top of page

III–V on Silicon Tandems (PP3.2)

PP3.2: III–V on Silicon Tandems

The aim of this package is to leverage Australia’s strengths in silicon-based solar research and to work with partners whose world-leading expertise on tandem solar cells and their complementary fabrication and characterization capabilities, to provide synergies for the development of a new generation of silicon wafer cell technologies.

By stacking top cells on silicon, the inherent efficiency advantages of tandem cell designs can be harnessed to substantially improve on current silicon performance through the addition of thin layers of high-performance materials.


The III-V materials system is one of the promising top cell candidates, with a proven track record of high-performance, long-term stability, and an existing deep knowledge base in terms of material properties and fabrication techniques. The major ways that III-V/silicon tandem cells can be realised are through direct growth, wafer bonding, mechanical stacking and spectrum splitting approaches.

PP3.2.1 Direct Growth

Investigators: A/Prof Stephen Bremner (UNSW), Prof Anita Ho-Baillie (Sydney University)

UNSW has been working closely with collaborators at Ohio State University on integrating a gallium phosphide buffer layer approach with high performance silicon designs. A key advantage of this approach is the demonstrated high quality of the grown III-V layers. This work has led to an official world record of 23.4% under 1 sun, with ample room for further improvements to beyond 30% linked mainly to optimising the silicon bottom cell.


A key outcome has been developing processes that preserve performance throughout disparate processing steps, as well as developing alternate silicon designs.

Future work in the direct growth activity includes adapting current and planned devices to CZ silicon to further reduce costs, and, working with collaborators at TNO, integrating poly-silicon contacting to the rear of the bottom silicon solar cell. Longer term, alternate designs, such as 3-terminal tandems, as well as novel integration approaches that are compatible with maintaining silicon cell performance will be key focuses.


The overarching aim of this activity is the development of high performance III-V/silicon tandems that offer a pathway to commercialisation by leveraging outstanding silicon cell expertise with world leading integration and novel processing approaches to lower costs and adapt to novel applications. 


  • Transfer of monolithic III-V/silicon tandem designs to CZ silicon to deliver lower costs

  • Implement low cost, industrially relevant rear structures for silicon bottom sub-cells such as polysilicon-based designs

  • Adaptation of devices including alternative III-V on silicon integration approaches, for further applications such as hydrogen generation.

PP3.2.2  Mechanically Stacked III-V on Silicon Tandem Cells

Investigators: N.Ekins-Daukes, Xiaojing Hao, Andrew Blakers, S. Bremner

The highest efficiency III-V cells will be achieved using a pristine III-V substrate, however the substrate cost will be prohibitive for all but concentrator and space applications. 


To retain the highest efficiency performance, while also integrating the solar cell with silicon, two mechanical stack processes are pursued. Both require III-V cells to be released from the substrate and then bonded onto a silicon cell either using an epoxy, or through physical wafer bonding. Both approaches have already demonstrated high efficiency, attaining 1-sun efficiencies in excess of 35% on silicon, the challenge remains to make these commercially attractive.

Present industrial processes involve lateral etching of a sacrificial layer in HF for many hours followed by chemical mechanical polishing of the wafer to recover an epi-ready surface. This allows for wafer reuse providing the potential for substantial cost saving, however, the time taken to undertake this process limits the value of the recovered wafers.   Other release and wafer reuse methods are presently under investigation, for example IQE-plc, the world’s largest compound semiconductor foundry, has recently demonstrated full 6-inch wafer lift-off with no requirement for lateral etching or wafer polishing.  With the lift-off process successfully demonstrated, realising III-V device architectures compatible with this process is required.

An alternative approach is to apply the ‘sliver’ technique to GaAs. In this approach a thick (~1mm) GaAs wafer is cut vertically into a myriad of long and thin rectangles (“slivers”), and then each sliver is rotated 90 degrees (i.e. laid flat) to vastly increase the effective surface area of the wafer.


  • Demonstrate cost effective methods for recovering an epi-ready surface for wafer reuse.

  • Demonstrate III-V devices fabricated from epitaxial templates composed of either silicon, germanium, gallium arsenide or indium phosphide.

  • Demonstrate ‘interstitial light-trapping’ that provides wavelength dependant light confinement using epitaxial lift-off devices.

  • Developing and optimising mechanically stacked III-V/silicon tandem designs and process flows to enhance reproducibility and commercial viability of this approach.

PP3.2.3 Spectrum Splitting

Investigators:  M. Keevers, J.Jiang, N. Ekins-Daukes

A tandem cell performs spectrum splitting naturally, with broadband sunlight being selectively absorbed by absorbers stacked one upon another.  Considerable freedom in design can be achieved by separating separate bands of sunlight and directing them towards different solar cells. The approach is particularly well suited for concentrator PV systems.

Working in partnership with RayGen, researchers at UNSW have set the world record for solar power conversion of 40.6% outdoor efficiency and are working to augment this to 50%. Achieving 50% electrical power conversion efficiency demands a high-performance optical system, tailored III-V photovoltaic devices and advanced cell metallisation.

A concentrator power plant inevitably produces both solar heat and electricity and is well suited to any application where both are desirable, such as desalination, hydro-production though electrolysis or medium grade process heat.

To date PV-Thermal schemes have generated modest electrical power and a great deal of low-grade heat with limited application. We can turn this around by enabling concentrated photovoltaic thermal (CPV-T) systems to access a larger range of applications.


  • Demonstrate >50% outdoor photovoltaic solar power conversion

  • Demonstrate a manufacturable spectrum splitting receiver that is configurable for high photovoltaic efficiency or hybrid PV-Thermal solar energy harvesting.

  • Demonstrate separation of the electrical and thermal generation in a high-efficiency, high temperature receiver.

bottom of page