Applied Physics Seminar | Computational Discovery of Ultra-Wide-Band-Gap Semiconductors

Semiconducting materials play a crucial role in modern society, from information technology and optical communications to renewable energy generation and vehicle electrification. In particular, ultra-wide-band-gap (UWBG) semiconductors, i.e., semiconductors with band gaps wider than the 3.5 eV value of GaN, are promising for higher efficiency, reduced size, and lower cost in high-power electronics applications. For example, materials such as diamond, cubic BN, β-Ga 2 O 3 , and AlGaN promise higher conversion efficiency and orders-of-magnitude improvements in power density compared to current technologies (Si, SiC, and GaN). However, none of the above UWBG
materials offers all the desired properties needed for high-performance electronics and, despite decades of research, very few alternative UWBG semiconductors have been realized to date. There is therefore a pressing need to develop synergistic methods that combine predictive theory with experimental synthesis and validation in order to discover and design new UWBG semiconductors that can surpass the limitations of current technologies.

In recent years, our team has advanced the development of rutile GeO 2 and its alloys with SnO 2 as a novel family of UWBG semiconductors that can surpass the state of the art in power electronics. Our predictive atomistic calculations demonstrate that these alloys exhibit superior fundamental properties that overcome the limitations of current materials. Their band gaps span from 3.6 eV for SnO 2 to 4.68 eV for rutile GeO 2 [1]. They are predicted to exhibit ambipolar dopability, with Sb Ge , As Ge , Ta Ge , H i , and F O acting as shallow donors, while Al Ge and Ga Ge acting as acceptors [2]. The predicted carrier mobilities are high [3], while the relatively light carrier effective masses prevent the formation of self-trapped polarons. The predicted thermal conductivity is also high and surpasses β-Ga 2 O 3 , a prediction that we verified experimentally in unoptimized polycrystalline bulk samples [4]. Overall, we find that the predicted Baliga figure of merit of rutile GeO 2 (i.e., a measure of the performance of materials in power electronics), modified to account for donor ionization, surpasses all known semiconductors, demonstrating its unique potential for energy-efficient power electronics [5].

Experimentally, we demonstrate the synthesis of single-crystalline GeO 2 -based thin films and substrates, which are prerequisites for epitaxial devices. Using suboxide olecular-beam epitaxy (MBE), we demonstrate the stability of GeSnO 2 alloy thin films over their entire composition range [6], while the development and epitaxy on single-crystalline rutile GeO 2 substrates enables the epitaxy of single-crystalline thin films [7]. Overall, our work demonstrates the unique promise of rutile GeO 2 -based materials for advancing the state of the art in power electronic devices.

[1] J. Appl. Phys. 126, 085703 (2019)
[2] Appl. Phys. Lett. 114, 102104 (2019)
[3] Appl. Phys. Lett. 117, 182104 (2020)
[4] Appl. Phys. Lett. 117, 102106 (2020)
[5] Appl. Phys. Lett. 118, 260501 (2021)
[6] Appl. Phys. Lett. 117, 072105 (2020)
[7] J. Vac. Sci. Technol. A 40, 050401 (2022)

Short Bio:
Emmanouil (Manos) Kioupakis is a Professor of Materials Science and Engineering and the Karl F. and Patricia J. Betz Family Faculty Scholar at the University of Michigan. His
research focuses on developing and applying predictive materials-modeling methods to explain and predict the synthesis and functionalities of new semiconductor materials for
electronics, optoelectronics, and energy applications. Highlights of his work include pioneering calculations of phonon-mediated quantum phenomena in materials, such as
optical absorption in silicon and non-radiative recombination in LEDs, as well as uncovering fundamental insights on the structure-property relationships of modern nitride and oxide semiconductors.