Thin Solid Films 531 (2013) Contents lists available at SciVerse ScienceDirect. Thin Solid Films

Transcription

1 Thin Solid Films 531 (2013) Contents lists available at SciVerse ScienceDirect Thin Solid Films journal homepage: Properties of carbon film deposited on stainless steel by close field unbalanced magnetron sputter ion plating Weihong Jin a,c, Kai Feng a, Zhuguo Li a,, Xun Cai a, Lei Yu b, Danhua Zhou b, Paul K. Chu c a Shanghai Key laboratory of Materials Laser Processing and Modification, School of Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai , PR China b Zhejiang Huijin-Teer Coatings Co., Ltd., Lin'an, , PR China c Department of Physics and Materials Science, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong, China article info abstract Article history: Received 15 March 2012 Received in revised form 8 January 2013 Accepted 11 January 2013 Available online 25 January 2013 Keywords: Carbon film Bipolar plates Magnetron sputtering Corrosion resistance Electrical conductivity Carbon films are deposited on 304 stainless steel (SS304) by close field unbalanced magnetron sputter ion plating using different substrate bias voltages and target currents to improve the corrosion resistance and electrical conductivity of bipolar plates made of SS304 in proton exchange membrane fuel cells (PEMFCs). The surface morphology, Raman scattering spectra, corrosion resistance, interfacial contact resistance (ICR), and contact angle with water of the carbon films are determined. A dense carbon film is produced on the SS304 by this technique and the corrosion resistance is improved significantly. The ICR value diminishes drastically and water contact angle increases after deposition. In addition, the passive current density in the simulated PEMFC environment decreases initially, increases as the substrate bias voltage is increased, and drops with decreasing target current. As the substrate bias is increased, the ICR between the carbon film and carbon paper exhibits an initial diminishing trend and then increases, but the effect of the target current on the ICR is not as substantial as that of the bias voltage Elsevier B.V. All rights reserved. 1. Introduction Although magnetron sputtering [1] is widely used in surface engineering of functional materials, the main limitation is the reduced ion current impacting the substrate at large distances. An unbalanced magnetron field [2] enables the plasma to flow towards the substrates thereby allowing some of the secondary electrons produced during sputtering to follow the field lines resulting in additional ionizing collisions. Close field unbalanced magnetron sputter ion plating (CFUBMSIP) was invented by Teer [3] in In this technique, the closely linked magnetic field lines trap all the electrons generated in the system to further increase the ion current density surrounding the substrates to about 7 ma cm 2 [4], which is one hundred times higher than that in traditional magnetron sputtering. Therefore, ion bombardment on the substrates becomes more intense enabling the deposition of dense and adherent coatings with less defects. According to previous studies [5 8], carbon films with a large sp 2 percentage prepared by chemical or physical vapor deposition have high electrical conductivity, chemical inertness, and hydrophobicity. However, in bipolar plate applications, the properties need to be further improved. Bipolar plates account for a large percent of the mass and cost [9] in the proton exchange membrane fuel cell (PEMFC) stack in electric vehicles [10]. The ideal bipolar plates should possess high electric conductivity, good corrosion resistance, high hydrophobicity, high Corresponding author. Tel.: ; fax: address: lizg@sjtu.edu.cn (Z. Li). mechanical strength, high gas impermeability, light weight, and low cost [11]. The properties of the carbon films fabricated by CFUBMSIP are affected by deposition parameters such as the substrate bias voltage, target current, gas pressure, as well as substrate temperature. In particular, the former two factors play an important role in the film performance. There have been studies on the influence of the substrate bias voltage on the microstructure and mechanical properties [12 15], corrosion resistance [16 18], and electrical conductivity [19 21], but the relationship between the corrosion resistance and electrical conductivity with the substrate bias voltage has not been studied. Furthermore, there have been few studies concerning the influence of the target current which is a very important deposition parameter. In our previous study, carbon-coated 304 stainless steel (SS304) samples were found to have high corrosion resistance, low interfacial contact resistance (ICR), and hydrophobicity and is thus a promising candidate as bipolar plates [8]. In this work, carbon films are produced by CFUBMSIP at different substratebiasvoltagesandtargetcurrents, and the effects of these deposition parameters on film properties such as the density, corrosion resistance, electrical conductivity, and surface hydrophobicity are investigated. 2. Experimental details 2.1. Preparation and characterization The SS304 substrates were polished with No SiC waterproof abrasive papers, cleaned with acetone in an ultrasonic cleaner for /$ see front matter 2013 Elsevier B.V. All rights reserved.

2 W. Jin et al. / Thin Solid Films 531 (2013) Table 1 Deposition parameters and thickness of the carbon films. Sample Bias voltage (V) Target current (A) Thickness of transition layer (μm) S S S S S S Thickness of surface layer (μm) 15 min, and dried. The carbon films were deposited on the substrates on a CFUBMSIP system consisting of two 99.99% pure graphite targets and two 99.99% pure chromium targets. Before deposition, the substrates were sputter-cleaned by a plasma at 500 V to obtain an active surface to improve the adhesion between the substrate and coating. To produce a uniform film, the substrates were rotated at 2.5 rmin 1.Achromium carbide was first deposited onto the substrates as a transition layer to enhance adhesion using all four targets. Afterwards, the carbon film was deposited using the two graphite targets at substrate bias voltages of 20, 60, 100, and 150 V while the target current was fixed as 7 A. After determining the optimal bias voltage based on the performance the carbon films, the carbon films were deposited using target currents of 5 and 3 A. The thickness of the carbon film prepared under each set of conditions was fixed at approximately 3 μm by using the proper deposition time. The thickness of the carbon film on the SS304 was measured on a crater machine (BC-2, Teer Coatings, Ltd.) [22] and the important deposition parameters and carbon film thicknesses are listed in Table 1. The surface morphology of the carbon films were examined by high-resolution field emission scanning electron microscope (SEM) (Sirion 200, FEI) and Raman scattering spectra were obtained on a dispersive Raman microscope (Senterra R200-L, Bruker Optics) Electrochemical tests Electrochemical measurements were conducted on an advanced electrochemical system (PARSTAT 2273, Princeton Applied Research, a subsidiary of AMETEK, Inc.) using a three-electrode cell. A saturated calomel electrode (SCE) was the reference electrode, a graphite rod served as the counter electrode, and the samples were the working electrode. Unless otherwise stated, the potential was referenced to the SCE. To simulate the operation conditions of the PEMFC, all the electrochemical experiments were conducted in a 0.5 M H 2 SO 4 +2 ppm HF solution at 80 C. Because bipolar plates in the PEMFC are exposed to air/oxygen on one side and hydrogen gas on the other side, the solution was purged with either pressurized air (to simulate the PEMFC cathode environment) or hydrogen (to simulate the PEMFC anode environment) during the measurements, as reported by Fukutsuka et al. [5] and Choi et al. [23]. To ensure the electrochemical stability of the system, the open circuit potential (OCP) was monitored for 1 h before the potentiodynamic tests and to clarify the electrochemical behavior, the potentiodynamic curves were acquired at a scanning rate of 1mV-s Interfacial contact resistance The ICR between the gas diffusion layer and bipolar plate is an important property of bipolar plates in PEMFCs and significantly affects the power output of the fuel cell stack. The ICR between the carbon-coated SS304 and conductive carbon paper (Toray TGP-H-060) was evaluated Fig. 1. SEMimagesofthesurfaceofcarbonfilms sputtered using a target current of 7 A and substrate bias voltages of (a) 20 (S1), (b) 60 (S2), (c) 100 (S3) and (d) 150 V (S4).

3 322 W. Jin et al. / Thin Solid Films 531 (2013) according to the method proposed by Wang et al. [24]. In the measurement, two pieces of carbon paper were sandwiched between the two copper plates and coated sample. An external force was applied to the two insulators above the upper and lower copper plates, respectively. The two copper plates were connected to an ohmmeter to measure the total resistance of the assembly. One piece of carbon paper was sandwiched between the two copper plates, and the ICR between the carbon paper and copper plate was monitored to calibrate the ICR between the carbon film and carbon paper Water contact angle measurements Because water is produced during the operation of the PEMFC stack and water flooding can compromise the fuel cell performance, water should be removed and bipolar plates should be hydrophobic. A larger surface contact angle corresponds to higher hydrophobicity which indicates water removal from the fuel cell stack to prevent accumulated water from flooding the system. In our experiments, the water contact angles on the carbon-coated SS304 were determined on a video-based contact angle measurement device (OCA 20, Dataphysics). To obtain better statistics, three measurements are taken and averaged for each specimen. ratios from 6.4 to 7.1 indicative of larger sp 2 fractions. When the target current is increased, the I D /I G ratio of the carbon film increase slightly initially and then decreases, indicating that the sp 2 fraction in the carbon films is similar. The target current does not appear to affect the sp 2 fraction as much as the substrate bias voltage Electrochemical measurements Fig. 7 shows the OCP of the uncoated SS304 and carbon films prepared at different substrate bias voltages. All four carbon films exhibit much more positive OCP in both the simulated cathode and anode environments than the uncoated SS304, indicating lower corrosion tendency and higher corrosion resistance. Each carbon film can reach a constant OCP after 1 h in the simulated cathode and anode environments implying 3. Results and discussion 3.1. Surface morphology Fig. 1 depicts the SEM images of the surface of the carbon films prepared at different substrate bias voltages. The initial increase in the bias voltage produces carbon films that are more dense and homogeneous. This is because the negative bias on the substrate increases ion bombardment [4] which not only enhances diffusion and the chemical reactions between the incoming atoms and ions, but also sputters off the weakly bonded molecules on the film [20]. Therefore, the density and homogeneity of the carbon film are improved. However, the density and homogeneity decrease if bias voltage is too high, because when the bias voltage exceeds a certain value, extensive ion bombardment on the film leads to excessive damage making the carbon film loose. In our work, sample 2 exhibits the best density and homogeneity, suggesting an optimal bias of 60 V in the CFUBMSIP process. Fig. 2 shows the SEM images of thesurface of carbon films prepared using different target currents. The carbon film becomes more dense and homogeneous with decreasing target current, and sample 6 is observed to be the best. This is because as the target current decreases, the number of atoms and ions sputtered from the target surface and reaching the sample surface per unit time decreases, consequently resulting in sample temperature reduction. As indicated by our experiments, a smaller deposition rate tends to form denser films Raman scattering spectra The Raman spectra acquired from the carbon films produced using different substrate bias voltages are displayed in Fig. 3. Gaussian fitting and peak deconvolution reveal two peaks, the G-band at approximately 1575 cm 1 corresponding to the graphite band and the D-band at about 1395 cm 1 corresponding to the disordered band [25]. The ratios of the two bands are shown in Fig. 4.Thesp 2 fraction in the carbon film can be inferred from the intensity of the D peak to that of the G peak (I D /I G )and alargei D /I G ratio corresponds to a larger sp 2 fraction [26,27]. According to Fig. 4, a higher bias voltage increase the I D /I G ratio indicating that the sp 2 fraction is larger. When the bias voltage is small, the I D /I G ratio decreases and so a small bias voltage does not improve the sp 2 fraction. The Raman spectra of the carbon films produced using different target currents are shown in Fig. 5 again revealing the G-band corresponding to the graphite band and the D-band corresponding to the disordered band. Fig. 6 shows that all the carbon films have larger I D /I G Fig. 2. SEM images of the surface of carbon films sputtered using a substrate bias voltage of 60 V and target currents of (a) 7 A (S2), (b) 5 A (S5) and (c) 3 A (S6).

4 W. Jin et al. / Thin Solid Films 531 (2013) Fig. 3. Raman spectra of carbon films sputtered using a target current of 7 A and substrate bias voltages of (a) 20 (S1), (b) 60 (S2), (c) 100 (S3) and (d) 150 V (S4). that the system has reached a stable state. Samples 1 to 4 have approximately stable OCP and sample 5 shows much lower OCP which is still more positive than that of the uncoated SS304. The OCPof the uncoated SS304 and carbon films fabricated at different target currents are shown in Fig. 8. The OCP of the three carbon films are much higher in both the simulated cathode and anode environments than the uncoated SS304 indicating that the carbon films have much lower corrosion tendency and better corrosion Fig. 4. Bond ratio of carbon films sputtered using a target current of 7 A and substrate bias voltages of 20 (S1), 60 (S2), 100 (S3) and 150 V (S4). resistance than the uncoated SS304. The OCP of the three carbon films prepared at different target currents shows a constant value after 1 h in the simulated cathode and anode environments, suggesting that the system has reached a stable state. The OCP of the three carbon films do not differ significantly. The potentiodynamic polarization curves of carbon films prepared at different substrate bias voltages and different target currents are shown in Figs. 9 and 10. Table 2 lists the corrosion potentials and passive current densities of the different carbon films at 0.6 V (cathodic operating potential) in the simulated cathode environment determined from these potentiodynamic polarization curves. Fig. 9 shows that the uncoated SS304 is in the active state, with an active/passive-transition appearing upon anodic polarization, whereas the coated samples show spontaneous passivity. Fig. 9(a) shows that the corrosion potential of the uncoated SS304 is 356 mv and the passive current density at the cathodic operating potential is μa cm 2. All the corrosion potentials of the carbon films prepared at different substrate bias voltages move to the right in the simulated cathode environment compared to the uncoated SS304. The passive current density of sample 1 is higher than that of the uncoated SS304 and samples 3 and 4 have passive current densities similar to the uncoated SS304. Among the carbon films and the uncoated SS304, sample 2 exhibits the most positive corrosion potential of 257 mv and the lowest passive current density at the cathodic operating potential of 5.76 μa cm 2 implying much better corrosion resistance than the uncoated SS304. Fig. 9(b) shows that the corrosion potential of the uncoated SS304 is 364 mv and the passive current density at 0.1 V (the anodic operating potential) is μa cm 2. All the corrosion potentials of the

5 324 W. Jin et al. / Thin Solid Films 531 (2013) Fig. 6. Bond ratio of carbon films sputtered using a substrate bias voltage of 60 V and target currents of 7 A (S2), 5 A (S5) and 3 A (S6). voltage exceeds 60 V, the corrosion potential is reduced and the passive current density rises with increasing substrate bias voltage. This is because the initial increase in the substrate bias voltage contributes to the production of a dense film which inhibits penetration of the corrosion solution into the substrate thus improving the corrosion resistance of the film. However, an excessively large bias voltage Fig. 5. Raman spectra of carbon films sputtered using a substrate bias voltage of 60 V and target currents of (a) 7 A (S2), (b) 5 A (S5) and (c) 3 A (S6). carbon films produced at different substrate bias voltages shift to the right and are more positive than 0.1 V in the simulated anode environment. This indicates that all the carbon films are in the cathodic state and can protect the SS304 substrate from corrosion in the simulated anode environment. In general, the carbon film deposited at a substrate bias voltage of 60 V exhibits the best corrosion resistance in the simulated PEMFC environment. Fig. 9 also discloses that when the substrate bias voltage is very low, a larger substrate bias voltage increases the corrosion potential and decreases the passive current density. When the substrate bias Fig. 7. Open circuit potential of the uncoated SS304 and carbon films sputtered using a target current of 7 A and substrate bias voltages of 20 (S1), 60 (S2), 100 (S3) and 150 V (S4) in 0.5 M H 2 SO 4 +2 ppm F solution at 80 C: (a) bubbled with air and (b) bubbled with hydrogen.

6 W. Jin et al. / Thin Solid Films 531 (2013) Fig. 8. Open circuit potential of the uncoated SS304 and carbon films sputtered using a substrate bias voltage of 60 V and target currents of 7 A (S2), 5 A (S5) and 3 A (S6) in 0.5 M H 2 SO 4 +2 ppm F solution at 80 C: (a) bubbled with air and (b) bubbled with hydrogen. Fig. 9. Potentiodynamic curves of the uncoated SS304 and carbon films sputtered using a target current of 7 A and substrate bias voltages of 20 (S1), 60 (S2), 100 (S3) and 150 V (S4) in 0.5 M H 2 SO 4 +2 ppm F solution at 80 C: (a) bubbled with air and (b) bubbled with hydrogen. produces a loose film and defects such as micro poles [28] which can become channels for the corrosion solution to traverse. Thus, an excessively high bias voltage does compromise the corrosion resistance of the carbon film. Fig. 10(a) shows that all the carbon films produced at different target currents have more positive corrosion potentials and smaller passive current densities than the uncoated SS304 in the simulated cathode environment. Samples 2 and 5 exhibit a passive current density at the cathodic operating potential, whereas sample 6 has a much lower passive current density of 2.10 μa cm 2 and the corrosion resistance of sample 6 is also improved in the simulated cathode environment. Fig. 10(b) shows that all the corrosion potentials of the carbon films prepared at different target currents move to the right and are much higher than 0.1 V in the simulated anode environment. Therefore, all the carbon films are in the cathodic state and SS304 is protected against corrosion in the simulated anode environment. Fig. 10 also shows that a smaller target current decreases the passive current density. Although the corrosion potential of sample 6 is more negative than those of samples 2 and 5, sample 6 has the lowest corrosion current and the best corrosion resistance. The carbon film deposited at a target current of 3 A exhibits the best corrosion resistance among the three types of carbon films produced at different target currents in the simulated PEMFC environment. This is because the amount of the target atoms sputtered is reduced and the number of ions in the chamber decreases with smaller target currents. Consequently, the film deposition rate decreases. Meanwhile, ion bombardment is still moderate when the target current decreases to 3 A. The small deposition rate and a certain degree of ion bombardment lead to the formation of a dense film with few defects and improved corrosion resistance Interfacial contact resistance TheICR of theuncoated SS304 and carbonfilms fabricated at different substrate bias voltages is shown in Fig. 11. The ICR between the SS304 and carbon paper is very high reaching mω cm 2 because of the weak electrical conductivity of the passive film on the surface. Compared to the uncoated SS304, all the carbon films produced at different bias voltages have lower ICR and it can be ascribed to the high conductivity of the carbon film. The ICR of the uncoated SS304 and carbon films decreases quickly at a small compaction force, mainly because the contact points between the tested sample and carbon paper increase as the compaction force increases and the ICR is mainly influenced by a small compaction force. Afterwards, the ICR decreases gradually finally reaching a steady state because the contact area between the carbon coated SS304 and carbon paper gradually reaches the maximum and the compact force is no longer the main influence on the ICR. Fig. 11 also shows that the ICR between the carbon film and carbon paper decreases initially and then increases with increasing substrate bias voltage. The ICR values of samples 2 and 3 are 5.37 mω cm 2 and 5.11 mω cm 2, respectively, under a compaction force of 135 N cm 2 and they are smaller than those of samples 1 and 4. This is because at a larger substrate bias voltage, the sp 2 fraction rises. The graphite component increases and the electrical conductivity of the carbon film goes up.

7 326 W. Jin et al. / Thin Solid Films 531 (2013) Fig. 11. Interfacial contact resistance of the uncoated SS304 and carbon films sputtered using a target current of 7 A and substrate bias voltages of 20 (S1), 60 (S2), 100 (S3) and 150 V (S4) with carbon paper. substrate bias voltage from the perspective of the electrical conductivity of the carbon films Contact angles Fig. 10. Potentiodynamic curves of the uncoated SS304 and carbon films sputtered using a substrate bias voltage of 60 V and target currents of 7 A (S2), 5 A (S5) and 3 A (S6) in 0.5 M H 2 SO 4 +2 ppm F solution at 80 C: (a) bubbled with air and (b) bubbled with hydrogen. However, when the substrate bias voltage is too high, although the sp 2 fraction increases, which is beneficial to the drop of the ICR value, excessive ion bombardment leads to the formation of a loose film and defects in the carbon film, finally compromising the electrical conductivity and reducing the ICR. Therefore, a moderate substrate bias voltage is very important in the film deposition in order to produce films with good properties. The ICR values of the uncoated SS304 and carbon films prepared at different target currents are shown in Fig. 12. All the carbon films have lower ICR than the uncoated SS304. Samples 2, 5, and 6 have approximately the same ICR, revealing that the ICR is not affected by the target current when the target current is in the range between 3 A and 7 A. The results are consistent with the Raman scattering spectra. Therefore, the influence of the target current is not as obvious as that of the Fig. 13 shows the water contact angles on the uncoated SS304 and carbon films produced at different substrate bias voltages. The average contact angle value of SS304 is 66.9 and all the carbon films prepared at different bias voltages have a much larger average contact angle than SS304. In general, the contact angle of the carbon film shows an upward trend with increasing substrate bias voltages. The water contact angles on the uncoated SS304 and carbon films produced at different target currents are shown in Fig. 14. The average contact angles on all the carbon films are similar, but much larger than that on SS304. The contact angle of sample 6 is The contact angles on all the carbon films produced at different target currents are similar disclosing that the contact angle is not affected significantly by the target current. Allinall,thecarbonfilm deposited at a substrate bias voltage of 60 V and target current of 3 A has the best corrosion resistance and high electrical conductivity. This carbon film also shows a large water contact angle. Our results suggest that that a 60 V substrate bias voltage and 3 A target current is the optimal values in the CFUBMSIP system. Table 2 Electrochemical characteristics of different types of carbon films. Sample Ecorr (mv) in the cathode environment Ecorr (mv) in the anode environment SS S S S S S S Icorr (μa cm 2 ) at 0.6 V in the cathode environment Fig. 12. Interfacial contact resistance of the uncoated SS304 and carbon films sputtered using a substrate bias voltage of 60 V and target currents of 7 A (S2), 5 A (S5) and 3 A (S6) with carbon paper.

8 W. Jin et al. / Thin Solid Films 531 (2013) trend as the target current decreases from 7 A to 3 A. The ICR decreases initially and increases with increasing substrate bias voltage, but the target current does not affect the ICR as much as the bias voltage. The water contact angle on the carbon film is influenced by the substrate bias voltage and target current. The passive current density of the carbon film prepared at 60 V substrate bias voltage and 3 A target current decreases to 2.10 μa cm 2 at 0.6 V in the simulated cathode PEMFC environment. This is the lowest level among all the carbon films. This carbon film also has an ICR of 4.92 mω cm 2 under a compaction force of 135 N cm 2 and a large water contact angle of close to 90. Our results indicate that it is imperative to use a moderate substrate bias voltage and target current in order to produce carbon films with high performance. Acknowledgments Fig. 13. Contact angle value of the uncoated SS304 and carbon films sputtered using a target current of 7 A and substrate bias voltages of 20 (S1), 60 (S2), 100 (S3) and 150 V (S4) with water. This work is financially supported by the National Natural Science Foundation of China (grant nos , and ), the Ministry of Science and Technology of the People's Republic of China (grant no. 2009DFB50350), and Hong Kong Research Grants Council (RGC) General Research Funds (GRF) nos and References Fig. 14. Contact angle value of the uncoated SS304 and carbon films sputtered using a substrate bias voltage of 60 V and target currents of 7 A (S2), 5 A (S5) and 3 A (S6) with water. 4. Conclusion In CFUBMSIP, the substrate bias voltage and target current are the two important parameters dictating the film properties. In this work, carbon films are sputtered on SS304 at different substrate bias voltages and target currents to produce high-performance materials and also to clarify the influence of the two parameters on the carbon film properties. The passive current density decreases as the substrate bias voltage increases from 20 V to 60 V and then increases when the bias voltage exceeds 60 V. It shows a decreasing [1] R.D. Arnell, P.J. Kelly, Surf. Coat. Technol. 112 (1999) 170. [2] P.J. Kelly, R.D. Arnell, Vacuum 56 (2000) 159. [3] D.G. Teer, US Patent 5,556,519 (1996). [4] K. Laing, J. Hampshire, D. Teer, G. Chester, Surf. Coat. Technol. 112 (1999) 177. [5] T. Fukutsuka, T. Yamaguchi, S.I. Miyanoa, Y. Matsuo, Y. Sugi, Z. Ogumib, J. Power Sources 174 (2007) 199. [6] Y. Show, Surf. Coat. Technol. 202 (2007) [7] K. Feng, X. Cai, H. Sun, Z. Li, P.K. Chu, Diamond Relat. Mater. 19 (2010) [8] W. Jin, K. Feng, Z. Li, X. Cai, L. Yu, D. Zhou, J. Power Sources 196 (2011) [9] H. Tsuchiya, O. Kobayashi, Int. J. Hydrogen Energy 29 (2004) 985. [10] B.C.H. Steele, A. Heinzel, Nature 414 (2001) 345. [11] X.G. Li, I. Sabir, Int. J. Hydrogen Energy 30 (2005) 359. [12] Y. Mikami, K. Yamada, A. Ohnari, T. Degawa, T. Migita, T. Tanaka, K. Kawabata, H. Kajioka, Surf. Coat. Technol. 133 (134) (2000) 295. [13] M. Weber, K. Bewilogua, H. Thomsen, R. Wittorf, Surf. Coat. Technol. 201 (2006) [14] C. Yu, L. Tian, Y. Wei, S. Wang, T. Li, B. Xu, Appl. Surf. Sci. 255 (2009) [15] D. Bhaduri, A. Ghosh, S. Gangopadhyay, S. Paul, Surf. Coat. Technol. 204 (2010) [16] J.H. Yoo, S.H. Ahn, J.G. Kim, S.Y. Lee, Surf. Coat. Technol. 157 (2002) 47. [17] M. Flores, L. Huerta, R. Escamilla, E. Andrade, S. Muhl, Appl. Surf. Sci. 253 (2007) [18] J.C. Caicedo, C. Amaya, L. Yate, W. Aperador, G. Zambrano, M.E. Go'mez, J. Alvarado-Rivera, J. Mun oz-saldan, P. Prieto, Appl. Surf. Sci. 256 (2010) [19] I. Ahmad, S.S. Roy, P.D. Maguire, P. Papakonstantinou, J.A. McLaughlin, Thin Solid Films 482 (2005) 45. [20] J.H. Lee, J.T. Song, Thin Solid Films 516 (2008) [21] D. Brassard, M.A. El Khakania, L. Ouellet, J. Appl. Phys. 102 (2007) [22] J. Zhang, Vacuum 5 (1992) 23. [23] H.S. Choi, D.H. Han, W.H. Hong, J.J. Lee, J. Power Sources 189 (2009) 966. [24] H. Wang, M.A. Sweikart, J.A. Turner, J. Power Sources 115 (2003) 243. [25] A.C. Ferrari, J. Robertson, Phys. Rev. B 61 (2000) [26] M.J. Matthews, M.A. Pimenta, G. Dresselhaus, M.S. Dresselhaus, M. Endo, Phys. Rev. B 59 (1999) R6585. [27] M.J. Paterson, K.G. Orrman-Rossiter, Diamond Relat. Mater. 2 (1993) [28] Q. Kong, L. Ji, H. Li, X. Liu, Y. Wang, J. Chen, H. Zhou, Mater. Sci. Eng. B 176 (2011) 850.