Invited: Advances in III-V/Active-Silicon Multijunction Photovoltaics: Progress Toward a Si-Plus Architecture

Multijunction solar cells consisting of III-V sub-cells monolithically integrated with Si growth substrates, which also serve as active sub-cells, hold the potential for the achievement of high conversion efficiencies that are on par with pure III-V multijunction structures, but at substantially lower costs.  In addition to high-efficiency photovoltaics, the III-V/Si materials system is of interest to a range of other applications, including Si-based light emitters and detectors and III-V enhanced microelectronics.  Although a “Si-plus” architecture such as this has been a goal within the photovoltaics research community for decades, progress toward this end has been effectively hampered by issues related to heterovalent (polar/nonpolar) epitaxy at the GaP/Si interface.  However, work in recent years on the epitaxy of high-quality GaP on Si(100) substrates with substantial control and suppression of nucleation-related defects, including antiphase domains, stacking faults, and microtwins, via both molecular beam epitaxy (MBE) and metalorganic chemical vapor deposition (MOCVD), has provided realistic pathways for the development of these structures.  Combined with an active GaP/Si 1.12 eV bottom junction, near-ideal 2- and 3-junction monolithic structures are possible via integration with metamorphic GaAsP top-cell and GaInP/GaAsP top-/middle-cell stacks using semi-transparent compositionally-graded GaAs_yP_1-ybuffers to bridge the lattice mismatch.

For the Si-plusarchitecture to succeed, every aspect of the multijunction structure must be fully optimized, as any weak point will reduce the overall performance accordingly.  This includes not only the metamorphic III-V sub-cells and associated tunnel junctions, but also the Si sub-cell.  Achievement of high performance from the metamorphic III-V sub-cells requires minimization of detrimental defects resulting from the lattice-mismatched heteroepitaxy (i.e. threading dislocations), as well as optimization of device structures and growth conditions at these relatively unexplored compositions.  Sub-cell interconnection requires high-performance tunnel diodes (optically transparent, low electrical resistance), made additionally difficult by the wider bandgaps and metamorphic nature of the III-V system of interest.  And finally, the Si sub-cell must be re-designed and optimized for multijunction application and integration with III-V materials, which, despite Si PV’s general high level of maturity, is a non-trivial undertaking due to the material’s indirect bandgap and high sensitivity to non-idealities and defects.

To this end, we will discuss progress made by our group toward the achievement of the Si-plus architecture and efforts toward addressing the full range of associated issues.  This will include discussion of methodologies developed for the heteroepitaxial growth of GaP on Si, via both MBE and MOCVD, free of nucleation-related defects, as well as basic materials studies of this integrated materials system.  We will also present on efforts toward the development of metamorphic III-V graded buffers and materials to support high-quality integrated devices.  The development and optimization of epitaxial GaInP, GaAsP, and Si sub-cells, as well as matching high-performance metamorphic III-V tunnel junctions, will be discussed.  Finally, we present results from efforts toward the integration of these many parts into functioning 2- and 3-junction solar cells, and directions for further optimization and refinement.