Search

This paper re-examines the impact of atmospheric absorption bands on series-connected multijunction cell design, motivated by the numerous local efficiency maxima that appear as the number of junctions is increased. Some of the local maxima are related to the bottom subcell bandgap and are already well understood: As the bottom subcell bandgap is varied, a local efficiency maximum is produced wherever the bottom cell bandgap crosses an atmospheric absorption band. The optimal cell designs at these local maxima are generally current matched, such that all subcells have nearly the same short-circuit current. Here, we systematically describe additional local maxima that occur wherever an upper subcell bandgap encounters an atmospheric absorption band. These local maxima are not current matched and become more prevalent as the number of junctions increases, complicating the solution space for five-junction and six-junction designs. A systematic framework for describing this complexity is developed, and implications for numerical convergence are discussed. Copyright © 2017 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

Photovoltaic conversion efficiencies of 32.6 ± 1.4% under the AM1.5 G173 global spectrum, and 35.5% ± 1.2% at 38-suns concentration under the direct spectrum, are demonstrated for a monolithic, dual-junction 1.7/1.1 eV solar cell. The tandem cell consists of a 1.7 eV GaInAsP top-junction grown lattice-matched to a GaAs substrate, followed by a metamorphic 1.1 eV GaInAs junction grown on a transparent, compositionally graded metamorphic AlGaInAs buffer. This bandgap combination is much closer to the dual-junction optimum and offers headroom for absolute 3% improvement in efficiency, in comparison to the incumbent lattice-matched GaInP/GaAs (∼1.86/1.41 eV) solar cells. The challenge of growing a high-quality 1.7 eV GaInAsP solar cell is the propensity for phase separation in the GaInAsP alloy. The challenge of lattice-mismatched GaInAs solar cell growth is that it requires minimizing the residual dislocation density during the growth of a transparent compositionally graded buffer to enable efficient metamorphic tandem cell integration. Transmission electron microscopy reveals relatively weak composition fluctuation present in the 1.7 eV GaInAsP alloy, attained through growth control. The threading dislocation density of the GaInAs junction is ∼1 × 106 cm−2, as determined from cathodoluminescence measurements, highlighting the quality of the graded buffer. These material advances have enabled the performance of both junctions to reach over 80% of their Shockley-Queisser limiting efficiencies, with both the subcells demonstrating a bandgap-voltage offset, WOC (=Eg/q-VOC), of ∼0.39 V. [ABSTRACT FROM AUTHOR]

The use of InGaAsP/InGaP quantum well structures is a promising approach for subcells in next generation multi-junction devices due to their tunable bandgap (1.50–1.80 eV) and for being aluminum-free. Despite these potentials, the accumulation of stress during the growth of these structures and high background doping in the quantum well region have previously limited the maximum number of quantum wells and barriers that can be included in the intrinsic region and the sub-bandgap external quantum efficiency to less than 30.0%. In this paper, we report on the use of in-situ curvature monitoring by multi-beam optical stress (MOS) sensor measurements during the growth of this quantum well structure to monitor the stress evolution in these thin films. A series of In 0.32 Ga 0.68 AsP/In 0.49 Ga 0.51 P quantum wells with various arsine to phosphine ratios have been analyzed by in-situ curvature monitoring and X-ray diffraction (XRD) to obtain nearly strain-free lattice matched structures. Sharp interfaces, as indicated by the XRD fringes, have been achieved by using triethyl-gallium and trimethyl-gallium as gallium precursors in InGaAsP and InGaP, respectively, with constant flows of trimethyl-indium and phosphine through the entire quantum well structure. The effect of the substrate miscut on quantum well growth was compared and analyzed using XRD, photoluminescence and time resolved photoluminescence. A 100 period quantum well device was successfully grown with minimal stress and approximately flat in-situ curvature. [ABSTRACT FROM AUTHOR]