Influence of Zr on structure, mechanical and thermal properties of Ti-Al-N.

Multinary Ti-Al-N thin films are used for various applications where hard, wear and oxidation resistant materials are needed. Here, we study the effect of Zr addition on structure, mechanical and thermal properties of Ti(1-x)Al(x)N based coatings under the guidance of ab initio calculations. The preparation of Ti(1-x-z)Al(x)Zr(z)N by magnetron sputtering verifies the suggested cubic (NaCl-type) structure for x below 0.6-0.7 and z ≤ 0.4. Increasing the Zr content from z = 0 to 0.17, while keeping x at ~ 0.5, results in a hardness increase from ~ 33 to 37 GPa, and a lattice parameter increase from 4.18 to 4.29 Å. The latter are in excellent agreement with ab initio data. Alloying with Zr also promotes the formation of cubic domains but retards the formation of stable wurtzite AlN during thermal annealing. This leads to high hardness values of ~ 40 GPa over a broad temperature range of 700-1100 °C for Ti(0.40)Al(0.55)Zr(0.05)N. Furthermore, Zr assists the formation of a dense oxide scale. After 20 h exposure in air at 950 °C, where Ti(0.48)Al(0.52)N is already completely oxidized, only a ~ 1 μm thin oxide scale is formed on top of the otherwise still intact ~ 2.5 μm thin film Ti(0.40)Al(0.55)Zr(0.05)N.

Introduction

Components like valves, tappets, and camshafts used in automotive industry, as well as tools for advanced machining processes (e.g., high-speed and dry cutn class="Chemical">ting) are exposed to severe tribological and thermal conditions. Aluminum based ternary transition metal nitrides Me–Al–N hard coatings (e.g., Ti–Al–N, Cr–Al–N, and Zr–Al–N) with cubic NaCl (c) structure, where Al substitutes for Me in the MeN based lattice (i.e., Me1-xAlxN), are in favor for such industrial applications due to their high hardness and wear resistance, together with good thermal stability, and oxidation resistance [1-11]. Among them, Ti–Al–N coatings are the most widely used to reduce tool wear. The crystal structure and mechanical properties of Ti–Al–N are depending on the Al content. Single phase cubic Ti–Al–N films with high Al contents exhibit excellent mechanical properties, age-hardening and oxidation resistance. For Al contents exceeding the maximum solubility (AlN mole fraction xmax ~ 0.7, depending on the deposition conditions, see Ref. [12]) in the cubic phase, a mixed cubic-NaCl and wurtzite-ZnS (w-AlN) structure is formed. The wurtzite configuration exhibits lower hardness, bulk-, elastic-, and shear-moduli, as well as wear resistance [13-17].

High thermal stability and oxidation ren class="Chemical">sistance of hard films are the key requirements of many industrial applications. Ti–Al–N films form a bi-layered Al2O3/TiO2 oxide scale, if exposed to air at elevated temperatures [18,19]. The oxidation resistance of Ti–Al–N films, with a reported onset temperature for oxidation of up to ~ 850 °C [19], is a result of the formation of dense Al2O3 scales, which retard the corresponding diffusion processes (simultaneous outward diffusion of Al and inward diffusion of O). However, the high growth rate of porous TiO2, when Ti–Al–N is exposed to temperatures exceeding 850 °C, results in crack formation also within the dense and protective Al2O3 outer-scale [18,19]. Consequently, break away oxidation occurs. In addition to the oxidation resistance also the age-hardening ability of Ti–Al–N films, which originates from spinodal decomposition to form cubic Al-rich and Ti-rich domains during annealing at temperatures below 1000 °C, [4-7,20] is attractive for industrial applications. Similar properties have led to the introduction of Zr1-xAlxN, where the reported metastable solubility limit for AlN in the c-Zr1-xAlxN is at 42–47 mol% [9,10,21-23]. Annealing of c-Zr1-xAlxN results in decomposing into their stable phases c-ZrN and w-AlN via the formation of cubic Al-rich domains and concomitant age-hardening up to an annealing temperature of 600 °C [21,22]. The transformation of the cubic Al-rich domains into the stable structure w-AlN for annealing treatments (or exposure during operation) above 1000 °C results in a rapid decline of the mechanical properties [4-6]. Furthermore, with increasing time of the exposure the corresponding transformation can occur already at lower temperatures.

Therefore, during severe industrial applications reaching or exceeding working temperatures of 1000 °C, or to increase the life-time of these coan class="Chemical">tings, further optimization of the oxidation resistance and the thermal stability in general is needed. An improvement of their properties can be achieved by the incorporation of large (substitutional) atoms (e.g., Nb, Ta and Y) into Ti–Al–N [24-28], as they effectively retard diffusional driven processes resulting in an increased thermal stability as well as improved mechanical properties. Alloying Zr to TiN and Ti–Al–N is reported to improve their hardness due to solid solution strengthening [29,30]. However, a detailed literature survey revealed that the impact of Zr incorporation on thermal stability and oxidation resistance of Ti–Al–N has not yet been explored.

In the present work we use differential scanning cn class="Chemical">alorimetry (DSC), thermal gravimetric analysis (TGA), stress measurements, X-ray diffraction analysis (XRD), and nanoindentation measurements to investigate the thermal stability, structure evolution, and mechanical properties of magnetron sputtered Ti–Al–Zr–N films for annealing temperatures up to 1500 °C. The effect of Zr-incorporation on the oxidation resistance of Ti–Al–N films is studied by means of DSC measurements in synthetic air, subsequent XRD, and cross-sectional scanning electron microscopy (SEM) investigations of coated polycrystalline Al2O3 substrates after isothermal oxidation for 20 h at 850 and 950 °C.

For a better understanding and explanation of the observed results we use density function theory (DFT) to cn class="Chemical">alculate the energy of formation, bulk moduli and lattice parameters of cubic and wurtzite modifications of the Ti1-x-zAlxZrzN solid solution with x between 0 and 1, and z between 0 and 0.3.

Experimental and calculation details

Ti1-x-zAlxZrzN films were deposited onto various substrates (see next paragraph) by unbalanced magnetron sputtering from a powder-metallurgically prepared Ti0.5Al0.5 compound target (diameter of 152 mm, thickness of 5 mm and purity of 99.9%, PLANSEE) in a mixed Ar + N2 (both of 99.999% purity) glow discharge. Ti1-x-zAlxZrzN films with different compositions were obtained by placing Ti and Zr platelets on the target race track (diameter of 5 mm and thickness of 1 mm). Prior to the deposition with a constant substrate temperature (Ta) of 500 °C, a base pressure ≤ 0.8 mPa, a bias potential of − 60 V, a N2-to-total pressure ratio of 17%, and a substrate-to-target distance of 85 mm, the substrates were etched for 20 min using an Ar+ glow discharge with − 1250 V and 25 mA, at a pressure of 3.0 Pa. More details on the magnetron sputtering system used are given in Ref. [31].

Different substrates were used for the individual investigations: austenitic n class="Chemical">stainless steel for hardness measurements of as deposited films; polished MgO (100) plates (10 × 10 × 1 mm3) for hardness measurements of as deposited and annealed films; Si stripes (20 × 7×0.3 mm³, both sides polished) for residual stress measurements; low alloyed steel for DSC and X-ray diffraction measurements of as deposited and annealed films and polycrystalline Al2O3 plates (10 × 10 × 1 mm3) for oxidation resistance measurements. Before loading the deposition chamber, the substrates were ultra-sonically cleaned in acetone and ethylene.

DSC with TGA was performed in a Netzch-STA 409C from room temperature (RT) to 1500 °C with a hean class="Chemical">ting rate of 20 K/min in flowing He (99.9% purity and 20 sccm flow rate) and synthetic air (79% N2, 21% O2, and 20 sccm flow rate). Prior to these measurements, the Ti–Al–Zr–N films were removed from their low alloyed steel substrates by chemical etching in 10 mol% nitric acid, in order to avoid substrate interference. For isothermal oxidation experiments, coated polycrystalline Al2O3 substrates were isothermally oxidized at 850 and 950 °C for 20 h in a conventional tube furnace, and afterwards investigated by fracture cross-sectional scanning electron microscopy (SEM) studies.

The chemical compositions of the films in their as deposited state and after the oxidation experiments were determined using energy dispersive X-ray analysis (EDX) with an Oxford Instruments INCA EDX unit attached to a SEM operated with 25 kV. The error of measurements for the metal atoms is below 2 at.%. Phase identification and structural investigations of the layers (after removal from their low alloy steel substrates) in their as deposited state and after thermal treatment with the DSC equipment in He or synthetic air were conducted by XRD with CuKα radiation using a Brucker D8 diffractometer in Bragg/Brentano mode. For classification of XRD peak, the JCDPS database was used [32]. Nanoindentation measurements of as deposited films and after vacuum annealing to 1100 °C were conducted with a CSIRO ultra-micro-indentation system (UMIS) using a Berkovich indenter. With respect to a proper statistic, at least 30 indents were performed for each sample with maximum loads ranging from 8 to 30 mN. Hardness and indentation moduli were calculated from the loading and unloading curves employing the Oliver–Pharr method [33]. Residual stresses σ of as deposited films were obtained using the substrate-curvature method. Detailed information on the measurement and the calculation of σ is described in Ref. [34].

Vacuum annealing of our films on n class="Chemical">MgO substrates was conducted in a vacuum furnace (pressure ≤ 0.1 mPa) to temperatures up to Ta = 1100 °C, using a heating rate of 20 K/min (corresponding to the DSC measurements).

DFT calculations were performed un class="Chemical">sing the VASP code [35] and employing the projector augmented wave pseudopotentials based on the generalized gradient approximation as parametrized by Perdew and Wang [36]. We used 500 eV for the plane wave cut-off energy and approximately 320 k-points/atom. Such parameters guarantee the calculation accuracy in the order of meV/atom. Special quasi-random structure (SQS) [37] based supercells (3 × 3 × 2 (36 atoms) for the cubic and 2 × 2 × 2 (32 atoms) for the wurtzite modification) were generated to model the ternary and quaternary alloys. This methodology allows to randomize the Ti and Al atoms on the metallic sublattice with respect to their short-range order (SRO) parameters. We used our own scripts for generating the supercells which are in detail described for the wurtzite phase elsewhere [38]. The resulting supercells used for the calculations in the present work are summarized in Appendix A (Tables A1 and A2). The deviation from ideal random-like behavior can be quantified by Warren–Cowley SRO parameters which are given in Table A3 in Appendix A. One can see that compared with a similar approach by Alling et al. [39], our cells exhibit in general worse SRO parameter values. This is likely to be cased by using considerably smaller supercells (36 and 32 atoms in the cubic and wurtzite phase, respectively compared with 64 atoms used by Alling et al. to model the cubic structure [39]). Nevertheless, even these rather small supercells have proved to be able to provide useful information on the structure and energetics of other similar ternary systems [40,41]. To obtain quaternary alloys we replaced one or two Ti or Al atoms by Zr. Thereby, we obtained cubic Ti1-x-zAlxZrzN structures with z = 0.056 and 0.111, and wurtzite Ti1-x-zAlxZrzN structures with z = 0.0625 and 0.125. Additionally, we also optimized an ad hoc cubic Ti0.33Al0.39Zr0.28N structure (the metal sublattice is composed of 6 Ti, 7 Al, and 5 Zr atoms).

Results and discussion

Structure and mechanical properties

Elementn class="Chemical">al analysis by EDX reveals that our Ti1-x-zAlxZrzN films are stoichiometric with N/metal ratios of 1 ± 0.1. As mentioned in the experimental section the composition of the metal sublattice is varied by adding Ti or Zr platelets at the race track of the Ti0.5Al0.5 compound target. Thereby the following films are obtained: Ti0.48Al0.52N with 21 Ti platelets, Ti0.40Al0.55Zr0.05N with 10 Zr platelets, Ti0.39Al0.51Zr0.10N with 20 Zr platelets, Ti0.36Al0.47Zr0.17N with 40 Zr platelets, and Ti0.34Al0.37Zr0.29N with 80 Zr platelets. XRD investigations of these as deposited films, as shown in Fig. 1, reveal a single phase cubic structure, which is in agreement with ab initio calculations. Fig. 2a presents the energy of formation, Ef, of the cubic and wurtzite Ti1-x-zAlxZrzN alloys with constant z = 0, 0.05 and 0.1 as a function of the AlN mole fraction x. The data suggest a transition from cubic to wurtzite structure Ti1-x-zAlxZrzN at x ~ 0.72, 0.70, and 0.68 for a ZrN mole fraction z of 0, 0.05, and 0.10, respectively. Since the compositional steps given by the supercell sizes are different for the cubic (1/18) and for the wurtzite (1/16) alloys, to evaluate the maximum solubility of AlN in the cubic phase we proceeded as follows. First, we fitted each set of data with constant ZrN mole fraction with a third order polynomial (a0 + a1·x + a2·x2 + a3·x3). For c-Ti1-x-zAlxZrzN, the AlN mole fraction x was varied 19 times for z = 0, 12 times for z = 0.055, and 10 times for z = 0.111. The calculations of w-Ti1-x-zAlxZrzN were obtained with 7, 6, and 5 variations in x for z = 0, 0.0625, and 0.125, respectively, in the composition range x = 0.5–1. Subsequently, for each phase (i.e. cubic or wurtzite) we fitted individually the coefficients (i.e., a0, a1, a2, and a3) of their third order polynomial for the three different ZrN mole fractions, z, with a linear expression in the ZrN contents. This way, two polynomial fits (one for the cubic and one for the wurtzite modification) as functions of x (AlN mole fraction) and z (ZrN mole fraction) were obtained. In the last step, we used these fits to estimate the cross-over between the formation energies of the cubic and wurtzite phases at fixed ZrN mole fractions (and thus to estimate the influence of Zr on the maximum AlN mole fraction in the cubic Ti1-x-zAlxZrzN). Fig. 2b shows this transition with a solid line connecting the data points (circles) as obtained for Ti1-x-zAlxZrzN with z = 0, 0.05, and 0.1 and ZrN-AlN, in the quasi-ternary TiN-AlN-ZrN diagram. It is worth noting, that the calculated transition point (maximum Al content in the cubic phase) is not too sensitive to a chosen order of the fitting polynomials: it ranges from 0.72 (z = 0) to 0.67 (z = 0.11) when using quadratic fits, from 0.705 (z = 0) to 0.68 (z = 0.11) for the third order polynomial, and from 0.70 (z = 0) to 0.68 (z = 0.11) for the fifth order polynomials. The chemical composition of our deposited films (indicated by asterisks) are deep within the single phase cubic region and hence in agreement to the XRD results. Additionally, we calculated the cubic-to-wurtzite transition for Zr1-zAlzN alloy to be ~ 0.5.

Fig. 1

XRD patterns of as deposited powdered Ti1-x-zAlxZrzN thin films.

Fig. 2

(a) Energy of formation, Ef, for c-Ti0.33Al0.39Zr0.28N and cubic and wurtzite phases Ti1-x-zAlxZrzN with z = 0, 0.05 and 0.1 as functions of AlN mole fraction. (b) Overall chemical compositions of our Ti1-x-zAlxZrzN films, in the as deposited state, plotted within the TiN–AlN–ZrN quasi-ternary phase diagram. The solid line indicates the transition between preferred cubic and wurtzite phases.

The shift of the XRD peak positions to lower 2θ angles indicate increan class="Chemical">sing lattice parameters from 4.18 to 4.19 to 4.23 to 4.31 Å with increasing ZrN mole fraction z for the films Ti0.48Al0.52N, Ti0.40Al0.55Zr0.05N, Ti0.39Al0.51Zr0.10N, and Ti0.34Al0.37Zr0.29N, respectively. These lattice parameters, obtained from powdered film samples after removing from their low alloy steel substrates, are in excellent agreement (maximum deviation of 0.2%) with the ab initio obtained lattice parameters of a = 4.184, 4.197, 4.238, and 4.320 Å for Ti0.50Al0.50N, Ti0.389Al0.556Zr0.056N, Ti0.389Al0.50Zr0.111N, and Ti0.33Al0.39Zr0.28N. Fig. 1 furthermore shows that the 111/200 peak intensity increases with increasing Zr content for the samples Ti0.40Al0.55Zr0.05N, Ti0.36Al0.47Zr0.17N, and Ti0.34Al0.37Zr0.29N, which is in agreement to our previous studies [3,16] and other reports [4,42] suggesting that with increasing Al content the 200 orientation is preferred. Stress measurements by the substrate-curvature method of films on Si substrates suggest increasing compressive stresses from around − 1.45 to − 3.53 GPa with increasing ZrN mole fractions z from 0 to 0.29, see Fig. 3.

Fig. 3

Hardness, H, indentation modulus, E, and residual stresses, σ, of Ti1-x-zAlxZrzN films as a function of z.

The nanoindentation obtained values for n class="Disease">hardness H and indentation moduli E of our Ti1-x-zAlxZrzN films are also presented as a function of their ZrN mole fraction in Fig. 3. While the compressive stresses scale with the ZrN mole fraction the maximum H and E values of ~ 38.1 and 499.7 GPa are obtained for the Ti0.40Al0.55Zr0.05N film which contains 5 mol% ZrN. A further increase in Zr content results in lower hardness values of ~ 36.0, 36.7, and 36.3 GPa for 10, 17, and 29% ZrN, respectively. This is connected with a decrease in the Al content and hence a decrease in the covalent character of bonding in the films [12], as the overall structure (see the XRD patterns presented in Fig. 1) and columnar morphology (not shown here) of the coatings are very similar. Nevertheless, the hardness values for the Ti0.36Al0.47Zr0.17N and Ti0.34Al0.37Zr0.29N films are by ~ 3–4 GPa above the hardness of ~ 32.7 GPa from the Zr-free reference film Ti0.48Al0.52N, which has a higher Al content. Consequently, Zr is very effective in increasing the strength of the materials. The indentation moduli E first increases from ~ 443 GPa for Ti0.48Al0.52N to 500 GPa for Ti0.40Al0.55Zr0.05N and then slightly decreases to 467 GPa for Ti0.34Al0.37Zr0.29N with increasing Zr, see Fig. 3. Ab initio calculations suggest a decrease in bulk moduli by ~ 6 and 12 GPa by the addition of z = 0.055 and 0.111 ZrN to c-Ti1-xAlxN to form c-Ti1-x-zAlxZrzN, respectively. In particular the calculated bulk moduli of Ti0.45Al0.55N, Ti0.395Al0.55Zr0.055N, Ti0.339Al0.55Zr0.111N, and Ti0.33Al0.39Zr0.28N are 261, 255, 249, and 245 GPa, respectively. Consequently, the experimentally observed rapid increase in H and E when adding 5% ZrN to Ti1-xAlxN is mainly attributed to microstructural changes.

Thermal stability

Fig. 4 shows simultaneous thermn class="Chemical">al analysis results of our Ti1-x-zAlxZrzN films, combining DSC (Fig. 4a) and TGA (Fig. 4b). The base line corrected dynamical DSC signal between 400 and 1500 °C consists of exothermic features from irreversible reactions. The shape of the broad DSC feature in the temperature range between 500 and 1100 °C suggests the contribution of several exothermic reactions for all films investigated, according to the results presented in Ref. [5]. With increasing ZrN mole fraction from 0 to 0.05 to 0.10 the peak position decreases from 800 °C for Ti0.48Al0.52N to 760 °C for Ti0.4Al0.55Zr0.05N to 740 °C for Ti0.39Al0.51Zr0.1N. However, a further increase of Zr content to z = 0.29 (Ti0.34Al0.37Zr0.29N) again results in a higher peak temperature of ~ 845 °C of this overall exothermic feature. For the exothermic feature at temperatures above 1100 °C, a shift in peak temperature from 1210 °C for Ti0.48Al0.52N to 1280 °C for Ti0.34Al0.37Zr0.29N is obtained with the incorporation of Zr. TGA measurements of these films indicate a decreasing mass loss of ~ 0.9, 0.6, 0.5 and 0.0% with increasing Zr content during annealing up to 1500 °C, see Fig. 4b. Based on previous studies, see Ref. [24], the mass loss can be correlated with N2-release, here, 1% mass loss corresponds to less than ~ 2 at.% N. Hence, the contribution to the exothermic output during DSC is mainly given by the decomposition of the metastable solid solution into their ‘stable’ nitrides, recovery and recrystallization. Ab initio obtained mixing enthalpies (from the binary constituents c-TiN, c-ZrN, and w-AlN) for c-Ti0.48Al0.52N, c-Ti0.4Al0.55Zr0.05N, c-Ti0.39Al0.51Zr0.1N, and c-Ti0.34Al0.37Zr0.29N are 104, 116, 119, and 120 meV/at, respectively, and hence increase with increasing ZrN content. When using a cubic solid solution between TiN and ZrN, i.e. c-Ti1-yZryN (with y = z/(1-x)) as a constituent next to w-AlN, we obtain mixing enthalpies of 104, 98, 87, and 72 meV/at for z = 0, 0.05, 0.10, and 0.29, respectively. In Ref. [43] it is reported that TiN and ZrN forms a solid solution, which is stable up to at least 1200 °C during thermal annealing in vacuum for 2 h, and which also exhibits almost no exothermic contribution during DSC. The latter reference and the comparison of the mixing enthalpies with the DSC experiments, which exhibit an overall exothermic contribution of 192, 232, 227, and 140 W·K/g for the coatings c-Ti0.48Al0.52N, c-Ti0.4Al0.55Zr0.05N, c-Ti0.39Al0.51Zr0.1N, and c-Ti0.34Al0.37Zr0.29N, suggests (by the overall trend of decreasing enthalpies with increasing Zr content) that there is no separation into the constituents c-TiN, c-ZrN, and w-AlN (increasing mixing enthalpies with increasing Zr) but into c-Ti1-yZryN and w-AlN (decreasing mixing enthalpies with increasing Zr), see above. This is verified by XRD analysis of samples annealed to various temperatures using the DSC equipment with the same setup, atmosphere, heating and cooling rates.

Fig. 4

(a) DSC and (b) TGA scans in inert atmosphere (He) of our Ti1-x-zAlxZrzN thin films.

Fig. 5 shows the structural evolution during annen class="Chemical">aling of our Ti0.48Al0.52N (a) Ti0.4Al0.55Zr0.05N (b) and Ti0.39Al0.51Zr0.1N (c) films by means of XRD patterns after annealing to 700, 850, 1100, 1200, and 1500 °C. The Zr-free Ti0.48Al0.52N film exhibits a small shift of the XRD peaks during annealing to 700 °C as compared with the as deposited state, see Fig. 5a, suggesting only minute structural changes like recovery and relaxation which contribute to the exothermic DSC feature in this temperature range. The XRD patterns of the Zr-containing films Ti0.4Al0.55Zr0.05N and Ti0.39Al0.51Zr0.1N annealed to 700 °C reveal also a shift in the peak position to higher diffraction angles but also an increase in peak broadening, see Fig. 5b and c. The latter is an indication for a reduction in grain size and/or an increase in microstresses which can result from the onset of a decomposition process [14,20]. This can better be seen after annealing at 850 °C, where the XRD peaks exhibit on both sides (lower and higher diffraction angles) an increase in intensity and width, suggesting the formation of Al-depleted and Al-enriched domains. After annealing at 1100 °C, a pronounced shoulder-formation on both sides of the ‘matrix’ XRD peak can be seen clearly. These shoulders indicate the formation of TiN- and AlN-rich cubic domains for Ti0.48Al0.52N, and TiN-, ZrN- and AlN-rich cubic domains for the Zr-containing films. While the Zr-free film, Ti0.48Al0.52N, exhibits the formation of w-AlN already after annealing at 1100 °C, no w-AlN formation can be detected for the Zr-containing films (compare Fig. 5a, b, and c), indicating that Zr effectively retards the formation of w-AlN. After annealing at 1200 °C, almost no intensity at the XRD peak positions of as deposited films can be detected, indicating close-to-complete decomposition of the original supersaturated matrix. The Ti0.48Al0.52N film is mainly composed of TiN- and AlN-rich cubic phases and w-AlN, whereas it is still hard to detect any w-AlN for the Zr containing films, Ti0.4Al0.55Zr0.05N and Ti0.39Al0.51Zr0.1N, which at this stage compose mainly of ZrN-, TiN- and AlN-rich cubic phases.

Fig. 5

XRD patterns after annealing in vacuum to temperatures Ta up to 1500 °C of (a) Ti0.48Al0.52N, (b) Ti0.40Al0.55Zr0.05N, and (c) Ti0.39Al0.51Zr0.10N.

Increasing the temperature further to 1500 °C results in a completed decomposition to form the stable counterparts c-TiN and w-AlN for the Zr-free film, and c-Ti1-yZryN and w-AlN for the Zr-containing films, see Fig. 5. The development of the XRD patterns of the Zr-containing films with temperature suggests that the TiN and ZrN rich cubic domains, formed at temperatures below 1200 °C, combine to form the solid solution c-Ti1-yZryN phases present after annealing at 1500 °C. The obtained cubic phases of Ti0.89Zr0.11N and Ti0.80Zr0.20N have lattice parameters a = 4.28 and 4.31 Å, respectively. This behavior is further studied with the structural development of the higher Zr containing film Ti0.34Al0.37Zr0.29N as a function of annealing temperature, see Fig. 6. The XRD patterns after annealing at 850, 950 and 1100 °C suggest the formation of cubic structure ZrN-, TiN- and AlN-rich domains. The small peaks at the positions for c-TiN and c-ZrN move towards c-Ti0.54Zr0.46N (a = 4.39 Å) with a further increase in temperature to 1200, 1300 and 1500 °C. The formation of w-AlN can only be detected for Ta ≥ 1300 °C, see the small peaks at 2θ = 33.2 and 38.2°.

Fig. 6

XRD patterns of Ti0.34Al0.37Zr0.29N after annealing in vacuum to temperatures Ta up to 1500 °C.

The structural changes occurring in the films during annen class="Chemical">aling result in corresponding changes of their mechanical properties. The hardness values of our Ti1-x-zAlxZrzN films for annealing temperatures Ta up to 1100 °C are presented in Fig. 7. The hardness of Ti0.48Al0.52N slightly decreases from ~ 31.9 to 31.4 GPa, within the error of the measurement of ± 1.5 GPa, during annealing to 700 °C. The previous structural investigations indicate that in this temperature range mainly recovery and relaxation processes occur, which can explain the small change in H. With a further increase in Ta, and hence the onset of the decomposition of the supersaturated cubic matrix to form cubic domains, the hardness increases to ~ 34.7 GPa with Ta = 1000 °C. The formation of w-AlN and coarsening of individual phases during annealing to 1100 °C leads to a decrease in H to ~ 30.3 GPa (compare Figs. 7 and 5). The 5 mol% ZrN containing film Ti0.40Al0.55Zr0.05N exhibits an almost steady increase in H from the as deposited value of 37.7 ± 2.6 GPa to 42.0 ± 2.3 GPa with Ta = 1000 °C within the error of measurement. After annealing at 1100 °C the hardness only slightly decreased to 39.7 ± 1.8 GPa. The higher Zr containing film Ti0.39Al0.51Zr0.1N shows also an increase in H from the as deposited value of 33.5 ± 2.0 GPa to 37.0 ± 2.7 GPa with Ta = 900 °C. A further increase in Ta results in a decrease of H to 34.6 ± 2.5 GPa at Ta = 1100 °C. Consequently, the higher Zr containing film reaches its peak hardness already at 900 °C. The earlier increase in hardness followed by a later decrease with increasing temperature is consistent with the structural investigations indicating an earlier onset of the decomposition process to form cubic domains but postponed formation of w-AlN for the Zr containing films as compared with Ti0.48Al0.52N, see Figs. 5 and 6.

Fig. 7

Hardness, H, of Ti1-x-zAlxZrzN layers on MgO (001) as a function of Ta.

Oxidation resistance

Fig. 8 shows the results of dynamical DSC experin class="Species">ments of powdered freestanding film samples up to 1500 °C in synthetic air, allowing for a qualitative comparison of the temperatures where distinct reactions occur. The Ti0.34Al0.37Zr0.29N film shows the lowest onset temperature (~ 800 °C) of exothermic heat flow of the films investigated, which is mainly attributed to the low Al content (x ~ 0.37) and the high amount of Zr, which is highly affine to O2. The other films with a higher Al content (x ~ 0.5) exhibit the pronounced oxidation reaction at temperatures above 1000 °C with the Zr-containing films exhibiting an additional pre-reaction in the temperature range 870–1030 °C. The latter can especially be observed for Ti0.40Al0.55Zr0.05N. The XRD analyses of Ti0.48Al0.52N and Ti0.40Al0.55Zr0.05N freestanding films after annealing in ambient air at 1000 °C for 1 h, Fig. 9a, indicate that Ti0.48Al0.52N is almost completely oxidized to form rutile, r-TiO2, (with some anatase, a-TiO2,) and α-Al2O3. On the contrary, the XRD pattern of Ti0.40Al0.55Zr0.05N after annealing in air is very similar to that after annealing in vacuum, Fig. 9b, and only small XRD peaks, suggesting the formation of TiO2, α-Al2O3, and Zr0.33Ti0.67O2 can be detected. Consequently, the pre-reaction to the strong exothermic contribution during DSC of Ti0.40Al0.55Zr0.05N (Fig. 8) is assigned to the formation of stable oxides which slow down a further oxidation.

Fig. 8

DSC spectra of powdered Ti1-x-zAlxZrzN films in synthetic air (20 sccm) with a 20 K/min heating rate.

Fig. 9

XRD scans of Ti0.48Al0.52N and Ti0.40Al0.55Zr0.05N after (a) annealing at 1000 °C in synthetic air (20 sccm) for 1 h, and (b) vacuum annealing to 950 °C with a 20 K/min heating rate (b).

Fig. 10 shows back scattered SEM fracture cross sections of Ti0.48n class="Chemical">Al0.52N (a), Ti0.4Al0.55Zr0.05N (b), and Ti0.39Al0.51Zr0.1N (c) films after oxidation at 850 °C for 20 h. On top of the Zr-free Ti0.48Al0.52N film, a multilayered, ~ 2 μm thick oxide scale composed of a dense Al2O3 outer-layer (~ 0.3 μm) and porous TiO2 containing sub-layer can be identified with the assistance of EDX. Here, already some crack formation within the sub-layer can be observed, see Fig. 10a. The oxide scale on top of the lower Zr-containing film, Ti0.4Al0.55Zr0.05N, is with 0.7 μm much thinner whereas that on the higher Zr-containing film, Ti0.39Al0.51Zr0.1N, is with ~ 2.3 μm even thicker than that on Ti0.48Al0.52N, see Fig. 10b and c. Nevertheless, the oxide scales on both Zr-containing films are dense and uniform, without any layered structure or crack formation. EDX analyses confirm the uniform (within the detection limit) distribution of Ti, Al, Zr, and O. After oxidation at 950 °C for 20 h, the Zr-free film, Ti0.48Al0.52N, is completely oxidized, see Fig. 11a. The Zr-containing films, Ti0.4Al0.55Zr0.05N and Ti0.39Al0.51Zr0.1N, still exhibit intact nitride layers underneath their oxide scales with thicknesses of ~ 1.0 and 3.4 μm, respectively, see Fig. 11b and c. These observations are in agreement with the XRD results presented in Fig. 10 and clearly suggest that the addition of ~ 5% ZrN to Ti1-xAlxN yields optimized oxidation resistance.

Fig. 10

SEM fracture cross-sectional images of (a) Ti0.48Al0.52N, (b) Ti0.40Al0.55Zr0.05N, and (c) Ti0.39Al0.51Zr0.10N after isothermal oxidation at 850 °C for 20 h. The arrow in (a) indicates the formation of a crack.

Fig. 11

SEM fracture cross-sectional images of (a) Ti0.48Al0.52N, (b) Ti0.40Al0.55Zr0.05N, and (c) Ti0.39Al0.51Zr0.10N after isothermal oxidation at 950 °C for 20 h.

Summary and conclusions

Here, we present a detailed study on the effect of Zr incorporation on structurn class="Chemical">al, mechanical and thermal properties of Ti1-xAlxN thin films prepared by reactive magnetron sputtering. All films with compositions of Ti0.48Al0.52N, Ti0.40Al0.55Zr0.05N, Ti0.39Al0.51Zr0.10N, Ti0.36Al0.47Zr0.17N, and Ti0.34Al0.37Zr0.29N exhibit a single phase cubic NaCl-type structure in the as deposited state, which is in agreement with ab initio calculations. Addition of Zr to Ti1-xAlxN results in an increased hardness, where the maximum value of ~ 38 GPa is obtained for Ti0.40Al0.55Zr0.05N.

During annealing in vacuum, the supersaturated matrix phase n class="Chemical">Ti1-x-zAlxZrzN decomposes to form Al-depleted and Al-enriched domains which cause an increase in hardness. Further annealing above 1000 °C results in the formation of w-AlN and hence a decreased hardness. Zirconium is very effective in influencing the age-hardening as it promotes the formation of cubic domains but retards the formation of w-AlN. Consequently, the hardness increases to values above ~ 35 GPa already after annealing at 700 °C and stays above ~ 35 GPa even after annealing to 1100 °C. The 5% ZrN containing layer, Ti0.40Al0.55Zr0.05N, exhibits the best results of the films investigated. Its hardness increases from the as deposited value of ~ 38 GPa to ~ 42 GPa with Ta = 1000 °C and ~ 40 GPa with Ta = 1100 °C. In addition to the improvement of the age-hardening behavior of Ti1-xAlxN by alloying with Zr, also their oxidation resistance increases. Zirconium promotes the formation of a dense oxide scale and thereby retards its growth rate. Also here, the best results are obtained for Ti0.40Al0.55Zr0.05N, which shows even after annealing in air at 950 °C for 20 h only a ~ 1 μm thick oxide scale on top of the otherwise still intact layer.

Based on the results presented, we can conclude that the coating Ti0.40n class="Chemical">Al0.55Zr0.05N which contains only 5% Zr at the metal sublattice exhibits the best performance during hardness measurements in the as deposited state and after thermal annealing up to 1100 °C as well as during oxidation tests for 20 h at 850 and 950 °C.