Surface & Coatings Technology 228 (2013) S210–S213 Contents lists available at SciVerse ScienceDirect Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat Synthesis and characterization of zirconium oxynitride coatings deposited by filtered cathodic vacuum arc technology Yi-Ming Chen, Bin Liao, Xian-Ying Wu, Hui-Xing Zhang, Xu Zhang ⁎ Key Laboratory of Beam Technology and Material Modification of Ministry of Education, College of Nuclear Science and Technology, Beijing Normal University, Beijing 100875, PR China a r t i c l e i n f o Available online 13 June 2012 Keyword: Filtered arc Zirconium oxynitride Air Mechanical property a b s t r a c t Thin films of zirconium oxynitride (ZrOxNy) were deposited onto glass and Si substrate at room temperature by filtered cathodic vacuum arc (FCVA) technology using air as a reactive gas. The compositions and structures of the zirconium oxynitride films influenced by air flow rate were investigated by scan electronic spectroscopes, X-ray diffraction and X-ray photoelectron spectroscopes. The results showed that crystal structure of the films transformed from ZrO and ZrN mixed phases to ZrN phase with the increasing air flow rate. The hardness, elastic modulus and elastic recovery parameter (ERP) of the zirconium oxynitride films were also determined by nano indentation tests. At the optimum deposition parameters, the hardness and elastic modulus values reached 28.94 GPa and 253.44 GPa respectively; good wear resistant properties were also achieved due to the high H/E ratio (0.115) and good elastic recovery (86.18%). The optical band gap of the film at the optimum deposition parameter was 1.91 eV, implying that the film had great potentials in many photocatalytic and optoelectronic applications. © 2012 Elsevier B.V. All rights reserved. 1. Introduction Transition metal oxynitride film, especially zirconium oxynitride (ZrOxNy), had attracted considerable interests in the last few years due to their remarkable properties. ZrOxNy thin films had been widely applied in gate dielectrics [1], temperature sensor element in magnetic fields up to 32 T operating at temperatures between 2 and 286 K [2], corrosion resistance coatings [3], and decorative films such as eyeglass frames, wristwatch casings, and wristbands [4]. The appearance of oxygen during deposition process leads to a formation of ionic metal–oxygen bonds in a matrix of covalent metal–nitrogen bonds; it generates a new structure (metal–oxynitride) with different properties [5]. Therefore, by variation of the oxygen/nitrogen ratio it allows us to tailor the structural, optical, mechanical and electronic properties of the transition metal oxynitride film. Various techniques were used to prepare zirconium oxynitride films like radiofrequency (RF) reactive magnetron sputtering, direct current (DC) reactive magnetron sputtering, and reactive cathodic arc evaporation [6]. Traditionally, N2/O2 mixtures were employed as reactive gases during deposition; other gases had also been used such as water vapor–nitrogen atmosphere [7] and air [8]. It is worth mentioning that air as a reactive gas can reduce the processing time because of the low vacuum. Reactive cathodic arc evaporation was adopted by M. Laurikaitis et al. to prepare ZrNxOy films, however, macroparticles were found in their work [19]. Filtered cathodic vacuum arc (FCVA) technology, free of ⁎ Corresponding author. Tel.: + 86 10 6220 8249; fax: + 86 10 6223 1765. E-mail address: zhangxu@bnu.edu.cn (X. Zhang). 0257-8972/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.surfcoat.2012.06.015 macroparticle contamination, had been widely confirmed as an excellent method for deposition films with better qualities [9]. So in our present work, ZrOxNy thin films were prepared by FCVA technology using air as a reactive gas. Furthermore, we investigated the effect of air flow rate on the compositional, structural, optical and mechanical characteristics of the ZrOxNy films. 2. Experimental details Zirconium oxynitride films were deposited onto glass and Si (100) substrate, a 90 mm in diameter zirconium plate of 99.8% purity was used as the cathodic arc source. The plasma was produced and then introduced into the processing chamber by 90° bent magnetic filter duct. This procedure removed the unwanted neutral particles and the macroparticles. The base pressure in the processing chamber was about 3 × 10 − 2 Pa during the experiment. Before deposition the zirconium was pre-implanted to get a zirconium seed layer, which could enhance the ZrOxNy thin films' adherence force to the substrate. The deposition parameters, such as substrate bias voltage (200 V), filter coil current (2.0 A), and substrate temperature (room temperature) were kept the same except for the air flow rates. The crystal structure of the films was determined by grazing incidence X-ray diffraction (improved D/max-RB type)(θ=1°) operated at 40 kV and 200 mA with a Cu Ka (λ=0.154 nm) excitation source. The average grain size of film (D) was calculated from the full width at half maximum (FWHM) using Scherrer's formula: D¼ kλ : B cosθ Y-M. Chen et al. / Surface & Coatings Technology 228 (2013) S210–S213 S211 where constant k=0.89, B is the FWHM (degree), λ is the wavelength (nm), and θ is the diffraction angle (degree). The thickness and curvature of the film were measured by the surface morphology device with the pattern of Talysurf 5P-120 from Rank Taylor Hobson. The elemental composition of the film was obtained from the energy disperse spectroscopy (EDS) on the Hitachi S-4800 system and X-ray photoelectron spectroscopy (XPS) on VG ESCALAB MKII spectrometer. Film's hardness and Young's modulus were determined from the loading and unloading curves on MML NanoTest UK. The maximum load used was 2 mN, with a loading time of 20 s, holding 5 s, and unloading in 20 s, producing an average number of 9 indentations per sample. UV–visible spectroscopy (SPECORD 200) was employed to investigate the absorption spectra of the film over the spectral range of 200–800 nm, the optical band gap was calculated from obtained spectra. 3. Results and discussions 3.1. Crystal structure Grazing incidence X-ray diffraction analysis was carried out to investigate the phase of deposited films. Fig. 1 showed the GIXRD patterns of the ZrOxNy films prepared with air flow rate varying from 10 to 70 sccm. At low air flow rate (10–30 sccm), the (200), (220) and (311) peaks of ZrO [PDF#20-0684] were observed, when the air flow rate increased, those peaks shifted slightly toward higher angles and transformed into (200), (220) and (311) of the ZrN [PDF#65-0961]. The atomic concentrations of films varying with air flow rate were shown in Fig. 2(a). Generally speaking, the formation of zirconium oxide was more favorable than zirconium nitride. However, in our present work, it seemed to be contrary except for the low flow rates (10–30 sccm), which was consistent with the results reported by Mu-Hsuan Chan [8]. The atomic concentrations in Fig. 2(a) were well consistent with the above-mentioned analysis. The content of oxygen increased and that of nitrogen decreased as the air flow rate varies from 10 to 30 sccm. On the contrary, at high air flow rate ranging from 40 to 70 sccm, the concentration of nitrogen enlarged rapidly. J. M. Ngaruiya et al. deduced that formation of ZrO2 reduced the Gibbs free energy the most, it meant that formation of ZrO2 was expected to be observed [10], however, there were no remarkable peaks of ZrO2 that exist in all our XRD patterns. From thermodynamic point of view, ZrO2 (ΔGf,0ZrO2: −1091.6 kJ/mol) is energetically more favorable than ZrN (ΔGf,0ZrN: −364.3 kJ/mol) [8]. While in our work, the nitridation of Zr is easier than the oxidation especially at high air flow rate, which may be attributed to molecule kinetics theory [8]. Oxygen Fig. 2. (a) The atomic concentrations of films varying with air flow rate. (b) The average grain size of deposited films varying with air flow rate. electronegativity is 3.5 and it is larger than that of nitrogen (3.0), which indicates that oxygen is more likely to attract electrons than nitrogen. A similar phase transition to ZrN was also found in Rizzo's work [12]. The mechanism of phase formation and transition is a complex issue, and it needs to be further investigated. The average grain size of deposited films was measured and also showed in Fig. 2(b). The average grain size increased from 5 nm to 11 nm as the air flow rate increases from 20 sccm to 70 sccm. The highest deposition rate of 160 nm·min− 1 was 10 times as fast as the data reported in the other paper [8]; high efficiency was one of the advantages of the FCVA technology [9]. 3.2. Chemical composition Fig. 1. The GIXRD patterns of the ZrOxNy films prepared with air flow rate varying from 10 to 70 sccm. The XPS technique was used to identify elements and their chemical states present in the outermost surface of zirconium oxynitride film. The XPS spectra of Zr 3d for the air flow rate of 10, 40 and 70 sccm were shown in Fig. 3. The Zr 3d peak had been decomposed into three components, corresponding to the nitride, oxynitride and oxide phase of zirconium. The oxynitride component (labled 2) was located in the BE range of 181.1–182.1 eV [7], the doublet labled 1 was associated with nitride phase which was positioned in BE ~178.8 eV [11], the third doublet 3 had the Zr 3d centered at 182.2 eV, which represented the oxide [12]. Peak positions of Zr 3d deconvoluted components and their percentage were collected in Table 1. As shown in Table 1, the percentage of nitride increased from 4.52% to 22.37% as the air flow rate increases from 10 to 70 sccm, which was consistent with the XRD analysis and atomic concentration results. Furthermore, a shift of about 1.0 eV of S212 Y-M. Chen et al. / Surface & Coatings Technology 228 (2013) S210–S213 30 70sccm Hardness 28 2 10 internal stress 3 Hardness/GPa Intensity /arb.units 26 40sccm 2 3 1 10sccm 8 24 6 22 20 4 18 2 2 16 3 1 10 180 185 40 50 60 70 0 Fig. 4. The hardness and internal stress of ZrOxNy films as a function of the air flow rate. Fig. 3. The Zr3d XPS spectra of the ZrOxNy films prepared with air flow rates of 10, 40 and 70 sccm, labels 1, 2 and 3 represented nitride phase, oxynitride phase and oxide phase, respectively. ZrN peaks to higher binding energy was observed. A similar tendency was reported by P. Carvalho et al. [11], the reason was correlated to the change in the ionic valences induced by the chemical environment. 3.3. Mechanical property and optical band gap The thickness of ZrOxNy thin film prepared with air flow rate varying from 10 to 70 sccm was within the range of 0.378–1.26 μm. The internal stress of the film (σ) can be calculated by the Stoney formula: σ¼ 30 air flow rate/sccm 190 Binding Energy /eV   1 Es t2s 1 1 ; − 6 ð1−νs Þt Rn R0 where Es (180 GPa) is Young's modulus of the silicon substrate, ts (0.5 mm) is the thickness of silicon substrate, νs (0.26) is the Poisson's ratio, t is the thickness of deposited film, and R0, and Rn are the radii of curvature before and after deposition respectively, usually R0 approximately equals to infinity. Fig. 4 showed the hardness and internal stress of ZrOxNy films as a function of the air flow rate, with the air flow rate increased from 10 to 60 sccm; the hardness of films increased to the maximum value of 28.94 GPa, and further increase of air flow rate would decrease the hardness. The situation for internal stress was almost the same except for the sample at air flow rate of 50 sccm. Based on the above analysis, the content of ZrN phase enriched with the increase of air flow rate, and covalent nitrides such as ZrN were the more attractive candidates for achieving high hardness than ionic compounds since electrostatic interactions were omnidirectional [13]. On the other hand, the average grain size in Fig. 2(b) showed the size of 9 nm in the 60 sccm, grain growth was also playing an important role in hardness [4]. The highest internal stress in 60 sccm could be associated with the incorporation of oxygen within the ZrN lattice, which would induce lattice defects in the film structure acting as obstacles for the dislocation motion, resulting in the enhanced hardness [14]. The hardness/elasticity (H/E) ratio was used to describe the failure mechanism of the materials. The ratio H/E of ZrOxNy films was measured between 0.088 and 0.115. The elastic recovery parameter (ERP) values were also determined in the nanoindentation test. Fig. 5 showed the elastic modulus and ERP of ZrOxNy films as a function of air flow rate. As can be seen from the figure, the elastic modulus and ERP value reached the maximum simultaneously when the air flow rate was 60 sccm. The large H/E ratio (0.115) and ERP value (0.8618) indicated good wear resistant properties of the ZrOxNy thin film prepared at the air flow rate of 60 sccm [15,16]. The UV–visible absorption spectra of the film prepared with the air flow rate of 60 sccm, filter coil current of 2.0 A and bias voltage of 200 V, were measured, as shown in Fig. 6(a). The optical band gap Eg of the film was determined by the following formula [20],  n αhν∝ hν−Eg Where α is the absorption coefficient related to A = αd [A is the absorption intensity, d is the film's thickness [17], hνis the incident photon energy, Eg is the optical band gap energy, and n, a number which characterizes the transition process, equals 2 for ZrOxNy [18]]. Plots of (αhν)1/2 versus hν for ZrOxNy film are shown in Fig. 6(b). Value of Eg is evaluated by extrapolating the linear portion of the curve to (αhν)1/2 =0. The Eg 260 1.2 elastic modulus 1.1 250 ERP Elastic modulus/GPa 175 20 240 1.0 230 0.9 0.8 220 ERP 1 12 internal stress/GPa Zr-3d 0.7 210 0.6 200 Table 1 Peak positions of Zr3d deconvoluted components and their percentage. 0.5 190 Air flow rate/ sccm 1 2 3 Zr–N % Zr–O–N % Zr–O % 10 40 70 178.31 178.85 179.59 4.52 8.46 22.37 181.51 181.18 181.77 68.73 77.09 62.88 182.18 182.11 182.22 26.75 14.45 14.75 0.4 180 10 20 30 40 50 60 70 air flow rate/sccm Fig. 5. The elastic modulus and ERP of ZrOxNy films as a function of air flow rate. Y-M. Chen et al. / Surface & Coatings Technology 228 (2013) S210–S213 a 3.0 ZrN phase as the air flow rate varying from 10 to 70 sccm. The XPS results were consistent with the change of the XRD patterns. Hardness measurements showed that the optimum experiment parameters were determined, hardness and elastic modulus of ZrOxNy films reached 28.94 GPa and 251.44 GPa respectively. The phase transition, average grain size and internal stress were the factors to interpret the hardness data. Good wear resistant properties were also achieved due to the high H/E ratio (0.115) and good elastic recovery (86.18%). The optical band gap of the film at the air flow rate of 60 sccm was 1.91 eV, implying that the film had great potentials in many photocatalytic and optoelectronic applications. 2.5 Abs 2.0 1.5 1.0 0.5 Acknowledgment 0.0 200 300 400 500 600 700 This work was financially supported by the National Science Foundation of China under Grant no. 10975020. 800 λ/ nm References b 2000 1/2 ( αhν) /[eV*cm-1]1/2 S213 1500 1000 500 0 1 2 3 4 5 6 7 hν/eV Fig. 6. (a) The UV–visible absorption spectra of the film prepared with the air flow rate of 60 sccm; (b) Plots of (αhν)1/2 versus hν for ZrOxNy film at the air flow rate of 60 sccm. value of the film is 1.91 eV, it is close to the value reported in the literature (Eg = 1.94 eV), which is possibly due to the high nitrogen incorporation in the film [8]. 4. Conclusions Zirconium oxynitride thin films were deposited by using filtered cathodic vacuum arc technique in the presence of air gas. The crystal structure of the films transformed from ZrO and ZrN mixed phases to [1] R.E. Nieh, C.S. Kang, H.J. Cho, K. Onishi, R. Choi, S. Krishnan, J.H. Han, Y.H. Kim, M.S. Akbar, J.C. Lee, IEEE Trans. Electron Devices 50 (2003) 333. [2] B.L. Brandt, D.W. Liu, L.G. Rubin, Rev. Sci. Instrum. 70 (1999) 104. [3] E. Ariza, L.A. Rocha, F. Vaz, L. Cunha, S.C. Ferreira, P. Carvalho, L. Rebouta, E. Alves, P. Goudeau, J.P. Riviere, Thin Solid Films (2004) 274. [4] P. Carvalho, F. Vaz, L. Rebouta, L. Cunha, C.J. Tavares, C. Moura, E. Alves, A. Cavaleiro, P. Goudeau, E. LeBourhis, J.P. Riviere, J.F. Pierson, O. Banakh, J. Appl. Phys. 98 (2005) (023715.1.). [5] N. Martin, O. Banakh, A.M.E. Santo, S. Springer, R. Sanjines, J. Takadoum, F. Levy, Appl. Surf. Sci. 185 (2001) 123. [6] S.H. Mohamed, A.M.A. El-Rahman, M.R. Ahmed, J. Phys. D: Appl. Phys. 40 (2007) 7057. [7] A. Rizzo, M.A. Signore, L. Mirenghi, A. Cappello, L. Tapfer, Surf. Coat. Technol. 204 (2010) 2019. [8] M.-H. Chan, P.-L. Wu, F.-H. Lu, Thin Solid Films 518 (2010) 7300. [9] P.K. Chu, B.Y. Tang, L.P. Wang, X.F. Wang, S.Y. Wang, N. Huang, Rev. Sci. Instrum. 72 (2001) 1660. [10] J.M. Ngaruiya, O. Kappertz, C. Liesch, P. Muller, R. Dronskowski, M. Wuttig, Phys. Status Solidi A (2004) 967. [11] P. Carvalho, J.M. Chappe, L. Cunha, S. Lanceros-Mendez, P. Alpuim, F. Vaz, E. Alves, C. Rousselot, J.P. Espinos, A.R. Gonzalez-Elipe, J. Appl. Phys. 103 (2008) (104907.1). [12] A. Rizzo, M.A. Signore, L. Mirenghi, E. Piscopiello, L. Tapfer, J. Phys. D: Appl. Phys. 42 (2009) 2354. [13] P. Carvalho, L. Cunha, E. Alves, N. Martin, E.L. Bourhis, F. Vaz, J. Phys. D: Appl. Phys. 42 (2009) 1955. [14] K. Carlson, L. Hultman, J.E. Sundgren, Thin Solid Films 371 (2000) 167. [15] K. Sedlackova, T. Ujvari, R. Grasin, P. Lobotka, I. Bertoti, G. Radnoczi, Vacuum 82 (2008) 214. [16] WangY.H. , ZhangX. , WuX.Y. , ZhangH.X. , X.J. Zhang, Appl. Surf. Sci. 254 (2008) 5085. [17] C. Xia, Z. Bingqing, S. Lin, M. Lili, H. Liyuan, H. Bing, Inform. Rec. Mater. 10 (2009) 26. [18] L.Q. Zhu, L.D. Zhang, G.H. Li, G. He, M. Liu, Q. Fang, Appl. Phys. Lett. 88 (2006) 232901. [19] M. Laurikaitis, J. Dudonis, D. Milcius, Thin Solid Films 516 (2008) 1549. [20] L.Q. Zhu, Q. Fang, G. He, M. Liu, L.D. Zhang, Nanotechnology 16 (2005) 2865.