Microscopic Views of Ge Segregation and Scavenging Ge on Thin Si on Epi-Ge(001)

Si capping layer is the most notable approach used in Ge metal-oxide-semiconductor (MOS);1-2 however, the Ge segregation and diffusion occurred during the growth of Si.3-4 The formation of undesirable GeOx is detrimental to the Ge nMOS reliability.5 This work focuses on using the scavenging process to reduce the segregated Ge atoms and to completely remove GeOx in the high-k/epi-Si/epi-Ge(001). We used high-resolution synchrotron radiation photoelectron spectroscopy (SRPES) to show the detailed development of Ge segregation and scavenging Ge using in-situ film growth, oxidation, and annealing. The Si films in 8Å thickness were grown on the epi-Ge(001) in a semiconductor molecular beam epitaxy (MBE) chamber.2 These samples were in-situ transferred to National Synchrotron Radiation Research Center in Taiwan for electronic structure studies using photoemission. Molecular oxygen was exposed to the epi-Si/epi-Ge(001) surfaces at 300°C in the photoemission chamber. The samples were then in-situ annealed at 500°C for 5 min under an ultra-high vacuum (UHV). The topmost surface of epi-Ge(001) is terminated with Ge-Ge buckled dimers.6 In the case of room-temperature grown amorphous Si (a-Si) film, the intensity of the Ge down-dimer component from the underlying epi-Ge remains as it is. The Ge up-dimer atoms were partly diffused into the a-Si film, and some of them were segregated to the top of the a-Si surface. The epi-Si grown at 260 - 280°C causes the rest of the Ge down-dimer atoms to move to the epi-Si surface to become both segregated Ge (segGe) and diffused Ge (diff-Ge). The growth of Si merely affects the topmost surface, and the Ge atoms in the second layer of the epi-Ge remain intact. A comparison of the amount of GeOx for HfO2/epi-Si/epi-Ge(001) and HfO2/epi-Ge(001) shows that the epi-Si has greatly reduced the amount of GeOx. However, the GeOx, segGe and diff-Ge components are still observed in the HfO2/epi-Si/Ge(001) samples. We have previously reported that three-time scavenging cycles have greatly reduced the amount of segGe and diff-Ge atoms in high-κ/epi-Si/n-Ge(001), thus decreasing electron traps.7 Each scavenging cycle includes room-temperature oxidation followed by thermal annealing. In this study, the oxidation of the as-grown epi-Si/epi-Ge(001) samples was performed at 300°C. It is worth noting that there is no GeOx formation on the surface after the thermal oxidation, which is different from the room-temperature oxidation of the epi-Si/epi-Ge(001) surfaces. In addition, the oxidation at 300°C affects part of the diff-Ge atoms to evaporate from the surface, which is also different from our previous work, where the diff-Ge component shows no change in intensity since the oxidation occurred at room temperature. The subsequent in-situ annealing at 500°C moved the residual Ge-boned Si (diff-Ge) inside the epi-Si to the surface to become part of the segGe atoms. In conclusion, we have used the aforementioned process to further reduce the segregated Ge, and thus the GeOx, on top of the epi-Si/epi-Ge(001). To whom the correspondence is addressed: mhong@phys.ntu.edu.tw (M. Hong), raynien@phys.nthu.edu.tw (J. Kwo), and pi@nsrrc.org.tw (T. W. Pi) Acknowledgments This work is supported by MOST 110-2112-M-002-036-, 110-2622-8-002-014-, 110-2923-M-002-001-, and 110-2112-M-213-012- of the Ministry of Science and Technology in Taiwan. Reference 1 H. Arimura, E. Capogreco, A. Vohra, C. Porret, R. Loo, E. Rosseel, A. Hikavyy, D. Cott, G. Boccardi, L. Witters, G. Eneman, J. Mitard, N. Collaert, and N. Horiguchi, IEEE Int. Electron Devices Meet., 2.1.1−2.1.4 (2020). 2 H. W. Wan, Y. J. Hong, Y. T. Cheng, C. K. Cheng, C. H. Hsu, C. T. Wu, T. W. Pi, J. Kwo, and M. Hong, M., ACS Appl. Electron. Mater. 3, 2164−2169 (2021). 3 R. Loo, H. Arimura, D. Cott, L. Witters, G. Pourtois, A. Schulze, B. Douhard, W. Vanherle, G. Eneman, O. Richard, P. Favia, J. Mitard, D. Mocuta, R. Langer, N. Collaert, ECS J. Solid State Sci. Technol. 7, 66−72 (2018). 4 Y. T. Cheng, H. W. Wan, C. K. Cheng, C. P. Cheng, J. Kwo, M. Hong, T. W. Pi, Appl. Phys. Express 13, 085504 (2020). 5 J. Franco, B. Kaczer, P. J. Roussel, J. Mitard, S. Sioncke, L. Witters, H. Mertens, T. Grasser, G. Groeseneken, IEEE Int. Electron Devices Meet., 15.2.1−15.2.4 (2013). 6 Y. T. Cheng, Y. H. Lin, W. S. Chen, K. Y. Lin, H. W. Wan, C. P. Cheng, H. H. Cheng, J. Kwo, M. Hong, T. W. Pi, Appl. Phys. Express 10, 075701 (2017). 7 Y. T. Cheng, H. W. Wan, T. Y. Chu, T. W. Pi, J. Kwo, M. Hong, ACS Appl. Electron. Mater. 3, 4484-4489 (2021).