Open Access
Add to BibSonomy
Abstract
In this work, optical constants of direct current (DC) and radio frequency (RF) co-sputtered Ti-SiO2 nanocomposite thin films were investigated and tuned by controlling the deposition power of Ti in the SiO2 host matrix. X-ray photoelectron spectroscopy (XPS) results confirmed that the metallic Ti was completely oxidized into different titanium oxide states while the Si4+ state was reduced to the Si2+ or Si0 state by observing the Ti 2p, O 1s and Si 2p line shapes changing under different deposition conditions. The optical constants of the composites were characterized with a spectroscopic ellipsometer (SE) and reduced by using the modified harmonic oscillator approximation (HOA) model. The results show that the metal-dielectric nanocomposite will have an advantage over natural materials because its optical properties of n and k can be properly tuned by adjusting the concentration of Ti in the Ti-SiO2 nanocomposites, thus satisfying the requirements of photonic device design and applications in broad spectral regions.
© 2017 Optical Society of America
1. Introduction
Extensive attention has been paid to metal-dielectric nanocomposite materials in the past decades because of the unique optical and electronic properties of these materials surpassing those of natural ones for potential applications including highly absorbing optical coatings for solar energy utilization [1–6], metamaterials operating in the optical frequency [7], sub-wavelength imaging structures [8], and photocatalysts [9]. When bringing the metal and dielectric materials together to form the nanocomposites, in principle, the electronic and optical properties of the metal and dielectric materials can be hybridized and easily tuned by controlling the metal nanoparticle filling fraction [2, 10]. Moreover, the effective dielectric functions of thin film metal-dielectric composites can even be tuned by controlling the deposition temperature [7]. The widely and relatively easily tailorable properties of metal-dielectric nanocomposites will be very useful for particular applications or design goals [2].
For the applications of nanocomposites in the optical field, accurate values of the optical constants of the composites must be known. Usually, the effective optical constants of nanocomposite thin films can be derived from reflectance and transmittance measurements [4] or spectroscopic ellipsometric (SE) parameters [2, 7, 10] with a suitable optical dispersion model. In consideration of the two or more phases existing in the metal-dielectric composites, effective medium approximation (EMA) theory [11] has often been used to describe the optical dispersion relations. However, it has been found that the standard EMA theory fails in describing the effective dielectric functions of Ag-SiO2 composites [7], which may be due to trying to describe the effective behavior of a complex mixture using only a simple mixing formula in approximation. Another approach to describe the effective optical constants of a metal-dielectric composite is to use a parametric formula by accounting for the dispersion of the effective optical constants with a commonly used dispersion model such as the classical Lorentz multiple oscillator model [7]. In consideration of the distributed resonances rather than a single resonance, the model can be improved further by employing the Gaussian-shaped distribution of the resonance [7].
Ti/SiO2 multilayered film structures are usually used as the solar absorber for solar-to-heat conversion [12–14]. Nevertheless, due to the limitations of the intrinsic optical constants of Ti and SiO2 materials existing in the nature, it tends to be required to add more Ti and SiO2 layers in the design of the metal-dielectric multilayered film structure to harvest more of solar energy in a much wider solar radiation spectral region [13, 15]. Accompanying with increasing of Ti-SiO2 layers in the Ti/SiO2-based multilayered film structure, the thickness of each individual Ti layer will decrease drastically [12, 13] to cause increasing difficulties for preparation of the Ti/SiO2-based multilayered film structures. However, this problem can be overcome by using a Ti-SiO2 nanocomposite with tunable optical constants instead of using only pure Ti film, which will greatly enhance the flexibility of the film structure design. In addition, Ti often exhibits very strong chemical activity, with different chemical states through oxidization during the co-sputtering process of Ti and SiO2 [16]. Hence, the chemical states of Ti, Si and O must be carefully studied before selecting an optical dispersion model properly approximating the Ti-SiO2 nanocomposite material.
In this work, the composition and chemical states of Ti-SiO2 nanocomposite thin films were first determined by X-ray photoelectron spectroscopy (XPS) analysis. Then, a modified harmonic oscillator approximation (HOA) model [17], derived from the quantum mechanical interpretation for electrons exciting from the valance band to the conduction band with different phases for each oscillator, was used to fit the ellipsometric parameters to obtain the effective optical constants of the composite thin films. The influence of the Ti concentration on the optical properties was also discussed.
2. Experimental details
Both of Ti and SiO2 were fabricated on the optically polished Si(100) substrate by using the electron beam assisted sputtering system (INFOVION, Seoul, Korea) at room temperature with a background pressure of 4.5 × 10−6 Torr. The direct current (DC) sputtering was used to deposit Ti, while SiO2 was deposited by the radio frequency (RF) sputtering method. The growth pressure was fixed at 2 × 10−3 Torr by a throttle valve with an argon (Ar) gas flow rate of 20 sccm (standard cubic centimeter per minute). The sputtering power for SiO2 was fixed at 200 W, while the sputtering power for Ti was set at 0 W, 30 W, 50 W, 80 W or 100 W to control the concentration of Ti in the nanocomposite thin films.
The compositions and chemical states of the films were measured by XPS analysis using a Kratos AXIS Ultra DLD spectrometer (Al Kα X-ray sources). High-resolution scans were performed over the energy ranges corresponding to the Ti 2p, Si 2p and O 1s element peaks. The XPS measurements were performed for the films after Ar ion sputtering for 90 s in vacuum to reduce surface contamination.
The surface morphologies of the film samples were characterized by atomic force microscopy (AFM Bruker Dimension Icon, ScanAsyst mode) in the tapping mode with a scanning area of 2.5 × 2.5 μm2. The film thickness of the samples was measured by using a step profiler (Bruker DektakXT).
A variable-angle spectroscopic ellipsometer (J.A. Woollam VASE) was used to characterize the optical properties of the composite thin films. The ellipsometric parameters (Ψ, Δ) were acquired over a wavelength range of 300-1200 nm at three different incident angles: 65°, 70° and 75°.
3. Results and discussions
3.1. XPS analysis of film compositions and chemical states
To determine the chemical element compositions and states of the Ti-SiO2 nanocomposites, XPS analysis was performed. Figure 1 shows the XPS survey spectra of the Ti-SiO2 film samples with Ti deposited at different sputtering powers. The surface contamination was removed by Ar ion etching for 90 s in vacuum, as confirmed by there being no C 1s peak observed in the spectra. Here, the XPS spectra for pure SiO2 are not shown, because we largely focus on the interaction between the Ti and the SiO2 host matrix. As sketched in the figure, the films are composed of Ti, Si, O and Ar. The weak Ar peaks seen in the spectra are due to the Ar ion etching. With the increase of Ti in the composites, the Ti signal intensity increases with the decreasing Si peak intensity shown in the spectra.
To extract the elemental ratios and states of Ti and Si, high-resolution XPS spectra of Ti 2p and Si 2p were peak-fitted, with the results shown in Fig. 2(a) and 2(b). The symbols represent the experimental data points, and the solid lines are reduced from the data curve fitting procedure.
Fig. 2 XPS spectra of (a) Ti 2p and (b) Si 2p at the Ti deposition powers (from top to bottom) of 30 W, 50 W, 80 W and 100 W.
For the composite fabricated at the Ti deposition power of 30 W, the Ti 2p spectra are decomposed into two components originating from the Ti4+ and Ti3+ states because of the asymmetric lines, as depicted in Fig. 2(a). The fitted energy difference between the Ti 2p1/2 and Ti 2p3/2 lines is approximately 5.57 eV, which somewhat corresponds to the reference value ΔEb = 5.67 eV for TiO2 [16, 18, 19]. The Ti3+ state is the intermediate oxidation state (Ti2O3) between TiO2 and silicon oxides. The peak at approximately 469.3 eV is attributed to the satellite state of Ti [19]. It can be further found that Ti is fully oxidized since no additional Ti 2p3/2 peak is observed at the binding energy value of metallic Ti at 454.1 eV [16]. With the increasing concentration of Ti, the Ti 2p profile is decomposed into the components of Ti4+ and Ti2+, demonstrating that the oxygen in the composite will be sufficient to support the formation of TiO but not that of Ti2O3. Moreover, no pure metallic Ti peak is observed, even when the concentration of Ti reaches its highest level at the Ti deposition power of 100 W.
From the chemical state analysis of Ti in the composite, all deposited Ti atoms react with O, forming the titanium oxides. Hence, it can be inferred that Si4+ should be reduced to the lower valence state due to the fixed oxygen content in the composite. This can be confirmed by detailed study of the high-resolution XPS spectrum of the Si 2p peak as shown in Fig. 2(b). For the composite with the lowest concentration of Ti as shown in Fig. 2(b), two peaks appear in the Si 2p region. The peak having lower intensity at the binding energy of 103.2 eV is attributed to the charged silicon atom (Si4+), while the higher intensity peak at 102.7 eV is assigned to the intermediate state (Sin+) [19]. When more Ti atoms are added into the SiO2 host matrix, an additional peak appears at 99.0 eV, which can be assigned to the neutral silicon atoms (Si0) [19, 20]. This result demonstrates that the Si4+ can be reduced to a lower chemical state by adding Ti atoms to the SiO2 host matrix, because Ti is less electronegative than Si [20].
O 1s spectra for Ti-SiO2 nanocomposites with Ti deposited at different powers are shown in Fig. 3. The peak can be de-convoluted into three components located at 530.8 eV, 531.6 eV and 532.3 eV, except for the composite with the lowest Ti concentration, for which only two components are observed, at 531.2 eV and 532.3 eV. These three distinct components are associated with Ti-O-Ti, Ti-O-Si and Si-O-Si bonds, respectively [19]. With the increasing concentration of Ti in the composite, the Ti-O-Ti peak intensity is enhanced.
Fig. 3 XPS spectra of O 1s with the Ti deposition powers (from top to bottom) of 30 W, 50 W, 80 W and 100 W.
The ratios of individual element concentrations in the Ti-SiO2 nanocomposites under different Ti deposition power conditions are also estimated from data reduction to fit the peak intensities of the Ti 2p, Si 2p and O 1s lines shown in the high-resolution XPS profiles. The results are listed in Table 1. With the increase in the Ti deposition power, the ratio of the relative amount of Ti in the Ti-SiO2 nanocomposites increases from 0 to 0.74. It can be seen that the atomic ratio for O/(Ti + Si) is approximately 1.91 rather than the 2 in the typical SiO2, which is because of oxygen loss during the sputtering deposition process [21].
Table 1. Ratios of individual element concentrations in Ti-SiO2 composites estimated through data reduction of the XPS analysis.
3.2. Film morphology
The surface morphologies of the nanocomposite samples are depicted in Fig. 4. The surface roughness of the films at different concentration of Ti is listed in Table 2. Here, we assume that the atomic ratio of the pure SiO2 host matrix is approximately equal to SiO1.91, as reduced from the high-resolution XPS analysis of the Ti-SiO2 composites. The surface roughness of the deposited samples is less than 1.0 nm, demonstrating the smooth surface of the composite thin films, which is in favor of the ellipsometry measurements and analysis.
Fig. 4 AFM images of the nanocomposite thin films (a) SiO1.91, (b) Ti0.38Si0.62O1.91, (c) Ti0.59Si0.41O1.91, (d) Ti0.70Si0.30O1.91 and (e) Ti0.74Si0.26O1.91.
Table 2. Surface roughness of the samples with different concentrations of Ti.
3.3. Optical analysis
Spectroscopic ellipsometry was used to investigate the effect of composition on the optical constants of the sputtered Ti-SiO2 nanocomposite thin films. The ellipsometric parameters (Ψ, Δ) were obtained in the wavelength range of 300-1200 nm at three incident angles: 65°, 70° and 75°.
In data reduction, a proper theoretical dispersion model should be chosen to obtain the optical constants of the nanocomposite films. However, effective medium theories are often used to depict the dispersive function of the materials composed of two or more media in different phases with the general assumption that the optical properties of each individual medium are not influenced by the other [2]. In this work, however, the incorporated Ti will be partially or fully oxidized in the Ti-SiO2 nanocomposites to have different optical constants depending on the film fabrication conditions in the experiment. To simplify the data fitting procedure, a modified HOA model [7, 17] was chosen to fit the experimental data.
The modified HOA model can be expressed as follows:
In Eq. (1), Aj, Ej and Γj are the amplitude, center energy and damping coefficient of the oscillator, respectively. The model was derived from the quantum mechanical interpretation and modified by adding a phase factor ϕ based on the consideration for each oscillator to have a different phase of ϕ j [17].A four-phase model consisting of air/roughness layer/Ti-SiO2 composite layer/Si substrate was employed in the SE parameter-fitting procedure. For the Ti-SiO2 layer, the modified HOA model with three oscillators was chosen. The Bruggeman EMA model was used to depict the roughness layer, which consists of the Ti-SiO2 and voids. To reduce the fitting parameters, the void fraction was set to 50%, and the rough-layer thickness was fixed to have a value of about 2.0 nm in approximation, corresponding to about two times of the measured surface roughness of the samples [22, 23]. The quality of fitting was verified by minimizing the difference between the measured data and the fitted results, defined by the root mean square error (RMSE) [24]:
where N is the number of the experimental data points, M is the number of parameters, σ is the standard deviation, and the superscripts mod and exp refer to the modeled and measured data, respectively.The film thickness for each sample obtained from the ellipsometric analysis was compared with the measured results from the step profiler measurements. As listed in Table 3, the fitted film thickness of each sample agrees well with the measured one, proving the accuracy of the data fitting.
Table 3. Measured and fitted film thickness of the Ti-SiO2 nanocomposites.
The measured and fitted ellipsometric parameters at the incident angle of 65° are presented in Fig. 5. The fitting of the Ti-SiO2 composites with different concentrations of Ti presents good agreement with the experimental data over the entire measured spectral range, with a factor of RMSE ≈ 0.37, demonstrating the accuracy and reliability of the data fitting procedure.
Fig. 5 Measured (symbol) and fitted (line) ellipsometry data (a) Ψ and (b) Δ of the composite films at an incident angle of 65°.
Figure 6 shows the optical constants reduced from the fitting of ellipsometric data using the model as described above. As sketched in Fig. 6, both the refractive index n and the extinction coefficient k can be properly tuned by adding Ti to the SiO2 host. With the increasing concentration of Ti in the composite, values of both n and k of the composites in the spectral range of 300-1200 nm are improved, which is largely due to the increasing fraction of the TiO, SiO and Si in the composites.
Fig. 6 Refractive index n (a) and extinction coefficient k (b) of the Ti-SiO2 nanocomposite thin films at different concentrations of Ti.
4. Summary
In conclusion, the optical properties of DC and RF co-sputtered Ti-SiO2 nanocomposite thin films were investigated in this work. By controlling the deposition power of Ti, the concentration of Ti in the composite can be effectively tuned. XPS analysis was used to determine the elemental compositions and states, with results showing that the metallic Ti atoms were completely oxidized into different titanium oxide states, while the Si4+ state was reduced to the chemical states of Si2+ or Si0. The optical properties of the composites were characterized by spectroscopic ellipsometry with the optical constants obtained by data fitting the ellipsometric parameters with the modified HOA dispersion model, rather than the EMA model, because of the interaction of Ti and SiO2 as demonstrated by XPS measurements. The results show that by varying the concentration of Ti in the Ti-SiO2 nanocomposites, both the refraction index n and the extinction coefficient k can be properly tuned to satisfy the required optical properties of materials for photonic device design and applications in broad spectral regions.
Funding
National Natural Science Foundation of China (Nos. 61427815, 61605089 and 61274054); the China Postdoctoral Science Foundation (No. 2017M611875); and the Innovation Program of SITP, CAS (No. CX-75).
References and links
1. F. Cao, K. McEnaney, G. Chen, and Z. Ren, “A review of cermet-based spectrally selective solar absorbers,” Energy Environ. Sci. 7(5), 1615–1627 (2014). [CrossRef]
2. S. Babar, A. U. Mane, A. Yanguas-Gil, E. Mohimi, R. T. Haasch, and J. W. Elam, “W:Al2O3 Nanocomposite Thin Films with Tunable Optical Properties Prepared by Atomic Layer Deposition,” J. Phys. Chem. C 120(27), 14681–14689 (2016). [CrossRef]
3. Z. Y. Nuru, D. E. Motaung, K. Kaviyarasu, and M. Maaza, “Optimization and preparation of Pt–Al2O3 double cermet as selective solar absorber coatings,” J. Alloys Compd. 664(6), 161–168 (2016). [CrossRef]
4. E. Wäckelgård, A. Mattsson, R. Bartali, R. Gerosa, G. Gottardi, F. Gustavsson, N. Laidani, V. Micheli, D. Primetzhofer, and B. Rivolta, “Development of W–SiO2 and Nb–TiO2 solar absorber coatings for combined heat and power systems at intermediate operation temperatures,” Sol. Energy Mater. Sol. Cells 133(0), 180–193 (2015). [CrossRef]
5. X. Xiao, G. Xu, B. Xiong, D. Chen, and L. Miao, “The film thickness dependent thermal stability of Al2O3:Ag thin films as high-temperature solar selective absorbers,” J. Nanopart. Res. 14(3), 746 (2012). [CrossRef]
6. A. Antonaia, A. Castaldo, M. L. Addonizio, and S. Esposito, “Stability of W-Al2O3 cermet based solar coating for receiver tube operating at high temperature,” Sol. Energy Mater. Sol. Cells 94(10), 1604–1611 (2010). [CrossRef]
7. J. Sancho-Parramon, “Tuning the effective dielectric function of thin film metal-dielectric composites by controlling the deposition temperature,” J. Nanophotonics 5(1), 051805 (2011). [CrossRef]
8. L. Shi and L. Gao, “Subwavelength imaging from a multilayered structure containing interleaved nonspherical metal-dielectric composites,” Phys. Rev. B 77(19), 195121 (2008). [CrossRef]
9. H. Sharma, R. Singhal, V. V. Siva Kumar, and K. Asokan, “Structural, optical and electronic properties of Ag–TiO2 nanocomposite thin film,” Appl. Phys. A-mater. 122(12), 1010 (2016).
10. J. Sancho-Parramon, V. Janicki, and H. Zorc, “On the dielectric function tuning of random metal-dielectric nanocomposites for metamaterial applications,” Opt. Express 18(26), 26915–26928 (2010). [CrossRef] [PubMed]
11. S. Bosch, J. Ferré-Borrull, N. Leinfellner, and A. Canillas, “Effective dielectric function of mixtures of three or more materials: a numerical procedure for computations,” Surf. Sci. 453(1–3), 9–17 (2000). [CrossRef]
12. X. F. Li, Y. R. Chen, J. Miao, P. Zhou, Y. X. Zheng, L. Y. Chen, and Y. P. Lee, “High solar absorption of a multilayered thin film structure,” Opt. Express 15(4), 1907–1912 (2007). [CrossRef] [PubMed]
13. M. H. Liu, E. T. Hu, Y. Yao, K. Y. Zang, N. He, J. Li, Y. X. Zheng, S. Y. Wang, O. Yoshie, Y. Lee, C. Z. Wang, D. W. Lynch, and L. Y. Chen, “High efficiency of photon-to-heat conversion with a 6-layered metal/dielectric film structure in the 250-1200 nm wavelength region,” Opt. Express 22(S7), A1843–A1852 (2014). [CrossRef] [PubMed]
14. J. N. Chen, J. P. Guo, and L. Y. Chen, “Super-wideband perfect solar light absorbers using titanium and silicon dioxide thin-film cascade optical nanocavities,” Opt. Mater. Express 6(12), 3804–3813 (2016). [CrossRef]
15. N. P. Sergeant, O. Pincon, M. Agrawal, and P. Peumans, “Design of wide-angle solar-selective absorbers using aperiodic metal-dielectric stacks,” Opt. Express 17(25), 22800–22812 (2009). [CrossRef] [PubMed]
16. V. Stranak, S. Drache, H. Wulff, Z. Hubicka, M. Tichy, A. Kruth, C. A. Helm, and R. Hippler, “Oxidation behavior of Cu nanoparticles embedded into semiconductive TiO2 matrix,” Thin Solid Films 589(9), 864–871 (2015). [CrossRef]
17. F. L. Terry Jr., “A modified harmonic oscillator approximation scheme for the dielectric constants of AlxGa1-xAs,” J. Appl. Phys. 70(1), 409–417 (1991). [CrossRef]
18. S. Y. Lee, C. Jeon, S. H. Kim, Y. Kim, W. Jung, K.-S. An, and C.-Y. Park, “In-situX-ray Photoemission Spectroscopy Study of Atomic Layer Deposition of TiO2 on Silicon Substrate,” Jpn. J. Appl. Phys. 51(3), 1102 (2012).
19. I. Iatsunskyi, M. Kempiński, G. Nowaczyk, M. Jancelewicz, M. Pavlenko, K. Załęski, and S. Jurga, “Structural and XPS studies of PSi/TiO2 nanocomposites prepared by ALD and Ag-assisted chemical etching,” Appl. Surf. Sci. 347(0), 777–783 (2015). [CrossRef]
20. R. Methaapanon and S. F. Bent, “Comparative Study of Titanium Dioxide Atomic Layer Deposition on Silicon Dioxide and Hydrogen-Terminated Silicon,” J. Phys. Chem. C 114(23), 10498–10504 (2010). [CrossRef]
21. T. Pencheva and M. Nenkov, “The properties of barium titanate RF sputtered in argon,” Vacuum 48(1), 43–48 (1997). [CrossRef]
22. Z. Song, B. R. Rogers, and N. D. Theodore, “Spectroscopic ellipsometry characterization of ZrO2 films on Si(100) deposited by high-vacuum-metalorganic chemical vapor deposition,” J. Vac. Sci. Technol. A 22(3), 711–718 (2004). [CrossRef]
23. E. T. Hu, R. J. Zhang, Q. Y. Cai, Z. Y. Wang, J. P. Xu, Y. X. Zheng, S. Y. Wang, Y. F. Wei, R. Z. Huang, and L. Y. Chen, “Study of the thickness effect on the dielectric functions by utilizing a wedge-shaped Ti film sample with continuously varied thickness,” Appl. Phys A-mater. 120(3), 875 (2015).
24. E. T. Hu, Q. Y. Cai, R. J. Zhang, Y. F. Wei, W. C. Zhou, S. Y. Wang, Y. X. Zheng, W. Wei, and L. Y. Chen, “Effective method to study the thickness-dependent dielectric functions of nanometal thin film,” Opt. Lett. 41(21), 4907–4910 (2016). [CrossRef] [PubMed]
