Conference PaperPDF Available

Development of advanced space solar cells at Spectrolab

Authors:

Abstract and Figures

High efficiency multi-junction solar cells utilizing inverted metamorphic and semiconductor bonding technology are being developed at Spectrolab for use in one-sun space and near-space applications. Recently that effort has been extended to include low-concentration space applications. This paper will review the present state-of-the-art cell technologies at Spectrolab, with an emphasis on performance characterization data at both 1-sun and low-concentration operating conditions that these cells will experience in flight.. A cell coupon utilizing IMM solar cells has been assembled and subjected to thermal cycling. Pre-and post thermal cycling data have been collected and there is no performance degradation or mechanical issues after the test.
Content may be subject to copyright.
DEVELOPMENT OF SPACE SOLAR CELLS AT SPECTROLAB
J. Boisvert, D. Law, R. King, D. Bhusari, X. Liu, S. Mesropian, D. Larrabee, R. Woo, K. Edmondson, D. Krut, D.
Peterson, K. Rouhani, B. Benedikt, and N. Karam
Boeing-Spectrolab, Inc., 12500 Gladstone Ave., Sylmar, CA 91342 U.S.A.
ABSTRACT
High efficiency Inverted Metamorphic (IMM) and
Semiconductor Bonded Technology (SBT) multi-junction
solar cells have been under development at Spectrolab for
use in space and near space applications. This paper will
review the present state-of-the-art of this technology at
Spectrolab with an emphasis on performance
characterization data at operating conditions that these
solar cells will experience in flight. Solar cell current-bias
characteristics under illumination (LIV) at AM0 28°C are
presented along with external quantum efficiency
measurements that are used to verify the X-25 solar
simulator LIV short circuit current density. A mechanical
and thermal stress model has been used to predict
mechanical stresses on a ultra-lightweight panel assembly
in orbit and will be discussed.
INTRODUCTION
High efficiency IMM and SBT multi-junction solar cells [1-5]
have been under development at Spectrolab for use in
space and near space applications. This paper reviews the
present state-of-the-art of this technology at Spectrolab
with an emphasis on performance characterization data at
operating conditions that these solar cells will experience
in flight.
Under the AFRL IBIS program a coupon utilizing large
area, low mass IMM solar cells has been assembled. A
cross section of an IMM solar cell is shown in Figure 1.
Figure 1. A typical IMM solar cell that is grown in an
inverted configuration on a Ge or GaAs substrate. The
grown structure is affixed to a handle and the growth
substrate removed.
In a typical 3-junction IMM space cell three constituent
GaAs-based subcells are grown in an inverted
configuration. Large volume production MOVPE reactors
are used to grow these solar cells on 100 mm substrates.
The widest bandgap alloy (top cell) is grown first followed
by the middle cell, buffer layers and finally a low bandgap
metamorphic cell lattice mismatched to the growth
substrate. Subsequent wafer processing places the
inverted multijunction solar cell in an upright configuration
and the growth substrate is removed. Processes typical of
standard, high-volume semiconductor wafer processing
are used to complete fabrication. Cell-Interconnect-
Coverglass (CICs) are then assembled based on typical
production assembly processes.
TEST ARTICLES
A variety of solar cell test articles have been constructed
for use in this technology development. Small area (1 cm2)
as well as large area (26 cm2) cells for use in 1 sun AM0
environments have been fabricated and tested. Low
concentration small area 2.5 cm2 cells have also been
fabricated and will be reported on in this paper. Because
specific metal grid patterns vary with design, performance
data for the concentrator cells is reported for cell aperture
areas only. CHARACTERIZATION DATA
Typical current-bias characteristics for Spectrolab 3J IMM
solar cell under 1 sun AM0 solar simulator illumination (LIV
data) are shown in Figure 2. These simulator data were
collected on an AX-25 solar simulator using calibrated IMM
Lear Jet flight standards; to date no IMM balloon flight
standards have been flown. Typical spectral response
measurements for these cells are shown in Figure 3 and
are agree with the measured Jsc data shown in Figure 2.
Low concentration IMM cells were subjected to additional
characterization at both 1 sun and ~12X concentration.
Typical LIV characteristics at 25 °C are shown in Figure 4.
Low mass 26.62 cm2 IMM cells and Coverglass-
Interconnect-Cells (CICs) have been fabricated for use on
the AFRL IBIS program. Figure 5 displays the backside
handle of this cell. A cavity structure has been fabricated in
the handle to reduce the total cell mass. A finite element
mechanical and thermal analysis of an IBIS panel
consisting of these CICs affixed to a metalized Kapton
substrate has been performed. This analysis predicts that
the structure will withstand exposure to -180 °C which a
space solar panel can be subjected to on orbit. The stress
distribution for the half plane of this structure is shown in
Figure 6. The maximum anticipated stresses on the
Growth
Direction
Ge or GaAs growth
substrate from which
active cells are removed
nucleation
GaInAs buffer
separation layer
Ge or GaAs growth
substrate from which
active cells are removed
nucleation
nucleation
GaInAs buffer
separation layer
Growth
Direction
Growth
Direction
Growth
Direction
Ge or GaAs growth
substrate from which
active cells are removed
nucleation
GaInAs buffer
separation layer
Ge or GaAs growth
substrate from which
active cells are removed
nucleation
nucleation
GaInAs buffer
separation layer
Figure 2. The LIV characteristic of a 1x1 cm2 3J IMM
cell at 28 °C. This cell was designed for 1 sun AM0
applications and has an AM0 conversion efficiency of
32.6%. The reported efficiency is based on Lear Jet
IMM calibration standards.
Figure 3. Typical external quantum efficiency
characteristics collected on the 3J IMM solar cells
shown in Figure 2.
Figure 4. 25 °C LIV characteristics of low concentration
IMM cells at 1X AM0 and ~12X AM0 concentration.
Figure 5. The low mass IBIS 3J IMM cell handle.
Figure 6. The low concentration 3JIMM half-plane
stress model.
Figure 7. An assembled IBIS coupon.
0
0.002
0.004
0.006
0.008
0.01
0.012
0.014
0.016
0.018
0.02
00.511.522.533.
5
V
olta
g
e
(
V
)
Jsc (Amp/cm2)
0
10
20
30
40
50
60
70
80
90
100
350 550 750 950 1150 1350
Wavelength (nm)
EQE (%)
0
0.1
0.2
0.3
0.4
0.5
01234
Voltage (V)
Current (A)
8e-11313-04
Figure 8. Photoluminescence maps of 1 eV
metamorphic wafers show recent progress in material
quality.
structure are on the order of 10 MPa, much less than the
failure strengths of the cell materials.
A coupon of low mass IMM CICs was assembled for the
AFRL IBIS program and is shown in Figure 7. Spectrolab
continues to make progress on MOVPE growth of
metamorphic material quality. Figure 8 displays
photoluminescence maps of two wafers demonstrating that
progress. The PL signal shows much better uniformity on
the more recent wafer. That uniformity is reflected in 3J
IMM cell performance shown in Figure 9 along with large
area cell performance data collected solar cells fabricated
from similar material. Large area 3J IMM cells presently
display about 31% AM0 efficiency as measured on an AX-
25 solar simulator set up to a Lear Jet flight calibration.
Figure 9. A wafer uniformity map of 1x1 cm2 3J IMM
cells and LIV characteristics of large area 26 cm2 cells
built from similar material.
Spectrolab is also pursuing semiconductor bonded solar
cells for space applications. A 4J SBT solar cell is shown in
Figure 10. This cell has the advantage that all subcells can
be grown lattice-matched to independent substrates which
leads to improved material quality and higher subsequent
performance. The trade is that large area wafer bonds
require very low surface roughness to be held over the
entire wafer. Spectrolab has succeeded in fabricating large
area bonds as shown in Figure 11. 4J SBT cells have been
characterized and shown to have 33.5% AM0 conversion
efficiency as measured on an AX-25 solar simulator set up
to 3J IMM Lear Jet standards – Figure 12. The external
quantum efficiency (as measured using a spectrometer set
up to NIST-traceable calibration standards) of one of these
cells is shown in Figure 13 and demonstrates that all 4
subcells exceed 90% EQE.
Figure 10. A 4J SBT utilizes lattice-matched subcells
grown on two different substrates (in this case GaAs
and InP) that are bonded together and subsequently
processed similarly to an IMM cell.
Figure 11. A 4J SBT wafer fabricated with 1x1 cm2
cells.
Figure 12. The AM0 LIV characteristics of 1x1 cm2 4J
SBT cells.
Target parameters: V
OC
3.6 V, J
SC
16 mA/cm
2
, FF 84%, eff. 35% AM0
0.5
0.4
0.3
0.2
0.1
0.0
Current Density [mA/cm
2
]
18001600140012001000800600400
Wavelength [nm]
2.0 1.5 1.1 0.7 eV
0.000
0.002
0.004
0.006
0.008
0.010
0.012
0.014
0.016
0.018
0.020
0 0.5 1 1.5 2 2.5 3 3 .5
4
Voltage (V)
Current density (A/cm2)
4J-SBT Target, 35% AM0
S1: 1.470 V, 15.6 mA, 0.856
S2: 1.081 V, 16.0 mA, 0.845
S3: 0.680 V, 16.8 mA, 0.817
S4: 0.350 V, 18.4 mA, 0.711
GaAs
s1
s2
InP
s4
s3
+=
GaAs
InP
+
bond
polish & reuse
remove
s2
s3
s4
s1
y
,y
Target parameters: V
OC
3.6 V, J
SC
16 mA/cm
2
, FF 84%, eff. 35% AM0
0.5
0.4
0.3
0.2
0.1
0.0
Current Density [mA/cm
2
]
18001600140012001000800600400
Wavelength [nm]
2.0 1.5 1.1 0.7 eV
0.000
0.002
0.004
0.006
0.008
0.010
0.012
0.014
0.016
0.018
0.020
0 0.5 1 1.5 2 2.5 3 3 .5
4
Voltage (V)
Current density (A/cm2)
4J-SBT Target, 35% AM0
S1: 1.470 V, 15.6 mA, 0.856
S2: 1.081 V, 16.0 mA, 0.845
S3: 0.680 V, 16.8 mA, 0.817
S4: 0.350 V, 18.4 mA, 0.711
GaAs
s1
s2
InP
s4
s3
+=
GaAs
InP
+
bond
polish & reuse
remove
s2
s3
s4
s1
s2
s3
s4
s1
y
,y
0.000
0.002
0.004
0.006
0.008
0.010
0.012
0.014
0.016
0.018
0.020
01234
Voltage [V]
Current Density [A/cm2]
AM0 4J-SBT_B12: 3.394V 15.99mA 0.835 33.5%
AM0 4J-SBT_B6: 3.393V 15.94mA 0.833 33.3%
0
10
20
30
40
50
60
70
80
90
100
350 450 550 650 750 850 950 1050 1150 1250 1350 1450 1550 1650 175 0 185
0
Wavelength (nm)
Extern al QE (%)
4J-SBT_s1
4J-SBT_s2
4J-SBT_s3
4J-SBT_s4
SBT4J Reflectance
Figure 13. The external quantum efficiency versus
wavelength as measured on a 4J SBT solar cell.
CONCLUSIONS
Spectrolab continues development of inverted
metamorphic solar cell technology for high efficiency space
and near space applications. IMM solar cells with 1X AM0
efficiency greater than 32.5% at 28 °C have been
demonstrated. These efforts have led to development of 3J
IMM low concentration cells. 3J IMM CICs have been
measured at 28 °C under concentration and have
demonstrated greater than 34% efficiency. A coupon with
3J IMM CICs has been fabricated under the IBIS program.
4J SBT solar cells have been fabricated and demonstrate
33.5% AM0 efficiency at 28 °C.
ACKNOWLEDGEMENTS
The authors would like to thank the entire R&D team at
Spectrolab. Support from the Air Force Research
Laboratory Space Vehicles Directorate under Contracts
FA9453-09-C-0373 and FA9453-04-2-0042; and funds
(support) from The Boeing Company is gratefully
acknowledged.
References
[1] M. W. Wanlass, S. P. Ahrenkiel, R. K. Ahrenkiel, D. S.
Albin, J. J. Carapella, A. Duda, J. F. Geisz, S. Kurtz, T.
Moriarty, R. J. Wehrer, and B.Wernsman, Proceedings of
the 31st IEEE Photovoltaic Specialists Conference, p. 530
(2005).
[2] R. R. King, D. C. Law, C. M. Fetzer, R. A. Sherif, K. M.
Edmondson, S. Kurtz, G. S. Kinsey, H. L. Cotal, D. D. Krut,
J. H. Ermer, and N. H. Karam, Proc. 20th European
Photovoltaic Solar Energy Conference, p. 118 (2005).
[3] D. C. Law, D. M. Bhusari, S. Mesropian, J. C. Boisvert,
W. D. Hong, A. Boca, D. C. Larrabee, C. M. Fetzer, R. R.
King, and N. H. Karam, 2009 34th IEEE Photovoltaic
Specialists Conference (PVSC), p. 2237-2239 (2009).
[4] Yoon, Hojun; Haddad, Moran; Mesropian, Shoghig;
Yen, Jason; Edmondson, Kenneth; Law, Daniel; King,
Richard R.; Bhusari, Dhananjay; Boca, Andreea; Karam,
Nasser H., Proceedings of the 33rd IEEE Photovoltaic
Specialists Conference, p. 1 (2006).
[5] J. Boisvert, D. Law, R. King, D. Bhusari, X. Liu, A.
Zakaria, W. Hong, S. Mesropian, D. Larrabee, R. Woo, A.
Boca, K. Edmondson, D. Krut, D. Peterson, K. Rouhani, B.
Benedikt, and N.H. Karam, Proceedings of the 35th IEEE
Photovoltaic Specialists Conference, p. 123 (2010).
... The single-junction QE is from Alta Devices' internal measurements. The triplejunction QE is from Boisvert [4]. We assume that the shapes of the quantum efficiency curves do not change with angle of incidence. ...
... Fig. 3. External quantum efficiency of single junction (1J) and triplejunction (3J) cells. The 3J QE is from [4]. The nominal AM0 current densities shown were used to normalize the results as discussed in the text. ...
Conference Paper
Full-text available
Solar-powered fixed-wing high-altitude long-endurance pseudosatellites (HALE or HAPS) are being developed to provide telecommunication and related services to markets across the globe. For technical, operational and regulatory reasons, these aircraft need to operate continuously at stratospheric altitudes, which is a challenge given the limited electrical energy available aloft. We have conducted SMARTS simulations to generate solar spectra at 20 km altitude on the Tropic of Cancer during the winter and summer solstices. Estimates of component cell currents for single junction and triple junction III-V cells were also calculated. These results and approach are expected to be of benefit to the calculation and optimization of performance of HALE aircraft.
Article
Full-text available
The spectral distribution of the solar irradiance incident on photovoltaic (PV) modules is a key variable controlling their power production. It is required to properly simulate the production and performance of PV plants based on technologies with different spectral characteristics. Spectroradiometers can only sense the solar spectrum within a wavelength range that is usually too short compared to the actual spectral response of some PV technologies. In this work, a new methodology based on the Simple Model of the Atmospheric Radiative Transfer of Sunshine (SMARTS) spectral code is proposed to extend the spectral range of measured direct irradiance spectra and to increase the spectral resolution of such experimental measurements. Satisfactory results were obtained for both clear and hazy sky conditions at a radiometric station in southern Spain. This approach constitutes the starting point of a general methodology to obtain the instantaneous spectral irradiance incident on the plane of array of PV modules and its temporal variations, while evaluating the magnitude and variability of the abundance of atmospheric constituents with the most impact on surface irradiance, most particularly aerosols and water vapor.
Chapter
Semiconductor heterostructures allow solving the problems of controlling the fundamental parameters of semiconductor devices owing to the possibility of change in the electronic band structure, bandgaps and refractive indices of a material itself during epitaxial growth, as well as effective masses and mobilities of charge carriers in it.
Conference Paper
Full-text available
We present a thorough evaluation of a first-of-its-kind quad-source AM 0 solar simulator developed by TS-Space Systems and installed at the AFRL Space Vehicles Directorate in Kirtland AFB, NM. With an underlying need to evaluate more complex and advanced space photovoltaics, the TS-Space Unisim 100 was chosen as replacement to the Spectrolab X-25. Several parameters of the Unisim were characterized including spectral irradiance, lamp stability (long-term and short-term), spatial uniformity, and repeatability. In addition, the impact of test plane translation due to variable cell holders was studied. The potential impact of AC input voltage to lamp ballast was also investigated. Further, comparisons are made to the X-25. Finally, operational considerations for the use of a multi-source simulator are discussed.
Chapter
Metal organic vapor phase epitaxy (MOVPE) has proved to be the primary materials growth technique for low‐cost, high‐volume III‐V multijunction (MJ) solar cells. This chapter shows the typical applications that drive volume production for MJ solar cells using MOVPE. The two commonly used structures for lattice‐mismatched MJ solar cells are the upright metamorphic (UMM) structure and the inverted metamorphic (IMM) structure. The chapter discusses the key aspects of reactor technology pertinent to UMM and IMM solar cell growth. The main commercial low‐pressure, As/P MOVPE reactor technology are vertical chambers where wafers are held on a planar rotating graphite susceptor, acting to aid the laminar flow or subplanetary rotation used to planarize the growth. The in situ monitoring becomes a critical component when considering lattice‐mismatched growth, in particular the measurement of wafer curvature. One may relate the wafer curvature to the net strain at growth temperature through the use of Stoney's equation.
Article
Multi-junction solar cells based on III–V compounds are the most efficient converters of solar energy to electricity and are widely used in space solar arrays and terrestrial photovoltaic modules with sunlight concentrators. All modern high-efficiency III–V solar cells are based on the long-developed triple-junction III–V GaInP/GaInAs/Ge heterostructure and have an almost limiting efficiency for a given architecture — 30 and 41.6% for space and terrestrial concentrated radiations, respectively. Currently, an increase in efficiency is achieved by converting from the 3-junction to the more efficient 4-, 5-, and even 6-junction III–V architectures: growth technologies and methods of post-growth treatment of structures have been developed, new materials with optimal bandgaps have been designed, and crystallographic parameters have been improved. In this review, we consider recent achievements and prospects for the main directions of research and improvement of architectures, technologies, and materials used in laboratories to develop solar cells with the best conversion efficiency: 35.8% for space, 38.8% for terrestrial, and 46.1% for concentrated sunlight. It is supposed that by 2020, the efficiency will approach 40% for direct space radiation and 50% for concentrated terrestrial solar radiation. This review considers the architecture and technologies of solar cells with record-breaking efficiency for terrestrial and space applications. It should be noted that in terrestrial power plants, the use of III–V SCs is economically advantageous in systems with sunlight concentrators.
Article
Full-text available
Multijunction solar cells for terrestrial concentrator applications have reached the point at which the next set of technology improvements are likely to push cell efficiencies over 40%. This paper discusses the semiconductor device research paths being investigated with the aim of reaching this efficiency milestone. Lattice-matched (LM) GaInP/ GaInAs/ Ge 3-junction cells have achieved the highest independently confirmed efficiency for a photovoltaic device, at 39.0% at 236 suns, 25°C under the standard AM1.5D, low-AOD terrestrial spectrum. Lattice-mismatched, or metamorphic (MM), materials offer still higher potential efficiencies, if the crystal quality can be maintained. Theoretical efficiencies well over 50% are possible for a MM GaInP/ 1.17-eV GaInAs/ Ge 3-junction cell limited by radiative recombination at 500 suns. The bandgap – open circuit voltage offset, (E g /q) – V oc , is used as a valuable theoretical and experimental tool to characterize multijunction cells with subcell bandgaps ranging from 0.7 to 2.1 eV. Experimental results are presented for prototype 6-junction AlGaInP/ GaInP/ AlGaInAs/ GaInAs/ GaInNAs/ Ge cells employing an active ~1.1-eV dilute nitride GaInNAs subcell, with active-area efficiency greater than 23% and over 5.3 V open-circuit voltage under the 1-sun AM0 space spectrum. Such cell designs have theoretical efficiencies under the terrestrial spectrum at 500 suns concentration exceeding 55% efficiency, even for lattice-matched designs. Through a combination of device structure advances under investigation in research groups around the world, the goal of a practical 40%-efficient photovoltaic cell is near.
Conference Paper
Full-text available
We discuss lattice-mismatched (LMM) approaches utilizing compositionally step-graded layers and buffer layers that yield III-V photovoltaic devices with performance parameters equaling those of similar lattice-matched (LM) devices. Our progress in developing high-performance, LMM, InP-based GaInAs/InAsP materials and devices for thermophotovoltaic (TPV) energy conversion is highlighted. A novel, monolithic, multi-bandgap, tandem device for solar PV (SPV) conversion involving LMM materials is also presented along with promising preliminary performance results.
Article
Procedures for measuring the spectral response of multi-junction cells in general require variation of the bias spectrum and voltage biasing. It is shown that a refined procedure including optimization of bias spectrum and voltage is necessary to minimize a measurement artifact, which appears if the subcell under test has non-ideal properties, such as a low shunt resistance or a low reverse breakdown voltage. This measurement artifact is often observed on measuring the spectral response of the Ge bottom cell of GaInP/Ga(In)As/Ge triple-junction cells. The main aspect of the measurement artifact is that the response of another subcell is simultaneously measured, while at the same time the signal of the Ge subcell is too low. Additionally, the shape of the spectral response curve is influenced under certain measurement conditions. In this paper the measurement artifact is thoroughly discussed by measurement results and simulation. Based on this analysis, a detailed procedure for the spectral response measurement of multi-junction cells is developed, specially designed to minimize such measurement artifacts. Copyright
Article
Experimental studies and numerical analyses are carried out to optimize A1GaAs‐GaAs heteroface solar cell structures. Carrier removal rate and damage constant for diffusion length in n‐GaAs due to 1‐MeV electron irradiation are found to be larger than those in p‐GaAs. These results are explained by taking into account deep‐level traps associated with radiation‐induced defects. Numerical analysis shows that n<sup>+</sup><sup>+</sup>‐n<sup>+</sup>‐p heteroface cell structure is relatively radiation resistant in a shallow junction solar cell (below 0.2 μm) with a substrate carrier concentration above 3×10<sup>1</sup><sup>5</sup>cm<sup>-</sup><sup>3</sup>. In the p<sup>+</sup><sup>+</sup>‐p<sup>+</sup>‐n heteroface solar cell, optimum junction depth and substrate carrier concentration are 0.2–0.3 μm and 2–5×10<sup>1</sup><sup>5</sup> cm<sup>-</sup><sup>3</sup>, while those in the n<sup>+</sup><sup>+</sup>‐n<sup>+</sup>‐p heteroface solar cell are less than 0.1 μm and 3–10×10<sup>1</sup><sup>5</sup> cm<sup>-</sup><sup>3</sup>.
Article
The maximum efficiencies of ideal solar cells are calculated for both single and multiple energy gap cells using a standard air mass 1.5 terrestrial solar spectrum. The calculations of efficiency are made by a simple graphical method, which clearly exhibits the contributions of the various intrinsic losses. The maximum efficiency, at a concentration of 1 sun, is 31%. At a concentration of 1000 suns with the cell at 300 K, the maximum efficiencies are 37, 50, 56, and 72% for cells with 1, 2, 3, and 36 energy gaps, respectively. The value of 72% is less than the limit of 93% imposed by thermodynamics for the conversion of direct solar radiation into work. Ideal multiple energy gap solar cells fall below the thermodynamic limit because of emission of light from the forward‐biased p‐n junctions. The light is radiated at all angles and causes an entropy increase as well as an energy loss.
Conference Paper
Inverted metamorphic (IMM) solar cells based on III–V materials have the potential to achieve solar conversion efficiencies that are significantly higher than today's state of the art solar cells which are based on the 3-junction GaInP/GaInAs/Ge design. The 3J IMM device architecture based on (Al)GaInP/GaInAs/GaInAs, for example, allows for a higher voltage solar cell by replacing the low bandgap Ge (0.67 eV) from the conventional 3J structure with the higher bandgap (∼1 eV) metamorphic GaInAs. The inverted growth simply allows the lattice-matched junctions (i.e., (Al)GaInP/GaInAs) to be grown first on the growth substrate, thereby minimizing or shielding them from the defects that arise from the metamorphic layers. Spectrolab has demonstrated 30.5% AM0 efficiency based on the 3J IMM cell architecture grown on a Ge substrate, with V oc = 2.963V, J sc = 16.9 mA/cm<sup>2</sup>, and FF = 82.5%. In addition, 4J IMM cells have been demonstrated with V oc of 4.072 V and AM0 efficiency approaching 25%. With additional development, demonstrating 33% AM0 efficiency is expected in the near future. However, the IMM devices demand more complex processing requirements than conventional solar cells, and we demonstrate the capability to fabricate large area solar cells from standard Ge solar cell substrates.
Conference Paper
Device-wafer bonding provides a platform for the implementation of ultra-high-efficiency multijunction solar cell designs, by allowing optimal subcell bandgap combinations to be attained while using only high-quality materials lattice-matched to their growth substrates. One promising new method for achieving wafer bonding is to use carbon nanotube composite thin films as the bonding agent between subcells grown on dissimilar substrates. In this paper we present the first demonstration of CNT-composite bonding of III-V materials, and evaluate its suitability for solar-cell integration in terms of optical transparency, electrical conductivity, bond uniformity and robustness, and bonded-device electrical performance. Another, relatively more mature method for device-wafer integration is that of direct semiconductor bonding technology. In order to provide a basis for comparison with CNT-bonding, we also summarize the latest achievements of the SBT solar cell development effort at Spectrolab.
Conference Paper
Boeing-Spectrolab recently demonstrated monolithic 5-junction space solar cells using direct semiconductor-bonding technique. The direct-bonded 5-junction cells consist of (Al)GaInP, AlGa(In)As, Ga(In)As, GaInPAs, and GaIn(P)As subcells deposited on GaAs or Ge and InP substrates. Large-area, high-mechanical strength, and low-electrical resistance direct-bonded interface was achieved to support the high-efficiency solar cell structure. Preliminary 1-sun AM0 testing of the 5-junction cells showed encouraging results. One of the direct-bonded solar cell achieved an open-circuit-voltage of 4.7 V, a short-circuit current-density of 11.7 mA/cm<sup>2</sup>, a fill factor of 0.79, and an efficiency of 31.7%. Spectral response measurement of the five-junction cell revealed excellent external quantum efficiency performance for each subcell and across the direct-bonded interface. Improvements in crystal growth and current density allocation among subcells can further raise the 1-sun, AM0 conversion efficiency of the direct-bonded 5-junction cell to 35 - 40%.
Ultra-High-Efficiency Multijunction Cell and Receiver Module
  • R R King