Van der Waals Ga2O3 functional materials epitaxy: Revolutionary power electronics
Lead Research Organisation:
University of Bristol
Department Name: Physics
Abstract
The imminent climate risks for our planet have been highlighted by the UN's Intergovernmental Panel on Climate Change (IPCC) calling it "code red for humanity" in its scientific report in August 2021. Presently, nearly all energy conversion power electronics use Silicon (Si), which is relatively inefficient, wasting energy as heat. In fact, 72% of global primary energy consumption is wasted, of which 20% could be saved with new power electronics! Wide bandgap devices using GaN and SiC are entering the market, but they either do not sustain high enough voltages or are too expensive for widespread use, e.g., in smart grids. Ultrawide bandgap Gallium Oxide (Ga2O3), with efficiency far exceeding that of narrow bandgap Si, has emerged as a transformative contender with low cost and >1-2 kV, even 10 kV voltage capability, with potential for a massive >100x reduction in power conversion efficiency losses. Considering Ga2O3's high Baliga figure of merit, the metric determining how beneficial a material is for power devices, it has potential to even exceed current wide bandgap power devices (GaN, SiC), now replacing incumbent Si power electronics, by more than a factor of 5-10 in performance.
In this project, we target high voltage power devices using van der Waals epitaxy of functional Ga2O3 as a transformative and revolutionary vehicle to open a new research field for low cost high voltage power devices. This form of epitaxy uses an intermediate layer, which reduces the substrate-epitaxial layer interaction, enabling growth of the epitaxial layer onto a foreign material with different crystal structure at the wafer-level, potentially followed by layer transfer to other beneficial substrates, which would otherwise not be possible. If successful this will enable for the first time even >10kV low cost power electronics devices, e.g. for smart grids, i.e., a high reward but the approach taken is highly speculative: (i) Do van der Waals materials survive the reactive ambient of a Ga2O3 MOCVD chamber? (ii) does potential damage to the van der Waals material impact the subsequent device quality? (iii) can we control Ga2O3's polytypes during growth? (iv) do van der Waals grown interfaces provide high enough electron and phonon transport through them, and can these be optimized or mitigated for e.g. using growth conditions or h-BN thickness, or during a layer transfer? If successful, subsequent layer transfer would allow heterogeneous integration of Ga2O3 with numerous low-cost high thermal conductivity substrates such as poly-AlN and poly-diamond to revolutionize device heat sinking. Use of p-type substrates would open the design space to bipolar devices, even superjunctions, i.e. device concepts which have transformed Si power electronics, concepts which have rarely being able to venture beyond Si, a major impact if we are successful. Lateral devices will be used to validate the effectiveness of the integration, with routes to possible commercialization of high performance devices to be explored through our industrial partnership with Dynex Semiconductor. Vertical devices (with improved power density over lateral configurations, attractive for power electronics applications) will also be demonstrated.
The programme also marks a key milestone from a sustainability perspective, within a circular economy; van der Waals epitaxy or epilayer transfer used for the active devices will minimize the need for Ga2O3 substrates, to transfer to substrates with less sustainability issues (elemental Ga may become scarce within 100 years), minimizing the amount of Ga being used in power devices. Successful demonstration of heterogeneous integration of Ga2O3 by van der Waals epitaxy will also pave the way for low-cost, wafer-level integration with other ultrawide bandgap materials (e.g. single crystalline AlGaN, AlN, diamond), offering the potential to strongly reduce the contribution of inefficiency in power electronics to global energy consumption.
In this project, we target high voltage power devices using van der Waals epitaxy of functional Ga2O3 as a transformative and revolutionary vehicle to open a new research field for low cost high voltage power devices. This form of epitaxy uses an intermediate layer, which reduces the substrate-epitaxial layer interaction, enabling growth of the epitaxial layer onto a foreign material with different crystal structure at the wafer-level, potentially followed by layer transfer to other beneficial substrates, which would otherwise not be possible. If successful this will enable for the first time even >10kV low cost power electronics devices, e.g. for smart grids, i.e., a high reward but the approach taken is highly speculative: (i) Do van der Waals materials survive the reactive ambient of a Ga2O3 MOCVD chamber? (ii) does potential damage to the van der Waals material impact the subsequent device quality? (iii) can we control Ga2O3's polytypes during growth? (iv) do van der Waals grown interfaces provide high enough electron and phonon transport through them, and can these be optimized or mitigated for e.g. using growth conditions or h-BN thickness, or during a layer transfer? If successful, subsequent layer transfer would allow heterogeneous integration of Ga2O3 with numerous low-cost high thermal conductivity substrates such as poly-AlN and poly-diamond to revolutionize device heat sinking. Use of p-type substrates would open the design space to bipolar devices, even superjunctions, i.e. device concepts which have transformed Si power electronics, concepts which have rarely being able to venture beyond Si, a major impact if we are successful. Lateral devices will be used to validate the effectiveness of the integration, with routes to possible commercialization of high performance devices to be explored through our industrial partnership with Dynex Semiconductor. Vertical devices (with improved power density over lateral configurations, attractive for power electronics applications) will also be demonstrated.
The programme also marks a key milestone from a sustainability perspective, within a circular economy; van der Waals epitaxy or epilayer transfer used for the active devices will minimize the need for Ga2O3 substrates, to transfer to substrates with less sustainability issues (elemental Ga may become scarce within 100 years), minimizing the amount of Ga being used in power devices. Successful demonstration of heterogeneous integration of Ga2O3 by van der Waals epitaxy will also pave the way for low-cost, wafer-level integration with other ultrawide bandgap materials (e.g. single crystalline AlGaN, AlN, diamond), offering the potential to strongly reduce the contribution of inefficiency in power electronics to global energy consumption.
Publications
Nandi A
(2023)
Epitaxial Growth of (-201) ß-Ga 2 O 3 on (001) Diamond Substrates
in Crystal Growth & Design