Open Access
Add to BibSonomy
Abstract
Optical and thermo-optical studies of hydrogenated amorphous silicon-rich nitride films were carried out. The films were produced by plasma-assisted chemical vapor deposition on glass. It is shown that the films deposited under appropriately selected processing conditions contain little nitrogen, as confirmed by Fourier-transform infrared spectroscopy therefore they are referred to as silicon-rich nitrides, a-SRN:H. Spectroscopic ellipsometry, reflectance, and transmittance spectroscopy were used to determine the optical indexes of the films and their thicknesses. It results from the ellipsometric measurements performed within a 190-1700nm spectral wavelength range that a-SRN:H films exhibit a high refractive index of about 3.7. It is also shown that post-deposition annealing up to 300°C does not affect the optical parameters of the films. Additionally, they are transparent in the near-infrared region, which makes them a good candidate for applications in various optoelectronic systems.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Films of high refractive index are useful in many optical devices, such as imaging/optical sensors, emissive displays, optical filters, and selective multilayer filters. The use of such coatings makes it possible to set the transmittance and reflectance to the desired values. In the paper, we show that hydrogenated amorphous silicon nitride films of high silicon content (so-called silicon-rich nitride, a-SRN:H) deposited on glass (fused quartz and borosilicate) have a high refractive index, about 3.7, which makes them attractive for many optical applications.
Since the eighties, non-stoichiometric, hydrogenated, amorphous silicon nitride films of various nitrogen content, abbreviated as a-SiNx:H, have become essential for various optical applications [1,2]. They can be used as antireflective and passivating coatings for solar cells and in integrated optics [3–5]. They are also known for possessing interesting nonlinear properties which are achieved in an isotropic way [6–9]. Thermal stability is a plus.
The silicon nitride films are commonly grown by plasma-assisted chemical vapor deposition, PACVD [10], and magnetron sputtering techniques [11]. The PACVD method makes it possible to produce high-quality materials at relatively low temperatures. Therefore, it is attractive for substrates that cannot be exposed to high temperatures. The advantage here is also the possibility to tailor the properties of the material by changing its chemical composition, i.e. the contents of the elements, silicon, nitrogen, and hydrogen. To ensure a given chemical composition the deposition conditions have to be adjusted. Although all the processing parameters affect the structure and properties of the films, the flows of the reactive gases have the most impact on the chemical composition and hence the optical properties [5].
Samuelson and Mar measured the [Si]/[N] and [Si-H]/[N-H] ratios in films deposited by PACVD at various SiH4 and NH3 flows [12]. They showed that the refractive indexes of the films were directly correlated to the [Si-H]/[N-H] ratio but less to the [Si]/[N] ratio. Similar studies were performed by Claassen et al., [1]. The authors used the Rutherford backscattering spectrometry technique for [Si]/[N] measurements and made ellipsometric measurements of the refractive index at λ = 632.8 nm. Based on the results, in the experimental regime, they provided a linear dependence between the [Si]/[N] ratio varied between 0.6 and 1.7 and the refractive index increasing between 1.8 and 2.6:
Consistent research on the effect of the plasma processing parameters on the optical parameters of the PACVD silicon nitride films grown from SiH4, NH3 and N2 mixture on Si wafers was performed by Karouta et al., in [2]. The authors showed that the films deposited at higher silane flows contained less nitrogen and hydrogen, and they were silicon rich nitrides, SRN. Simultaneously, the refractive index increased. The possibility to grow silicon nitride films of high refractive index was confirmed by us but only in the case of films deposited on crystalline silicon [13].
Similar results, i.e. that the refractive index increases as the nitrogen content decreases has been confirmed by Debieu et al., [11] who studied various compositions of amorphous O- and H-free SiNx films deposited on silicon wafers by way of reactive sputtering and co-sputtering techniques.
The present research is aimed at studying the optical and thermo-optical properties of a-SRN:H films deposited on glass substrates, fused quartz and borosilicate glass. The films are fabricated by chemical vapor deposition operating at radio frequency, i.e. the RF-PACVD technique. The approach explored in this work is based on a combination of the technology of silicon-rich nitride films with ellipsometric studies in a wide spectral range and temperatures up to 300°C. We also show that post-deposition annealing does not affect the optical properties of the a-SRN:H films, which expands their possible applications. Such extensive research has not been performed for a-SRN:H, and such films have never been deposited on glass substrate.
2. Experimental
Non-stoichiometric amorphous hydrogenated silicon nitride films of high silicon content were fabricated by plasma-assisted chemical vapor deposition with the use of plasma excited at the radio-frequency of 13.56 MHz, rf-PACVD. The deposition was performed at the RF (radio frequency) power of 50 W, under the working pressure of about 53 Pa and at the temperature of 220°C. The films were deposited from gaseous silane SiH4 and ammonia NH3 at the flows of 40 sccm and 10 sccm, respectively, on fused quartz and borosilicate glass for optical studies, and on silicon wafers for Fourier-transform infrared measurements, Table 1. Before deposition, the substrates were preliminarily washed in acetone and isopropyl alcohol and then placed in the reactor on the anode and subjected to 10-minute cleaning in an argon plasma environment.
Table 1. Processing parameters used to deposit the a-SRN:H films on fused quartz, borosilicate glass and crystalline silicon.
The films were systematically characterized with the use of the following techniques.
- • Fourier transform infrared spectroscopy, FTIR, was applied to study the atomic structure. The measurements were carried out for the films deposited on silicon wafers with the use of a Bruker Vertex 70 vacuum spectrometer working in a transmittance mode with 4 cm-1 resolution. The FTIR, reflectance and transmittance spectra were measured at room temperature before annealing.
- • Atomic force microscopy, AFM, was used to provide the roughness of the films deposited on fused quartz and borosilicate glass. The measurements were made before annealing and repeated after annealing. Bruker AFM MULTIMODE 8 using the PeakForce tapping mode with a silicon tip was used.
- • Spectroscopic ellipsometry was used to provide the thicknesses, as well as optical and thermo-optical parameters of the layers. The measurements were made with the use of a Woollam Co., Inc, M-2000 J.A. RAE ellipsometer (rotating analyzer). The angles Ψ and Δ were measured at three different angles of incidence, 65, 70 and 75°, within a 190 ÷ 1700nm spectral wavelength range.
- • The models were fitted to the ellipsometric data with the use of the Complete EASE 6.0 software. The degree of compliance between the model and experimental data was evaluated by the mean-squared error (MSE) method.
- • For thermal ellipsometric studies, the samples were placed into a heating chamber. To remove the moisture, the chamber was connected to a single-stage oil vacuum pump (V-i160SV). The achieved vacuum was not controlled but it should be in the range of 20 - 30 Pa. During the measurement, the temperature was increased from 25 to 300°C for 180 minutes and then reduced to room temperature during the subsequent 180 minutes. The temperature was changed in an interval manner with the step of 20°C and stabilization periods, 10 min each.
- • Spectrometric measurements of the total transmittance and reflectance were carried out with the use of Perkin Elmer Lambda 900 equipped with an integrating sphere in the spectral range of 250 to 2500 nm.
3. Results and discussion
In Fig. 1, we present the FTIR spectrum for one, arbitrarily chosen a-SiNx-H film deposited on a crystalline silicon wafer, (001) oriented.
Fig. 1. The FTIR spectrum of the arbitrarily chosen a-SiNx:H film deposited on crystalline silicon at the conditions listed in Table 1.
The spectrum reveals the features typical for a-SiNx:H films. The strongest peak at about 825 cm-1 corresponds to the stretching of the Si-N bonds. A weaker band with the maximum at about 650 cm-1 can be assigned to the Si-H deforming vibrations. The Si-H stretching vibrations give the absorption at about 2015cm-1. The position of this band, ${\nu _{Si - H}}$, depends on the environs of the SiH group (N and/or Si atoms). Therefore, it corresponds to the nitrogen content. According to Afanasyev-Charkin et al., [14], every additional nitrogen atom bonded to the Si in Si-H increases the energy of the stretching Si-H vibrations by about 24 cm-1. This result has been confirmed by Weeber et al., [15], who have shown that when νSi-H is close to 2000cm-1, the film contains less nitrogen. To evaluate the nitrogen content in the a-SiNx:H film from the FTIR results we have used an empirical model presented and discussed in [10]. The model combines the frequency νSi-H with an average electronegativity of the atoms coordinating the SiH group. By fitting the model to the experimental value, νSi-H = 2015cm-1, we have calculated that the content of nitrogen in the film deposited in this research was extremely low, [N]/[S] = 0.08. Therefore, we call the layers hydrogenated amorphous silicon-rich nitride, a-SRN:H.
The roughness of the films on fused quartz and borosilicate glass was measured by way of AFM measurements. The provided roughness parameters are: Rq = 0.790 ± 0.090 nm, Ra = 0.603 ± 0.050 nm for the pristine film on fused quartz and Rq = 1.71 ± 0.54 nm, Ra = 1.22 ± 0.17 nm after annealing. For borosilicate glass the following parameters have been obtained: Rq = 0.633 ± 0.046 nm, Ra = 0.481 ± 0.029 nm for the pristine film on fused quartz and Rq = 1.59 ± 0.12 nm, Ra = 1.244 ± 0.079 nm after annealing. An increase of the roughness after post-deposition annealing can be caused by the molecular hydrogen release from the surface which appears at relatively low temperatures (the low-temperature effusion peak below 400°C) [16]. Although due to the annealing, the roughness has increased it remains low enough to confirm that the films are optically flat and there are no scattering loses.
The measured spectra of transmittance, reflectance and absorbance of the film deposited on fused quartz are presented in Fig. 2. As they are similar for all the deposited films, we show here the results for only one arbitrarily chosen film. Both reflectance and transmittance exhibit strong interference patterns at the wavelength above 500 nm. The results show that the films are opaque for light of photon energies higher than 2 eV. The main absorption edge at about 2.45 eV of the photon energy (510 nm) is observed. With the use of the envelope method [17] with regard to the reflectance spectra, a refractive index of 3.5 within the 700 to 2500 nm spectral range has been estimated.
Fig. 2. Spectral transmittance (T), reflectance (R) and absorbance (A) of the a-SRN:H film on fused quartz.
In the ellipsometric studies, the Tauc-Lorentz optical model was fitted to the experimental Ψ(λ) and Δ(λ) angles. In the first step, the Kramers-Krönig relations [18,19], were fitted to the data, in order to obtain physical consistency between the real and imaginary parts of the optical indexes, $n(E)$ and $k(E)$. The fitting was performed for three incident angles 60°, 65° and 75°, which made it possible to obtain better reliability of the used theoretical model. The Tauc-Lorentz’s model is based on the combination of the Tauc absorption edge [20,21] and oscillator broadening as given by the Lorentz oscillator [22]. The Tauc model predicts the absorption coefficient, α
where M contains all the constants described by the simple Gaussian model. The energy ETauc is the Tauc gap energy which is a measure of the band gap in the material. The power m = 2 describes the direct transition typical for amorphous materials.The results of fitting the Tauc-Lorentz model to the experimental data collected before and after the post-deposition annealing cycle at the incidence angle 70° are shown in Fig. 3. The fitting is perfect with MSE = ∼25.
Fig. 3. The ellipsometric angles, Ψ (Psi) and Δ (Delta), measured for the a-SRN:H film before (a) and after annealing (b) at the incidence angle 65° (stars) together with the Tauc-Lorentz model fitting (solid line).
The dispersion relations of the refractive index, n(E) and extinction coefficient, k(E), derived from the ellipsometric measurements for the a-SRN:H films on fused quartz and borosilicate glass are shown in Fig. 4 together with the data for crystalline silicon [10] and amorphous silicon on silicon wafer [23]. It is necessary to notice here that hydrogenated amorphous silicon-rich nitride exhibits high transparency for the wavelengths above 800 nm, i.e. in visible, infrared and red light, which is an advantage here.
Fig. 4. The dispersion spectra of (a) refractive index, (b) extinction coefficient of the a-SRN:H films on 1) fused quartz and 2) borosilicate glass in comparison with the data for 3) amorphous silicon, a-Si:H, film on silicon wafer and 4) crystalline silicon, c-Si.
To study the thermal behavior of the films, we performed ellipsometric measurements at the temperature increasing to 300°C and then cooling down to 25°C. Hence temperature changes of the refractive index and the extinction coefficient at heating and cooling have been determined at various wavelengths between 400 nm and 900 nm, Fig. 5. The results show perfect linear changes of both the n and k indexes upon the temperature increase and decrease. It is due to the thermal stresses appearing at higher temperatures. It is also seen that the values of the optical constants at heating and cooling overlap, which means, that all the changes due to the temperature variation are reversible. In particular, the dispersion functions measured after annealing overlap perfectly with those for pristine films and therefore they are not shown in Fig. 4. Moreover the thermal dependencies at different wavelengths are parallel. It means that the thermo-optical coefficients TOC = (∂n/∂T)λ=const [23–26], are the same at various wavelengths. The calculated values of TOC for λ = 900 nm are presented in Table 2.
Fig. 5. Temperature dependences of a) refractive index, b) extinction coefficient at chosen wavelengths for the a-SRN:H film on fused quartz, measured upon heating (pink) and cooling (blue) between 25°C and 300°C.
Table 2. Comparison of optical parameters of a-SRN:H films on glass and crystalline silicon.
In the studied temperature interval, the calculated TOCs are positive, TOC > 0, and constant, i.e. independent of temperature. In such a case, the TOC can be expressed as:
where $\Phi $ is the temperature coefficient of electric polarization, α is the thermal expansion coefficient and $f(n) = \,{{({{n^2} - 1} )\,\,({{n^2} + 2} )} / {6\,n}}$.As one may notice, the TOC is controlled by two factors, which compete to give positive or negative ${{dn} / {dT}}$. The contribution from the thermal expansion coefficient is negative because α is normally positive. Its contribution is usually small. The thermo-optical coefficient is positive when the polarization changes dominate over the volume change. The polarization changes are caused by thermal stresses. At the same time, the absorption increases.
It also results from the ellipsometric measurements that the studied films are thick of about 700 nm and more, Table 1. The post-deposition annealing does not affect the thickness.
By way of fitting the Tauc-Lorentz formula to the real and imaginary parts of the dielectric constant the optical gaps, Eg, for the studied films have been calculated, Table 2. Together, with Eg, we present the absorption features, amplitude, A and broadening, B. It is seen that the films on various substrates are similar. The only differences are various thicknesses of the films. This is due to the different adhesive properties of the substrates.
4. Summary
In the paper, we present the optical and thermo-optical properties of hydrogenated amorphous silicon-rich nitride films manufactured on glass in a plasma-assisted chemical vapor deposition process. The films are hydrogenated, amorphous, and they contain little nitrogen as confirmed in the FTIR studies and therefore are referred to as silicon-rich nitride. They exhibit a high refractive index with the average value of 3.7 and a lack of absorption within the 800-1700nm spectral range and therefore are good candidates for potential applications in the near infra-red region. The studied films are thermally stable within the temperature range of up to 300°C with a low thermo-optical coefficient, ∼7–9·10−4 1/K, the same at heating and cooling. The process of thermal treatment is reversible. The heating and cooling temperature dependences of the optical parameters do not show hysteresis. At room temperature, after the post-deposition annealing cycle, the constants reach the initial values. Additionally all the parameters are reached in an amorphous, isotropic way.
The films with such properties can be considered for future application in photonics, optoelectronics, and photovoltaics. They can also be important in the designing of multilayered optical systems for operations at various temperatures. In applications, the good transparency and thermal stability are superior to amorphous and crystalline silicon.
Disclosures
The authors declare that there are no conflicts of interest related to this article.
References
1. W. A. P. Claassen, W. G. J. N. Valkenburg, F. H. P. M. Habraken, and Y. Tamminga, “Characterization of Plasma Silicon Nitride Layers,” J. Electrochem. Soc. 130(12), 2419 (1983). [CrossRef]
2. F. Karouta, K. Vora, J. Tian, and C. Jagadish, “Structural, compositional and optical properties of PECVD silicon nitride layers,” J. Phys. D: Appl. Phys. 45(44), 445301 (2012). [CrossRef]
3. Y.-T. Lin and C. Chen, “Trap-state density in continuous-wave laser-crystallized single-grainlike silicon transistors,” Appl. Phys. Lett. 88(23), 233511 (2006). [CrossRef]
4. A. T. Hatzopoulos, N. Arpatzanis, D. H. Tassis, and C. A. Dimitriadis, “Electrical and noise characterization of bottom-gated nanocrystalline silicon thin-film transistors,” J. Appl. Phys. 100(11), 114311 (2006). [CrossRef]
5. J. F. Lelièvre, B. Kafle, P. Saint-Cast, P. Brunet, R. Magnan, E. Hernandez, S. Pouliquen, and F. Massines, “Efficient silicon nitride SiNx:H antireflective and passivation layers deposited by atmospheric pressure PECVD for silicon solar cells,” Prog. Photovoltaics 27(11), 1007–1019 (2019). [CrossRef]
6. J. Hong, W. M. M. Kessels, W. J. Soppe, A. W. Weeber, W. M. Arnoldbik, and M. C. M. van de Sanden, “Influence of the high-temperature ‘firing’ step on high-rate plasma deposited silicon nitride films used as bulk passivating antireflection coatings on silicon solar cells,” J. Vac. Sci. Technol., B: Microelectron. Process. Phenom. 21(5), 2123 (2003). [CrossRef]
7. D. S. Kim, S. G. Yoon, G. E. Jang, S. J. Suh, H. Kim, and D. H. Yoon, “Refractive index properties of SiN thin films and fabrication of SiN optical waveguide,” J. Electroceram. 17(2-4), 315–318 (2006). [CrossRef]
8. W. A. P. Claassen, W. G. J. N. Valkenburg, M. F. C. Willemsen, and W. M. Wijgert, “Influence of deposition temperature, gas pressure, gas phase composition and RF frequency on composition and mechanical stress of plasma silicon nitride layers,” J. Electrochem. Soc. 132(4), 893 (1985). [CrossRef]
9. H. Gleskova, S. Wagner, V. Gasparik, and P. Kovac, “Low-temperature silicon nitride for thin-film electronics on polyimide foil substrates,” Appl. Surf. Sci. 175-176, 12–16 (2001). [CrossRef]
10. M. Jurzecka-Szymacha, P. Boszkowicz, and K. Tkacz-Śmiech, “Silicon nitride layers of various N-content: technology, properties and structure,” Thin Solid Films 520(4), 1308–1312 (2011). [CrossRef]
11. O. Debieu, R. P. Nalini, J. Cardin, X. Portier, J. Perrière, and F. Gourbilleau, “Structural and optical characterization of pure Si-rich nitride thin films,” Nanoscale Res. Lett. 8(1), 31 (2013). [CrossRef]
12. G. M. Samuelson and K. M. Mar, “The Correlations Between Physical and Electrical Properties of PECVD SiN with Their Composition Ratios,” J. Electrochem. Soc. 129(8), 1773–1778 (1982). [CrossRef]
13. J. Jaglarz, M. Jurzecka-Szymacha, and S. Kluska, “Film materials based on a-SiNx:H with high refractive index obtained by plasma enhanced chemical vapour deposition technology,” Thin Solid Films 669, 564–570 (2019). [CrossRef]
14. I. V. Afanasyev-Charkin, L. G. Jacobson, R. D. Averitt, and M. Nastasi, “Amorphous silicon nitride films of different composition deposited at room temperature by pulsed glow discharge plasma immersion ion implantation and deposition,” J. Vac. Sci. Technol., A 22(6), 2342–2346 (2004). [CrossRef]
15. A. W. Weeber, H. C. Rieffe, W. C. Sinke, and W. J. Soppe, “Structural and Passivating Properties of SiNx:H Deposited Rusing Different Precursor Gases,” 19th European Photovoltaic Solar Energy Conference, 2004, France, 2004.
16. S. Jafari, J. Hirsch, D. Lausch, M. John, N. Bernhard, and S. Meyer, “Composition limited hydrogen effusion rate of a-SiNx:H passivation stack,” AIP Conf. Proc. 2149, 050004 (2019). [CrossRef]
17. J. C. Martinez-Anton and E. Bernabeu, “Interferential specrtrogoniometry for simultaneous determination on thickness and refractive index of films,” Opt. Commun. 132(3-4), 321–328 (1996). [CrossRef]
18. J. A. Woollam, CompleteEASE Data Analysis Manual (J.A Woollam Co, Inc., 2011).
19. J. Workman and A. Springsteen, Applied Spectroscopy, (Academic Press, New York, 1998).
20. G. E. Jellison and F. A. Modine, “Parameterization of the optical functions of amorphous materials in the interband region,” Appl. Phys. Lett. 69(3), 371–373 (1996). [CrossRef]
21. H. Shahrokhabadi, A. Bananej, and M. Vaezzadeh, “Investigation of Cody-Lorentz and Tauc-Lorentz Models in Characterizing Dielectric Function of (HfO2)x(ZrO2)1-x Mixed, Thin Film,” J. Appl. Spectrosc. 84(5), 915–922 (2017). [CrossRef]
22. J. Tauc, R. Grigorovici, and A. Vancu, “Optical Properties and Electronic Structure of Amorphous Germanium,” Phys. Stat. Sol. (b) 15(2), 627–637 (1966). [CrossRef]
23. J. Jaglarz, M. Jurzecka-Szymacha, K. Tkacz-Śmiech, and B. Sahraoui, “Optical characteristics of a-Si:H layers deposited by PACVD at various temperatures,” Opt. Mater. 39, 26–33 (2015). [CrossRef]
24. J. Jaglarz, M. Kępińska, and J. Sanetra, “Thermo-optical properties of 1H[3,4-b] quinolone films used in electroluminescent devices,” Opt. Mater. 36(8), 1275–1280 (2014). [CrossRef]
25. M. Marszałek, K. Śmiech, and J. Jaglarz, “Thermo-optical properties of hydrogenated amorphous a-C:N PACVD films deposited on silicon substrate,” Acta Phys. Pol., A 126(6), 1241–1245 (2014). [CrossRef]
26. E. E. Havinga, “The temperature dependence of dielectric constants,” J. Phys. Chem. Solids 18(2-3), 253–255 (1961). [CrossRef]
