Degradation of subcells and tunnel junctions during growth of GaInP/Ga(In)As/GaNAsSb/Ge 4‐junction solar cells

Abstract

A GaInP/Ga(In)As/GaNAsSb/Ge 4J solar cell grown using the combined MOVPE + MBE method is presented. This structure is used as a test bench to assess the effects caused by the integration of subcells and tunnel junctions into the full 4J structure. A significant degradation of the Ge bottom subcell emitter is observed during the growth of the GaNAsSb subcell, with a drop in the carrier collection efficiency at the high energy photon range that causes a~15% lower J sc and a V oc drop of~50 mV at 1-sun. The V oc of the GaNAsSb subcell is shown to drop by as much as~140 mV at 1-sun. No degradation in performance is observed in the tunnel junctions, and no further degradation is neither observed for the Ge subcell during the growth of the GaInP/Ga(In)As subcells. The hindered efficiency potential in this lattice-matched 4J architecture due to the degradation of the Ge and GaNAsSb subcells is discussed. At this stage of development, it is pertinent to have a closer look at the integration of components forming the 4J structure. For example, the insertion of the dilute nitride subcell in the 3J structure brings about added thermal loads to the Ge bottom subcell, an effect accentuated by the fact that dilute nitrides usually require an annealing step to improve their electronic properties. The dilute-nitride subcell itself suffers annealing during the growth of the upper subcells, which could be expected to have an impact on its performance. The focus has to be put in identifying and quantifying these effects, and redesigning the growth process to minimize their impact.In this work, we present a detailed characterization of our 4J solar cell, based on a prototype structure achieved by a combination of MOVPE + MBE growth methods, and using a GaNAsSb junction.This solar cell succeeds in integrating 4 component junctions into a monolithic, lattice matched structure, but it is still far from being a high efficiency device, mainly due to sub-optimum characteristics of the dilute nitride subcell. However, the device provides valuable insight into the integration of the dilute nitride subcell into a 4J solar cell. We focus mainly on the effect of thermal load on the performance of the Ge and GaNAsSb bottom subcells, and the tunnel junctions. It is found that significant losses are at stake, mainly in the Ge and GaNAsSb subcells, which can limit the potential of this 4J solar cell structure to compete in efficiency with other architectures.

Development of the Lattice Matched GaInP/GaInAs/Ge Triple Junction Solar Cell with an Efficiency Over 40%

“…This study has confirmed that the V OC obtained in Ge cells grown as single junction devices is significantly higher than that of a Ge BC in a 3JSC. This V OC loss can be as high as 55 mV [19,20] together with a drop in short-circuit current density in the BC of ~2.3 mA/cm 2 (due to thicker emitters as a result of a deeper III-V elements diffusion during the thermal load). Simulations confirm that all these losses can be explained by the degradation in the minority carrier properties at the emitter region, especially close to the GaInP/Ge interface [19].…”

Impact of the III-V/Ge nucleation routine on the performance of high efficiency multijunction solar cells

“…Additionally, Figure 3 IQE results reveal that despite as-grown Ge subcells (1JSCs) show evident differences in Jsc ( Figure 2), the thermal load associated with the growth of the middle and top subcells even out such differences reducing the J SC of the cells to similar final levels, as evidenced by Figure 4. This was also observed during the development of four-junction structures based on Ge bottom cells [32] where similar IQEs of the Ge BC were obtained between 2J and 4J. In conclusion, it does not make sense to optimize the IQE performance of the Ge BC out of a 3JSC structure (i.e; it makes no sense to optimize it in a 1JSC structure that is going to be implemented in a 3JSC).…”

Theoretical and experimental assessment of thinned germanium substrates for III–V multijunction solar cells

“…The Ge solar cell considered is a state‐of‐the‐art device with a 6·10 17 cm −3 constant doping level for the base. The emitter doping profile has been assumed to be uniform at 10 19 cm −3 , as a reasonable approximation 1,41 . Minority‐carrier mobilities and lifetimes reported in the literature 42,43 are used as input for the simulations.…”