Effect of deposition temperature on microstructure and corrosion resistance of ZrN thin films deposited by DC reactive magnetron sputtering

Highlights • The deposition conditions determine the thickness, composition and microstructure of the ZrN thin films and are fundamental to controlling the corrosion resistance of the material. • When the resulting film is relatively thin, the microstructure of the film shows no preferred orientation. When the resulting film is relatively thick, there is a strong preferred orientation in the (1 1 1) direction as indicated by the XRD results. • The air exposure after the deposition with relative high temperatures (500 °C), promotes the formation of a thin zirconium oxide and oxinitride surface layer that promotes the pitting corrosion resistance. Abstract Thin films of zirconium nitride were deposited on different substrates by direct current reactive magnetron sputtering, varying the deposition time, Ar/N 2 partial pressure ratio and substrate temperature. The physicochemical, crystalline structure and corrosion resistance of the thin films were studied by glancing angle X-ray diffraction, Rutherford backscattering spectrometry, scanning electron microscopy, X-ray photoelectron spectroscopy and potentiodynamic polarization tests in artificial saliva solution. The results show that the thin films presents high texture in [1 1 1] direction verified by X-ray diffraction measurements which indicated the lack of a Bragg peak for (2 0 0) crystallographic planes for a lower deposition temperatures. The XPS analysis showed the presence of ZrN and also the oxide species (ZrN x O y and ZrO 2 ) at the surface, with chemical states changing with deposition temperatures. In addition, the thin ZrN films were found to be stable in an electrochemical cell over a large potential range and the pitting potential increases with increasing the deposition temperature. For deposition at 500 °C, the pitting potential was found to be E p = 1.5 V/SCE. The corrosion behavior is attributed to the formation of thin ZrN x O y and ZrO 2 layer on the top surface of the films, with increasing of the deposition temperature. Keywords Zirconium nitride Zirconium oxynitride ZrN thin films Strain energy 1 Introduction Numerous different thin film coatings are currently available for specific applications depending on the coating properties [1,2] . Titanium nitride (TiN), vanadium carbide (VC), zirconium nitride (ZrN), aluminum–itanium nitride (AlTiN), and aluminum–zirconium nitride (AlZrN) are examples of compounds that have recently been applied as hard coatings in surface engineering for the wear protection and friction reduction of mechanical components in industry. Thin films of TiN and VC are widely used due to their good hardness and wear resistance properties [3,4] . Thin ZrN films have potential application as cutting tools and molds due to their wear, corrosion and abrasion resistance [5] . Also, ZrN thin films and compounds present high hardness and thermal stability [6,7] . Furthermore, ZrN thin films are biocompatible and can be used as coating for biomaterials [8] . However, even with high corrosion resistance, the dissolution process that occurs in aggressive media containing chlorine ions can lead to contamination of the organism with metal ions, affecting the biocompatibility and the mechanical properties of the prosthesis [9] . The corrosion protection feature of ZrN, an electrical conductor, can lead to the potential application in bioelectronic devices. In such applications, as in biomedical implants, severe restrictions are imposed on the materials used with respect to their stability in aqueous solutions with high concentrations of electrolytes. In particular, all kinds of solid electrodes, through which integrated microelectronic devices will interact with the surrounding medium of aqueous solutions of biological molecules, are highly susceptible to corrosion and other undesirable electrochemical interactions. Recently, a TiN layer has been proposed to be used as electrical contact in bioelectronic devices, where corrosion resistance is needed [10] . Although TiN were already studied and proposed as electrical contact in corrosive media, the use of ZrN as electrical contact in bioelectronic devices has not been explored despite its good electrical conductivity and corrosion resistance. The aim of this work is to study the physicochemical, crystalline structure and corrosion resistance in biological medium of ZrN thin films deposited by DC reactive magnetron sputtering. The pitting corrosion behavior of stoichiometric ZrN thin films were investigated aiming to study it's dependency of the substrate temperature and the chemical state of the phases at the surface. 2 Experimental procedures In order to prepare and characterize the ZrN films, which involved preparing the samples for Rutherford Backscattering Spectrometry (RBS), X-ray diffraction (XRD) and corrosion tests, we chose C (graphite), Si(1 0 0) wafer and Ti as the substrate, respectively. Once we want to study a thin film by means of several characterization techniques, we use different substrates due to the effect that each one of the substrates have over the results of different characterization techniques. The silicon substrate provides a flat surface where the ZrN thin film can be deposited avoiding the effects of a rough surface. Also, on SEM analysis, the silicon substrate is easy to cleavage and the preparation does not need any cutting or polishing processes. For these reasons, the Si substrate was used for XRD and SEM analysis. However, when analyzing a thin film of ZrN with RBS, the Si substrate is not recommended once that the technique depends on the element mass. Therefore, Si atoms hides the signals of the N and the O atoms on the RBS spectra due to its larger mass. For this reason, we choose the carbon as substrate, commonly used for analysis of light elements like N or O on RBS. The use of Ti substrate was employed to corrosion tests, due to its biomaterial characteristics and once we want to study the corrosion resistance of the thin film of ZrN at biological medium. The Ti samples were ground after being cut into various sizes and then polished with an alumina suspension with up to 0.1 μm of particle size. The samples were then cleaned with acetone in an ultrasonic bath for 30 min. The films were deposited simultaneously on the different substrates by DC reactive magnetron sputtering from a Zr target in Ar/N 2 plasma, with a total pressure of 2 × 10 −2 mbar and a power of 100 W at the target. The base pressure in the chamber was 10 −7 mbar. Three sets of samples were produced. In the first group the partial pressure of N 2 ranged from 0.2 × 10 −3 to 1.7 × 10 −2 mbar. In the second group the deposition time was varied from 10 to 80 min. In the third group the substrate temperature was varied from 25 to 500 °C. The thickness and Zr/N average atomic ratios of the zirconium nitride films were determined by RBS with incident 2 MeV He+ ions. The crystalline structure of the zirconium nitride films was studied by X-ray diffraction at room temperature with a grazing incidence of 0.6° on an X-ray diffractometer (Shimadzu XRD-6000) with Cu Kα ( U = 40 kV and I = 30 mA) radiation ( λ = 1.54064 Å), using as the reference standard an XRD powder standard sample of ZrN with a 2-956 JCPDS-ICDD file that retains an Fm3m cubic structure. The chemical bonds at the surface of the films were characterized by X-ray photoelectron spectroscopy (XPS). These measurements were performed on an Omicron SPHERA station using Mg Kα radiation (1253.6 eV) at a take-off angle of 60° with an energy resolution of 0.9 eV. For the scanning electron microscopy (SEM) analysis, the cross-section of the samples was studied with a Shimadzu SSX-550 electron microscope. The thickness of the thin films was directly measured from the SEM images and also by RBS analysis. Corrosion tests were performed in a VoltaLab model PGZ100 potentiostat/galvanostat. All electrochemical experiments were carried out at 37 °C in non-deaerated solutions. The electrolyte was an artificial saliva solution AFNOR S90-701 prepared using analytical grade reagents without previous purification. A ZrN-coated or uncoated titanium substrate of circular geometry was employed as the working electrode. For the corrosion tests, the thin film area to be studied was defined by partially covering the sample with insulating resin. A 0.95 cm 2 area of the sample was exposed to the electrolytic solution. Prior to each experiment the working electrode was degreased with acetone, cleaned with ethanol, and rinsed with distilled and deionized water. A graphite rod was used as the auxiliary electrode and a saturated calomel electrode (SCE) was used as the reference electrode connected to the cell by a salt bridge and a Luggin–Habber capillary. All potentials mentioned in the text are quoted versus this reference electrode. The potentiodynamic polarization technique was performed with a 1 mV s −1 scan rate, after recording the open circuit potential (OCP) for 60 min. The potentiodynamic polarization started at a value of 250 mV more negative than the OCP and proceeded toward the positive direction of potentials. The pitting potential was taken as the value at which the current density sharply increases during the positive potential scan. 3 Results and discussion 3.1 RBS analysis Fig. 1 (a) shows a typical RBS spectrum for the 2 MeV He+ incident ions of a zirconium nitride film on a carbon substrate. The Zr and the N signals are clearly seen, as well as the carbon signal for the substrate and a small O signal. Fig. 1 (b) shows the thickness of the zirconium nitride films as a function of the N 2 partial pressure during the deposition process as determined by RBS with 5% accuracy, assuming the nominal bulk density of zirconium nitride. The deposition was carried out for 20 min at 25 °C varying the N 2 partial pressure. The film thickness clearly increases as the partial pressure decreases. This behavior can be attributed to an increase in the mean free path of the plasma species under low pressure conditions, leading to an increase in the deposition rate which is reflected in the film thickness. Fig. 2 (a) shows the Zr/N atomic ratio of the thin films as a function of the N 2 partial pressure. Although the thickness varies according to the N 2 partial pressure, the Zr/N atomic ratio remains close to 1 presenting stoichiometric behavior (Zr/N = 1). In order to study the dependence on the deposition time a set of samples were prepared maintaining the substrate temperature fixed at 25 °C and the N 2 partial pressure at 0.5 × 10 −3 mbar. The deposition time was varied from 10 to 80 min. Fig. 2 (b) shows the Zr/N atomic ratio of the thin films as a function of the deposition time and the stoichiometric characteristic of the films (Zr/N = 1) can be observed. The exception in the case of the film deposited over 10 min is attributed to the initial growth process of the film. Thus, it is clear that the film thickness can be controlled by varying the deposition time. 3.2 XRD results The phase analysis of the thin ZrN films was studied by XRD. Fig. 3 shows the diffraction pattern for the ZrN films deposited on a Si substrate at 25 °C over 20 min for different N 2 partial pressures. Clearly, the diffraction lines can be attributed to the ZrN phase (NaCl-type structure, ICDD PDF #35-0753) with small peaks related to the Si substrate. As shown in the figure, the intensity of the peaks of the ZrN phase becomes higher with decreasing N 2 partial pressure. The peak intensity increases as the crystalline phase concentration increases. These results corroborate the observation by RBS concerning the deposition rate and the film thickness for different N 2 partial pressures. For higher pressures, the mean free path is shorter, the deposition rate is slower and, consequently, the film thickness is lower. For thinner films (<5 nm), the diffraction pattern shows low intensity or even no diffraction lines. On the other hand, for thicker films, the diffraction lines are more intense and the crystalline structure becomes evident. In comparison with the ZrN diffraction powder pattern the ZrN (1 1 1) peak is the more intense, indicating a strongly (1 1 1) preferred orientation of the films deposited with lower N 2 partial pressure. Fig. 4 (a) shows the X-ray diffraction patterns for ZrN films deposited over several deposition times, with a fixed deposition temperature of 25 °C and the N 2 partial pressure fixed at 0.5 × 10 −3 mbar. The well known consequence of the increase in deposition time is increasing the thickness of the deposited film that reflects on the XRD peak intensities. The peak integrated intensities for the ZrN (1 1 1) and (2 2 0) planes clearly increase as a function of time during the deposition. The texture coefficient P ( h k l ) i which characterizes the preferred orientation of the films can be calculated using the Harris method [11] . The calculated texture coefficient for the (1 1 1) plane in samples deposited at different deposition time is shown in Fig. 4 (b). As revealed in the figure, the thin ZrN films exhibited a (1 1 1) preferred orientation [ P (1 1 1) > 1], and the preferred orientation became stronger with increasing deposition time. This behavior is attributed to the increase in the ZrN film thickness for longer deposition times. At shorter deposition times (10 and 20 min), the diffraction patterns show a low intensity peak for the ZrN (2 0 0) planes which vanishes for longer deposition times. This is due to the reduced thickness when shorter deposition times are used. Under these conditions, the deposition process leads to the formation of a powder-like film structure with random ZrN grain orientation [12] . This random orientation is confirmed by the intensity of the various peaks attributed to the ZrN phase, which are similar to those observed in a powder pattern of the ZrN phase. As the deposition times increase, the ZrN films grow with preferred orientation leading to the absence of the peak attributed to the (2 0 0) plane for longer deposition times. At constant N 2 partial pressure and deposition time, the substrate temperature was varied. Fig. 5 (a) shows the XRD patterns of zirconium nitride films for different substrate temperatures in the interval of 100–500 °C, for 60 min deposition with N 2 partial pressure fixed at 0.5 × 10 −3 mbar. The diffraction lines for the powder pattern of the ZrN structure are also present for comparison. The calculated texture coefficients for the (1 1 1) plane of samples deposited at different temperatures are shown in Fig. 5 (b). As revealed in the figure, for this set of samples, the thin ZrN films also exhibited a (1 1 1) preferred orientation [ P (1 1 1) > 1], and the preferred orientation weakened with increasing deposition temperature. As the intensity of the XRD Bragg peaks is dependent on the constructive interference of the outgoing X-ray radiation after scattering [13] , the peak intensity increases with increasing atomic concentration in each plane. In an FCC structure the (1 1 1) and (2 0 0) planes intersect at tetrahedral and octahedral interstitial sites, respectively. Taking into account that nitrogen atoms can occupy, at least in principle, both tetrahedral and octahedral interstitial sites, the experimental observation that the intensity ratio I (111) / I (200) varies with substrate temperature could be explained by the inter-xchange of nitrogen atoms between the two interstitial sites. This observation was previously reported for the deposition of VC on a Si substrate, presenting the same crystalline structure [14] . Assuming the same effect, the inter-exchange is governed by the temperature. Other authors have previously observed an increase in the (2 0 0) peak intensity, for thin ZrN films deposited at high temperatures [15] . 3.3 SEM analysis The microstructure of the films was observed using SEM. Fig. 6 displays the scanning electron microscopy cross-section images of thin films deposited on Si substrates over 60 min at 500 °C with a N 2 partial pressure of 1.0 × 10 −3 mbar. In the sample, a thin film of about 790 nm is clearly observed (outermost layer) over the substrate. The film has a columnar characteristic indicating the preferential direction of the growth process perpendicular to the surface plane of the substrate. This characteristic columnar structure has previously been observed for thin ZrN films [16] . 3.4 XPS results Chemical bonding at the zirconium nitride thin film surface was studied by XPS. Fig. 7 shows the N 1s and Zr 3d photoelectron energy regions for films deposited at different substrate temperatures (100 °C, 300 °C and 500 °C). The spectra presented here correspond quite well to those obtained in the literature [17] . Since no special care was taken to prevent surface oxidation of the thin ZrN films, the exposure to air induced the formation of an oxide layer a few nanometers thick on top of the nitride coatings. Fig. 7 (a) presents the N 1s XPS spectra for air-exposed ZrN coatings, deposited at different temperatures. There are three components commonly assigned to the bonding configurations of N in ZrN at a binding energy of 397.4 eV [18] , at 396.6 eV in N 3− for ZrON [19] and of ZrN x O y at 399 eV [17,20] . The reference value for the N 1s binding energy in the ZrN compound obtained by XPS is indicated in the literature as being 397.3 eV [18] . In XPS studies on nitrogen containing zirconia, binding energies of 397.6 eV have been reported and assigned to the N 3− state in ZrN and the value of 396.4 eV to the N 3− state in Zr 3 N 4 [18,21] . Also, investigations on the in situ oxidation of ZrN to ZrO 2 have indicated values of 396.3 eV and 400.05 eV for the nitrogen peaks of the oxynitride Zr(N,O) [22] . Another study on oxidized ZrN coatings found the same contribution at 396 eV, attributed to an oxynitrided compound [8] . This low binding energy attributed to the zirconium oxynitride is controversial because, in general, an increase in the valence state of N 3− towards less negative values (oxidation) is related to an increase in the binding energy. The same authors assigned a higher binding energy of 403.2 eV to adsorbed N 2 species [22] . This high binding energy for the N 1s peak has also been registered at 402.8 eV representing dinitrogen species with high interaction in the solid [23,24] . On the other hand, the value of the N 1s peak at 395.8 eV is attributed to a ZrN 2 compound [25] . Based on our results, for the N 1s photoelectron picture we assume two components assigned to zirconium oxynitride especies, one at 399.9 eV for ZrN x O y and another one at 396.6 eV for ZrON. The nitrogen attributed to ZrN component was assigned at 397.5 eV. Another component at 403.5 eV was assigned to adsorbed N 2 species. Fig. 7 (b) shows the Zr 3d XPS spectra for the air-exposed ZrN coatings, deposited at different temperatures. The shift to high binding energies with increasing deposition temperature is clear. According to the literature, the reference value for the Zr 3d binding energy of ZrN in XPS lies between 179.6 eV and 180.3 eV [18] . The Zr 3d energy region for the 100 °C sample presents two doublets assigned to ZrN at 180 eV and zirconium oxynitride at 182.2 eV [15] . For the 300 °C sample a shift to high binding energies is observed and the two assigned doublets are present at 180.3 eV for ZrN and 182.5 eV for zirconium oxynitride [22] . For the 500 °C sample, again the shift to high binding energies is more expressive and peaks assigned to ZrN x O y and ZrO 2 at 182.2 eV and at 183.9 eV, respectively, were observed. The shift to high binding energies with an increase in the substrate temperature is due to the increase in the oxidation state of the Zr atoms. As indicated by different authors, the high number of species of Zr with different oxidation states does not allow an accurate deconvolution of the Zr 3d XPS spectra obtained to differentiate with accuracy the nitride, oxynitride and oxide chemical states [15] . Even with a careful cleaning process, oxygen contamination occurs due to its high reactivity with zirconium atoms [15] . Also, the observed increase in the intensity of the doublets associated with oxide and oxynitride species at the surface with increasing substrate deposition temperature, confirms the high O concentration in near surface regions seen in Fig. 7 (b) [26] . Furthermore, the intensity of the Zr 3d component assigned to the Zr–N bond is lower for a higher deposition temperature and vanishes with deposition at 500 °C. On analyzing the picture presented in the discussion of the N 1s photoelectrons, we interpret the Zr 3d photoelectron region of the nitride samples as being due to Zr in zirconium oxynitride. The O1s photoelectron region for the ZrN films (not shown) displays two components: one at a binding energy of 530 eV, which is assigned in the literature [14,26] to the O–Zr bond in ZrO 2 ; and a second at 532 eV in the photoelectron energy regions, which can be associated to N–O and OH groups adsorbed at the surface. As noted in the literature, the most probable explanation for the BE shift of both N 1s and Zr 3d to high energies is that oxygen atoms replace the nitrogen atoms in the Zr–N network. Due to the higher electronegativity of oxygen in comparison to that of nitrogen, the valence electron of N atoms will be moved further from the nuclei by the surrounding O atoms. As a result, the core electrons will be attracted more strongly by the protons in the nuclei. Analogously, the binding energy of O1s also increases as a result of the decrease in the surrounding N atoms [15,17,27] . Besides this explanation, some studies have attributed the nitrogen and zirconium chemical state variation to a large scale N defect structure rather than a different compound formation. 3.5 Corrosion tests In order to evaluate how the changes observed in the thin ZrN film morphology affect its corrosion resistance, potentiodynamic polarization experiments were performed in artificial saliva solutions. Fig. 8 shows the potentiodynamic polarization curves obtained for ZrN films deposited on Ti substrates at different temperatures. The corrosion behavior of thin ZrN films in artificial saliva solutions is clearly dependent on the variations in the deposition parameters in terms of the microstructure and phase composition of the films described above. The curve corresponding to the uncoated substrate indicates that Ti is more easily corroded in artificial saliva solution than ZrN-coated titanium. On the other hand, the current density of deposited samples is lower at relatively high potentials (above −0.5 V). The polarization curves observed in Fig. 8 indicate that deposition temperatures profoundly influence the corrosion resistance of the material. The ZrN thin film breaks down, as previously suggested, at more positive potentials with increasing deposition temperature. Samples deposited at higher temperatures present a greater passivation potential range than untreated Ti. The breakdown potential of +1.43 V for the sample deposited at 500 °C is more than 1.0 V higher than that for the uncoated titanium (+0.25 V), indicating an increased corrosion resistance. These results corroborate the variations in the nature of the surface formed at different deposition temperatures. As indicated by the results of the XPS analysis, for higher temperatures the surface was formed primarily of ZrN x O y and ZrO 2 , leading to a more stable characteristic in comparison with ZrN x obtained for the samples deposited at lower temperatures. The long-term corrosion behavior of both ZrN and TiN coatings has been investigated in 0.5 N NaCl [28,29] . It was proposed that during exposure the ZrN coating reacts with the environment causing the replacement of nitrogen with oxygen. This process results in the formation of a near-surface layer of zirconium oxide (ZrO 2 ), which represents a good protective barrier against corrosion. Long-term titanium exposure to artificial saliva leads to a more corrosion-resistant surface due to the abundant compact precipitates formed by the titanium oxidation [30] . 4 Discussion The ZrN thin films show different characteristics depending on the deposition parameters. There are two ways to control the film thickness. One is to set different times to obtain different thickness, as the thickness increases when longer deposition times are used. The other way is to set different N 2 partial pressures, changing the mean free path of species in the chamber leading to changes in the deposition rate. In both cases, when the resulting film is relatively thin, the microstructure of the film shows no preferred orientation. When the resulting film is relatively thick, there is a strongly preferred orientation in the (1 1 1) direction as indicated by the XRD results. This texture is also pronounced for deposition performed at temperatures up to 300 °C. At a deposition temperature of 500 °C, this preference toward orientation in the (1 1 1) direction tends to weaken and the intensity of the (2 0 0) diffraction peak increases. Several authors observed the preferred orientation (1 1 1) or (2 0 0) on TiN and ZrN thin films with NaCl-type structure [31,32] . The preferred orientation of such structure was determined by the lowest overall total energy conditions, resulting from a critical competition between surface energy and strain energy [33] . In NaCl-type structure, the (2 0 0) plane has the lowest surface energy while the (1 1 1) plane has the lowest strain energy [32] . Therefore, (2 0 0) preferred orientation is predicted in the deposition conditions using low kinectic energy particles in which the strain energy is small and the surface energy is dominant. On the other hand, the (1 1 1) preferred orientation is predicted in the deposition conditions using high kinetic energy particles in which the strain energy is large and dominant [33,34] . However, in a NaCl-type structure the thickness also plays an important role on the preferred orientation. As pointed by OH, the surface energy is dominant in the early stage of film growth, in which the (2 0 0) plane has the lowest surface energy and grows faster than other planes, resulting in the (2 0 0) preferred orientation. Furthermore, beyond a critical thickness, the strain energy begins to be dominant and the (1 1 1) preferred orientation is developed [33] . In our results, we obtain different film thickness for different deposition times. For sample deposited for 10 min the presence of (1 1 1) and (2 0 0) peaks is clear, as shown in Fig. 4 (a). As the deposition time increases, the (2 0 0) peak vanishes and only the (1 1 1) is present. These results are in agreement with those observed by OH [33] , where the strain energy is dominant beyond a critical thickness and the (1 1 1) preferred orientation is observed. The XPS spectra obtained for the most representative samples of the thin ZrN films deposited at different temperatures reveals interesting behavior due to the shifts of the core level binding energies of Zr 3d and N 1s electrons. These shifts occurs when the nitride surface is exposed to atmospheres containing oxygen. As expected, this exposure led to the formation of zirconium oxynitrides and zirconium oxide at the outermost surface. These oxynitrides are proven to be the first known materials with high nitrogen ion mobility [27] . This ion mobility has some influence on the XRD and XPS result. Depending on the site occupancy of the N and O atoms in the Zr–N crystalline structure, the XRD peaks show different intensities for the (1 0 0) and (2 0 0) planes. For substrate deposition temperatures near 500 °C, the (2 0 0) planes show an increasing diffraction intensity and for lower temperatures the (1 0 0) diffraction peak is the most intense. This behavior is suggested by the high mobility of the N atoms on applying higher temperatures. In the XPS results, the high binding energy shift in the Zr 3d and N 1s region for high-temperature samples also suggests that the chemical binding differs according to the site occupancy of N and O atoms. The high ion mobility of nitrogen atoms in ZrN x O y shifts its core level binding energy to high values due to the different site occupancy. This high binding energy shift also indicates the formation of zirconium oxide and oxynitride in the near surface region. The corrosion behavior for the samples deposited at different temperatures reflects its dependency on the presence of a thin oxynitride and oxide layer at the outermost surface. The samples deposited at relatively high temperatures present better corrosion protection in comparison with the low-temperature samples as observed by the pitting potential. This behavior is in agreement with XPS results showing the presence of more stable species like ZrN x O y and ZrO 2 at the outermost surface. 5 Conclusions The microstructure, composition and corrosion behavior of thin zirconium nitride films deposited under different conditions was studied. The deposition conditions which determine the thin film thickness, composition and microstructure are fundamental to controlling the corrosion resistance of the material. The film thickness increases when the N 2 partial pressure decreases and is also proportional to the deposition time. When the resulting film is relatively thin, the microstructure of the deposited film shows no preferred orientation. When the resulting film is relatively thick, there is a strongly preferred orientation in the (1 1 1) direction as indicated by the XRD results. XPS results revealed the presence of ZrN but also oxide and oxynitride phases attributed to the air exposure after the deposition process. The corrosion behavior was evaluated by potentiodynamic polarization experiments in artificial saliva. The deposition temperature and the oxidation of the surface play important roles in the corrosion behavior. As expected, for low-temperature deposition, the thin ZrN films have good corrosion resistance in comparison with the pure Ti substrate. For relatively high deposition temperatures, the exposure of the thin ZrN films to the atmosphere leads to the formation of the oxides and oxynitrides at the surface and the pitting corrosion resistance is increased. Acknowledgements The authors would like to acknowledge MCT/CNPq and CAPES for financial support. CLGA and JB receive grants from CAPES and DR, CAF, AS and IJRB from CNPq. References [1] K.L. Choy Prog. Mat. Sci. 48 2003 57 170 [2] P.J. Kelly R.D. Arnell Vacuum 56 2000 159 172 [3] H.G. Prengel W.R. Pfouts A.T. Santhanam Surf. Coat. Technol. 102 3 1998 183 190 [4] D.I. Bazhanov A.A. Knizhnik A.A. Safonov A.A. Bagatur’yants M.W. Stoker A.A. Korkin J. Appl. Phys. 97 2005 044108 044114 [5] E.W. Niu L. Li G.H. Lv H. Chen X.Z. Li X.Z. Yang S.Z. Yang Appl. Surf. Sci. 254 2008 3909 3914 [6] M. Chhowalla H.E. Unalan Nat. Mater. 4 4 2005 317 322 [7] J.V. Ramana S. Kumar C. David A.K. Ray V.S. Raju Mater. Lett. 43 2000 73 76 [8] Y. Cheng Y.F. Zheng IEEE Trans. Plasma Sci. 34 4 2006 1105 1108 [9] X. Liu P.K. Chu C. Ding Mater. Sci. Eng. R: Rep. 47 2004 49 121 [10] M. Birkholz K.-E. Ehwald D. Wolansky I. Costina C. Baristiran-Kaynak M. Fröhlich H. Beyer A. Kapp F. Lisdat Surf. Coat. Technol. 204 2010 2055 2059 [11] C.S. Barrett T.B. Massalski Structure of Metals 1980 Pergamon Oxford p. 204 [12] H. Jiménez E. Restrepo A. Devia Surf. Coat. Technol. 201 2006 1594 1601 [13] C. Kittel Introduction to Solid State Physics 7th ed. 1996 John Wiley & Sons Inc. New York [14] E. Portolan C.L.G. Amorim G.V. Soares C. Aguzzoli C.A. Perottoni I.J.R. Baumvol C.A. Figueroa Thin Solid Films 517 2009 6493 6496 [15] 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 pp. 104907-1–104907-15 [16] J.H. Huang K.-H. Chang G.-P. Yu Surf. Coat. Technol. 201 2007 6404 6413 [17] M.A. Signore A. Rizzo L. Mirenghi M.A. Tangliente A. Cappello Thin Solid Films 515 2007 6798 6804 [18] P. Pietro L. Galan J. Sanz Phys. Rev. B 47 3 1993 1613 1615 [19] L.I. Johansson H.I.P. Johansson Mat. Sci. Forum 325–326 2000 335 340 [20] A. Rizzo M.A. Signore L. Mirenghi D. Dimaio Thin Solid Films 515 2006 1486 1493 [21] G. Soto W. de la Cruz M.H. Farías J. Electron. Spectrosc. Relat. Phenom. 135 1 2004 27 39 [22] I. Milosev H.H. Strehblow M. Gaberscek B. Navinsek Surf. Interfaces Anal. 24 1996 448 458 [23] L.L. Gendre R. Marchand Y. Laurent J. Eur. Ceram. Soc. 17 1997 1813 1818 [24] H. Wiame M.-A. Centeno S. Picard P. Bastians P. Grange J. Eur. Ceram. Soc. 18 1998 1293 1299 [25] M. Del Re R. Gouttebaron J.-P. Dauchot P. Leclère G. Terwagne M. Hecq Surf. Coat. Technol. 174–175 2003 240 245 [26] I. Valov R.A. de Souza C.Z. Wang A. Borger C. Korte M. Martin K.-D. Becker J. Janek J. Mater. Sci. 42 2007 1931 1941 [27] M. Lerch J. Janek K.D. Becker S. Berendts H. Boysen T. Bredow R. Dronskowski S.G. Ebbinghaus M. Kilo M.W. Lumey M. Martin C. Reimann E. Schweda I. Valov H.D. Wiemhöfer Prog. Solid State Chem. 37 2009 81 131 [28] L. van Leaven M.N. Alias R. Brown Surf. Coat. Technol. 53 1992 25 34 [29] R. Brown M.N. Alias R. Fontana Surf. Coat. Technol. 62 1993 467 473 [30] D. Krupa J. Baszkiewicz J.W. Sobczak A. Biliński A. Barcz J. Mater. Process. Technol. 143–144 2003 158 163 [31] G. Abadias Surf. Coat. Technol. 202 2008 2223 2235 [32] J.L. Ruan D.F. Lii J.S. Chen J.L. Huang Ceram. Int. 35 2009 1999 2005 [33] U.C. Oh J.H. Je J. Appl. Phys. 74 3 1993 1692 1696 [34] J. Pelleg L.Z. Zevin S. Lungo Thin Solid Films 197 1991 117 128