Wafer-scale Ge epitaxial foils grown at high growth rates and released from porous substrates for triple-junction solar cells

Abstract

Germanium is listed as a critical raw material, and for environmental and economic sustainability reasons, strategies for lower consumption must be implemented. A promising approach is Ge lift-off concepts, which enable to re-use the substrate multiple times. Our concept is based on the Ge-on-Nothing approach that is the controlled restructuring at high temperature of a macroporous Ge surface, forming a Ge foil weakly attached to its parent wafer. Its suitability as III-V epitaxy seed and support substrate has previously been demonstrated with proof-of-concept solar cells. This work focuses on bringing this concept to the next level, by upscaling the detachable area to a full 200-mm wafer scale, increasing foil thickness for sufficient light absorption in the Ge bottom cell, and improving the control on the strength that is bonding the suspended foil to its parent. By introducing a new high growth-rate epitaxy process from GeCl4, and by engineering the GeON structure to introduce pillars with ad hoc density and shape, we fabricated P-type foils with tunable boron doping up to 15 mu m in thickness and 11 cm x 11 cm in area, for which the detachment strength could be adapted to the stresses induced by the solar cell process steps. The surface roughness and the electrical and crystal qualities of these foils were inspected by AFM, SIMS, SRP, ECCI, and TEM to check the GeCl4-based epitaxy conditions and to check that the ad hoc pillars were not introducing any damage. Small-area triple-junction lattice-matched GaInP/GaInAs/Ge solar cells were fabricated on 7-mu m-thick Ge foils with various pillar densities and on a standard reference Ge wafer. The III-V layer nucleation was virtually the same on both substrates and the solar cells on the GeON foils performed in the same way as the cells on the Ge wafer, albeit a small loss in short-circuit current and open-circuit voltage that can be attributed to the thickness reduction and absence of rear-side passivation. We conclude that it is possible to gain control on the GeON detachability and upscale the concept to areas relevant for the space PV industry, proving that porous germanium is a serious candidate for replacement of bulk Ge wafers in view of a more sustainable multijunction solar cell process.