Suppression of GeO_x interfacial layer and enhancement of the electrical performance of the high-K gate stack by the atomic-layer-deposited AlN buffer layer on Ge metal-oxide-semiconductor devices

Abstract

For high-performance nanoscale Ge-based transistors, one important point of focus is interfacial germanium oxide (GeO_x), which is thermodynamically unstable and easily desorbed. In this study, an atomic-layer-deposited AlN buffer layer was introduced between the crystalline ZrO₂ high-K gate dielectrics and epitaxial Ge, in order to reduce the formation of interfacial GeO_x. The results of X-ray photoelectron spectroscopy and high-resolution transmission electron microscopy demonstrate that the AlN buffer layer suppressed the formation of interfacial GeO_x. Hence, significant enhancement of the electrical characteristics of Ge metal-oxide-semiconductor (MOS) capacitors was achieved with a two-orders-of-magnitude reduction in the gate leakage current, a 34% enhancement of the MOS capacitance, and a lower interfacial state density. The results indicate that the AlN buffer layer is effective in providing a high-quality interface to improve the electrical performance of advanced Ge MOS devices.

---

1. Introduction

Over the last decade, it has become increasingly difficult to improve the performance of silicon (Si) metal-oxide-semiconductor (MOS) field-effect transistors (FETs) via conventional device scaling. Due to high carrier mobility, germanium (Ge) and III–V compound MOSFETs have been regarded as very promising candidates for the further improvement of device performance and scaling.^1–4 However, the lack of high-quality and thermodynamically stable gate dielectrics is a major problem in implementing Ge and III–V semiconductors as the channel materials.^5–10 For Ge, it is difficult to suppress the formation of a low-K germanium oxide (GeO_x) interfacial layer (IL) at the high-K and Ge interface, which limits the minimum achievable equivalent oxide thickness.^11,12 In contrast to the SiO₂/Si system, interfacial GeO_x has been reported to be thermodynamically unstable.⁵ The smaller conduction band offset at the GeO_x/Ge interface also results in an increase in the gate leakage current.¹³ Therefore, it is essential to prevent the formation of interfacial GeO_x in high-performance Ge-based MOSFETs.^13,14

Many methods have been used to passivate the defects and achieve a stable high-K/Ge interface, such as the use of ultrathin Si, high-quality Ge oxides, and rare-earth oxides (Y₂O₃, SmGeO_x, etc.).^6,15–17 One of the most widely used methods is the insertion of an Al₂O₃ buffer layer between the high-K oxide and Ge because of the high bandgap and good thermal stability of Al₂O₃.^18,19 Besides, it has been reported that nitrogen incorporation into GeO_x yields an improvement in the thermal stability and the dielectric constant.^20,21 However, an unstable GeO_x is usually formed at the Al₂O₃/Ge and Ge-oxynitride/Ge interfaces. Furthermore, the use of germanium nitride (Ge₃N₄ or GeN_x) as the gate dielectric and the buffer layer has been demonstrated to exhibit well-behaved capacitance–voltage characteristics.^22–25 This improvement in the electrical performance may be the result of the suppression of interfacial GeO_x because the preparation of the nitride does not involve the use of oxygen.

AlN is a good material for gate dielectrics and buffer layers because it has a higher dielectric constant than GeO_x, good chemical stability, and a wide bandgap of ∼6.2 eV.^26–28 AlN thin films are conventionally prepared by chemical vapor deposition (CVD) and sputtering.^29,30 However, the typical deposition temperature for AlN prepared using CVD is greater than 700 °C,³¹ which is unfavorable for the integration of semiconductor processing. Besides, it is difficult to deposit high-quality nanoscale thin films using sputtering.³² Recently, atomic layer deposition (ALD) has been reported for the preparation of high-quality nanoscale AlN thin films at low temperatures.^26,31 In this paper, remote plasma ALD (RP-ALD) was used to deposit a nanoscale AlN buffer layer between the crystalline ZrO₂ high-K gate dielectrics and Ge. The interfacial GeO_x was suppressed by the AlN buffer layer, so the capacitance equivalent thickness (CET), the interfacial state density (D_it) and the gate leakage current (J_g) of Ge MOS capacitors were significantly improved.

2. Experiments

The gate stacks in MOS capacitors are plotted schematically in Fig. 1. A Ge epitaxial layer was grown by remote plasma chemical vapor deposition using GeH₄ at 375 °C on a highly-doped (0.001–0.003 Ω cm) n-type Si substrate. After pre-cleaning, the samples were immersed in dilute HF (1 min) and then rinsed with de-ionized water (30 s) for several cycles, in order to remove the native oxide on the Ge. The gate stacks, which were composed of ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂, were then immediately deposited by RP-ALD (Fiji, Ultratech) at 250 °C. Tetrakis(dimethylamino)zirconium (TDMAZ), trimethylaluminum (TMA), O₂ plasma, and N₂/H₂ plasma were respectively used as the precursors and reactants for Zr, Al, O, and N. A platinum (Pt) top electrode with an area of 3 × 10⁻⁴ cm² was deposited by radio-frequency sputtering, and then an aluminum (Al) back contact was then produced by thermal evaporation. Finally, all samples were annealed in a furnace at 450 °C in N₂ ambient for 30 minutes.

Fig. 1 Schematic of the (a) ZrO₂, (b) Al₂O₃/ZrO₂, and (c) AlN/ZrO₂ gate stacks.

The chemical bonding in the gate stacks was determined by X-ray photoelectron spectroscopy (XPS) using Al Kα radiation at 1486.6 eV. The cross-sectional images of the gate stacks were observed using high-resolution transmission electron microscopy (HRTEM, Philips Tecnai F20 G2 FEI-TEM, 200 kV). The crystalline phases of ZrO₂ were measured by grazing incident angle X-ray diffraction (GIXRD) using Cu Kα radiation at an incident angle of 0.5°. The capacitance density versus voltage (C–V), the leakage current density versus voltage (I–V), and the conductance density versus the frequency (G–f) curves of the gate stacks were characterized using an Agilent B1500A semiconductor device analyzer at room temperature.

3. Results and discussions

Fig. 2 shows the Ge 3d XPS spectra of the ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ gate stacks with and without the annealing treatment, which reveals the oxidation states of Ge. The XPS spectra can be decomposed into the peaks corresponding to the binding energies at 29.4 eV (Ge), 31.7 eV (GeO), 33.6 eV (GeO₂), and 30.9 eV (Ge–N).¹⁹ The XPS spectra of the samples without the annealing treatment are shown in Fig. 2(a)–(c). The presence of the GeO oxidation state in the ZrO₂ sample (Fig. 2(a)) is attributed to the supply of oxygen from ZrO₂. Fig. 2(b) reveals that the interfacial GeO was suppressed in the Al₂O₃/ZrO₂ gate stack compared to that in the ZrO₂ sample, which can be deduced from the lower Gibbs free energy of Al₂O₃ than that of ZrO₂.³³ It can be seen in Fig. 2(c) that the AlN buffer layer led to the formation of Ge–N bonds, and the GeO oxidation states were almost absent in the AlN/ZrO₂ sample because oxygen was not involved during AlN deposition. The result also suggests that the diffusion of oxygen from ZrO₂ toward the interface was effectively blocked by the AlN buffer layer, and so the formation of interfacial GeO_x was suppressed. Moreover, Fig. 2(d)–(f) are the XPS spectra of the samples treated with the annealing process. As compared with Fig. 2(a) and (b), the annealing treatment gave rise to an increase in the GeO signals and the formation of GeO₂ oxidation states in the ZrO₂ and Al₂O₃/ZrO₂ gate stacks:³⁴

2GeO → GeO₂ + Ge

Fig. 2 The Ge 3d XPS spectra of the (a) ZrO₂, (b) Al₂O₃/ZrO₂, and (c) AlN/ZrO₂ gate stacks without the annealing treatment and the (d) ZrO₂, (e) Al₂O₃/ZrO₂, and (f) AlN/ZrO₂ gate stacks treated with the annealing process (in a furnace at 450 °C for 30 minutes in N₂ ambient). The intensity of the Ge peak of each sample was normalized to one. The annealing treatment facilitated the formation of interfacial GeO_x in the ZrO₂ and Al₂O₃/ZrO₂ gate stacks. The interfacial GeO_x was significantly suppressed in the AlN/ZrO₂ gate stack.

The decomposition of GeO into GeO₂ and Ge is ascribed to the lower Gibbs free energy of GeO₂ than that of GeO.³⁴ The outcome indicates that the annealing process facilitated the formation of interfacial GeO_x. Notice that no obvious interfacial GeO_x appeared in the XPS spectrum of the AlN/ZrO₂ gate stack as shown in Fig. 2(f). The result confirms again that the AlN buffer layer could effectively restrain oxygen diffusion during the ZrO₂ deposition and annealing processes.

The Al 2p XPS spectra of the Al₂O₃/ZrO₂ and AlN/ZrO₂ gate stacks are shown in Fig. 3. The Al–O bond (75.6 eV) was present in the Al₂O₃/ZrO₂ sample due to the introduction of the Al₂O₃ buffer layer. The Al 2p peak of the AlN/ZrO₂ sample was located near 74.6 eV, which is associated with the Al–N bond because AlN was substituted for the Al₂O₃ buffer layer.³⁵ Depositing ZrO₂ on the AlN buffer layer results in a slight shift in the Al 2p peak (∼75 eV) from the standard Al–N bond at 74.6 eV toward the Al–O bond at 75.6 eV, indicating the partial oxidation of AlN.

Fig. 3 Al 2p XPS spectra of the Al₂O₃/ZrO₂ and AlN/ZrO₂ gate stacks. The Al–O (75.6 eV) and Al–N (74.6 eV) bonds were present in the Al₂O₃/ZrO₂ and AlN/ZrO₂ samples.

The HRTEM images of the ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ gate stacks are shown in Fig. 4. It is seen that the total physical thickness of all the samples was approximately 7.5 nm. The ZrO₂ layer featured a crystalline lattice in the images for all of the samples, which demonstrates that ZrO₂ was crystallized during thermal annealing.³⁶ In Fig. 4(a), there is an obvious and rough interfacial GeO_x layer in the ZrO₂ sample. An Al₂O₃ buffer layer and interfacial GeO_x are observed in the Al₂O₃/ZrO₂ gate stack as shown in Fig. 4(b). Fig. 4(c) reveals a sharp interface without interfacial GeO_x between AlN and Ge, indicating that the AlN buffer layer was capable of suppressing interfacial GeO_x. The HRTEM results are in good agreement with the Ge 3d XPS spectra shown in Fig. 2.

Fig. 4 Cross-sectional HRTEM images of the (a) ZrO₂, (b) Al₂O₃/ZrO₂, and (c) AlN/ZrO₂ gate stacks. There is a sharp interface between AlN and Ge in the AlN/ZrO₂ sample.

The GIXRD patterns of the ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ gate stacks are shown in Fig. 5. There is an obvious diffraction peak at 2θ = 30.4° in all of the samples, which corresponds to the tetragonal (101) phase (88-1007 JCPDS) or the cubic (111) phase (49-1642 JCPDS) of ZrO₂.³⁷ The tetragonal/cubic phase in ZrO₂ had a much higher dielectric constant than amorphous ZrO₂, which allows further CET scaling.³⁸ Although tetragonal/cubic ZrO₂ was thermodynamically stable at temperatures greater than 1170 °C,^39,40 the tetragonal/cubic phase had been observed in nanoscale ZrO₂ thin films.⁴¹ The presence of tetragonal/cubic ZrO₂ in the nanoscale layers, as shown in Fig. 5, is consistent with the results of previous studies.^36,41

Fig. 5 GIXRD patterns of ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ gate stacks. The obvious diffraction peak at 2θ = 30.4° is associated with tetragonal/cubic ZrO₂.

The C–V curves of the ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ gate stacks are shown in Fig. 6(a). The capacitance of the AlN/ZrO₂ sample was much greater than that of the ZrO₂ and Al₂O₃/ZrO₂ samples. Table 1 shows the CET and the effective dielectric constant (k_eff) of the ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ gate stacks. These values were derived from the C–V curves. The CET and k_eff values of the ZrO₂ sample are 2.35 nm and 12.42 nm, respectively. The k_eff of the ZrO₂ sample is much lower than the dielectric constant of tetragonal/cubic ZrO₂,^42,43 which can be deduced from the presence of low-K interfacial GeO_x in the ZrO₂ sample, as shown in the XPS spectrum and the HRTEM image (Fig. 2(d) and 4(a)). Although the low-K interfacial GeO_x was suppressed by the Al₂O₃ buffer layer, as demonstrated by the XPS spectrum (Fig. 2(e)), the introduction of the Al₂O₃ buffer layer between ZrO₂ and Ge still caused a slight degradation of the CET (2.38 nm) and k_eff (12.29) of the Al₂O₃/ZrO₂ gate stack because Al₂O₃ has a lower dielectric constant than ZrO₂. The substitution of AlN for Al₂O₃ as the buffer layer led to a significant decrease in the CET (1.75 nm) and an increase in the k_eff (16.72) of the AlN/ZrO₂ sample, which is mainly attributed to the suppression of the interfacial GeO_x, as clearly shown in Fig. 2(f) and 4(c).

Fig. 6 (a) C–V and (b) I–V characteristics of the ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ gate stacks. The AlN/ZrO₂ gate stack had a higher capacitance and a lower J_g than the ZrO₂ or Al₂O₃/ZrO₂ samples.

Table 1 The capacitance equivalent thickness (CET), the effective dielectric constant (k_eff), the gate leakage current (J_g), and the interfacial state density (D_it) of the ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ gate stacks^a

Sample	CET (nm)	keff	Jg (A cm^−2)	Dit (cm^−2 eV^−1)
a These values were extracted from the C–V, I–V, and G–f curves as shown in Fig. 6 and 7. The leakage current density Jg was determined at Vfb (flat-band voltage) +1 V. The AlN/ZrO2 gate stack exhibited a significant improvement in CET, keff, Jg, and Dit over the ZrO2 and Al2O3/ZrO2 samples.
ZrO2	2.35	12.42	1.82 × 10^−1	9.08 × 10^12
Al2O3/ZrO2	2.38	12.29	2.60 × 10^−2	6.46 × 10^12
AlN/ZrO2	1.75	16.72	1.12 × 10^−3	3.85 × 10^12

---

Fig. 6(b) and Table 1 show the I–V curves and the J_g of the ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ gate stacks. The ZrO₂ sample had a high J_g value of 1.82 × 10⁻¹ A cm⁻² as a result of the leakage current path via the grain boundaries in crystalline ZrO₂. Due to the high bandgap of Al₂O₃, the presence of the Al₂O₃ buffer layer resulted in a decrease in the J_g (2.60 × 10⁻² A cm⁻²) of the Al₂O₃/ZrO₂ sample. There was a significant reduction in the J_g (1.12 × 10⁻³ A cm⁻²) of the AlN/ZrO₂ gate stack, which is two orders of magnitude lower than that of the ZrO₂ sample because of the insertion of the AlN buffer layer. This can be understood from the suppressed growth of interfacial GeO_x, which has a low conduction band offset of only ∼0.8 eV.⁴⁴

The D_it value was measured by the conductance (Nicollian–Goetzberger) method.^45,46 Fig. 7(a) shows the equivalent parallel conductance over the angular frequency (G_P/ω) as a function of the frequency of the ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ gate stacks. The value of D_it was estimated from the maximum value of G_P/ω as:⁴⁵

where A is the area of the MOS capacitors. The D_it distributions in the bandgap of the ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ gate stacks are shown in Fig. 7(b). Since the weak inversion response results in the overestimation of D_it for Ge at room temperature,⁴⁷ the minimum D_it values near the mid-gap were adopted to characterize the interfacial quality of the ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ gate stacks.⁴⁸ As shown in Table 1, the minimum D_it of the ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ samples were 9.08 × 10¹² cm⁻² eV⁻¹, 6.46 × 10¹² cm⁻² eV⁻¹, and 3.85 × 10¹² cm⁻² eV⁻¹, respectively. Owing to the high thermal stability of the Al₂O₃ buffer layer and the reduced growth of interfacial GeO_x, the D_it of the Al₂O₃/ZrO₂ gate stack was lower than that of the ZrO₂ sample. The introduction of the AlN buffer layer caused a further decrease in D_it of the AlN/ZrO₂ gate stack, which is ascribed to the suppression of interfacial GeO_x by the AlN buffer layer. Because remote N₂/H₂ plasma was used as the reactant in the deposition of the AlN buffer layer, hydrogen passivation of the interfacial states also led to decrease in D_it of the AlN/ZrO₂ sample.⁴⁹

Fig. 7 (a) G_P/ω versus the frequency of the ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ gate stacks. The AlN/ZrO₂ gate stack had the lowest value of (G_P/ω)_max, and so the D_it of the AlN/ZrO₂ sample was lower than that of the ZrO₂ and Al₂O₃/ZrO₂ gate stacks. (b) D_it distributions versus bandgap energy, which were extracted from the conductance method, of the ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ gate stacks.

The C–V curves of the ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ gate stacks at various frequencies are shown in Fig. 8. In the frequency range from 1 kHz to 100 kHz, the frequency dispersion is 4.2%, 2.2%, and 1.4% for the ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ gate stacks at an accumulation bias of 2 V. The larger capacitance at a high frequency of 1 MHz was deduced from the parasitic inductance.⁵⁰ The relatively low-frequency dispersion in the AlN/ZrO₂ gate stack was correlated with the lower D_it as compared with the ZrO₂ and Al₂O₃/ZrO₂ samples.⁵¹

Fig. 8 The frequency dependence of the C–V curves of the (a) ZrO₂, (b) Al₂O₃/ZrO₂, and (c) AlN/ZrO₂ gate stacks. Low-frequency dispersion was observed in the AlN/ZrO₂ sample.

4. Conclusion

In this study, the electrical and structural characteristics of the ZrO₂, Al₂O₃/ZrO₂, and AlN/ZrO₂ gate stacks on Ge were investigated carefully. The introduction of an Al₂O₃ buffer layer between ZrO₂ and Ge resulted in a decrease of J_g value due to the high bandgap of Al₂O₃. However, it is difficult to prevent the growth of unstable, low-K interfacial GeO_x using an Al₂O₃ buffer layer because oxygen was involved during the deposition of Al₂O₃. The formation of interfacial GeO_x was significantly suppressed by an AlN buffer layer, as evidenced by the XPS and HRTEM characterizations. This produced a significant enhancement in the electrical characteristics, including the CET, k_eff, J_g, and D_it of the AlN/ZrO₂ gate stack. The results of this study show that an AlN buffer layer is an effective approach to high-quality interfacial engineering for high-performance high-K gate stacks in advanced Ge MOS transistors.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors gratefully acknowledge the financial support of the Taiwan Semiconductor Manufacturing Company (TSMC) and the Ministry of Science and Technology (MOST 107-2622-8-002-018), Taiwan.

References

1. S. Takagi, et al., IEDM'03 Technical Digest, IEEE International, 2003, pp. 3.3.1–3.3.4.
2. M. V. Fischetti and S. E. Laux, J. Appl. Phys., 1996, 80(4), 2234–2252.
3. K. Saraswat, C. O. Chui, T. Krishnamohan, D. Kim, A. Nayfeh and A. Pethe, Mater. Sci. Eng., B, 2006, 135(3), 242–249.
4. S. Takagi, M. Noguchi, M. Kim, S.-H. Kim, C.-Y. Chang and M. Yokoyama, et al., Solid-State Electron., 2016, 125, 82–102.
5. K. Prabhakaran, F. Maeda, Y. Watanabe and T. Ogino, Appl. Phys. Lett., 2000, 76(16), 2244–2246.
6. Y. Seo, T. I. Lee, C. M. Yoon, B.-E. Park, W. S. Hwang and H. Kim, et al., IEEE Trans. Electron Devices, 2017, 64(8), 3303–3307.
7. G. He, X. Chen and Z. Sun, Surf. Sci. Rep., 2013, 68(1), 68–107.
8. G. He, J. Gao, H. Chen, J. Cui, Z. Sun and X. Chen, ACS Appl. Mater. Interfaces, 2014, 6(24), 22013–22025.
9. G. He, B. Deng, H. Chen, X. Chen, J. Lv and Y. Ma, et al., APL Mater., 2013, 1(1), 012104.
10. G. He, J. Liu, H. Chen, Y. Liu, Z. Sun and X. Chen, et al., J. Mater. Chem. C, 2014, 2(27), 5299–5308.
11. Q. Xie, S. Deng, M. Schaekers, D. Lin, M. Caymax and A. Delabie, et al., Semicond. Sci. Technol., 2012, 27(7), 074012.
12. J. Zhang, G. He, L. Zhou, H. Chen, X. Chen and X. Chen, et al., J. Alloys Compd., 2014, 611, 253–259.
13. V. Afanas'ev and A. Stesmans, Appl. Phys. Lett., 2004, 84(13), 2319–2321.
14. Y. Fukuda, K. Kato, H. Toyota, T. Ono, Y. Nagasato and T. Ueno, Jpn. J. Appl. Phys., 2006, 45(9S), 7351.
15. R. Zhang, T. Iwasaki, N. Taoka, M. Takenaka and S. Takagi, Appl. Phys. Lett., 2011, 98(11), 112902.
16. P. Zimmerman, et al., IEDM'06, IEEE International, 2006, pp. 1–4.
17. C.-C. Lin, Y.-H. Wu, C.-Y. Wu and C.-W. Lee, IEEE Electron Device Lett., 2014, 35(3), 384–386.
18. S. Iwauchi and T. Tanaka, Jpn. J. Appl. Phys., 1971, 10(2), 260.
19. R. Zhang, T. Iwasaki, N. Taoka, M. Takenaka and S. Takagi, IEEE Trans. Electron Devices, 2012, 59(2), 335–341.
20. C. O. Chui, F. Ito and K. C. Saraswat, IEEE Electron Device Lett., 2004, 25(9), 613–615.
21. C. O. Chui, F. Ito and K. C. Saraswat, IEEE Trans. Electron Devices, 2006, 53(7), 1501–1508.
22. T. Maeda, T. Yasuda, M. Nishizawa, N. Miyata, Y. Morita and S. Takagi, Appl. Phys. Lett., 2004, 85(15), 3181–3183.
23. T. Maeda, M. Nishizawa, Y. Morita and S. Takagi, Appl. Phys. Lett., 2007, 90(7), 072911.
24. Y. Otani, Y. Itayama, T. Tanaka, Y. Fukuda, H. Toyota and T. Ono, et al., Appl. Phys. Lett., 2007, 90(14), 142114.
25. G. V. Rao, M. Kumar, T. Rajesh, D. R. K. Reddy, D. Anjaneyulu and B. Sainath, et al., Materials Today: Proceedings, 2018, 5(1), 650–656.
26. Y. J. Lee, J. Cryst. Growth, 2004, 266(4), 568–572.
27. A. Ahmed, A. Rys, N. Singh, J. Edgar and Z. Yu, J. Electrochem. Soc., 1992, 139(4), 1146–1151.
28. M. Morita, S. Isogai, K. Tsubouchi and N. Mikoshiba, Appl. Phys. Lett., 1981, 38(1), 50–52.
29. X.-H. Xu, H.-S. Wu, C.-J. Zhang and Z.-H. Jin, Thin Solid Films, 2001, 388(1–2), 62–67.
30. A. C. Jones, C. R. Whitehouse and J. S. Roberts, Chem. Vap. Deposition, 1995, 1(3), 65–74.
31. D. Riihelä, M. Ritala, R. Matero, M. Leskelä, J. Jokinen and P. Haussalo, Chem. Vap. Deposition, 1996, 2(6), 277–283.
32. P. Kelly and R. Arnell, Vacuum, 2000, 56(3), 159–172.
33. G. Thompson, P. Skeldon, X. Zhou, K. Shimizu, H. Habazaki and C. Smith, Aircr. Eng., 2003, 75(4), 372–379.
34. S. Kai Wang, H.-G. Liu and A. Toriumi, Appl. Phys. Lett., 2012, 101(6), 061907.
35. P. Motamedi and K. Cadien, Appl. Surf. Sci., 2014, 315, 104–109.
36. J.-J. Huang, L.-T. Huang, M.-C. Tsai, M.-H. Lee and M.-J. Chen, Appl. Surf. Sci., 2014, 305, 214–220.
37. C. Zhao, S. Taylor, M. Werner, P. Chalker, R. Murray and J. Gaskell, et al., J. Appl. Phys., 2009, 105(4), 044102.
38. D. Vanderbilt, X. Zhao and D. Ceresoli, Thin Solid Films, 2005, 486(1–2), 125–128.
39. J. Robertson, Rep. Prog. Phys., 2005, 69(2), 327.
40. S. Bang, S. Lee, S. Jeon, S. Kwon, W. Jeong and S. Kim, et al., J. Electrochem. Soc., 2008, 155(9), H633–H637.
41. C. Scanlan, M. Gajdardziska-Josifovska and C. Aita, Appl. Phys. Lett., 1994, 64(26), 3548–3550.
42. Y.-H. Wu, L.-L. Chen, R.-J. Lyu, M.-Y. Li and H.-C. Wu, IEEE Electron Device Lett., 2010, 31(9), 1014–1016.
43. R. Hegde, D. Triyoso, S. Samavedam and B. White Jr, J. Appl. Phys., 2007, 101(7), 074113.
44. J. Robertson and R. M. Wallace, Mater. Sci. Eng., R, 2015, 88, 1–41.
45. R. Engel-Herbert, Y. Hwang and S. Stemmer, J. Appl. Phys., 2010, 108(12), 124101.
46. E. H. Nicollian, J. R. Brews and E. H. Nicollian, MOS (metal oxide semiconductor) physics and technology, Wiley, New York, 1982.
47. K. Martens, C. O. Chui, G. Brammertz, B. De Jaeger, D. Kuzum and M. Meuris, et al., IEEE Trans. Electron Devices, 2008, 55(2), 547.
48. C.-C. Hsu, W.-C. Chi, Y.-H. Tsai, M.-L. Tsai, S.-Y. Wang and C.-H. Chou, et al., J. Vac. Sci. Technol., B, 2018, 36(5), 051204.
49. A. Nayfeh, C. O. Chui, K. C. Saraswat and T. Yonehara, Appl. Phys. Lett., 2004, 85(14), 2815–2817.
50. C.-M. Lin, H.-C. Chang, I.-H. Wong, S.-J. Luo, C. Liu and C. Hu, Appl. Phys. Lett., 2013, 102(23), 232906.
51. F. Ji, J. Xu, P. Lai, C. Li and J. Liu, IEEE Electron Device Lett., 2011, 32(2), 122–124.

---