Research and Technology | NASA Glenn Research Center

High efficiency photovoltaic devices (~ 30%) have exclusively been fabricated from crystalline III-V semiconductor materials. These materials, such as gallium arsenide (GaAs) and indium gallium phosphide (InGaP) possess nearly optimum electrical and optical properties. Under monochromatic illumination, diode efficiencies well in excess of 50% are regularly demonstrated in these materials.

Unfortunately, the growth of these high efficiency devices has been restricted to lattice matched, crystalline substrates such as GaAs and germanium (Ge). Ge has been the substrate material of choice for commercial multi-junction (MJ) III-V devices because it offers the opportunity to form a bottom photovoltaic junction as well as being more robust than GaAs, thereby allowing the use of a thinner substrate, resulting in a lower mass solar cell. Unfortunately, crystalline Ge offers no capability to integrate the finished photovoltaic device into a flexible module or a pathway to achieve the high mass specific powers (>1000 W/kg) required for Exploration missions.

Figure 1 – Comparison of mass specific power (at the cell level) for III-V multi-junction devices on various substrates. As shown, we assume that moving to more optimum substrates generally involves a slight reduction in device efficiency, but a tremendous increase in mass specific power, critical for beyond Earth orbit applications. In addition, the flexibility offered by polycrystalline III-V’s will enable novel planetary power systems (i.e. inflatable tent solar arrays, etc).

An opportunity exists to incorporate the strengths of high efficiency III-V MJ cell technology with substrates that offer significant benefits compared to the Ge typically used today (figure 1). GRC, in collaboration with the Ohio State University (OSU) and the Massachusetts Institute of Technology (MIT), have taken the first step on this path by demonstrating III-V MJ solar cells on silicon (Si) substrates. Si, by virtue of its superior strength, lower mass and higher thermal conductivity, offers a tremendous advantage compared to Ge.

Figure 2 – Nomarski micrograph of p/i/n GaAs cell

Figure 3 – AM0 I-V measurements of poly crystalline p/i/n GaAs cells with two different buffer thicknesses. A single crystalline p/i/n GaAs control cell is shown for reference.

To accomplish the demonstration, the 4% lattice mismatch and > 60% thermal coefficient of expansion (TCE) mismatch between GaAs and Si had to be addressed. To date, our research team holds the World record for single junction GaAs/Si device performance as well as the World’s only demonstration of III-V lattice mismatched MJ cells on Si substrates. [1, 2] This technology has recently begun on-orbit testing aboard MISSE5.

The path towards high mass specific power density and flexibility, which is critical to missions beyond Earth orbit, continues with the development of poly-crystalline III-V MJ solar cells on flexible substrates. Initial activities on the proposed development path have recently begun at NASA Glenn Research Center in two parallel efforts: (1) the development and demonstration of polycrystalline III-V photovoltaic devices on polycrystalline Ge substrates and (2) the development of recrystallized Ge films on molybdenum (Mo or moly) substrates. It is anticipated that the devices developed in (1) will be readily transitioned to the substrates developed in (2), thereby facilitating the demonstration of a III-M MJ device on a metal foil. In area (1), initial p/i/n GaAs photovoltaic devices were grown on poly Ge substrates (figures 2 and 3).

References

1. D. Wilt, et al, Proceedings 31st IEEE Photovoltaic Specialist Conference (2005)
2. S. Ringel, et.al., Proceedings 31st IEEE Photovoltaic Specialist Conference (2005)

Contact:
David Wilt
david.wilt@nasa.gov
(216) 433-6293