polycrystalline silicon cell efficiency has been improved slightly to 17.8% for a 1-cm2 cell fabricated at the Georgia Institute of Technology and measured at Sandia National Laboratories. Rapid progress has been made also in the crystalline silicon module area (Table 11). Sandia National Laboratories, in outdoor testing, have measured an efficiency of 21.6X5 for a module fabricated by Honda Research and Development using rear point contact cells supplied by the SunPower Corporation. The module consists of 48 cells connected in series, each cell being 17.7 cm2 in area. The cells and the technology used in fabricating the panel are the same as used by the Honda car in the 1993 World Solar Challenge, described in an earlier issue of this journaL6 The cell technology also has been described re~ently.~ The cells are encapsulated in a laminate with a grooved acrylic top cover laminated by an ethylene-vinyljacetatel(EVA) to the cells and to the Tedlar rear. The module described displayed a very low nominal operating cell temperature of 39°C. In the non-silicon polycrstalline thin-film cell area, improved results have been obtained recently with copper-indium-gallium diselenide (CIGS) cells, with 16.4% confirmed for a 0.4-cm2 cell fabricated at the National Renewable Energy Laboratory (NREL). Unfortunately, the cell area is too small to qualify CCC 1062-7995/94/03023 1-4 0 1994 by John Wiley 8c Sons, Ltd. Received 28 April 1994 Revised 28 April 1994 A 232 M. A. GREEN AND K. EMERY Table I. Confirmed terrestrial cell and submodule efficiencies measured under the global AM 1.5 spectrum (lo00 W m-2) at 25°C Effie.* Areac Classification“ Silicon cells Si (crystalline) 23.5 4.00 (ap) KC (%) (cm2) (V) 0.702 0.628 FFd Test centree (mA cm-*) (%) (and date) J*C Description UNSW PERL UNSW PERL Georgia Tech Sharp (mech. textured) AstroPower (Si-film) Mitsubishi (60 pm on SiOJ Kopin, AlGaAs window ASEC. AlGaAs window Kopin, 5-pm CLEFT Kopin (4 CLEFT cells) Spire, epitaxial South Florida CSVT Solar Cells Inc. NREL, CIGS on glass Siemens (prism cover) 41.2 36.2 81.2 Si (moderate area) Si (multicrystalline) Si (large multicrystalline) Si (supported film) Si (large thin film) GaAs (crystalline cell) GaAs (Ge substrate cell) GaAs (thin-film cell) GaAs (submodule) InP (crystalline cell) Polycrysfalline thin ,film CdTe (cell) CdTe (submodule) ClGS (cell) Cu(In, Ga) (S, Se) (submodule) Amorphous Si a-Si (cell)# a-Si (submodule)@ Multijunction cells GalnP/GaAs GaAIAs/GaAs GaAs/CIS (thin film) a-Si/CIGS (thin film)@ a-Si/a-SiGe@ a-Si/a-Si/a-SiGe@ a-Si/a-SiGe/a-SiCe# a 21.6 17.8 17.2 14.9 14.2 25.1 24.3 23.3 21.0 21.9 15.8 9.8 13.9 12.7 45.7 (t) l.O(ap) 100 (t) 1.02 (ap) 100 0) 0.694 39.4 0.610 36.4 0.600 31.4 0.608 30.0 1.022 1.035 1.011 4.04 0.878 28.2 27.6 27.6 6.6 29.3 78.1 78.5 77.7 79.2 78.1 87.1 85.3 83.8 80 85.4 74.5 69 72.2 68.0 Sandia (3/94) Sandia (4/93) Sandia (3/94) JQA (3/93) Sandial (12/88) JQA (3/93) NREL (3/90) NREL (3/89) NREL (4/90) NREL (4/90) NREL (4/90) NREL (6/92) NREL (5/93) NREL (8/93) NREL (4/94) III-v 3.91 (t) 4.00 (t) 4.00(ap) 16 0) 4.02 (t) 1.05 (ap) 63.6 (ap) 6.636 (t) 69.1 (ap) 0.843 25.1 2.2 6.62 0.644 29.9 7.49 2.49 12.7 12.0 29.5 27.6 25.8 14.6 12.5 12.4 12.4 1.0 (da) 100(ap) 0.25 (t) 0.50 (t) 4.00(t) 0.887 19.4 1.3 12.5 2.385 14.0 2.403 14.0 - - - - 74.1 JQA (4/92) 73.5 JQA (12/92) 88.5 NREL (6/93) 83.4 NREL (3/89) - NREL (1 1/89) - NREL(6/88) 65.8 NREL (12/92) 70.0 NREL (2/88) 68.5 JQA (12/92) Sanyo Sanyo NREL (monolithic) Varian (monolithic) Kopin/Boeing (Cterminal) ARC0 (Cterminal) USSC/Cannon (monolithic) ECD (monolithic) Sharp (monolithic) 2.40(ap) 0.26(ap) 0.27 (t) I.O(da) 1.621 2.541 2.289 11.7 7.0 7.9 JQA Japan Quality Assurance. ’ Measurements corrected from originally measured values due to Sandia recalibration in January 1991. Unstabilized results. = ’ Effic. = efficiency. ‘ (ap) = aperture area; (t) = total area; (da) = designated illumination area. ‘ FF = RII factor. ClGS = CuInGaSe,: a-Si = amorphous silicon/hydrogen alloy; CIS = CulnSe,. for inclusion in these tables. ‘Active area’ efficiencies of up to 17.5% also have been reported recently for this technology for small-area cells but apparently they have not been confirmed independently, making these results ineligible for inclusion on several count^.^ In fairness to colleagues and to maintain an appropriate level of professionalism, researchers are encouraged to seek independent confirmation before reporting such ‘record-breaking’ results. An additional result in this area that has been confirmed submodule fabricated by independently is a 12.7% efficiency for a 69-cmZ CuIn,Ga, -,S,Se, -, Siemens Solar and measured at NREL, increasing from 12.4% after the application of a prismatic cover (Table I). An improved efficiency of 7.8% also has been measured at NREL for a large-area (0.7 m2) CdTe module fabricated by Solar Cells, Inc. (Table 11). In the amorphous silicon area, a milestone of 10% ‘stabilized’ module efficiency has now been exceeded. Reported in Table I1 are the results measured at NREL upon a module fabricated and stabilized at the United Solar Systems Corporation (USSC), the module consisting of a three-cell tandem stack. SHORT COMMUNICATION 233 Table 11. Confirmed terrestrial module efficiencies measured under the global AM1.5 spectrum (lo00 W m-2) at a cell temperature of 25°C Classification\" Si (crystalline) CIGS CIGS (large) CdTe' CdTe (large) a-Si/a-SiGe/a-SiGe (tandem)-' Effi~.~ Area' (%I (cm? 21.6 11.1 9.7 8.1 7.8 10.2 KC 4c (V) 32.6 25.9 37.8 21.0 92.0 232 (A) 0.703 0.637 2.44 0.573 0.969 6.47 (%I FFd Test centre (and date) Sandia (2/94) NREL (6/88) NREL (5/91) NREL (9/91) NREL (10/93) NREL (12/93) Description Honda/SunPower, 48 cells ARCO, 55 cells Siemens Solar Photon Energy Solar Cells, Inc. USSC 862(ap) 938 (ap) 3883 (ap) 838 (ap) 6838(ap) 903 (ad 813 64 64 55 60 61.2 * CIGS = CulnGaSe,; a-Si = amorphous silicon/hydrogen alloy. ' ' Effic. = efficiency. '(ap) = aperture area. FF = fill factor. ' Output varies with premeasurement conditions and bias rate. Efficiency taken with maximum power point tracking. Stabilized results. Table 111. Terrestrial concentrator cell and module efficiencies measured under the direct beam AM 1.5 spectrum at a cell temperature of 25°C. Classification Single cells GaAs Si Si (moderate area) Si (large) Multijunction cells GaAs/GaSb InP/GaInAs GaAs/GaInAsP GalnP/GaAs GaAs/Si Submodules GaAs/GaSb Modules Si a PA) 27.5 26.5 25.6 21.6 32.6 31.8 30.2 30.2 29.6 Effic.\" Areab (cm2) 0.126 (da) 0.150 (da) 1.21 (da) 20.0 (da) 0.053 (da) 0.063 (da) 0.053 (da) 0.103 (da) 0.317 (da) Concentration' (suns) 205 149 93 11 Test centre (and date) Sandiad (4/88) Sandiad (5/87) Sandia ( /93) Sandiad ( /90) Sandiad (10/89) NREL (8/90) NREL (10/90) Sandia (3/94) Sandiad (9/88) Description Varian (prism cover) Stanford point contact SunPower point contact UNSW laser-grooved Boeing, mechanical stack NREL, monolithic 3-terminal NREL, stacked 4-terminal NREL, monolithic 2-terminal Varian/Stanford/Sandia, mech. stack 100 50 40 180 350 25.1 20.3 41.4 (ap) 1875 (ap) 57 80 Sandia (3/93) Sandia (4/89) Boeing, 3 mech. stack units Sandia/UNSW/ENTECH (1 2 cells) Effic. = efficiency. (da) = designated illumination area; (ap) = aperture area. m-'. 'One sun corresponds to an intensity of loo0 W Measurements corrected from originally measured values due to Sandia recalibration in January 1991. Other modules supplied by USSC had an initial, unstabilized efficiency of 11.7%, which is also an improvement on previous results. This provides the opportunity to change the procedure for reporting amorphous silicon module results. In the future, amorphous silicon-based module results will be included only if they have been measured after stabilization. Finally, in the tandem cell area another milestone has been surpassed with the demonstration of an efficiency above 30% for a 0.103-cm' two-terminal monolithic GaInP/GaAs concentrator cell fabricated at NREL (Table 111). Also, a submodule fabricated by Boeing consisting of three units, each 234 M. A. GREEN AND K. EMERY incorporating a GaAs/GaSb mechanically stacked tandem cell, has been measured at Sandia to demonstrate an efficiency of 25.1%. While the information contained in the tables is provided in good faith, the authors, editors and publishers cannot accept direct responsibility for any errors or omissions. Acknowledgements One of the authors (M. A. G.) would like to thank David King of Sandia National Laboratories and F. Nagamine of the Japan Quality Assurance Organization for their contributions to this article. He would also like to thank Dr R. Swanson for providing a copy of Sandia’s test report for the 21.6% silicon module. REFERENCES 1. M. A. Green and K. Emery, ‘Solar cell efficiency tables’, Progr. Photovolt., 1, 25-29 (1993). 2. M. A. Green and K. Emery, ‘Solar cell efficiency tables (version 2)’, Progr. Photovolt., 1, 225-227 (1993). 3. M. A. Green and K. Emery, ‘Solar cell efficiency tables (version 3)’, Progr. Photovolt., 2, 27-34 (1994). 4. J. Zhao, A. Wang and M. A. Green, ‘23.5% Efficient silicon solar cell’, Progr. Photovolt., 2, 227-230 (1994). 5. P. J. Verlinden, R. M. Swanson and R. A. Crane, ‘21.6% Efficient silicon solar cell module’, paper presented at 12th EC Photovoltaic Solar Energy Conference, Amsterdam, April 1994. 6. M. A. Green, ‘World Solar Challenge 1993: the trans-Australian solar car race’, Progr. Photovolt., 2,73-79 (1994). 7. P. J. Verlinden, R. M. Swanson and R. A. Crane, ‘7000 High-efficiency cells for a dream’, Progr. Photovolt., 2, 143-152 (1994).