Open Access
Add to BibSonomy
Abstract
Herein, X-ray photoelectron spectrometer (XPS), angle-resolved XPS (ARXPS), and atomic force microscopy (AFM) are used to study the surface changes of Ta2O5 bombarded by Ar+ ions with different energies. The results reveal that the Ar+ bombardment of Ta2O5 leads to a preferential sputtering of O atoms, which results in an imbalance in the Ta/O ratio on the material surface; and the formation of an “altered layer” composed of Ta2O5, Ta1+, Ta2+, Ta3+, and Ta4+. The Ta/O ratio increases from 0.34 to 0.55 with the sputtering time; however, it does not vary with ion energy. Before reaching a steady-state, the thickness of the altered layer increases with the sputtering time; however, after reaching a steady-state, the thickness of the altered layer does not exceed 3 nm. Concurrently, it increases with increasing sputtering energy. Further, AFM measurements reveal that low-energy Ar+ bombardment leads to a slight increased surface roughness, which does not exceed the initial value (0.41 nm) by 25%.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Ta2O5 has been widely used in optical thin films, the microelectronics industry, and heat-resistant coatings because of its low absorptivity, high refractive index, high dielectric constant, wide spectral transmission range (0.3–10 µm), excellent mechanical properties, and thermal stability [1–4]. TaOx has a wide range of applications in the field of oxide-based resistive random access memory(RRAM) owing to its excellent durability and low power consumption [5,6]. With the continuous development of science and technology, the requirements for the preparation of Ta2O5 and TaOx are continuously increasing. Therefore, accurate characterization of material structure has important guiding significance for material preparation and application.
Sputtering depth profiling is a composite measurement and analysis technique that combines ion sputtering and surface element composition characterization; it can be used to obtain the depth distribution of elements on thin films and characterize the interfacial bonding properties between layers in optical thin films [7–11]. This is an important test method for characterizing the structures of thin films. However, argon ion (Ar+) sputtering leads to preferential sputtering of O atoms on the Ta2O5 surface, causing the Ta/O ratio to deviate from reality. In addition, it forms an altered layer and changes the surface roughness of the sample, affecting the surface concentration and electron escape depth in the mixed area, which will lead to large errors between measurement and theoretical results [12–16]. Hofmann and Sanz were the first to perform XPS studies on the selective sputtering of O on Ta2O5. They concluded that preferential sputtering was caused by a large difference in the atomic masses and surface binding energies (BEs) between Ta and O [17]. Additionally, they determined the composition of the surface (Ta, TaO, TaO2, and Ta2O5) in steady-state (3 keV Ar+). Holloway bombarded Ta2O5 with 0.5–5 keV Ar+ and found that the preferential sputtering of O was most significant at 0.5 keV [18]. G. D. Wang found that under a low-energy argon ion beam bombardment of CeO2, an altered layer with a different composition was formed on the surface of the sample. This layer had a greater impact on the results of sputtering depth analysis, and its thickness increased from 0 to 0.7 nm over time [19]. S. Hofmann discussed the effect of preferential sputtering on the depth resolution and profile shape of sputtering depth profiles using a mixing-roughness-information-depth (MRI) model, and found that the surface altered layer and roughness variation had a significant effect on the sputtering depth profile test results and depth resolution [16,20,21]. Although scholars have carried out research on argon ion beam sputtering of Ta2O5, the research conditions have been relatively single, and the surface modification effect of different argon ion beam sputtering parameters on Ta2O5 has not been systematically summarized [22–27].
In this study, the effect of surface modification on Ta2O5 by Ar+ sputtering according to different parameters was systematically and comprehensively investigated using XPS, ARXPS, and AFM. The results provide error correction for Ta2O5 depth profiling measurements and data support for fabrication damage of the Ta2O5 material surface by ion beam processing. Simultaneously, we provide technical methods for the analysis of other preferential sputtered material surfaces.
2. Experimental
2.1 Sample preparation
Ta2O5 was deposited by double ion beam sputtering. Argon was ionized by a 16-cm ion source for the formation of an Ar+ beam. The ion beam was focused and accelerated through a grid and the Ta target was bombarded. The purity of the Ta target was 99.999%, and a 100-nm thick Ta2O5 film was deposited onto the clean substrates. The background vacuum was 2 × 10−6 Torr, vacuum was 3.4 × 10−4 Torr, oxygen content was 60 sccm, and the film thickness was controlled by varying the deposition time.
2.2 Experiment and characterization
Component analysis was conducted through X-ray photoelectron spectrometer (KAlpha, Thermo Scientific) using a monochromatic Al Kα source, with a characteristic emission line of 1486.6 eV. It was equipped with a vacuum system, and hemispherical energy analyzer with an energy resolution higher than 0.5 eV. The Ar ion gun in the X-ray photoelectron spectrometer was used to sputter Ta2O5 with 0.5, 1, and 2 keV energy ions at an incident angle of 30°. The ion gun scanning range was 4 mm ${\times} $ 2 mm. The x-ray spot diameter was 200 µm.
The energy scale is usually calibrated according to ISO 15472:2001, with an uncertainty of 0.2 eV or less. The method involves the use of copper, gold and silver samples. After calibration, the XPS device was set to use a pass energy of 30 eV, step size of 0.1 eV, and dwell time of 100 ms for obtaining the XPS spectra. Thermo Scientific Avantage software was used for peak fitting and quantitative analysis. The Gauss/Lorentz mix value was 30%. Dimension-3100 AFM was used to analyze the surface roughness over three repeated measurements.
3. Results and discussion
3.1 XPS spectra and analysis
Figure.1 shows the spectral evolution of the Ta 4f spectrum measured at different times with incident Ar + energies of 0.5 keV, 1 keV, and 2 keV after subtracting the Shirley background. With increased sputtering time, the Ta 4f peak gradually broadened, and two new characteristic peaks appeared at a low binding energy (BE). This change was due to the preferential sputtering of Ta2O5 under Ar+ bombardment, whereby Ta5+ was reduced to a Ta oxidizing state [24,28]. After 100 s, the Ta 4f spectra under different energy bombardments remained unchanged, which indicates that the surface of the material had reached a steady-state. The trends of each Ta 4f spectrum were similar and did not differ according to Ar+ bombardment with different energies.
Figure 2 shows similar depth profiles for Ta2O5 bombarded by different energies of Ar + . Ta/O ratio was calculated after a Shirley background subtraction and used Scofield sensitivity factors, modified to account for the instrument transmission function. Owing to the adsorption of C and O from the air by the surface of Ta2O5 during storage, the measured initial Ta/O ratio was found to be lower than the theoretical value of 0.4. The results show that the Ta2O5 ratio increased from 0.34 to 0.55 with sputtering time for different energies of Ar+ bombardment of Ta2O5. Finally, a steady-state was reached, and the Ta/O ratio was determined to have no strong relationship with ion energy. The relative atomic masses of Ta and O are different, thereby resulting in the preferential sputtering of O atoms. With an increase in sputtering time, the surface Ta atom concentration increased and exceeded the O atom concentration, and the final Ta/O ratio tended to be steady-state owing to interactions between the relative atomic masses and surface concentrations. To better understand the evolution of the Ta2O5 reduction, the Ta oxidation state composition of the surface should be further analyzed.
Fig. 1. Ta 4f spectra measured during Ar+ bombardment at different bombardment times: (a) 0.5 keV; (b) 1 keV; (c) 2 keV
To determine the chemical state of Ta after bombarding the Ta2O5 layer using different ion beam sputtering parameters, Thermo Scientific Avantage software was used for fitting the Ta 4f spectrum. Before sputtering, all spectra were charge referenced with C 1s = 284.8 eV, whereby the Ta 4f spectrum in Ta2O5 consisted of Ta 4f 7/2 = 26.2 eV and Ta 4f 5/2 = 28.1 eV with a spin-orbit separation of 1.9 eV (Fig. 3). This result is consistent with data for fully oxidized Ta2O5 in the literature [25,27]. After Ar+ sputtering, C was removed from the sample surface. From Fig. 1, it can be observed that Ta5+ is always present on the sample surface; therefore, Ta 4f 7/2 = 26.2 eV of Ta5+ was used as the charge reference for the oxide spectra [26]. When fitting the XPS spectrum, the Ta 4f7/2 and Ta 4f5/2 peaks of Tan + (0 ≤ n ≤ 5) were constrained to allow a spin-orbit splitting of 1.9 eV and peak area of 4:3. The full width at half maximum (FWHM) of the Tan+ (0 ≤ n ≤ 5) spin-orbit doublet and the Lorentz/Gaussian (L/G) mixing ratio were the same to ensure that the only difference between the sub-oxidation states was the peak area.
Fig. 3. Experimental and fitted spectra of Ta2O5 measured by different Ar+ ion sputtering: (a) 2 keV for 0s; (b) 2 keV for 16s; (c) 2 keV for 50s; (d) 0.5 keV for 300s; (e)1 keV for 300s; (f) 2 keV for 300s. The red dots represent the experimental spectrum, and the blue solid line represents the fitting envelope.
In this study, Ta1+, Ta2+, Ta3+, Ta4+, and Ta5+ were used to fit the Ta 4f peaks obtained after ion bombardment (Fig. 3). To obtain a reliable and correct fitting for the Ta 4f spectra at different sputtering stages, the BEs of all Tan+ should be known. The BEs of Tan+ used in this study are based on values that were previously reported in the literature [22,24]. The BE and FWHM of all Tan+ are listed in Table 1. Figure 3 shows the peak fits for the Ta 4f spectra according to different Ar+ sputtering energies. In the initial stage of sputtering (0–30 s), Ta5+ was dominant and Ta5+ was primarily reduced to Ta4+ and Ta3+. With an increase in the sputtering time (30–120 s), the suboxidation state of Ta was dominant, whereas the Ta3+ and Ta4+ amounts remained unchanged. For a sputtering time longer than 120 s, the surface of the sample reached a steady-state and the amount of each oxidation state of Ta remained essentially unchanged. In contrast, the percentage of Tan+ on the sample surface did not exhibit a strong relationship with the incident ion energy once the steady-state was reached, and this experimental result further verifies the conclusion shown in Fig. 2.
Table 1. XPS Ta 4f binding energies and FWHM employed in the peak fits of Ta2O5
3.2 ARXPS and analysis
In the conventional XPS test, the signal strength of a certain depth, x, is as follow:
where θ is the emission angle of the photoelectrons; and λ is the average free path of photoelectron inelastic scattering in the material, which is not dependent on the matrix material. ARXPS can obtain signal intensity changes from different depths, x, by varying the sweep angle, θ. Therefore, ARXPS can be used to non-destructively analyze data of solid surfaces and thin films. Figure 4 shows the peak fits for the Ta 4f spectra at different grazing angles. With an increase in grazing angle, the amount of Ta in the suboxide state increased, which proves that the etched layer is not uniformly distributed over the Ta2O5 surface.Fig. 4. XPS spectra measured at different glancing angles after 0.5 keV Ar + bombardment of Ta2O5: (a) 0°; (b) 55°.
Figure 5 shows the Ar ion ratio concentrations after Ar+ bombardment of the Ta2O5 surface at 0.5 keV for 300 s. After deconvolution of the Ta 4f peak, the Tan+/Ta5+ intensity ratio was determined to be a function of the photoelectron grazing angle. The results show that Ar+ bombardment led to the formation of an alteration layer composed of Ta2O, TaO, Ta2O3, and TaO2 on the Ta2O5 substrate surface. Assuming that the altered layer is the same as the photoelectron cross section of Ta2O5, the altered layer model can be expressed using Eq. (2), as [29]:
Fig. 5. After 300 s of sputtering with Ar+ at 0.5 keV, Ta 4f was deconvoluted, and the intensity ratio Tan+/Ta5+ (Tan+ = Ta° + Ta1+ + Ta2+ + Ta3+ + Ta4+) was obtained as a function of the emission angle. The solid line represents the best fit between the experimental data and the model from Eq. (2). The dotted line represents the best fit of ɛ = 1 in Eq. (2), that is, the surface is a uniform film.
3.3 AFM measure and analysis
In sputtering depth analysis, the surface roughness of the material is changed due to Ar+ sputtering; therefore, there is an error between the measurement information and the actual Ref. [21,32]. To further study the surface modification effect of Ar+ bombardment on Ta2O5, AFM was used to measure the surface roughness of the materials before and after Ar+ bombardment. Figure 7 shows the relationship between the Ta2O5 surface roughness and the Ar+ sputtering energy. After Ta2O5 was bombarded with Ar+, the surface roughness became slightly larger, the growth range did not exceed 25% of the initial roughness. The relationship between the Ta2O5 surface roughness and the Ar+ sputtering energy is presented in Fig. 7. These results reveal a change in the surface roughness of the material before and after argon ion sputtering and provide data support for optimizing the sputtering resolution.
4. Conclusion
The surface modification of Ta2O5 by Ar+ sputtering at different energies was studied using XPS, ARXPS, and AFM. The results revealed that the Ta/O ratio increased from 0.34 to 0.55 owing to preferential sputtering on the surface of Ta2O5. Additionally, an alteration layer composed of different valence oxides, such as Ta1+, Ta2+, Ta3+, Ta4+, and Ta5+, was formed on the surface. The thickness of the altered layer increased from 0 to 3 nm with sputtering time before reaching a steady-state; additionally, it increased with an increase in the incident Ar+ energy. Furthermore, AFM tests showed that Ar+ bombardment increased the surface roughness of Ta2O5, but that the range of increase in roughness did not exceed 25% of the original roughness.
Funding
Shanghai Institute of Optics and Fine Mechanics, Chinese Academy of Sciences.
Acknowledgments
The author thanks Shanghai Institute of Optics and Fine Mechanics Chinese Academy of Sciences for this work
Disclosures
The Authors declare no conflict of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. T. Sertel, N. A. Sonmez, S. S. Cetin, and S. Ozcelik, “Influences of annealing temperature on anti-reflective performance of amorphous Ta2O5 thin films,” Ceram. Int. 45(1), 11–18 (2019). [CrossRef]
2. C.-L. Tien, “Influence of ejection angle on residual stress and optical properties of sputtering Ta2O5 thin films,” Appl. Surf. Sci. 255(5), 2890–2895 (2008). [CrossRef]
3. A.-K. Chu, Y.-Y. Lu, and Y.-Y. Lin, “Fabrication and analysis of tantalum pentoxide optical waveguide resonator of high thermal stability,” Opt. Express 27(5), 6629–6639 (2019). [CrossRef]
4. S. Sathasivam, B. A. D. Williamson, A. Kafizas, S. A. Althabaiti, A. Y. Obaid, S. N. Basahel, D. O. Scanlon, C. J. Carmalt, and I. P. Parkin, “Computational and Experimental Study of Ta2O5 Thin Films,” J. Phys. Chem. C 121(1), 202–210 (2017). [CrossRef]
5. F. Pan, S. Gao, C. Chen, C. Song, and F. Zeng, “Recent progress in resistive random access memories: Materials, switching mechanisms, and performance,” Mater. Sci. Eng., R 83, 1–59 (2014). [CrossRef]
6. Y. B. Zhu, K. Zheng, X. Wu, and L. K. Ang, “Enhanced stability of filament-type resistive switching by interface engineering,” Sci. Rep. 7(1), 43664 (2017). [CrossRef]
7. S. Hofmann, “Compositional depth profiling by sputtering,” Prog. Surf. Sci. 36(1), 35–87 (1991). [CrossRef]
8. S. Hofmann, “Quantitative depth profiling in surface analysis: A review,” Surf. Interface Anal. 2(4), 148–160 (1980). [CrossRef]
9. R. Shimizu, “Preferential sputtering,” Nucl. Instrum. Methods Phys. Res., Sect. B 18(1-6), 486–495 (1986). [CrossRef]
10. T. Zeng, T. Zeng, M. Zhu, M. Zhu, M. Zhu, Y. Chai, J. Li, J. Shao, J. Shao, J. Shao, J. Shao, and J. Shao, “Dichroic laser mirrors with mixture layers and sandwich-like-structure interfaces,” Photonics Res. 9(2), 229–236 (2021). [CrossRef]
11. Q. Xiong, Q. Xiong, Q. Xiong, Q. Xiong, S. Huang, Z. Zhan, J. Du, J. Du, J. Du, J. Du, X. Tang, X. Tang, Z. Hu, Z. Liu, Z. Zhang, W. Chen, Y. Leng, Y. Leng, Y. Leng, and Y. Leng, “Surface ligand modified cesium lead bromide/silica sphere composites for low-threshold upconversion lasing,” Photonics Res. 10(3), 628–636 (2022). [CrossRef]
12. Yu. Kudryavtsev, “Ion beam sputtering of multicomponent targets: Surface composition change and cluster emission,” Tech. Phys. 57(2), 247–250 (2012). [CrossRef]
13. R. Kelly, “An attempt to understand preferential sputtering,” Nucl. Instrum. Methods 149(1-3), 553–558 (1978). [CrossRef]
14. X. Liang, X. Wang, J. Guo, and X. Wang, “Preferential sputtering on multicomponent optical surfaces,” Nucl. Instrum. Methods Phys. Res., Sect. B 410, 153–157 (2017). [CrossRef]
15. N. Zhang, X. Qu, X. Qu, Q. Lyu, K. Wang, K. Wang, K. Wang, X. W. Sun, X. W. Sun, and X. W. Sun, “Highly efficient transparent quantum-dot light-emitting diodes based on inorganic double electron-transport layers,” Photonics Res. 9(10), 1979–1983 (2021). [CrossRef]
16. S. Hofmann, Y. S. Han, and J. Y. Wang, “Depth resolution and preferential sputtering in depth profiling of sharp interfaces,” Appl. Surf. Sci. 410, 354–362 (2017). [CrossRef]
17. S. Hofmann and J. Sanz, “Quantitative XPS analysis of the Surface Layer of Anodic Oxides Obtained during Depth Profiling by Sputtering with 3 keV Ar+ ions,” J. Controlled Release 138(1), 40–44 (2009). [CrossRef]
18. P. H. Holloway and G. C. Nelson, “Preferential sputtering of Ta2O5 by argon ions,” 5 (1979).
19. G. D. Wang, D. D. Kong, Y. H. Pan, H. B. Pan, and J. F. Zhu, “Low energy Ar-ion bombardment effects on the CeO2 surface,” Appl. Surf. Sci. 258(6), 2057–2061 (2012). [CrossRef]
20. S. Hofmann, S. Y. Lian, Y. S. Han, Y. Liu, and J. Y. Wang, “Depth resolution and preferential sputtering in depth profiling of delta layers,” Appl. Surf. Sci. 455, 1045–1056 (2018). [CrossRef]
21. S. Hofmann, G. Zhou, J. Kovac, S. Drev, S. Y. Lian, B. Lin, Y. Liu, and J. Y. Wang, “Preferential sputtering effects in depth profiling of multilayers with SIMS, XPS and AES,” Appl. Surf. Sci. 483, 140–155 (2019). [CrossRef]
22. R. Simpson, R. G. White, J. F. Watts, and M. A. Baker, “XPS investigation of monatomic and cluster argon ion sputtering of tantalum pentoxide,” Appl. Surf. Sci. 405, 79–87 (2017). [CrossRef]
23. F. J. Himpsel, J. F. Morar, F. R. McFeely, R. A. Pollak, and G. Hollinger, “Core-level shifts and oxidation states of Ta and W: electron spectroscopy for chemical analysis applied to surfaces,” Phys. Rev. B 30(12), 7236–7241 (1984). [CrossRef]
24. S. Kasatikov, E. Filatova, S. Sakhonenkov, A. Konashuk, and A. Makarova, “Relationship between Ta oxidation state and its local atomic coordination symmetry in a wide range of oxygen nonstoichiometry extent of TaOx,” J. Phys. Chem. C 123(11), 6849–6860 (2019). [CrossRef]
25. M. V. Ivanov, T. V. Perevalov, V. S. Aliev, V. A. Gritsenko, and V. V. Kaichev, “Electronic structure of δ -Ta2O5 with oxygen vacancy: ab initio calculations and comparison with experiment,” J. Appl. Phys. 110(2), 024115 (2011). [CrossRef]
26. R. Grilli, R. Simpson, C. F. Mallinson, and M. A. Baker, “Comparison of Ar+ Monoatomic and Cluster Ion Sputtering of Ta2O5 at Different Ion Energies, by XPS: Part 1 - Monoatomic Ions,” Surf. Sci. Spectra 21(1), 50–67 (2014). [CrossRef]
27. R. Grilli, R. Simpson, C. F. Mallinson, and M. A. Baker, “Comparison of Ar+ Monoatomic and Cluster Ion Sputtering of Ta2O5 at Different Ion Energies, by XPS: Part 2 - Cluster Ions,” Surf. Sci. Spectra 21(1), 68–83 (2014). [CrossRef]
28. E. Atanassova, D. Spassov, A. Paskaleva, and K. Kostov, “Composition of Ta2O5 stacked films on N2O- and NH3-nitrided Si,” Appl. Surf. Sci. 253(5), 2841–2851 (2006). [CrossRef]
29. C. Palacio, P. Ocón, P. Herrasti, D. Díaz, and A. Arranz, “XPS and ARXPS study of silver underpotential deposition on platinum in acid solution,” J. Electroanal. Chem. 545, 53–58 (2003). [CrossRef]
30. G. V. Samsonov, The Oxide Handbook (Springer Science & Business Media, 2013).
31. S. Tanuma, C. J. Powell, and D. R. Penn, “Calculation of electron inelastic mean free paths (IMFPs) VII. Reliability of the TPP-2M IMFP predictive equation,” Surf. Interface Anal. 35(3), 268–275 (2003). [CrossRef]
32. S. Hofmann, “Sputter-depth profiling for thin-film analysis,” Philos. Trans. R. Soc., A 362(1814), 55–75 (2004). [CrossRef]
