Structural properties and corrosion resistance of tantalum nitride coatings produced by reactive DC magnetron sputtering

Abstract

In this study, tantalum nitride (TaN) thin films were deposited on Si(100) and 316L stainless steel (SS) substrates by reactive DC magnetron sputtering. The effect of the nitrogen fraction ([N₂]) in the gas mixture on the composition, phase formation, roughness and corrosion resistance was investigated. The films were characterized by Rutherford backscattering spectrometry (RBS), X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS) and atomic force microscopy (AFM). The results reveal a transition from Ta₂N to N-rich phases by increasing [N₂] from 2 to 50%, with a dominance of cubic TaN (c-TaN) at intermediate values (5–20%). Moreover, the surface roughness for films with a c-TaN structure is significantly higher. Potentiodynamic polarization and electrochemical impedance spectroscopy (EIS) were also employed to evaluate the corrosion behavior of bare and coated SS. The results show that all TaN films increase the corrosion resistance of SS, irrespective of their bonding structure, which is attributed to the formation of a protective surface oxide layer. However, films with a c-TaN structure deposited at [N₂] ∼20% provide higher protection efficiency (∼93%), which can be related to a lower density of pinhole defects as derived from the EIS analysis.

---

Introduction

One of the most popular and economical methods for improving the corrosion resistance of metallic structures is to deposit a corrosion resistant coating on their surfaces.¹ Magnetron sputtering deposition is one of the physical vapor deposition (PVD) techniques which has been developed rapidly over the last few decades for industrial applications including the processing of hard, corrosion-resistant, wear-resistant, low-friction and decorative coatings as well as other systems in thin-solid-form with specific electrical or optical properties.^2–8 The sputtering technique is a cost-efficient method and can deposit high-purity and homogeneous thin films over large areas with strong adhesion to the substrate.² In the sputtering process, compound materials such as oxides, nitrides, carbides, fluorides or arsenides can be deposited by introducing the corresponding reactive gas into the deposition chamber along with an inert background gas (typically, Ar).^7–11 This working mode is called reactive sputter deposition and is able to control stoichiometry, structure, morphology and, subsequently, the properties of the deposited films.^9,10

Among various reactive sputtered coatings, transition metal nitrides (TiN, ZrN, TaN, CrN, etc.) are found to be suitable coatings for many industrial applications due to their unique combination of physical, mechanical and chemical properties.^3,12–16 Here, their properties depend strongly on the nitrogen (N₂) partial pressure since it is the basic parameter governing the reactive mode during sputtering. In this way, composition, structure and morphology of the deposited coatings can be tuned.^14–16

Tantalum (Ta) and Ta-based compounds exhibit outstanding corrosion resistance in most corrosive environments^17–19 and, thus, they evidence to be suitable coatings for protecting metallic structures against corrosion. However, there are a few studies about the corrosion behavior of Ta-based coatings showing their suitability for both acidic^20–24 and alkaline^25–27 environments. The high corrosion resistance of Ta-based alloys and coatings is usually attributed to the formation of an extremely stable surface oxide layer during the exposure to the corrosive environment.^20–25 This oxide layer is very thin, compact (pinhole free), providing strong and stable chemical bonds and, remarkably, capable of self-curing spontaneously in the case of damage.¹⁸

Among Ta-based coatings, reactive sputtered tantalum nitride (TaN) films are known for their good electrical, mechanical and thermal properties.^16,28–35 Recently, TaN films are attracting considerable interest as hard-coatings,³⁶ film resistors,³⁵ diffusion barrier layers³⁷ and high-speed thermal printing heads³⁸ in different industrial applications. The Ta–N binary system presents multiple stable phases such as Ta₂N, Ta₅N₆ and Ta₃N₅ as well as some metastable phases such as TaN, Ta₄N₅, Ta₃N₄.^35,39,40 The resulting TaN phase in reactive sputter deposition depends strongly on the deposition conditions, particularly, the N₂ partial pressure.^16,28–35

In addition to the structure, the impact of N₂ partial pressure on the mechanical,^32,36 electrical,^16,28–35 optical,³² tribological⁴¹ and acoustic³⁴ properties has extensively been studied. However, there are a few reports about the corrosion behavior of TaN films.^26,27 Flores et al.²⁷ studied the corrosion resistance of TaN PVD thin-films deposited on steel against a sodium chloride solution. They found that the coated substrates displayed better corrosion resistance than bare ones, although the electrolyte can diffuse toward the substrates through surface defects such as pores, pinholes or droplets. Furthermore, it has been suggested that TaN coatings significantly improves the microbial corrosion resistance of implants.²⁶ However, up to now, scarce information is available in the literature concerning the effect of the N₂ partial pressure on the corrosion behavior of TaN coatings.

In the present study, TaN coatings were deposited on Si(100) and stainless steel (SS) substrates by reactive direct-current (DC) magnetron sputtering at different N₂/Ar flow ratios. The structural properties and electrochemical behavior of the deposited TaN coatings were investigated. An overall and significant increase in the corrosion resistance with respect to bare substrates is observed for all TaN films irrespective of their original structure. In addition, the best performance is correlated with the highest compactness of the coatings (lowest pinhole defects) as derived from electrochemical impedance spectroscopy (EIS).

Experimental procedure

Thin-film growth

TaN coatings were deposited simultaneously on Si(100) and 316L SS substrates by DC reactive magnetron sputtering from a 3′′ circular Ta (99.95% purity) target. The nominal composition of 316L SS substrate in wt% was: Cr: 17.08; Ni: 11.23; Mo: 2.91; C: <0.03; Mn: <2; Si: <1.00; N: <0.1; P: <0.05; S: <0.02 and Fe: balance. The 316L SS substrates were cut into sheets of 15 mm × 15 mm × 1.5 mm, grinded with silicon carbide emery papers, polished with diamond suspensions and cleaned with acetone and ethanol in an ultrasonic bath and, finally, dried-out in N₂ before loading in the deposition chamber. The Ta cathode was facing the grounded substrates at a working distance of about 150 mm. The base and working pressure in the deposition chamber were ∼10⁻⁷ and 2.5 × 10⁻³ mbar, respectively. The deposition was carried out with a mixture of high-purity (99.99%) Ar and N₂ at a total gas flow of 25 sccm. For TaN growth, the N₂ content in the gas mixture ([N₂]) was varied between 2% and 50%. Prior to deposition, the substrates were sputter-cleaned by exposure to a low-power Ar plasma (to avoid unwanted deposition) for 30 min in order to improve the coating adhesion. During this initial step, a negative bias voltage of −400 V was applied to the substrates to enhance Ar⁺ bombardment through the plasma sheath. For plasma generation during growth, a DC voltage with an overall power of 100 W was applied to the cathode and the substrate temperature was kept constant at ∼20 °C by water cooling. No external bias was applied to the substrates during deposition, which were kept at ground potential. The growth time was set to 2 hours for all the coatings. A Dektak-150 (Veeco®) mechanical profilometer was used to measure the thickness of the deposited films. For this purpose, the substrates were partially masked during deposition to create a step between deposited and non-deposited areas.

Compositional and structural characterization

Rutherford backscattering spectrometry (RBS) was used to determine the composition of the coatings deposited on Si(100) substrates. The measurements were done with a 1.7 He⁺ probing beam and the energy of the backscattered projectiles was measured at a scattering angle of 170° with a solid-state Si detector (energy resolution of 15 keV). A circular batch-scan around the substrate normal (the so-called random spectrum) with a total charge of 20 μC was performed to avoid channeling effects in the substrate signal. Grazing-incidence X-ray diffraction (XRD) analysis was employed to analyze the resulting structural phases. The mean crystallite size (D) was calculated using Scherrer formula⁴² according:where λ is the wavelength of the Cu Kα X-ray (1.54056 Å), β is the full-width-at-half-maximum (FWHM) of the diffraction peak and θ is the angle of diffraction. The FWHM value was obtained from (002), (220), (300) and (223) reflections of Ta₂N, TaN, Ta₅N₆ and Ta₃N₅ phases, respectively.

The bonding structure of the TaN films was analyzed by XPS on an ESCALAB 250Xi (Thermo Fisher Scientific) equipped with a 500 mm Rowland circle monochromator and a micro-focused Al Kα X-ray source. Prior to analysis, the samples were sputter cleaned gently to remove surface contamination. All spectra were recorded at a 90° take-off angle and all binding energy values were corrected for the charging effect with reference to the adventitious C1s peak at 284.8 eV. The CasaXPS Software⁴³ was used for spectra analysis and fitting.

The surface morphology of the TaN films was studied at room temperature by atomic force microscopy (AFM) using a commercial set-up (Nanotec Electrónica S.L.). The images were acquired in amplitude modulation (AM-AFM) operation mode, with a Phase Locked Loop (PLL) dynamic board, using aluminum reflex coated silicon tips (Budget Sensors, k = 3 N m⁻¹ and f = 75 kHz). Freely available WSxM software⁴⁴ has been used for image acquisition and the free-ware Gwyddion code⁴⁵ was used for image processing and analysis.

Electrochemical measurements

The electrochemical behavior of bare and TaN coated 316L SS was investigated through potentiodynamic polarization and EIS measurements by a conventional three-electrode system. The sample acted as working electrode and the auxiliary and reference electrodes were a platinum sheet and a saturated calomel electrode (SCE), respectively. All experiments were conducted in 0.5 M H₂SO₄ + 2 ppm HF solution at room temperature using an SP-200 Potentiostat/Galvanostat/FRA system (BioLogic Scientific Instruments). This solution is usually used to simulate the proton exchange membrane fuel-cells (PEMFCs) environment.⁴⁶ Before the electrochemical measurements, the samples were kept at their open circuit potential (OCP) until the electrochemical stability was established. Potentiodynamic polarization tests were performed at a potential scanning rate of 1 mV s⁻¹. For EIS measurements, frequency scans were carried out by applying sinusoidal wave perturbations of ±10 mV in amplitude. The impedance values were recorded in the frequency range of 100 kHz to 1 mHz at their respective open circuit potentials. The analysis of the spectra was performed using Zview/Zplot software.⁴⁷

Results and discussion

Thickness and composition (RBS)

The thickness of the films was found to vary from 960 to 300 nm depending on [N₂] as measured by profilometry (Table 1). The decrease in the film thickness by increasing [N₂] is related to the decay in the sputtering rate caused by target poisoning, which is a well-known effect in reactive sputter deposition.¹⁰

Table 1 Film thickness, composition and calculated grain size for the TaN films deposited at different [N₂] values

a Groups are defined according to the main contributing phase from XPS: (I) Ta2N, (II) c-TaN, (III) N-rich nitrides.
[N2] (%)	2	5	10	15	20	25	35	50
Film thickness (nm)	960	780	620	520	470	430	380	300
N/Ta	0.57	0.98	1.22	1.35	1.44	1.63	1.75	2.13
Density (g cm^−3)	15.6	14.68	13.91	13.77	13.01	12.98	11.84	11.01
DTa2N (nm)	<3	—	—	—	—	—	—	—
DTaN (nm)	—	7.9	<3	<3	<3	—	—	—
DTa5N6 (nm)	—	—	9.8	8.7	7	3.9	3.4	3.1
DTa3N5 (nm)	—	—	—	—	—	<3	<3	<3
Group^a	I	II	II	II	II	III	III	III

---

The stoichiometric N/Ta atomic ratio of different TaN phases is also tabulated in Table 1. It is clear that increasing [N₂] causes the rise in the incorporation of nitrogen in the TaN films. The increase in the amount of incorporated nitrogen may be related to the progressive formation of N-rich phases with [N₂], although non-bonded N (e.g., N₂ trapped molecules) can also be accounted for. According to RBS data, for [N₂] of 2% and 5% the N/Ta atomic ratio is very close to the stoichiometric value of Ta₂N and c-TaN phases, respectively. Therefore, this result may suggest the eventual formation of single-phase films with the corresponding composition for each working conditions. The films deposited at [N₂] > 5% are N-rich with a N/Ta atomic ratio higher than unity. Here, the increase in the incorporated N could be attributed to the presence of N-rich (mixed) phases (e.g. Ta₅N₆ and Ta₃N₅) in the deposited films. The previous assumptions related to the phase formation are evaluated by XPS and XRD in the following sections.

The combination of the areal density extracted from the RBS spectra and the thickness values can be used to estimate the film density. The result of this analysis is also displayed in Table 1. Clearly, the progressive increase (reduction) of the N (Ta) content with [N₂] results in a net decrease in the film density. It should be noted that the values fall within those of α-Ta (16.4 g cm⁻³) and that of the N-richest TaN phase, Ta₃N₅ (10 g cm⁻³). The density of the different nitride phases was calculated from the unit cell volume extracted from the crystallographic information file (CIF) of each structure as done in ref. 46.

Bonding structure (XPS)

As shown in the previous section, the RBS analysis revealed that nitrogen incorporation in the sputtered TaN film increases with [N₂]. However, there are different nitride phases in Ta–N system and complementary analyzes is needed for phase identification. In this regard, further information about the phase formation can be gained at short-range order by studying the bonding structure by XPS.

Fig. 1(a) and (b) show the evolution the Ta 4f and Ta 4p/N 1s XPS core-level spectra, respectively, of TaN films deposited at various [N₂]. On the one hand, the Ta 4f core-level (Fig. 1(a)) presents a double-peak structure due to the spin–orbit splitting (∼1.92 eV) of the Ta 4f_7/2 and Ta 4f_5/2 states. On the other hand, the Ta 4p/N 1s (Fig. 1(b)) spectra show the superposition of the N 1s and Ta 4p_3/2 core-levels.

Fig. 1 XPS spectra of (a) Ta 4f and (b) Ta 4p/N 1s core-levels for the TaN films deposited at various nitrogen flow ratios.

It is obvious that with increasing [N₂], the Ta 4f core-level shifts to higher binding energy (Fig. 1(a)). This chemical shift indicates the progressive formation of N-richer phases, which is in agreement with RBS results. Depending on the position of the XPS features, the samples can be catalogued into three groups (I, II and III) as a function of [N₂]. In order to determine the bands associated with different Ta and N species, fitting of the Ta 4f and Ta 4p/N 1s core-level has been performed from spectra of each group (see lower graphs of each group in Fig. 1(a) and (b)). In all cases, the background has been determined using the Shirley method.⁴⁸

According to Fig. 1(a), the Ta 4f spectra can be deconvoluted into two components (each one contributing with a corresponding doublet with the same spin–orbit splitting). For group I ([N₂] = 2%), the Ta 4f_7/2 components are at 22.4 eV and 23.2 eV, which are in the good agreement with the values for Ta₂N and c-TaN phases, respectively.²⁹ It should be noted that the Ta 4f core-level position from hexagonal TaN coincides with that of c-TaN²⁹ but, as shown in the next section, only the latter phase has been hinted by XRD. In group II (5% ≤ [N₂] ≤ 20%), the components are at about 23.2 eV (c-TaN) and 24.1 eV, the latter located in between those of c-TaN and Ta₃N₅ (25.3 eV (ref. 49)). This might be an indication of the partial contribution of Ta₅N₆. Finally, for the films deposited in the group III (25% ≤ [N₂] ≤ 50%), the components appear at 24.1 and 25.2 eV, respectively. In this case, the former matches the expected value for the Ta₅N₆ phase and the latter is very close to the Ta 4f_7/2 binding energy of Ta₃N₅ (25.3 eV (ref. 49)). Consequently, the deconvolution of the Ta 4f spectra indicates that TaN films may consist of Ta₂N with a small contribution of c-TaN in group I, a mixture of dominant c-TaN with pseudo-Ta₅N₆ phases in group II and a mixture of Ta₅N₆ and Ta₃N₅ phases in group III.

The Ta–N bonding in the deposited films can also be studied by the Ta 4p/N 1s spectra (Fig. 1(b)). Niu et al.⁴⁹ reported that the binding energies of Ta 4p_3/2 and N 1s core-levels for the Ta–N bond are about 398 eV and 402 eV, respectively. As shown in Fig. 1(b), there is a shift in the N 1s and Ta 4p_3/2 core levels binding energies of the films assigned to the different groups, which can be attributed to different coordination in the TaN films.⁵⁰ However, in the literature there is no agreement concerning the impact of nitrogen content in the TaN film structure on the position of N 1s core level binding energy. Baba et al.⁵¹ and Li et al.⁵² found that the position of N 1s peak is not affected by the nitrogen content in the TaN film. On the contrary, Lamour et al.⁵⁰ reported that increasing the nitrogen fraction in the coating causes a chemical blue-shift in both the binding energies of Ta 4p_3/2 and N 1s bands. In addition, Arshi et al.¹⁶ and Arranz et al.⁵³ reported the opposite trend with the N content. It should be noted that this controversial results may be also affected not only by the N content in the films but also by the particular nitride phase formation (including amorphous state) by the different deposition methods.

The variation of the relative peak contributions in the Ta 4f_7/2, Ta 4p_3/2 and N 1s core levels extracted from the XPS fitting results is plotted as a function of [N₂] in Fig. 2. These values can be considered to be proportional to the corresponding phase fraction in the deposited film. First, it is obvious that the TaN films deposited within group I mostly contain the Ta₂N phase. Second, c-TaN and Ta₅N₆ are the dominant phases in groups II and III, respectively. Fig. 2(b) shows the variation in the area of the Ta 4p_3/2 and N 1s peaks as a function of nitrogen content. This analysis provides additional information on the relative content of Ta and N atoms in the deposited films.¹⁶ A linear relation is found between area of N 1s peak and N content. This trend is in agreement with the RBS results and shows that nitrogen incorporation in the deposited film increases with [N₂].

Fig. 2 Variation of the peak area of (a) Ta 4f_7/2 and (b) Ta 4p/N 1s for the deposited TaN films at the different nitrogen flow ratios.

Structural characterization (XRD)

In order to evaluate the order range of the different nitride phases identified by XPS, we have performed XRD analysis. The diffractographs from the TaN films deposited on Si(100) substrates at different [N₂] are shown in Fig. 3(a). The broad features in the XRD patterns suggest that the films are highly disordered, with extremely fine crystal grains (see below).

Fig. 3 (a) XRD patterns of TaN films deposited at different N₂ flow ratios; (b) enlarged view of the diffraction pattern for the selected TaN film.

The XRD pattern of the film deposited at [N₂] = 2% shows two broad peaks centered at around 36.7° and 67.2°. The former peak can be assigned to the (002) plane of the hexagonal Ta₂N phase (PDF, record 01-089-4764), while as it shown in Fig. 3(b), the latter may be the result of the superposition of (110), (103), (112) and (201) reflections of Ta₂N phase. This assignment may also support the negligible contribution from c-TaN (PDF, record 01-089-5196), as anticipated by XPS. In addition, RBS provides additional support for the major contribution of Ta₂N under this condition because the N/Ta atomic ratio is ∼0.57 (Table 1). In line with XPS, by increasing [N₂] to 5%, the formation of c-TaN can be identified by the appearance of (111), (200), (220), (311) and (222) reflections. Here, an additional weak reflection from the Si substrate can also be observed at around 56°. The presence of the c-TaN phase is also supported by RBS results since there is a good agreement between the stoichiometry of c-TaN and the N/Ta atomic ratio (∼0.98). By further increase in [N₂] between 10% and 20%, the (111) and (200) reflections of c-TaN are replaced with a broad reflection between 32° and 47°, together with the appearance of a new feature at 62.3°. In conjunction with the high N/Ta atomic ratio (1.22) and XPS results, the evolution of the XRD scans at high [N₂] points towards the presence of N-rich nitride phases. In fact, the deposition of N-rich Ta nitride phases at high [N₂] was previously reported for sputtered TaN films.^16,28,29 The reflections at 34.7° and 62.3° can be attributed to the (004) and (300) planes of hexagonal Ta₅N₆ (PDF, record 01-075-0628). As indexed in the middle graph of Fig. 3(b), it is also plausible that this phase coexist with c-TaN since the broad feature between 32° and 47° may be the result of the superposition of (111) and (200) reflections of c-TaN together with the (004), (111), (112) and (113) reflections of Ta₅N₆. Consequently, within this parameter region, it can be stated from XPS and XRD that the films may consist of a mixture c-TaN and Ta₅N₆ phases. According to Table 1, the amount of incorporated N rises with [N₂] and, hence, it may be plausible that the fraction of Ta₅N₆ could also increase at higher [N₂]. Furthermore, some features in the XRD pattern of the deposited film for [N₂] equal or above 25% suggest the eventual growth of grains with the monoclinic Ta₃N₅ phase (PDF, record 01-079-1533). This statement can be supported by the clear intensity increase in the region where the (023) and (223) reflections from this phase should appear (see Fig. 3(b)) and note that no other TaN phases may contribute there. Remarkably, at [N₂] > 25% there is a further intensity increase close to the position of the (023) reflection of Ta₃N₅, as well as the negligible intensity around the (200) reflection of c-TaN. Thus, probably the contribution from c-TaN is reduced at high [N₂] and the films are mostly composed of a mixture of Ta₅N₆ and Ta₃N₅ phases. Although this interpretation may be questionable from the present XRD data, it can be supported by the XPS results. It is worth noting that phase identification within the TaN system may be ambiguous when more than one phase coexists together with additional peak broadening in nanocrystalline films. This trend can also be affected at high [N₂] by the accommodation of N-excess in the films at interstitial sites and/or trapping of N₂ molecules during growth. As listed in Table 1, the N/Ta atomic ratio for [N₂] ≥ 35% is even higher than the stoichiometric value of the Ta₃N₅ phase (1.67), which is known to be the N-richest TaN phase.⁵⁴ In addition, the Ar background gas could also be incorporated during the growth process up to several atomic percent in films deposited at low substrate temperature,^55,56 which may induce additional lattice distortions.

XRD analysis was also performed on selected samples grown on SS substrates. The results indicate that deposition of TaN coatings on both Si and SS substrates provides similar diffraction features and, hence, the substrate does not significantly alter the structural properties of the as-produced TaN coatings.

Assuming the previous phase identification scheme, the crystal size (D) of each contributing phase can be calculated as listed in Table 1. First, the D values clearly indicate the disordered or quasi-amorphous structure of the deposited TaN films. It is also extracted that the D value of each phase decreases with increasing [N₂]. The decrease in the grain size of c-TaN phase with [N₂] could be attributed to the grain growth inhibition due to the concurrent nucleation of the Ta₅N₆ phase. On the other hand, for N-rich nitride phases, the decrease in D may be related to the lattice strain induced by the excess nitrogen atoms.⁵⁷ The latter trend has been also previously reported for reactive sputter deposition nanocrystalline of titanium,¹⁵ tungsten⁵⁷ and molybdenum⁵⁸ nitrides.

Surface morphology (AFM)

The two-dimensional (2D) and three-dimensional (3D) AFM images of TaN films deposited at different [N₂] are shown in Fig. 4. The 2D images (Fig. 4(a)) show that the surface of the deposited TaN films is rather homogenous with the presence of a granular morphology. These morphological features seem to decrease in size with [N₂]. In addition, the grains for films deposited at higher [N₂] (e.g. 35% and 50%) are better defined. Furthermore, the 3D images (Fig. 4(b)) show that, as [N₂] increases, the surface roughness firstly rises and then starts to fall. In order to further discuss about the impact of [N₂] on the surface topography of TaN films, the mean height and average roughness (R_a) as well as the root mean square roughness (R_rms) were measured using AFM topographical images. The results are displayed in Table 2. The surface of the film deposited at [N₂] = 2% is quite smooth and its mean height and average roughness are about 1.1 and ∼0.4 nm, respectively. With increasing [N₂], the mean height firstly increases until reaching a maximum value of 30.4 nm at [N₂] = 20%. Above that value of [N₂], it decreases until reaching a minimum of 6.1 nm at [N₂] = 50%. The trend in the average roughness is almost similar to the mean height but the average roughness of the TaN film deposited at [N₂] = 35% is slightly lower than the one deposited at [N₂] = 50%. According to XRD and XPS interpretation, the concurrent and competitive growth of c-TaN and N-rich nitride phases with distinct growth mechanisms could be responsible for the observed relatively rougher surface at intermediate [N₂] values. On the other hand, at low and high [N₂] the deposition of one single phase is dominant (see Fig. 2(a)). In fact, Ponon et al.¹⁴ have reported that the surface roughness of reactive sputtered TiN_x films depends on the deposited phases.

Fig. 4 The (a) 2D and (b) 3D AFM images of TaN films deposited at different N₂ flow ratios.

Table 2 Topographical parameters for the TaN films using AFM images

[N2] (%)	2	5	15	20	25	35	50
Mean height (nm)	1.1	4.3	28	30.4	26.5	10.2	6.1
Ra (nm)	0.4	1.4	11.8	11.6	8.4	1.9	2.1
Rrms (nm)	0.5	1.7	14.7	14.1	10.9	2.4	2.6
ζ (nm)	63	53	40	42	30	28	22

---

Moreover, the decrease in the surface roughness at high [N₂] can also attributed to the lower film thickness (decrease in the deposition rate). In this regard, the relatively low working pressure used in this work is expected to yield a columnar structure as corresponding to the zone T of the Thornton phenomenological growth model.⁵⁹ Such microstructure has indeed been previously observed in Ta-based coatings.^60–62 In zone T, the film in-depth structure is inhomogeneous as a result of grain coarsening with film thickness. Near the substrate, the growth starts with nucleation of randomly oriented nanocrystalline grains followed by emerging of V-shaped columns with favored orientations and the eventual texture development.^63,64 The faceted column tops result in surface roughness. Thick coatings may present overgrown columns that yield a rougher surface and leave (detrimental) open pores at the grain boundaries as a result of atomic shadowed deposition.⁶⁴

Finally, the average mound size of the sputtered TaN coatings can be determined by analyzing the AFM topographic images. For this purpose, the corresponding power spectral density (PSD) function was extracted by means of the Gwyddion free-ware software.⁴⁵ As discussed in Auger et al.,⁶⁵ a cross-over in the PSD curve is an indication of lateral correlation length (ζ), which here can be related to the average mound size formed on the surface.

The variation of ζ as extracted from the PSD is listed in Table 2 and indicates a decrease in the mound size with [N₂], which correlates with the D trend extracted from XRD. This result supports the assignment of the mound topological structures observed by AFM with the (nano)crystalline grains of the corresponding TaN phases.

Electrochemical measurements: potentiodynamic polarization

The potentiodynamic polarization curves for bare 316L SS substrate and TaN coatings deposited at different [N₂] in a 0.5 M H₂SO₄ + 2 ppm HF solution obtained at ambient temperature are presented in Fig. 5. The film deposited at [N₂] = 2% was delaminated after exposure to electrolyte, and hence, its electrochemical behavior was not studied. As shown in Fig. 5, bare and TaN coated substrates get passivated rapidly and exhibit a passivation behavior without active–passive transition region. This linear and quick transformation from active to passive state with increasing potential is indicative of the high material chemical stability.⁶⁶ In addition, the curves corresponding to TaN coated samples were moved to nobler potentials and to lower current densities, implying higher chemical inertness and better corrosion resistance, respectively. Furthermore, the corrosion potential (E_corr) and corrosion current density (i_corr) were obtained from the polarization curves through the Tafel extrapolation method⁶⁷ and summarized in Table 3. In addition, polarization resistance (R_p) and protective efficiency (P_i) were calculated as described elsewhere⁶⁸ and are also tabulated in Table 3. The results indicate very high R_p values (∼1 MΩ cm⁻²) of TaN films, which provides a good protective efficiency for 316L SS substrates and increase its corrosion resistance. The high polarization resistance of TaN coated samples can be attributed to a thin, impervious and inert passive layer of tantalum oxide (probably Ta₂O₅ amorphous oxide^22,69,70). Such layer is highly resistant to degradation by H₂SO₄ and dilute acids have no effect even at their boiling temperature.^18,19,22,70 Moreover, this passive film forms on the surface of Ta alloys upon exposure to the oxidizing solution (e.g., sulfuric acid) according to eqn (2):^22,70

2Ta + 5H₂O → Ta₂O₅ + 10H⁺ + 10e⁻ (2)

Fig. 5 Potentiodynamic polarization curves for 316L SS and TaN coatings deposited at different nitrogen flow ratios.

Table 3 Electrochemical parameters for 316L SS and TaN coatings deposited at different nitrogen flow ratios, obtained through potentiodynamic polarization curves

[N2] (%)	Ecorr (VSCE)	icorr (A cm^−2)	Rp (kΩ cm^−2)	Pi (%)	P (%)
Bare SS	−0.070	1.41 × 10^−7	261	—	—
5	0.113	2.58 × 10^−8	720	81.7	4.1
10	0.110	1.37 × 10^−8	1028	90.3	3.0
15	0.151	1.61 × 10^−8	931	88.6	2.0
20	0.182	9.40 × 10^−9	1438	93.3	0.9
25	0.161	1.32 × 10^−8	1051	90.6	1.6
50	0.059	1.54 × 10^−8	766	89.1	7.4

---

The discrepancy in the protective efficiency of the coating deposited at different [N₂] can be related to the difference in their porosity. As discussed before, the TaN coatings are chemically inert due to the quick formation of a protective passive oxide film, and hence, the polarization resistance of the coated samples measured at open circuit potential is related to the resistance of the 316L SS substrate exposed to the electrolyte through pinholes (open pores from the surface to the substrate). Actually, PVD thin films are not defect-free and always include certain porosity, pinholes, impurities, etc., depending on their chemical composition, structure and deposition parameters. In aggressive environments, these local defects can form direct paths between the substrate and the corrosive environment. The polarization results showed that the TaN films possess nobler corrosion potential than 316L SS substrate (Table 3), and thus they are cathodic to their substrates. Therefore, contact of the electrolyte with the substrate through pinholes causes a rapid localized galvanic attack and pitting corrosion of the 316L substrate (anode), while the TaN film (cathode) is cathodically protected. As shown above, the variation of [N₂] results in coatings with different composition, structure and surface roughness. This parameter may also affect the defect density, for example, through variation of the energy of bombardment ions arriving on the growing film surface.⁷¹ Therefore, the amount of [N₂] in the deposition chamber is a determinative parameter for the quality, porosity/pinhole content and subsequently corrosion resistance of the coating. In this regard, the porosity content of the coatings can be determined through electrochemical method according to eqn (3):^72–74

where P is the porosity index; R^c_p and R^s_p are the polarization resistance of the coated sample and bare substrate, respectively; ΔE_corr is the difference potential between the OCP of the coated sample and the bare substrate; and β^s_a is the anodic Tafel slope for the bare substrate. The calculated P values for the coating deposited at different [N₂] are summarized in Table 3. The results indicate that by increasing [N₂] the porosity index of 316L SS coated samples firstly decreases until reaching a minimum value of 0.9% at [N₂] = 20%. After that minimum, P increases until reaching a maximum of 7.4% at [N₂] = 50%. Therefore, there is a correlation between the structure, corrosion resistance, roughness and porosity index of the TaN coatings. In this regards, the film deposited at intermediate [N₂] provide the highest roughness and corrosion resistance and the lowest porosity index. It should be noted that the increase in the porosity index at high [N₂] may also be attributed to the decrease in the thickness of the coatings. Notter and Gabe⁷⁵ showed that the amount of open porosity in the thin films is inversely proportional with the film thickness.

Electrochemical measurements: impedance spectroscopy

For a deeper understanding of the corrosion behavior of the TaN coatings, bare and coated 316L SS substrates were examined by EIS and their Nyquist and bode plots are shown in Fig. 6. It can be seen in the Nyquist plots (Fig. 6(a)) that the real impedance of the TaN coated samples is much higher than that of the 316L SS substrate, which is in accordance with potentiodynamic polarization results stating that TaN coatings improve the corrosion resistance of the 316L SS substrate. In addition, the maximum phase angle for bare and TaN coated 316L SS is close to π/2 (Fig. 6(b)), representing a strong capacitance of their passive films.^23,25,76 The high frequency (HF) region of bode plots is related to local surface defects, while the processes within the film and those occurring at the metal/coating interface can be studied through the medium and low frequency (LF) regions, respectively.⁷⁷

Fig. 6 (a) Nyquist and (b) bode plots of bare and TaN-coated 316L SS in the 0.5 M H₂SO₄ + 2 ppm HF solution.

In order to account for the electrochemical behavior of the substrate and coating system, two equivalent electrical circuits (EECs) were assumed to simulate the metal/solution interface and to analyze the EIS plots. These EECs are plotted in Fig. 7. Fig. 7(a) shows the most common EEC used for analyzing the localized corrosion of SS.^78,79 In this circuit, R_s represents the uncompensated resistance of the solution between the working and reference electrodes. The time constant τ₁ = C_pR_p is associated with the HF region of the impedance spectrum of 316L SS, which is indicative of the electrolyte resistance in the pores of passive film and the capacitance properties of the passivated (intact) area of the films. In addition, the time constant associated with the LF region (τ₂ = C_dlR_ct) is ascribed to the charge transfer resistance and electrical double layer capacitance at the passive film pores. The physical meaning of these elements is discussed in more detail in Galván et al.⁷⁸ In this EEC, a constant phase element (CPE) was used instead of ideal capacitors because often the measured capacitance shows deviation from real capacitor behavior. The impedance representation of CPE is given by:^79,80

where Y₀ and n are constants; j = √−1; and ω is the angular frequency. Eqn (4) represents the pure capacitance (Y₀ = C) for n = 1, an infinite Warburg (diffusion) impedance for n = 0.5, pure resistance for n = 0 and pure inductance for n = −1.

Fig. 7 Equivalent electrical circuit model used to analyze the EIS data for (a) 316L SS substrate and (b) TaN coated samples.

The diffusion of aggressive electrolyte through the PVD coatings is usually characterized by introducing diffusion-related elements into the EECs, which describes the diffusion processes under certain boundary conditions.⁷⁶ Liu et al. reported that depending on the electrochemical properties of the substrate, the thin film microstructure as well as the size and shape of the structural defects, different EECs can be used to analyze EIS spectra of PVD thin films.^76,81 Fig. 7(b) shows the EEC used for analyzing the impedance spectra of TaN thin films. In this figure, C_t is the total capacitance of the passive films; R_po is the resistance of the electrolyte in the pinholes; and W_po is the diffusion element for the area at which pinholes exist. The W_po element has the same physical meaning as a CPE element (eqn (3)) and describes the localized corrosion, which is recognized by the presence of a transmission line in the LF region of the EIS spectra.⁷⁸

As shown in Fig. 5, the 316L SS is passivated once exposed to the electrolyte and a dense passive film with high dielectric constant forms on its surface. In TaN coated 316L SS, this passive film forms also at the pinholes, protects the substrate from degradation and generates a capacitive response.⁷⁶ It is hard to distinguish between the dielectric properties of the tantalum oxide passive film formed on top of the intact area of TaN layer and the chromium oxide passive film formed on the 316L SS substrate surface at the pinholes. Therefore, the element of C_t inserted in the corresponding EEC instead of these two capacitance according to eqn (5):⁷⁸

C_t = fC_s + (1 − f)C_f (5)

where C_s is the capacitance of 316L passive film; C_f is the capacitance of TaN passive film; and f is the area fraction of the pinholes (0 ≤ f ≤ 1).

Using the EECs shown in Fig. 7, the EIS spectra presented in Fig. 6 could be fitted by with the Zview/Zplot software.⁴⁷ The fitting curves for each spectrum are presented in Fig. 6 as solid lines and the calculated values for each one electrical element in the EECs are summarized in Table 4. This table shows that the capacitance of the substrate passive film is much higher than in TaN coated samples (C_s ≫ C_f). Therefore, according to eqn (5), C_t should be proportional to the surface fraction of pinholes. Consequently, the variation of C_t in Table 4 indicate that the pinhole density in the deposited TaN film initially decreases as [N₂] increases to 20%, and then rises with further increase in [N₂]. This result is in accordance with porosity index results (Table 3). Moreover, the exponent factors n_dl and n_t is usually related to the surface roughness, texture and inhomogeneity.^81,82 It is evident that the exponent n_t is close to unity and remains almost constant for bare and TaN coated 316L SS, suggesting surface homogeneity of these samples as well as capacitive characteristic of the interface Liu et al.⁸¹ reported that in the EEC shown in Fig. 7(b), the R_po element becomes the polarization resistance for the coated system and is inversely proportional to the pinhole density. Therefore, as shown in Table 4, by increasing [N₂] the polarization resistance of TaN coating initially rises as a result of decrease in the pinhole density and then decrease because of the increase in the pinhole density. The W_po element in the EEC reflects the mass transport of electroactive species (including aggressive ions) through the TaN coating. The smaller impedance values of W_po (i.e. smaller C_po and n_po values according to eqn (4)) indicate slower diffusion/adsorption of reactant agents,^79,83 which could result from longer diffusion path in a denser structure and/or plugging of corrosion products through pinholes. For instance, the straight diffusion path in a columnar structure shows smaller diffusion resistance than the zig-zag diffusion path in an equiaxed structure.⁸¹ Hence according to the values displayed in Table 4, the variation of C_po and n_po parameters indicate that the mass transport of electroactive species through the film deposited at [N₂] = 20% is slower than those of other films. Consequently, this coating provides the highest corrosion resistance among the TaN coatings, which is in the accordance with the potentiodynamic polarization results.

Table 4 Corrosion characteristics of bare and TaN-coated 316L SS as evaluated by EIS

	Rs (Ω cm^2)	Cp or Ct (μF s^n−1 cm^−2)	np or nt	Rp or Rpo (kΩ cm^2)	Cdl or Wpo (μF s^n−1 cm^−2)	ndl or npo	Rct (kΩ cm^2)	χ^2^a (×10^−4)
a In all samples and for all parameters, the fitting error was less than 4%.
Bare SS	1.96	94.4	0.93	67.2	425.5	0.81	50.1	34.8
5	4.3	12.0	0.96	128	18.4	0.61	—	13.8
10	3.8	11.0	0.96	163	18.1	0.69	—	9.7
15	4.0	11.0	0.96	167	17.6	0.56	—	6.7
20	4.3	10.3	0.96	185	13.2	0.44	—	8.5
25	4.0	10.4	0.95	176	10.6	0.50	—	5.0
50	4.0	34.4	0.95	156	19.9	0.65	—	4.8

---

Conclusions

TaN films were successfully deposited on Si(100) and 316L SS substrates by reactive DC magnetron sputtering technique at various N₂ flow ratios ([N₂]). It was found that the composition, structure, roughness and corrosion resistance of the deposited films depend strongly on [N₂]. The combination of XRD and XPS measurements reveal that the TaN films can be classified into three different groups according to the main contributing phase(s). For low [N₂] (<5%), group I, Ta₂N is the dominant phase. In addition, c-TaN and N-rich TaN phases are the main contributing phases for intermediate (5–20%), group II, and high (20–50%), group III, [N₂], respectively. Furthermore, AFM analysis indicate that the concurrent growth of c-TaN and N-rich nitride phases results in relatively rough surfaces at intermediate [N₂], while the films deposited at low and high [N₂] present a smooth surface. Moreover, potentiodynamic polarization measurements show a high polarization resistance for all deposited films except the film deposited at [N₂] = 2%. This response may be ascribed to the formation of a protective surface oxide layer. The calculation of porosity index through the electrochemical measurements indicates that, by increasing [N₂], the polarization resistance of TaN-coated SS initially rises and then falls as a result of a lower and higher pinhole density, respectively. The corrosion resistance, as well as corrosion mechanisms, of bare and TaN coated 316L SS samples were investigated by EIS measurement. It was found that the corrosion resistance of TaN films correlates with the roughness and porosity index, being in agreement with the polarization test.

Acknowledgements

Financial support from grants FIS2012-38866-C05-05 and CSD2010-00024 (Ministerio de Economía y Competitividad, Spain), P2013/MIT-2775 (Comunidad Autónoma de Madrid, Spain) as well as the Ministry of Science, Research and Technology of Iran are greatly acknowledged. The RBS experiments were carried out at the Ion Beam Center (IBC) of the Helmholtz-Zentrum Dresden-Rossendorf (HZDR) and the assistance of Irene Heras is greatly acknowledged. We also thank José Bartolomé for his technical support to carry out part of this work at ICMM-CSIC as well as Miguel Gómez for his assistance in the electrochemical measurements at ICV-CSIC. Authors would also like to acknowledge the support of Petr Vašina, Pavel Souček and, Monika Stupavská at CEPLANT-Masaryk University (Brno, Czech Republic). Additionally, Prof. Juan Carlos Galván (CENIM-CSIC) is gratefully acknowledged for valuable discussions about EIS results.

References

1. D. E. J. Talbot and J. D. R. Talbot, Corrosion Science and Technology, CRC Press, 2nd edn, 2007.
2. P. J. Kelly and R. D. Arnell, Vacuum, 2000, 56, 159–172.
3. C. Petrogalli, L. Montesano, M. Gelfi, G. M. La Vecchia and L. Solazzi, Surf. Coat. Technol., 2014, 258, 878–885.
4. S. S. Firouzabadi, K. Dehghani, M. Naderi and F. Mahboubi, Appl. Surf. Sci., 2016, 367, 197–204.
5. R. Gago, M. Vinnichenko, R. Hübner and A. Redondo-Cubero, J. Alloys Compd., 2016, 672, 529–535.
6. H. Gao, Y. Li, C. Li, F. Ma, Z. Song and K. Xu, RSC Adv., 2016, 6, 30998–31004.
7. N. Srinatha, Y. S. No, V. B. Kamble, S. Chakravarty, N. Suriyamurthy, B. Angadi, A. M. Umarji and W. K. Choi, RSC Adv., 2016, 6, 9779–9788.
8. D. Zheng, J. Xiong, P. Guo, S. Wang and H. Gu, RSC Adv., 2016, 6, 4908–4913.
9. I. Safi, Surf. Coat. Technol., 2000, 127, 203–218.
10. D. Depla and S. Mahieu, Reactive Sputter Deposition, Springer, Berlin, Heidelberg, 2008.
11. L. Zhao, Y. Di, C. Yan, F. Liu, Z. Cheng, L. Jiang, X. Hao, Y. Lai and J. Li, RSC Adv., 2016, 6, 4108–4115.
12. J. Xu, S. Xu, P. Munroe and Z.-H. Xie, RSC Adv., 2015, 5, 67348–67356.
13. M. Benegra, D. G. Lamas, M. E. Fernández de Rapp, N. Mingolo, A. O. Kunrath and R. M. Souza, Thin Solid Films, 2006, 494, 146–150.
14. N. K. Ponon, D. J. R. Appleby, E. Arac, P. J. King, S. Ganti, K. S. K. Kwa and A. O'Neill, Thin Solid Films, 2015, 578, 31–37.
15. N. Arshi, J. Lu, Y. K. Joo, C. G. Lee, J. H. Yoon and F. Ahmed, J. Mater. Sci.: Mater. Electron., 2013, 24, 1194–1202.
16. N. Arshi, J. Lu, Y. K. Joo, J. H. Yoon and B. H. Koo, Surf. Interface Anal., 2015, 47, 154–160.
17. S. D. Cramer, B. S. Covino and C. Moosbrugger, ASM Handbook Volume 13b: Corrosion: Materials, ASM International, 2005.
18. U. Gramberg, M. Renner and H. Diekmann, Mater. Corros., 1995, 46, 689–700.
19. P. E. P. A. Schweitzer, Fundamentals of Metallic Corrosion: Atmospheric and Media Corrosion of Metals, CRC Press, 2006.
20. H. Yu, L. Yang, L. Zhu, X. Jian, Z. Wang and L. Jiang, J. Power Sources, 2009, 191, 495–500.
21. C. Choe, H. Choi, W. Hong and J.-J. Lee, Int. J. Hydrogen Energy, 2012, 37, 405–411.
22. A. Robin and J. L. Rosa, Int. J. Refract. Met. Hard Mater., 2000, 18, 13–21.
23. S. Maeng, L. Axe, T. A. Tyson, L. Gladczuk and M. Sosnowski, Corros. Sci., 2006, 48, 2154–2171.
24. S. M. Maeng, L. Axe, T. A. Tyson, L. Gladczuk and M. Sosnowski, Surf. Coat. Technol., 2006, 200, 5717–5724.
25. A. Robin, J. Appl. Electrochem., 2003, 33, 37–42.
26. Y. Zhang, Y. Zheng, Y. Li, L. Wang, Y. Bai, Q. Zhao, X. Xiong, Y. Cheng, Z. Tang, Y. Deng and S. Wei, PLoS One, 2015, 10, e0130774.
27. J. F. Flores, J. J. Olaya, R. Colás, S. E. Rodil, B. S. Valdez and I. G. Fuente, Corros. Eng., Sci. Technol., 2006, 41, 168–176.
28. Y. M. Lu, R. J. Weng, W. S. Hwang and Y. S. Yang, Thin Solid Films, 2001, 398–399, 356–360.
29. C.-C. Chang, J. S. Jeng and J. S. Chen, Thin Solid Films, 2002, 413, 46–51.
30. X. Sun, E. Kolawa, J.-S. Chen, J. S. Reid and M.-A. Nicolet, Thin Solid Films, 1993, 236, 347–351.
31. H. B. Nie, S. Y. Xu, S. J. Wang, L. P. You, Z. Yang, C. K. Ong, J. Li and T. Y. F. Liew, Appl. Phys. A, 2001, 73, 229–236.
32. T.-C. Li, B.-J. Lwo, N.-W. Pu, S.-P. Yu and C.-H. Kao, Surf. Coat. Technol., 2006, 201, 1031–1036.
33. K. Radhakrishnan, N. Geok Ing and R. Gopalakrishnan, Mater. Sci. Eng., B, 1999, 57, 224–227.
34. S. M. Aouadi and M. Debessai, J. Vac. Sci. Technol., A, 2004, 22, 1975.
35. A. Scandurra, G. F. Indelli, B. Pignataro, S. Di Marco, M. A. Di Stefano, S. Ravesi and S. Pignataro, Surf. Interface Anal., 2008, 40, 758–762.
36. D. Bernoulli, U. Müller, M. Schwarzenberger, R. Hauert and R. Spolenak, Thin Solid Films, 2013, 548, 157–161.
37. J. H. Han, H. Y. Kim, S. C. Lee, D. H. Kim, B. K. Park, J.-S. Park, D. J. Jeon, T.-M. Chung and C. G. Kim, Appl. Surf. Sci., 2016, 362, 176–181.
38. W. Lei, D. Liu, L. Shen, J. Zhang, P. Zhu, Q. Cui and G. Zou, J. Cryst. Growth, 2007, 306, 413–417.
39. C. Stampfl and A. J. Freeman, Phys. Rev. B: Condens. Matter Mater. Phys., 2005, 71, 024111.
40. D. Li, F. Tian, D. Duan, K. Bao, B. Chu, X. Sha, B. Liu and T. Cui, RSC Adv., 2014, 4, 10133–10139.
41. K.-Y. Liu, J.-W. Lee and F.-B. Wu, Surf. Coat. Technol., 2014, 259, 123–128.
42. B. D. Cullity, Elements of X-ray Diffraction, Adisson-Wesley Publishing, Reading, Massachusetts, 1956.
43. N. Fairley, CasaXPS VAMAS processing software, http://www.casaxps.com/.
44. I. Horcas, R. Fernández, J. M. Gómez-Rodríguez, J. Colchero, J. Gómez-Herrero and A. M. Baro, Rev. Sci. Instrum., 2007, 78, 013705.
45. D. Nečas and P. Klapetek, Cent. Eur. J. Phys., 2011, 10, 181–188.
46. M. Alishahi, F. Mahboubi, S. M. Mousavi Khoie, M. Aparicio, R. Hübner, F. Soldera and R. Gago, J. Power Sources, 2016, 322, 1–9.
47. ZPlot®/ZView™(version 3.3f), A software for EIS measurements and data analysis, written by D. Johnson, Scribner Associates Inc, Southern Pines, NC (USA), Copyright© 1990-2013, http://www.scribner.com.
48. D. A. Shirley, Phys. Rev. B: Solid State, 1972, 5, 4709–4714.
49. E. Niu, L. Li, G. Lv, W. Feng, H. Chen, S. Fan, S. Yang and X. Yang, Appl. Surf. Sci., 2007, 253, 5223–5227.
50. P. Lamour, P. Fioux, A. Ponche, M. Nardin, M.-F. Vallat, P. Dugay, J.-P. Brun, N. Moreaud and J.-M. Pinvidic, Surf. Interface Anal., 2008, 40, 1430–1437.
51. K. Baba and R. Hatada, Surf. Coat. Technol., 1996, 84, 429–433.
52. L. Li, E. Niu, G. Lv, X. Zhang, H. Chen, S. Fan, C. Liu and S.-Z. Yang, Appl. Surf. Sci., 2007, 253, 6811–6816.
53. A. Arranz and C. Palacio, Appl. Phys. A, 2005, 81, 1405–1410.
54. Crystallography Open Database (COD), http://www.crystallography.net, accessed 18 November, 2015.
55. K. S. Havey, J. S. Zabinski and S. D. Walck, Thin Solid Films, 1997, 303, 238–245.
56. L. Hultman, J. Vac. Sci. Technol., A, 1989, 7, 1187.
57. M. L. Addonizio, A. Castaldo, A. Antonaia, E. Gambale and L. Iemmo, J. Vac. Sci. Technol., A, 2012, 30, 031506.
58. S.-Y. Lin and Y.-S. Lai, ECS J. Solid State Sci. Technol., 2014, 3, N161–N165.
59. J. A. Thornton, J. Vac. Sci. Technol., 1974, 11, 666.
60. A. A. Navid and A. M. Hodge, Mater. Sci. Eng., A, 2012, 536, 49–56.
61. G. S. Chen, S. T. Chen, S. C. Huang and H. Y. Lee, Appl. Surf. Sci., 2001, 169, 353–357.
62. J.-C. Lin, G. Chen and C. Lee, J. Electrochem. Soc., 1999, 146, 1835–1839.
63. P. B. Barna and M. Adamik, Thin Solid Films, 1998, 317, 27–33.
64. I. Petrov, P. B. Barna, L. Hultman and J. E. Greene, J. Vac. Sci. Technol., A, 2003, 21, S117–S128.
65. M. A. Auger, L. Vázquez, O. Sánchez, M. Jergel, R. Cuerno and M. Castro, J. Appl. Phys., 2005, 97, 123528.
66. H. Koivuluoto, J. Näkki and P. Vuoristo, J. Therm. Spray Technol., 2009, 18, 75–82.
67. E. McCafferty, Corros. Sci., 2005, 47, 3202–3215.
68. S. Pugal Mani, A. Srinivasan and N. Rajendran, Int. J. Hydrogen Energy, 2015, 40, 3359–3369.
69. J. F. Flores, B. Valdez-Salas, M. Schorr and J. J. Olaya, Anti-Corros. Methods Mater., 2006, 53, 88–94.
70. A. Robin, Int. J. Refract. Met. Hard Mater., 1997, 15, 317–323.
71. Y. G. Shen, Y. W. Mai, D. R. McKenzie, Q. C. Zhang, W. D. McFall and W. E. McBride, J. Appl. Phys., 2000, 88, 1380.
72. J. Creus, H. Mazille and H. Idrissi, Surf. Coat. Technol., 2000, 130, 224–232.
73. I. M. Notter and D. R. Gabe, Corros. Sci., 1993, 34, 851–870.
74. F. C. Walsh, C. Ponce de León, C. Kerr, S. Court and B. D. Barker, Surf. Coat. Technol., 2008, 202, 5092–5102.
75. I. M. Notter and D. R. Gabe, Corros. Rev., 1992, 10, 217–280.
76. C. Liu, Q. Bi, A. Leyland and A. Matthews, Corros. Sci., 2003, 45, 1243–1256.
77. N. D. Nam, M. J. Kim, D. S. Jo, J. G. Kim and D. H. Yoon, Thin Solid Films, 2013, 545, 380–384.
78. J. C. Galván, M. T. Larrea, I. Braceras, M. Multigner and J. L. González-Carrasco, J. Alloys Compd., 2016, 676, 414–427.
79. J. Flis and A. Gajek, J. Electroanal. Chem., 2001, 515, 82–90.
80. D. Wallinder, J. Pan, C. Leygraf and A. Delblanc-Bauer, Corros. Sci., 1998, 41, 275–289.
81. C. Liu, Q. Bi, A. Leyland and A. Matthews, Corros. Sci., 2003, 45, 1257–1273.
82. M. Alishahi, S. M. Monirvaghefi, A. Saatchi and S. M. Hosseini, Appl. Surf. Sci., 2012, 258, 2439–2446.
83. Z. Bou-Saleh, A. Shahryari and S. Omanovic, Thin Solid Films, 2007, 515, 4727–4737.

---