High Efficiency Multi-Junction III-V Semiconductor Photovoltaics on a Low-Cost Silicon Platform: A New Integration Technology

Problem Statement:
Current state-of-the-art, high-efficiency photovoltaics are used in high-performance, space-based power systems for satellites, space probes and space stations. Due to the physics of light absorption, carrier generation and carrier transport, these photovoltaics are particularly well suited for high concentration solar applications. More specifically, the performance of these cells is exceptional at 500x to 1000x sun concentration. Excellent performance under high solar concentration makes these cells extremely well suited for use in earth-bound, solar-electric power generation using solar concentration technologies.

Solar concentrators are an extremely promising solution for supplier-scale generation facilities in high-sun areas such as the southwestern United States, Australia, western China and desert regions of the Middle East and Africa. The limitations of direct sunlight and the requirements for solar tracking makes these systems less desirable for rooftop applications, but delivery on the promise of extremely low-cost module solutions could improve their attractiveness for broader consumer applications. In an effort to serve broader markets, several companies throughout Europe and the United States are realizing high-volume manufacturing of reliable optics and tracking systems.

The high-efficiency photovoltaic cell itself is based on multijunction technology developed by the U.S. government’s National Renewable Energy Laboratory (NREL) during the 1980s and ‘90s. This technology has traditionally been a multi-layer stack of III-V compound semiconductors grown on germanium (Ge) substrates. Germanium presents two problems of ongoing concern for this technology. First, 100mm diameter Ge substrates are expensive compared to other semiconductor materials such as silicon (Si). Second, the worldwide supply of Ge is primarily a byproduct of zinc (Zn) production, which is currently produced at about 100,000 kg per year. Politically and economically unstable regions of the world, including Africa and Central Asia, control this supply.

Ultimately, the cost of photovoltaic cell material will be a limiting factor on the implementation of concentrator systems on a broad scale. Nevertheless, concentrator-based solar energy production will be a dominant technology for certain regions of the world. With a low-cost, low-surface area cell material, concentrator systems could actually find broad appeal across the globe. For example, concentrators could be as much as 100 percent more productive per square-meter of solar module coverage for moderately sunny regions. Within regions where surface area is a premium, such as Japan, concentrators could also prove to be a significant technology.

AmberWave’s Solution:
AmberWave Systems operates a 3000 square-meter facility that houses 24 employees near Boston, Massachusetts, USA. Expertise in growing groups IV, III-V and III-N semiconductors directly on silicon, along with epitaxy and metrology tools, makes AmberWave the premier research and development company in the field of silicon- III-V heteroepitaxy.

AmberWave Systems has been a major contributor to the phenomenal success of introducing new materials to the integrated circuits (IC) industry. The company’s expertise in managing the integration of dissimilar semiconductor materials, while maintaining the semiconductor’s critical electrical properties, has proven to be essential to the successful advancement of IC technology over the last four years. AmberWave’s patented, silicon germanium-based strained silicon will continue to be one of the principle IC technologies for at least the next 10 years.

Unique in the materials science community, AmberWave’s understanding of the critical control of defects during the growth of semiconductor layers with high crystal lattice mismatch between the layers is significant to solving several advanced semiconductor problems. Applying this understanding to the photovoltaic issues described earlier, shows promise for a significant reduction in the cost of multijunction photovoltaics for concentrator-based systems.

Recent developments at AmberWave have shown great promise in creating III-V multijunction solar cells directly on Si. After more than 30 years of industry research, several paths have emerged for marrying the low cost of Si substrates and the performance of III-V semiconductors. Direct bonding of III-V compound semiconductors through a layer-transfer process has shown promise, but this technique is challenged by low yields and an inability to compensate for high thermal expansion differences in the multi-layer structures. As an alternative, growth of a graded composition interface layer has been used to create the desired structures. Yet, defect control in both graded buffer layers and thermal expansion mismatch continues to be unresolved barriers to success. Selective epitaxy is, in principle, very promising, but achieving low-defect layers with areas large enough for device applications has not previously been demonstrated.

AmberWave has invented several critical integration techniques, which hold great promise in simultaneously solving the defect and thermal expansion mismatch problems. There are separate strategies for both graded buffer and selective epitaxy technologies. With these techniques now well under control, AmberWave is ready to pursue R&D partnerships with premier solutions providers to develop commercial solutions for III-V based multijunction solar cells on Si.

AmberWave is now seeking research and development partners. Cooperation fees are in the range of US$1.0M to US$2.0M per year, depending on the scope and goals of the project. Licenses to AmberWave’s existing and/or future patent portfolio in these technologies are available, but not a requirement for initial pursuit of research objectives.