Volume 1 photovoltaic solar energy 1 22 – multiple junction solar cells

Volume 1 photovoltaic solar energy 1 22 – multiple junction solar cells Volume 1 photovoltaic solar energy 1 22 – multiple junction solar cells Volume 1 photovoltaic solar energy 1 22 – multiple junction solar cells Volume 1 photovoltaic solar energy 1 22 – multiple junction solar cells Volume 1 photovoltaic solar energy 1 22 – multiple junction solar cells

1.22 Multiple Junction Solar Cells M Yamaguchi, Toyota Technological Institute, Tempaku, Nagoya, Japan © 2012 Elsevier Ltd All rights reserved 1.22.1 1.22.2 1.22.2.1 1.22.2.2 1.22.2.3 1.22.2.4 1.22.3 1.22.3.1 1.22.3.1.1 1.22.3.1.2 1.22.3.1.3 1.22.3.2 1.22.3.3 1.22.4 Introduction Key Issues for Realizing High-Efficiency MJ Solar Cells Selection of Cell Materials and Improving Quality Lattice Matching Between Cell Materials and Substrates Effectiveness of Wide Bandgap Back-Surface-Field Layer Low-Loss Tunnel Junction for Intercell Connection and Preventing Impurity Diffusion from Tunnel Junction High-Efficiency InGaP/GaAs/Ge 3-Junction Solar Cells and their Space Applications Development of High-Efficiency InGaP/GaAs/Ge 3-Junction Solar Cells Wide bandgap tunnel junction Heteroface structure Ge bottom cell Precise lattice matching to Ge substrate Radiation Resistance of InGaP-Based MJ Solar Cells Space Applications of InGaP/GaAs/Ge 3-Junction Solar Cells Low-Cost Potential of Concentrator MJ Solar Cell Modules and High-Efficiency Concentrator InGaP/GaAs/Ge 3-Junction Solar Cell Modules and Terrestrial Applications 1.22.5 Most Recent Results of MJ Cells 1.22.6 Future Directions Acknowledgments References 497 498 498 500 501 502 503 503 504 504 504 505 506 506 509 509 513 514 1.22.1 Introduction Multi-junction (MJ, Tandem) solar cells are composed of multilayers with different bandgap energies; an example is shown in Figure They have the potential for achieving high conversion efficiencies of over 50% and are promising for space and terrestrial applications due to their wide photo response Figure shows theoretical conversion efficiencies of single-junction and MJ solar cells in comparison with experimentally realized efficiencies Tandem solar cells were proposed in 1955 by Jackson [1] and in 1960 by Wolf [2] Table shows progress of the III–V compound MJ solar cell technologies The MIT group [3] encouraged R&D of tandem cells based on their computer analysis Although AlGaAs/GaAs tandem cells, including tunnel junctions and metal interconnectors, were developed in the early years, a high efficiency close to 20% was not obtained [4] This is because of difficulties in making high-performance and stable tunnel junctions, and the defects related to the oxygen in the AlGaAs materials [5] A double-hetero (DH) structure tunnel junction was found to be useful for preventing diffusion from the tunnel junction and improving the tunnel junction performance by the authors [6] The authors demonstrated 20.2% efficiency AlGaAs/GaAs 2-junction cells [7] An InGaP material for the top cell was proposed by the NREL group [8] As a result of performance improvements in tunnel junction and top cell, over 30% efficiency has been obtained with InGaP/GaAs 2-junction cells by the authors [9] InGaP/GaAs-based MJ solar cells have drawn increased attention for space applications because of the superior radiation resistance of InGaP top cells and materials, as has been discovered by the authors [10]; in addition, they also have the possibility of reaching high conversion efficiency of over 30% In fact, the commercial satellite (HS 601HP) with 2-junction GaInP/GaAs-on-Ge solar cell arrays was launched in 1997 [11] More recently, InGaP/GaAs-based MJ solar cells have drawn increased attention for terrestrial applications because concentrator operation of MJ cells has great potential of providing high-performance and low-cost solar cell modules For concentrator applications, the cell contact grid structure has been designed in order to reduce the energy loss due to series resistance, and 38.9% (AM1.5G, 489 suns) efficiency has been demonstrated by Sharp [12] Most recently, 41.6% efficiency has been reported with InGaP/GaAs/Ge 3-junction concentrator cells by Spectrolab [13] In addition, the authors have realized high-efficiency and large-area (5445 cm2) concentrator InGaP/InGaAs/Ge 3-junction solar cell modules of an outdoor efficiency of 31.5% [14] as a result of developing high-efficiency InGaP/InGaAs/Ge 3-junction cells, low-optical-loss Fresnel lens and homogenizers, and designing low-thermal-conductivity modules Some companies including Sharp [15] have announced to commercialize InGaP/ GaAs/Ge 3-junction concentrator cell modules for terrestrial use Liquid-phase epitaxy (LPE) was used to fabricate GaAs single-junction solar cells in the 1960s and 1970s because it produces high-quality epitaxial film and has a simple growth system However, it is not as useful for devices that involve multilayers because of difficulty of control over layer thickness, doping, composition, and speed of throughput Since 1977, metal-organic chemical vapor deposition (MOCVD) has been used to fabricate large-area, GaAs single-junction solar cells and recent MJ solar cells because it is capable of large-scale, large-area production and has good reproducibility and controllability Molecular beam epitaxy (MBE) and Comprehensive Renewable Energy, Volume doi:10.1016/B978-0-08-087872-0.00126-8 497 498 Technology Light InGaP cell Eg = 1.85 eV GaAs cell 1.4 eV Ge cell 0.7 eV Figure A schematic structure of a multilayer solar cell, illustrating that high-energy light is absorbed by the top cell (InGaP), medium energy by the middle cell (GaAs), and low energy by the bottom cell (Ge) Conversion efficiency (%) 60 50 40 30 20 Theory (Conc.) Theory (1-sun) Realized (Conc.) Realized (1-sun) Number of junction Figure Theoretical conversion efficiencies of single-junction and multi-junction solar cells in comparison with experimentally realized efficiencies for 1-sun intensity and under concentration chemical beam epitaxy (CBE) have been used for laboratory stage’s research and development of MJ solar cells because they provide excellent of monolayer abruptness and thickness due to the nature of beam Especially, CBEs [16] are advantages of realizing new materials such as III–V-N materials and solar cells and novel multilayer structures such as quantum nanostructure solar cells 1.22.2 Key Issues for Realizing High-Efficiency MJ Solar Cells The key issues for realizing high-efficiency MJ tandem cells are discussed based on results obtained in our laboratory 1.22.2.1 Selection of Cell Materials and Improving Quality MJ cells with different bandgaps are stacked in tandem (see Figure 1) so that the cells cover a wide wavelength region from 300 to 1800 nm Cell materials are selected by considering bandgap energies close to the optimal bandgap energy combination based on theoretical efficiency calculation, by considering lattice matching to substrates, and by less impurity problems such as high-concentration residual impurities and impurities acting as recombination centers Figure shows minority-carrier diffusion length dependence of GaAs single-junction solar cell efficiency It is clear that a high minority-carrier diffusion length, L (minority-carrier lifetime, τ = L2/D, where D is the minority-carrier diffusion coefficient) is substantially necessary to realize high-efficiency solar cells Figure shows carrier concentration dependence of minority-carrier lifetime in p-type and n-type GaAs [17] Minority-carrier lifetime τ depends on carrier concentration N of solar cell layers as expressed by Multiple Junction Solar Cells Table Progress of the III–V compound multi-junction solar cell technologies 1955 Jackson 1960 Proposal of multi-junction solar cell Wolf 1982 1982 1987 1987 1989 Efficiency calculation of tandem cells 15.1% AlGaAs/GaAs 2-junction (2-J) cell Proposal of double-hetero structure tunnel junction for multi-junction interconnection 20.2% AlGaAs/GaAs 2-J cell 32.6% GaAs//GaSb concentrator 2-J cell (mechanical-stacked, 100 suns concentration) Proposal of InGaP as top a cell material 27.3% InGaP/GaAs 2-J cell 30.3% InGaP/GaAs 2-J cell Discovery of radiation resistance of InGaP top cell 33.3% InGaP/GaAs//InGaAs 3-J cell (mechanical-stacked) MIT RTI NTT NTT Boeing 1990 1990 1996 1997 1997 1997 2000 2003 2004 2006 Commercial satellite with 2-J cells 31.7% InGaP/InGaAs/Ge 3-J cell 37.4% InGaP/InGaAs/Ge 3-J cell (200 suns concentration) 38.9% InGaP/InGaAs/Ge 3-J cell (489 suns concentration) 31.5% large-area (5445 cm2) InGaP/InGaAs/Ge 3-J cell module (outdoor) 2006 2009 2009 2009 40.7% InGaP/GaAs/Ge 3-J cell (236 suns concentration) 41.1% InGaP/InGaAs/Ge 3-J cell (454 suns concentration) 41.6% InGaP/InGaAs/Ge 3-J cell (364 suns concentration) 35.8% InGaP/GaAs/InGaAs 3-J cell (1 sun) NREL NREL Jpn Energy Toyota Tech Inst Jpn Energy, Sumitomo & Toyota Tech Inst Hughes Jpn Energy Sharp Sharp & Toyota T.I Daido Steel, Daido Metal, Sharp &Toyota T.I Spectrolab Fraunhofer ISE Spectrolab Sharp 30 Efficiency (%) 25 20 Junction depth 0.3 μm 15 10 0.1 10 Minority-carrier diffusion length (μm) 100 Figure Minority-carrier diffusion length dependence of GaAs single-junction solar cell efficiency 104 p-GaAs n-GaAs τ (ns) 103 102 101 100 1015 1016 1017 1018 1019 N (cm–3) Figure Carrier concentration dependence of minority-carrier lifetime in p-type and n-type GaAs [17] 499 500 Technology 1000 S = (cm s–1) 5800 Relative PL intensity (Arb Unit) Growth temp (°C) 100 104 700 650 105 750 10 106 Without buffer layer × 106 Without AlInP Window 0.1 0.01 0.01 0.1 10 Minority-carrier lifetime τ (ns) 100 Figure Changes in photoluminescence (PL) intensity of the solar cell active layer as a function of the minority-carrier lifetime in GaAs, grown by MOCVD and surface recombination velocity (S) τ ¼ BN ½1Š where B is the radiative recombination coefficient Therefore, carrier concentration of cell layers must be optimized by considering minority-carrier lifetime, build-in potential and series resistance of p-n junction diodes Selection of cell materials, especially selection of top cell materials, is also important for high-efficiency tandem cells It has been found by the authors [5] that an oxygen-related defect in the AlGaAs top cell materials acts as a recombination centre As a top cell material lattice matched to GaAs or Ge substrates, InGaP has some advantages [8] such as lower interface recombination velocity, a lower oxygen-related defect, and a good window layer material compared to AlGaAs The top cell characteristics depend on the minority-carrier lifetime in the top cell layers Figure shows changes in photoluminescence (PL) intensity of the solar cell active layer as a function of the minority-carrier lifetime τ of the p-InGaP base layer grown by MOCVD and surface recombination velocity (S) The lowest S was obtained by introducing an AlInP window layer and the highest τ was obtained by introducing a buffer layer and optimizing the growth temperature The best conversion efficiency of the InGaP single-junction cell was 18.5% [18] 1.22.2.2 Lattice Matching Between Cell Materials and Substrates Lattice mismatching of cell materials to substrates should be decreased because miss-fit dislocations are generated in the upper cell layers due to lattice mismatch, which deteriorates cell efficiency Figure shows calculated and experimental dislocation density 1000 GaAs/Si Minority-carrier lifetime τ (ns) 1016 cm–3 16 –3 100 × 10 cm InP/Si NTT model InGaP/Si GaAs/GaAs 10 × 1017 cm–3 S.A Ringel et al InP/InP 1018 cm–3 NTT data 0.1 104 105 106 107 108 –2 Dislocation density Nd (cm ) Figure Calculated and experimental dislocation density dependence of minority-carrier lifetime in GaAs, InP, and InGaP NTT model (eqn [2]), NTT data and Ringel’s data refer to references 19–21, respectively Multiple Junction Solar Cells 501 dependence of minority-carrier lifetime in GaAs [19] Dislocation density Nd dependence of minority-carrier lifetime τ is expressed by the following equation [19]: 1 DNd ẳ ỵ þ τ τ r τ ½2Š where τr is radiative recombination lifetime, τ0 is minority-carrier lifetime associated with recombination at other unknown defects, and D is the minority-carrier diffusion coefficient Application of InGaAs middle cell [22] lattice matched to Ge substrates has been demonstrated to increase open-circuit voltage (Voc) due to lattice matching and short-circuit current density (Jsc) due to decrease in bandgap energy of the middle cell 1.22.2.3 Effectiveness of Wide Bandgap Back-Surface-Field Layer Figure shows the surface recombination effect on short-circuit current density Jsc of In0.14Ga0.86As homo-junction solar cells as a function of junction depth; an optimum (29 mA cm−2) exists for surface recombination velocity of 104 cm s−1 and a junction depth of ∼0.5 μm Therefore, in order to improve efficiency drop attributed from front and rear surface, recombination formation of heteroface or DH structure is necessary Figure shows changes in Voc and Jsc of InGaP single-junction cells as a function of potential barrier ΔE A wide bandgap back-surface-field (BSF) layer [22] is found to be most effective for confinement of minority carriers compared to highly doped BSF layers Short-circuit current density (mA cm−2) 35 S = 1E + cm/s S =1E + cm/s 30 S = 1E + cm/s S = 1E + cm/s 25 20 15 10 0 0.5 1.5 2.5 Junction depth (μm) Figure Surface recombination effect on short-circuit current density Jsc of In0.14Ga0.86As homo-junction solar cells as a function of junction depth Recombination velocity at rear surface (cm s–1) 106 11 104 102 100 Jsc (mA) p+ InGaP layer 10 AllnP Voc (V) 1.5 1.4 1.3 1.2 100 200 ΔE (meV) 300 Figure Changes in Voc and Jsc of InGaP single-junction cells as a function of potential barrier ΔE 502 1.22.2.4 Technology Low-Loss Tunnel Junction for Intercell Connection and Preventing Impurity Diffusion from Tunnel Junction 10000 1000 15 100 10 10 0.1 0.01 0.001 JSC Of bottom cell (mA cm–2) Tunnel peak current density (A cm–2) Another important issue for realizing high-efficiency monolithic-cascade-type tandem cells is the achievement of optically and electrically low-loss interconnection of two or more cells A degenerately doped tunnel junction is attractive because it involves only one extra step in the growth process To minimize optical absorption and achieve higher Jsc for bottom cell as shown in Figure 9, formation of thin and wide bandgap tunnel junctions is necessary However, the formation of a wide bandgap tunnel junction is very difficult because the tunneling current decreases exponentially with increase in bandgap energy In addition, impurity diffusion from a highly doped tunnel junction during overgrowth of the top cell increases the resistivity of the tunnel junction As shown in Figure 10, a DH structure was found to be useful for preventing diffusion by the authors [6] An InGaP tunnel junction has been for the first time tried for an InGaP/GaAs tandem cell in our work [9] As p-type and n-type dopants Zn and Si were used, respectively The peak tunneling current of the InGaP tunnel junction is found to increase from mA cm−2 up to A cm−2 by making a DH structure with AlInP barriers Effective suppression of the Zn diffusion from the tunnel junction by the InGaP tunnel junction with the AlInP DH structure is thought to be attributed to the lower diffusion coefficient [23] for Zn in the wider bandgap energy materials such as the AlInP barrier layer and InGaP tunnel junction layer Therefore, the InGaP tunnel junction has been observed to be very effective for obtaining high tunneling current, and the DH structure has also been confirmed to be useful for preventing diffusion Table summarizes key issues of realizing super high-efficiency MJ solar cells 0.0001 0.5 1.5 Bandgap energy (eV) Figure Calculated tunnel peak current density and short-circuit current density Jsc of GaAs bottom cell as a function of bandgap energy of tunnel junction Tunnel peak current density (A cm–2) 10 X = 0.9 X=0 10–1 X = 0.6 Top cell PAI×Ga1–×AS ++ p GaAS ++ n GaAS nAI×Ga1–×AS nGaAS sub Bottom cell 10–2 10–3 10–4 500 600 700 800 Annealing temperature (°C) Figure 10 Annealing temperature dependence of tunnel peak current densities for double-hetero structure tunnel diodes X is the Al mole fraction in AlxGa1−x As barrier layers Multiple Junction Solar Cells Table 503 Key issues for realizing super high-efficiency multi-junction solar cells Key issue Past Present Future Top cell materials Third cell materials Substrate Tunnel junction AlGaAs InGaP AlInGaP None Ge InGaAsN, etc GaAs DH-structure GaAs tunnel junction (TJ) GaAs middle cell Ge DH-structure InGaP TJ Si DH-structure InGaP or GaAs TJ InGaAs middle cell InGaP-BSF AlInP-BSF None None None Inverted Epi (Epi from top cell to bottom cell layers) (In)GaAs middle cell Widegap-BSF; Quantum Dots (QDs) Bragg reflector, QDs, etc Inverted Epi.; Epitaxial lift off (peeled thin film off technique) Lattice matching Carrier confinement Photon confinement Others 1.22.3 High-Efficiency InGaP/GaAs/Ge 3-Junction Solar Cells and their Space Applications 1.22.3.1 Development of High-Efficiency InGaP/GaAs/Ge 3-Junction Solar Cells As one of the projects in the Sunshine Program in Japan, an R&D project for super high-efficiency MJ solar cells was started in 1990 Conversion efficiency of InGaP/GaAs-based MJ solar cells has been improved by the following technologies: Selection and high-quality growth of InGaP as a top cell material Proposal of DH structure and wide bandgap tunnel junction for cell interconnection Precise lattice matching of InGaP top cell and InGaAs middle cell with Ge substrate Proposal of AlInP as a BSF layer for the InGaP top cell Proposal of InGaP-Ge heteroface structure bottom cell As a result of the above proposals and performance improvements, we have demonstrated a world-record efficiency (33.3% at 1-sun AM1.5G) of InGaP/GaAs/InGaAs 3-junction solar cells in 1997 [24] The conversion efficiency of InGaP/(In)GaAs/Ge 3-junction solar cells has been improved to 31.7% (AM1.5G) [25] A schematic illustration of the InGaP/(In)GaAs/Ge 3-junction solar cell and key technologies for improving conversion efficiency are shown in Figure 11 Approaches for high-efficiency, triple-junction cells (1) Wide-gap tunnel junction with double-hetero structure High transmittance High potential barrier (2) Combination of Ge cell with InGaP 1st hetero-layer Shallow junction (3) Precise lattice matching by adding 1%-indium No misfit dislocations (4) Widening top cellband-gap (developing 1.96 eV AllnGaP) Increase of Voc rode lect act e g t A ting aAs coa n-G AR P InGa ll n-AlInP window e c p to n-InGaP emitter V) 86 e ( p-InGaP base on ncti p-AlInP BSF el ju n n Tu p-AlGaAs/n-InGaP aAs n-InGaP window (In)G cell e l d n-InGaAs emitter mid eV) (1.4 p-InGaAs base n ctio p-InGaP BSF l jun e n Tun p-GaAs/n-GaAs yer er la Buff n-InGaAs buffer InGaP 1st hetero-layer Ge cell om n-Ge Bott 65 eV) (0 p-Ge ode lectr Ag e Figure 11 Schematic illustration of a 3-junction cell and approaches for improving the efficiency of the cell 504 Technology 1.22.3.1.1 Wide bandgap tunnel junction A wide bandgap tunnel junction, which consists of a DH structure p-Al(Ga)InP/p-AlGaAs/n-(Al)InGaP/n-Al(Ga)InP, increases the amount of incident light into the (In)GaAs middle cell and produces effective potential barriers for both minority carriers generated in the top and middle cells Both Voc and Isc of the cells are improved by the wide bandgap tunnel junction without absorption and recombination losses [9] It is difficult to obtain a high-tunneling peak current with wide gap tunnel junction, so thinning the depletion layer width by formation of a highly doped junction is quite necessary Since impurity diffusion occurs during growth of the top cell [6], both carbon and silicon having a low-diffusion coefficient are used for p-type AlGaAs and n-type (Al)InGaP, respectively Furthermore, the DH structure supposes to suppress impurity diffusion from the highly doped tunnel junction [23] The second tunnel junction between the middle and bottom cells consists of p-InGaP/p-(In)GaAs/n-(In)GaAs/n-InGaP, which has a wider bandgap than the middle-cell materials 1.22.3.1.2 Heteroface structure Ge bottom cell InGaP/GaAs cell layers are grown on a p-type Ge substrate A p–n junction is formed automatically during MOCVD growth by diffusion of the V-group atom from the first layer grown on the Ge substrate So, the material of the first hetero layer is important for the performance of the Ge bottom cell An InGaP layer is thought to be suitable as material for the first hetero layer, because phosphor has a lower-diffusion coefficient in Ge than arsenic and indium has a lower solubility in Ge than gallium Figure 12 shows the change in spectral response of the 3-junction cell by changing the first hetero growth layer on Ge from GaAs to InGaP The quantum efficiency of the Ge bottom cell was improved by using the InGaP hetero-growth layer In the case of the GaAs hetero-growth layer, the junction depth was measured to be around μm On the other hand, the thickness of the n-type layer produced by phosphor from the InGaP layer was 0.1 μm An increase in Ge quantum efficiency was confirmed to be due to a reduction in junction depth 1.22.3.1.3 Precise lattice matching to Ge substrate Although the 0.08% lattice mismatch between GaAs and Ge is thought to be negligibly small, misfit dislocations are found to be generated in thick GaAs layers, which deteriorate cell performance By adding about 1% indium into the InGaP/GaAs cell layers, all cell layers are lattice matched precisely to the Ge substrate As a result, crosshatch pattern caused by misfit dislocations due to lattice mismatch disappeared in the surface morphology of the cell with 1% indium, as shown in Figure 13 The misfit dislocations were found to influence the Voc of the cell and not the Isc Voc was improved by eliminating misfit dislocations for the cell with 1% indium In addition, wavelength of the absorption edge became longer and Isc of both the top and middle cells increased by adding 1% indium 0.7 Quantum efficiency 0.6 InGaP 0.5 0.4 GaAs 0.3 0.2 0.1 0.0 300 500 700 900 1100 Wavelength (nm) 1300 1500 1700 Figure 12 Change in the spectral response due to modification of the 1st hetero layer from GaAs to InGaP (without antireflection coating) In: 0% In: 1% In: 2% 100 μm Figure 13 Surface morphology of InGaAs with various indium compositions grown on Ge Multiple Junction Solar Cells 1.22.3.2 505 Radiation Resistance of InGaP-Based MJ Solar Cells Since radiation in space is severe, particularly in the Van Allen radiation belt, lattice defects are induced in semiconductors due to high-energy electron and proton irradiations, and the defects cause a decrease in the output power of solar cells Figure 14 shows effectiveness of radiation resistance and high conversion efficiency of space cells from the point view of power density (W m−2) for space missions Further improvements in conversion efficiency and radiation resistance of space cells are necessary for widespread applications of space missions Recently, InGaP/GaAs-based MJ solar cells have drawn increased attention because of the possibility of high conversion efficiency of over 40% and radiation resistance An AM0 efficiency of 29.2% has been demonstrated for an InGaP/InGaAs/Ge 3-junction cell (4 cm2) [12] Figure 15 shows the maximum power recovery due to light illumination of 100 mW cm−2 at various temperatures for 1-MeV electron-irradiated InGaP/GaAs tandem cells [10] The ratios of maximum power after injection, PI, to maximum power before irradiation, P0, are shown as a function of injection time Even at room temperature, photo injection-enhanced annealing of radiation damage to InGaP/GaAs tandem cells was observed The recovery ratio increases with an increase in ambient temperature within the operating range for space use Such a recovery is found to be attributed from damage recovery in the InGaP top cell layer [10] Therefore, the results show that InGaP/GaAs tandem cells under device operation conditions have superior radiation-resistant properties Figure 16 shows deep level transient spectroscopy (DLTS) spectra of Trap H2 (Ev + 0.55 eV) for various injection times at 25 °C with an AM1.5 light intensity of 100 mA cm−2 It is also found [26] by DLTS measurements that a major defect level H2 (Ev + 0.55 eV) recovers by forward bias or light illumination, that is, the signal decreases with prolonged light exposure Moreover, the H2 center is confirmed to act as a recombination center by using the double carrier pulse DLTS method The enhancement of defect annealing in InGaP top cell layer under minority-carrier injection conditions is thought to occur as a result of the nonradiative electron-hole recombination process [27] whose energy ER enhances the defect motion The thermal activation energy EA (1.1 eV) of the defect is reduced to EI (0.48–0.54 eV) by an amount ER (0.56–0.62 eV) Thus, electronic energy from a recombination event can be channeled into the lattice vibration mode that drives the defect motion: EI = EA − ER Power density (W m−2) Degradation rate 0% Degradation rate 40% Degradation rate 20% 600 500 400 300 200 100 10 15 20 25 30 BOL efficiency (%) 35 40 Figure 14 Effectiveness of radiation resistance and high conversion efficiency of space cells from the point view of power density (BOL = beginning of life) 0.8 MeV-1015 cm−2 Photo-injection: 100 mW cm−2 Power ratio (Pl/P0) 0.78 75 °C 0.76 25 °C 0.74 50 °C 0.72 0.7 10 20 30 40 Injection time (min) 50 Figure 15 The maximum power recovery of the InGaP/GaAs tandem cell due to light illumination at various temperatures 506 Technology Emission rate = 1005 s−1 p-InGaP DLTS signal (a.u.) 0.5 min 10 20 200 250 300 Temperature (K) 350 Figure 16 DLTS spectrum of Trap H2 (Ev + 0.55 eV) for various injection times at 25 °C with an injection density of 100 mA cm−2 1.22.3.3 Space Applications of InGaP/GaAs/Ge 3-Junction Solar Cells Advanced technologies for high-efficiency cells and discovery of superior radiation resistance of InGaP-based materials are thought to contribute to industrialization of InGaP-based MJ space solar cells in Japan Figure 17 shows the Sharp space solar cell conversion efficiency heritage Since 2002, InGaP/GaAs/Ge 3-junction solar cells have been commercialized for space use in Japan 1.22.4 Low-Cost Potential of Concentrator MJ Solar Cell Modules and High-Efficiency Concentrator InGaP/GaAs/Ge 3-Junction Solar Cell Modules and Terrestrial Applications Concentrator operation is very effective for cost reduction of solar cell modules and thus also of photovoltaic (PV) systems Figure 18 shows a configuration of a concentrator PV (CPV) system composed of a solar cell, optics, and tracker Concentrator Sharp Solar Cell Conversion Efficiency Heritage (Space use) Efficiency (%) MJG-2 30 GaAs −3J 26 MJG-1 BOL (29%) 27% EOL (25%) 24 23.5% Silicon NRS/BSF AHES-1 AHES-2 20 IBF BSFR (2) 16 12 BSF (1) 10.6% (EOL) 1970 1975 IBF IBF 17% (BOL) BSF (2) BSFR (1) CONV BSFR (3) 1980 1985 1990 1995 Year Figure 17 Sharp space solar cell conversion efficiency heritage 12.2% (EOL) 2000 13.1% (EOL) 2002 13.5% (EOL) 2003 Multiple Junction Solar Cells 507 Optics Cell Tracker M Figure 18 Configuration of PV system composed of solar cell, optics and tracker Cost ($ W–1) 100 10 Cell cost Module cost 0.1 10 100 Concentration ratio 1000 Figure 19 Summary of estimated cost for the concentrator PV systems vs concentration ratio [28] operation of the MJ cells is essential for economic viability of their terrestrial applications Since CPV systems using MJ solar cells have great potential of cost reduction as shown in Figure 19 [28], R&D on concentrator technologies including MJ cells was started in Japan In order to apply a high-efficiency MJ cell developed for 1-sun condition to a concentrator cell operating under ∼500-sun condition, reduction in energy loss due to series resistance is the most important issue Cell size was designed to be  mm considering the total current flow Grid electrode pitching, height, and width were designed in order to reduce series resistance Figure 20 shows the fill factor (FF) of the cell with various grid pitchings under 250 suns The grid electrode with μm height and μm width was made of Ag Grid pitching influences lateral resistance between two grids (RL) and total electrodes resistance (RE) Series resistance of the cell (Rs), RE, and RL are also shown in Figure 20 RE was measured directly after removing electrode from the cell by chemical etching RL was calculated by using the sheet resistance of window and emitter layers Based on the data in Figure 20, the best grid pitching is determined to be 0.12 mm at this time In order to reduce series resistance down to 0.01 Ω and obtain high FF under 500 suns, grid height should be increased twofold High efficiency under < 500 suns can be obtained by an optimal grid design without modification of the cell layer For concentrator applications, the grid structure has been designed in order to reduce the energy loss due to series resistance as was shown in Figure 20 Most recently, we have successfully fabricated high-efficiency concentrator InGaP/InGaAs/Ge 3-junction solar cells designed for 500-sun application The efficiencies by in-house measurements are 39.2 % at 200 suns and 38.9% at 489 suns as shown in Figure 21 [12] The solar simulator was equipped with both a Xe lamp and a halogen lamp and was adjusted for AM1.5G spectrum Concentrator InGaP/GaAs/Ge 3-junction solar cell modules have also been developed for terrestrial use [14] A new concentrator optic is introduced, consisting of a nonimaging dome-shaped Fresnel lens, and a kaleidoscope homogenizer The nonimaging Fresnel lens allows a wide acceptance half angle with keeping the same optical efficiency with minimum chromatic aberration The homogenizer reshapes the concentrated light into the square solar cell aperture, and mixes rays to yield a uniform flux Injection molding is capable of manufacturing thousands of lenses in a single day and by a single machine The drawback of this method is the difficulty in creating precise prism angles and flat facets The maximum efficiency was measured to be a little above 80%; the overall efficiency was 73% After improvement of the process conditions, the averaged efficiency raised to 85.4% A new packaging structure for III–V concentrator solar cells is developed, applicable mainly to Fresnel lens concentrator modules but it may also be used in a dish concentrator systems The solar cell used in the new receiver package is a specially developed III–V 508 Technology 0.1 0.9 0.08 0.85 0.06 0.8 FF Resistance (Ω) FF 0.04 0.75 Rs 0.02 0.7 RL RE 0.08 0.1 0.12 0.14 0.17 0.2 0.25 0.28 0.65 Grid pitch (mm) Figure 20 FF of the concentrator cells with various grids pitching under 250-sun light Series resistance (Rs), lateral resistance (RL), and total electrodes resistance (RE) are also shown 40 39 Efficieney (%) 38 37 36 35 34 33 32 31 30 10 100 Concentration ratio 1000 Figure 21 Efficiency of a high-efficiency InGaP/InGaAs/Ge 3-junction cell vs number of suns [12] 3-junction concentrator solar cell It is grown on a fragile Ge substrate with thickness of only 150 um The overall size was  mm with mm square aperture area In addition, the following technologies have been developed: Super high-pressure and vacuum-free lamination of the solar cell that suppresses the temperature rise to 20 °C under 550  geometrical concentration illumination of sunbeam Direct and voids-free soldering technologies of the fat metal ribbon to the solar cell, suppressing hot-spots and reducing the resistance, thereby allowing a current 400 times higher than normal nonconcentration operation to be passed with negligible voltage loss A new encapsulating polymer that survives exposure to high-concentration UV and heat cycles Beam-shaping technologies that illuminate the square aperture of the solar cell, from a round concentration spot Homogenizer technologies that give a uniform flux and prevent the conversion losses that stem from chromatic aberration and flux intensity distribution The concentrator module is designed with ease of assembly in mind All the technologically complex components are packaged into a receiver so that a series of receivers and lenses can be assembled with standard tools, using local materials and workforce The Multiple Junction Solar Cells 509 Figure 22 Inside view of the 400  concentrator module with 36 receivers connected in series concept is similar to the computer and automobile assembly industries, where key components are imported but the product assembled locally It is anticipated that this approach will reduce the manufacturing cost of the module as shown in Figure 22 The peak uncorrected efficiency for the 7056 cm2 400  module with 36 solar cells connected in series was 27.6%, measured in house The peak uncorrected efficiencies for the same type of module with six solar cells connected in series and 1176 cm2 area measured by Fraunhofer ISE and NREL were 27.4% and 24.9%, respectively The 5445 cm2 550  modules have also demonstrated 27–28.9%, measured in house Table summarizes the measured efficiency in three different sites New 400  and 550  (geometrical concentration ratio) modules are developed and show the highest efficiency in any types of PV as well as more than 20 years of accelerated lifetime This achievement is due to new innovative concentrator technologies The new concentrator system is expected to open a door to a new age of high-efficiency PV 1.22.5 Most Recent Results of MJ Cells Recently, more than 40% efficiency cells were reported by Fraunhofer ISE [29] and Spectrolab [30] Concentrator 4-junction or 5-junction solar cells have great potential for realizing super high efficiency of over 50% We have been studying concentrator MJ solar cells under the Japanese Innovative Photovoltaic R&D program started since fiscal year (FY) 2008 [31] Figure 23 shows an overview of NEDO’s PV R&D Program The Japanese Innovative Photovoltaics R&D Program has started in FY 2008 and the target in this program is to develop high-efficiency solar cells with conversion efficiency of more than 40% and low electricity cost of less than JPY kWh−1 until 2050 Most recently, world-record efficiency (35.8%) at sun (AM1.5G) has been realized with inverted epitaxial grown InGaP/GaAs/ InGaAs 3-junction cells by Sharp [32] Figure 24 shows the fabrication process of InGaP/GaAs/InGaAs 3-J solar cell by inverted epitaxial growth, and in Figure 25 the I–V curve of this world-record efficiency InGaP/GaAs/InGaAs solar cell is depicted Figure 26 shows chronological improvements in conversion efficiencies of III–V compound MJ solar cells under 1-sun and concentrator conditions Most recently, 42.1% efficiency under 230 suns has been obtained with InGaP/GaAs/InGaAs cell as shown in Figure 27 [32] 1.22.6 Future Directions MJ solar cells will be widely used in space because of their high conversion efficiency and better radiation resistance In order to apply super high-efficiency cells widely, it is necessary to improve their conversion efficiency and reduce their cost The new CPV Table Uncorrected peak efficiency measurement Concentration Area (cm2) 400  400  400  7056 7056 1176 400  550  550  1176 5445 5445 Site Inuyama, Japan, manufacturer Toyohashi, Japan, independent Fraunhofer ISE, Germany, independent NREL, USA, independent Inuyama, Japan, manufacturer Toyohashi, Japan, independent Ambient temperature (°C) Uncorrected efficiency (%) DNI (W m−2) 29 19 27.6 25.9 27.4 810 645 839 29 33 28 24.9 28.9 27 940 741 777 510 Technology Overview of NEDO’s PV R&D program May, 2010 Commercialization Accelerated commercialization R&D Advanced PV R&D Next generation PV R&D Next Generation PV R&D PL:M Yamaguchi Toward major energy source 4.08 BJPY (FY2010) 28.25 BJPY (5yrs) PL: M Yamaguchi & K Kutrokawa 5–10% of National PES Infrastructural R&D Infrastruc­ tural R&D Funda­ mental 2004 Mid/long-term: high performance next generation PV [14–7 JPY/KWh Target] Ultra-Long-Term: Innovative PV [40%˜PV Seeds Exploration R&D] 2006 2008 2010 2012 NEDO 2014 Fiscal year Figure 23 Overview of NEDO’s PV R&D Program MOCVD epitaxial growth Cell fabrication Growth direction Bottom InGaAs bottom buffer layer GaAs middle Buffer layer Middle In G Bu aAs ffe bo G rl aA ay ttom s er m In id G dl aP e to p InGaP top InGaP top GaAs middle Top Buffer layer GaAs sub Sub GaAs sub InGaAs bottom Support sub Invert growth Lattice mismatch bottom cell should be grown at the last in order to keep good crystal quality for top and middle cells Layer transfer Cell layers should be mounted on a support substrate for handling Figure 24 Fabrication process of InGaP/GaAs/InGaAs 3-J solar cell inverted epitaxial grown [29] system with times more annual power generation than the conventional crystalline silicon flat-plate system will open a new market for apartment or building rooftop applications Another interesting application is what we call the tree planting PV and large-scale PV power plant applications Now, we are approaching 40% efficiency by developing concentrator MJ solar cells as shown in Figure 28 Concentrator 4-junction or 5-junction solar cells have great potential for realizing super high-efficiency of over 50% [33, 34] Therefore, concentrator 3-junction and 4-junction solar cells have great potential for realizing super high efficiency of over 40% As a 3-junction combination, InGaP/ InGaAs/Ge cell on a Ge substrate will be widely used because this system has already been developed The 4-junction combination of an Eg = 2.0 eV top cell, a GaAs second-layer cell, a material third-layer cell with an Eg of 1.05 eV, and a Ge bottom cell is lattice matched to Ge substrates and has a theoretical efficiency of about 42% under 1-sun AM0 This system has a potential efficiency of over 47% under 500-sun AM1.5 condition Since concentrator MJ and crystalline Si solar cells are expected to contribute to electricity cost reduction for widespread PV applications as shown in Figure 29 [28], we would like to contribute to commercialization of CPV technologies as the third-generation PV technologies in addition to the first-generation crystalline Si PV and the second–generation, thin-film PV technologies The scale of the CPV industry lags that of flat-plate PV by about one or two decades It is expected, however, to make up this delay to the point where, in 2013, the cumulative installed capacity will lie in the region of several hundred MWp R&D work has to be undertaken particularly in the area of large-scale production, that is, high throughput, to realize this ambition Material consumption must also be reduced A projection for future turnkey CPV system prices extrapolated from current prices is shown in Multiple Junction Solar Cells I–V CURVE IEC 60904–3Ed.2 0.880 cm2 (aperture area) WHSS WHSS 40 12 30 20 Power (mw) Current (mA) 10 511 Date : Sep 2009 Date No : sc 9112–01 Sample No : sc 9112 Repeat Times : Isc Voc Pmax Ipmax Vpmax F F Eff (ap) DTemp MTemp DIrr MIrr 10 12 27 mA 012 V 31 54 mW 12 05 mA 617 V 85 % 35 % 25 °C 25 °C 100 mW / cm2 100 mW / cm2 (top) 100 mW / cm2 (middle) 100 mW / cm2 (bottom) Scan Mode Isc to Voc 0 0.5 1.5 Voltage (V) 2.5 Figure 25 Current(I)–voltage(V) curve of world-record efficiency InGaP/GaAs/InGaAs measured by the AIST [29] Target : 45% 45 Number of Junctions Efficiency (%) 40 III–V, concentrator 1-junction 2-junction 3-junction 35 30.3% JE 33.3% JE, Sumitomo TTI 35.8% This time 31.7% JE 30 III–V, 1-sun, AM1.5G 25 20.2% NTT 20 1985 SHARP Sharp Previous record Silicon, 1-sun, AM1.5G 1990 1995 2000 2005 2010 69 Year Figure 26 Chronological improvements in conversion efficiencies of III–V compound multi-junction solar cells under 1-sun and concentrator conditions [32] Figure 30 This figure shows the expected yearly production of CPV systems (red) and the price of turnkey installed CPV systems in Euro/Wp (black) [35] The most important R&D in CPV manufacturing will aim at [35]: Improving the efficiency of mass-produced cells to the levels currently seen in the laboratory (over 26%) and to 35–45% efficiency in the longer term Improving optical elements (optical efficiency, lifetime, and product engineering) 512 Technology 46 44 42.1% (× 230) 42 Efficiency (%) 40 38 36 34 32 30 28 10 100 Concentration ratio 1000 Figure 27 Concentration ration dependence of efficiency of a high-efficiency InGaP/GaAs/InGaAs 3-junction cell [32] 50 Conc III–V Multi-junction Efficiency (%) 40 30 Thin-film Si Cryst.Si CIS 20 Dye-sensitized Organic 10 a-Si 1940 1960 1980 2000 Year 2020 2040 2060 Figure 28 Future predictions of solar cell efficiencies [33, 34] 1:Si crystal 2:Thin-film 3:Concentrator 100 Electricity cost (yen kWh–1) 10 1995 2000 2005 2010 2015 Year 2020 Figure 29 Scenario of electricity cost reduction by developing concentrator solar cells [28] 2025 2030 105 10 104 103 102 101 100 2005 2010 2015 2020 Year 2025 2030 513 System price ( wP–1) Yearly prodcution (MWp) Multiple Junction Solar Cells Figure 30 The expected yearly production of concentrator PV (CPV) systems (red) and the price of turnkey installed CPV systems in Euro/Wp (black) [35] WBGU’s World Energy Vision 2100 1,600 Geothermal Other REs WBGU: German Advisory Council on Global Change http://www.wbgu.de/ Primary energy supply [EJ/Y] 1,400 Solar heat >70% 1,200 Solar electricity 1,000 20% 800 Wind 600 Biomass adv 400 Biomass trad 200 Hydro-PW Nuclear PW Gas Coal Oil 2000 2010 2020 2030 2040 2050 2100 YEAR Figure 31 World Energy Vision 2100 presented by the German Council on Global Change [36] Automated industrial module assembly (adjustment of elements, packaging, and sealing), high-throughput manufacturing with high yield, resulting in products with long lifetimes Construction of light, robust, and precise trackers for all outdoor climate conditions Set-up and monitoring of demonstration systems and large plants, in the range several hundred kWp (short term) to multi-MWp (medium term) Techniques for guaranteeing the quality of products with intended lifetimes of over 20 years, development of standards, in-line testing, and recycling methods for the modules Figure 31 shows the World Energy Vision 2100 presented by the German Council on Global Change [36] As a result of further development of higher efficiency, lower cost, and highly reliable solar cells, larger contribution to world energy by PV is expected, including concentrator MJ solar cells Acknowledgments This work was supported by the New Energy and Industrial Technology Development Organization (NEDO) under the Ministry of Economy, Trade and Industry (METI) The author thanks members of Toyota Technological Institute, Sharp, Daido Steel, JAXA, JAEA, University of Tokyo, Meijo University, Kyushu University, and Miyazaki University for their collaboration and cooperation 514 Technology References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] Jackson ED (1955) Transactions of the Conference on the Use of Solar Energy, Vol 5, p.122 Tucson: University of Arizona Press Wolf M (1960) Proceedings of the Institute of Radio Engineers 48: 1246 Fan JCC, Tsaur B-Y, and Palm BJ (1982) Proceedings of the 16th IEEE Photovoltaic Specialists Conference, p 692 New York: IEEE Hutchby JA, Markunas RJ, and Bedair SM (1985) In: Ded SK (eds.) Proceedings of the SPIE “Photovoltaics”, vol 543, p 543 Ando K, Amano C, Sugiura H, et al (1987) Japanese Journal of Applied Physics 26: L266 Sugiura H, Amano C, Yamamoto A, and Yamaguchi M (1988) Japanese Journal of Applied Physics 27: 269 Amano C, Sugiura H, Yamamoto A, and Yamaguchi M (1987) Applied Physics Letters 51: 1998 Olson J, Kurtz S, and Kibbler K (1990) Applied Physics Letters 56: 623 Takamoto T, Ikeda E, Kurita H, et al (1997) Japanese Journal of Applied Physics 36: 6215 Yamaguchi M, Okuda T, Taylor SJ, et al (1997) Applied Physics Letters 70: 1566 Brown MR, Goldhammer LJ, Goodelle GS, et al (1997) Proceedings of the 26th IEEE Photovoltaic Specialists Conference, p 805 New York: IEEE Takamoto T, Kaneiwa M, Imaizumi M, and Yamaguchi M (2005) Progress in Photovoltaics 13: 495 King RR, Law DC, Fetzer CM, et al (2005) Proceedings of the 20th European Photovoltaic Solar Energy Conference, p 118 Munich: WIP Araki K, Uozumi H, Egami T, et al (2005) Progress in Photovoltaics 13: 513 Tomita T (2006) Proceedings of the 4th World Conference on Photovoltaic Energy Conversion, p 2450 New York: IEEE Yamaguchi M, Warabisako T, and Sugiura H (1994) Journal of Crystal Growth 29: 136 Ahrenkiel RK, Keyes BM, Durbin SM, and Gray JL (1993) Proceedings of the 23rd IEEE Photovoltaic Specialists Conference, p 42 New York: IEEE Yang M-J, Yamaguchi M, Takamoto T, et al (1997) Solar Energy Materials and Solar Cell 45: 331 Yamaguchi M and Amano C (1985) Journal of Applied Physics 58: 3601 Yamaguchi M, Ohmachi Y, Oh’hara T, et al (2001) Progress in Photovoltaics 9: 191 Ringel SA, Sieg RM, Ting S, et al (1998) Proceedings of the 2nd World Conference on Photovoltaic Solar Energy Conversion, p.3594 European Commission Takamoto T, Agui T, Kamimura K, et al (2003) Proceedings of the 3rd World Conference on Photovoltaic Energy Conversion, p 581 Osaka: WCPEC-3 Takamoto T, Yamaguchi M, Ikeda E, et al (1999) Journal of Applied Physics 85: 1481 Takamoto T, Ikeda E, Agui T, et al (1997) Proceedings of the 26th IEEE Photovoltaic Specialists Conference, p 1031 New York: IEEE Takamoto T, Agui T, Ikeda T, and Kurita E (2000) Proceedings of the 28th IEEE Photovoltaic Specialists Conference, p 976 New York: IEEE Khan A, Yamaguchi M, Bourgoin JC, and Takamoto T (2000) Applied Physics Letters 76: 2559 Lang DV, Kimerling LC, and Leung SY (1976) Journal of Applied Physics 47: 3587 Yamaguchi M (2003) Solar Energy Materials and Solar Cells 75: 261 Bett AW, et al (2009) Proceedings of the 24th European Photovoltaic Solar Energy Conference, p WIP King RR, Boca A, Hong W, et al (2009) Proceedings of the 24th European Photovoltaic Solar Energy Conference, p 55 Munich: WIP Yamaguchi M, Ohshita Y, Kojima N, et al (2009) Proceedings of the 24th European Photovoltaic Solar Energy Conference, p 218 WIP Takamoto T, et al (2010) Proceedings of the 35th IEEE PVSC IEEE (to be published) Goetzberger A, Luther J, and Willeke G (2002) Solar Energy Materials and Solar Cells 74: Yamaguchi M (2004) Proceedings of the 19th European Photovoltaic Solar Energy Conference, pp l Munich: WIP EC (2007) A strategic research agenda for photovoltaic solar energy technology Report of the EU PV Technology Platform http://www.eupvplatform.org/fileadmin/Documents/ PVPT_SRA_Complete_070604.pdf [36] WBGU (German Advisory Council on Global Change) (2003) World in Transition – Towards Sustainable Energy Systems London: Earthsan ISBN 1-85383-882-9 http://www wbgu.de/ ... density 10 00 GaAs/Si Minority-carrier lifetime τ (ns) 10 16 cm–3 16 –3 10 0 × 10 cm InP/Si NTT model InGaP/Si GaAs/GaAs 10 × 10 17 cm–3 S.A Ringel et al InP/InP 10 18 cm–3 NTT data 0 .1 104 10 5 10 6 10 7 10 8... (ns) 10 3 10 2 10 1 10 0 10 15 10 16 10 17 10 18 10 19 N (cm–3) Figure Carrier concentration dependence of minority-carrier lifetime in p-type and n-type GaAs [17 ] 499 500 Technology 10 00 S = (cm s 1) 5800... developing concentrator solar cells [28] 2025 2030 10 5 10 10 4 10 3 10 2 10 1 10 0 2005 2 010 2 015 2020 Year 2025 2030 513 System price ( wP 1) Yearly prodcution (MWp) Multiple Junction Solar Cells Figure 30