j ourna l homepage: www.e lsev ie r .com/ locate /sur fcoat
Mechanisms of topography formation of magnetron-sputteredchromium-based coatings on epoxy polymer composites☆
Juergen M. Lackner a,⁎, Wolfgang Waldhauser a, Christian Ganser b,c, Christian Teichert b,Marcin Kot d, Lukasz Major e
a JOANNEUM RESEARCH Forschungsges.m.b.H., Institute of Surface Technologies and Photonics, Functional Surfaces, Leobner-Strasse 94, A-8712 Niklasdorf, Austriab Montanuniversitaet Leoben, Institute of Physics, Franz-Josef-Strasse 18, A-8700 Leoben, Austriac Christian Doppler Laboratory for Surface Chemical and Physical Fundamentals of Paper Strength, Graz University of Technology, Petersgasse 16/2, A-8010 Graz, Austriad AGH University of Science and Technology, AlejaAdamaMickiewicza 30, 30-059 Krakow, Polande Polish Academy of Sciences, Institute of Metallurgy and Materials Sciences, ul. Reymonta 25, 30-059 Krakow, Poland
☆ This manuscript is based on work presented at the SoAnnual Technical Conference in Providence, Rhode Island⁎ Corresponding author. Tel.: +43 316 876 2305; fax: +
The tribological protection of stiff technical polymers like epoxy resins, being main component in fiberstrengthened composites (CFC), gains increasing interest to enable their use in mechanical engineering ap-plications. Besides hardening of the surfaces to achieve similar properties like low carbon steels, smoothsurfaces are essential in sliding contact. However, the conditions of film growth during magnetronsputtering including thermal stresses result in high intrinsic film stresses, which trigger the formation ofdistinct surface topography. Especially chromium nitride (CrNx) single layer coatings show a fragmentationof the coating by cracking during film growth. Preferential growth of crystallites in subsequent depositioncloses these cracks, provides complete covering of the surface, and bulged topographical features. Goal ofthis work is describing the mechanisms of formation of these topographical features as well as emphasizinginfluences of higher film toughness by Cr-CrNx multilayer coatings on the density and height of the bulges.
Carbon fiber polymer composites (CFC) are used for a wide range ofindustrial products (aeronautics, automotive, medical devices, etc.).Their advantage is rather easy fabricationwith complex shapes. However,CFCperformance in termsof abrasion, sliding, and impactwear resistanceis very limited. Thus, a replacement of steel components by light-weightCFC demands effective surface wear protection by appropriate coatings.Applicable coating techniques are electrochemical/galvanic depositionor thermal and plasma spraying, which provide thick coatings of highload support [1–3]. However, they lack in adhesion during overloading,since their stiffness is too high and their plastic deformability is too lowto follow substrate deflection. Alternatively, thick soft polymer coatings(pure andmicro-/nanoparticle strengthened lacquers) possess high elas-ticity for bending to follow substrate deflection, but they fail in tribologi-cal resistance [4]. Thin films of materials combining hardness and wearresistance with high compliance as well as high resistance to cohesiveand adhesive crack propagation (high toughness) are future candidatesfor surface protection of high performance polymers [5–7]. Kääriäinenet al. [8] proposed chromium nitride (CrNx) single layer coatings on
ciety of Vacuum Coaters 56th, April 20-25, 2013.43 316 8769 2305.
ckner).
ghts reserved.
CFC, but they were not effective in tribological protection. Althoughthey achieved high adhesion of the magnetron-sputtered PVD coatingsdue to surface activation by high ion doses before film deposition orhigh substrate temperature, the realization of dense coatings was impos-sible. This was assigned to a very inhomogeneous and rough substratesurface, consisting of fibers with loose ends and polymer filler. For im-proved coatings, the authors stated that the surface of CFC should haveafiber-free layer to obtain a sufficiently smooth surface before deposition.We observed roughening of the CFC surfacewith thick, smooth epoxy toplayer at the micrometer scale during coating deposition in ownwork [9].This rougheningoccurs not homogenously over thewhole coated surface,but is dominated by domed lines (bulges) between fragmented, tablet-like structures.
Here, we describe this bulping effect as it is observed formagnetron-sputtered Cr-based coatings onCFC in detail and explain its deformationmechanisms. Additionally, influences of tougher multilayer vs. brittlesingle layer coatings on the formation of these topographical featuresare emphasized
2. Experimental details
2.1. Coating deposition
For deposition of single layer CrNx and multilayer Cr-CrNx coatingswith low friction top layers of diamond-like carbon (amorphous carbon,
a-C), we used epoxy-based CFC plates (manufactured by SecarTechnologie Ges.m.b.H., Hönigsberg, Austria), silicon (100) wafers, andpolished austenitic steel substrates (DIN EN 1.4301, AISI 304). Theywere cleaned in an industrial washing machine by using commercialtenside agents, followed by a final cleaning with ethanol before mount-ing in the vacuum chamber. After pumping down to high vacuum(2×10−3 Pa), organic contaminations were removed and the polymersurfaces were chemically activated using plasma of an anode layer ionsource [10,11]. Unbalanced magnetron sputtering from 4 rectangularcathodes (432 mm height) was afterwards applied to deposit Cr, CrNx,and a-C coatings from high purity Cr (99.99%, RHP Technology GmbH,Seibersdorf, Austria) and electrographite carbon targets (99.9%, SchunkGroup, Bad Goisern, Austria), respectively. The following coating archi-tectures were developed based on former works [12,13]:
(1) single layer CrNx coating with 3.94 ± 0.03 μm thickness,(2) Cr-CrNx multilayer coatings with totally 16 bilayers and a modu-
(3) Cr-CrNxmultilayer coatingwith totally 32 bilayers and amodula-tion ratio Cr:CrNx = 1:2 (42 nm Cr + 84 nm CrNx), total thick-ness 4.18 ± 0.03 μm.
To improve adhesion, we started deposition with a 50 nm Cr inter-face. 1 μma-Cfilmswere used as top layer on all three coating types, ex-cept for samples for dedicated topography and hardness investigations.Cr aswell as a-C deposition occurred in an inert Ar process gas. For CrNx,a mixture of Ar and N2 with a flow ratio N2:Ar = 3:1 was used in reac-tive deposition. The working pressure during deposition was set to2.9 Pa. Bias voltage of –50 V was applied to increase the energy densityon CFC substrates during deposition. The DC magnetron power, set to12 – 18 W cm−2, was controlled to prevent high substrate heating
Fig. 1. SEM surface topography and cross-section images of CrNx-based coatings with ~1 μm thCrNx single layer coating and (c) a Cr-CrNxmultilayer coating on Si substrate. (d) Surface of a 32on epoxy CFC substrate. The gap between the substrate and coating is due to delamination durchanically introduced notch at the back side of the sample).
exceeding the low thermal stability of the epoxy matrix (degradationstarting above ~140 °C). Finally, we achieved deposition ratesof ~3.4 μm h−1 for Cr and ~2.3 μm h−1 for CrN.
The temperaturemeasurement during depositionwas performed bytemperature measuring tapes (Testo, Testoterm), attached on the backside of the samples in the chamber. These tapes indicate, if a distinctmaximum temperature is reached during deposition.
2.2. Coating characterization
To study the topographical formation phenomena in comparison be-tween coated CFC and Si wafers, atomic force microscopy (AFM) wasapplied on an Asylum Research MFP-3D equipment in soft tappingmode. Olympus AC160TS Type 3 silicon probes with a nominal tip radi-us of 7 nmand anopening angle of about 35°were used for allmeasure-ments. From each surface a 30 × 30 μm2 area was scanned as overviewwith additional 10 × 10 and 5 × 5 μm2 scans for detailed analysis.Roughness analysis to obtain the root mean square (RMS) roughness(standard deviation of height values) was applied on whole imagesand selected areas.
Scanning electron microscopy (SEM, Zeiss EVO 50) was used tostudy surface morphology and growth structure of cross sections. Forcross-section sample preparation, cuttingwith an ATMBrillant 221 sys-tem was performed up to 2/3 of the sample thickness from backside.Fracturing of the coated CFC substrates occurred manually by tensileloadingwith pincers. For increasing the conductivity duringmicroscopyimaging, the specimens were evaporated with thin gold layers.
Structural investigations for obtaining crystallographic phases andphase orientation on coated epoxy CFC and Si substrateswere performedon a Bruker AXS D8 Advance diffractometer with CuKα radiation,equipped with Sol-X detector and Goebel mirror. Bragg–Brentano
ick a-C top layers: (a) Surface of a 32 bilayer coating on Si substrate. (b) Cross section of abilayer coating on epoxy CFC substrate. (e) Cross section of a 32-bilayer Cr-CrNxmultilayering sample preparation for cross-section imaging (fracture in liquid nitrogen along a me-
Fig. 2. AFM images (5 × 5 μm2 scan size) of (a, d) a single layer CrNx coating, (b, e) a 16 bilayer and (c, f) a 32 bilayer Cr-CrNx multilayer coating of ~4 μm thickness with ~1 μm a-C toplayer. Substrates for (a–c) are Si wafers and for (d–f) epoxy CFC plates. Furthermore, height profiles along marked lines are given for (d–f) below the images.
(locked couple) scanning geometry was applied between 20 and 80°with a step width of 0.02° and 1.2 s measurement duration per step.
Film thickness (given above) onmasked stepswasmeasured by stylesprofilometry (Veeco Dektak 150). The elastic modulus (E) of the coatingswas determined by a nanoindentation setup according to DIN EN ISO14577-1:2002 with a Vickers indenter (Fisherscope H100C). The appliedmaximum loads were 50 mN, the loading rates 20 nm s−1. Siliconwafersubstrates were employed to prevent any influences of high substratecompliance inmeasuringmechanical coating properties. The indentationresults were afterwards analyzed using Oliver and Pharr theory [14] inorder to obtain E. Vickers indentation with 9.8 N force (HV1) was usedto achieve defined deformation in coated epoxy CFC to deduce qualita-tively the toughness from scanning electron microscope images (SEM,Zeiss Auriga). A cone-shaped diamond indenter with 20 μm tip radiuswas used for indentation at 2 N normal force to achieve fracture in thesingle andmultilayer coatings for analyzing the deformation and coatingfracture on focused ion beam (FIB) cut cross sections. FIB cutting wasdone by Ga ion beam on a Quanta 200 3D DualBeam microscope withan in-situ OmniProbe micro manipulator. Subsequent imaging occurredby transmission electron microscopy (TEM, Tecnai G2 F20, 200 kV FEG).
Table 1RMS roughness of uncoated substrates (Si wafers, epoxy CFC) for CrNx single layer and Cr-CrNstrates are given for whole surfaces and selected areas (area size given in brackets, see e.g. Fig
Substrate Measurement area Scan size [μm2] Uncoated sub
Depositing CrNx-based coatings by magnetron sputtering at lowtemperatures on Si substrates results in the well-known formation ofa domed-shaped topography (Fig. 1a). The size of the sphericalmicrodomes, which are tops of growing columnar micro cones, wasfound to be in the 500 nm range. In the cross sections, the CrNx singlelayer on Si (Fig. 1b) shows a thin columnar structure of tapered crystalsand fibrous grains, which can be assigned to the “Zone 1” structure inthe Thornton structure model [15]. Such structure is generally foundfor PVD coatings deposited at low temperatures with features ofmicrocracks between the columns, pinholes, transient grain boundariesand often a high degree of through-coating porosity. At higher ion bom-bardment during deposition, this “Zone 1”merges to a transition “ZoneT” of fibrous grains with decreased porosity. The refinement of grainsoccurs in the deposited multilayer Cr-CrNx coating (Fig. 1c), althoughthe ion bombardment is rather similar to the single layer CrNx coating.Thus, both the ion bombardment, which triggers surface diffusion dur-ing deposition, and the repeated nucleation at the layer boundaries den-sify the microstructure similarly and diminish loosely bond cone
x multilayer coatings with 1 μm a-C:H top layers in nm. Values for coated epoxy CFC sub-. 4a) in between domed lines (statistics: 3 specimens).
strate CrNx single layer 16 bilayer Cr-CrNx 32 bilayer Cr-CrNx
Fig. 3. AFM overview images (30 × 30 μm2 scan size) of the surface topographies on (a) a CrNx single layer coating (an example for a “selected area”, used for roughness calculation inTable 1, is marked), (b) a 16 bilayer, and (c) a 32 bilayer Cr-CrNx multilayer coating of ~4 μm thickness with 1 μm a-C top coating on epoxy CFC substrate.
boundaries [16], as later discussed based on XRD results. As expected,the a-C layer on the top (Fig. 1a,b) is amorphouswithout anydistinct co-lumnar growth features, which depict on its surface the topography ofthe underlying Cr-CrNx hard coating without significant changes.
The micrometer scale topography of CrNx coatings on epoxy CFC isastonishingly different to Si or metal/ceramic substrates (Fig. 1a vs. d),while the growth structures in the cross sections are only slightly coars-er (Fig. 1c vs. e). Additionally to the sub-micron range features of spher-ical micro-domes on the surface (Fig. 1a), “pits” are found due to poreson the substrate surface and “lines” cover the whole surface, dividingthe coating to segments (Fig. 1d). These “lines” have dome-shaped,bulged tops. These domes are one order of magnitude larger than themicrodomes, occurring at the low temperature growth both on Si andepoxy CFC. Furthermore, they are elongated from triple point to triplepoint, where the lines hit. Film growth is much coarser in the conebelow the dome (Fig. 1e), and a crack is present at the bottom of thecone. The imprint of the crack is visible on the epoxy surface too, reveal-ing its formation during film growth and not as artifact duringpreparation.
The AFM images in Fig. 3 show this topography difference for CrNx
single and multilayer coatings on stiff Si (Fig. 2a–c) and rather flexibleepoxy CFC substrates (Fig. 2d–f). Images on coated Si were taken fromrandom sites, while especially triple points and their surroundingwere chosen for investigation on coated CFC. As visible, the smallestmicrodome size and most regular distribution are found for the 16 Cr-CrNx bilayer coatings (Fig. 2b, e). Furthermore, the feature size of the
Fig. 4. Locked-couple X-ray diffraction scans of a single 4 μmCrNx single layer and 16 and32 bilayer Cr-CrNx multilayer coatings with 1 μm a-C top layer on epoxy CFC (black spec-tra) and on Si (100) (grey spectra) including assignment of peaks to standard peak posi-tion from Powder Diffraction File (ICDD) (Cr (bcc): 00-027-1402, CrN (fcc):03-065-2899, β-Cr2N: 00-035-0803)) [19].
microdomes for both multilayer coatings is rather similar (compareFig. 2b and e, Fig. 2c and f). Homogeneity in the size distribution is lostfor the CrNx single layer coating only on Si (Fig. 2a), but not for the sim-ilar coating on the CFC substrate (Fig. 2d). Accompanying quantitativedata on RMS roughness, revealing the qualitative impression of Fig. 2,is given in Table 1: Comparing the average values after analyzing thewhole surface of the AFM scans (30×30 μm2, shown for coatings onepoxy CFC in Fig. 3), roughness is increasing by a factor of at least 5for the 32 bilayer Cr-CrNx coatings compared to Si. Higher increasesoccur for the 16 bilayer multilayer (factor 6) and single layer CrNx coat-ings (factor 16). In contrast, the analysis of RMS roughness for bothselected areas in between the domed lines for CFC and on Si (Table 1)reveals quite similar results for these areas with microdomes(RMS ~ 10–13 nm). The shape analysis of these microdomes on CFC,shown by the height profiles in Fig. 2d–f, indicates the round-shapedtops with steep flanks, which are well-known (Zone-T structure) forhard coatings deposited at low temperatures on hard substrates (steels,ceramics, hard metals) [15].
The density of domed lines is qualitatively observable in the AFMoverview images in Fig. 3: It increases slightly for the 16 bilayer Cr-CrNx multilayer (Fig. 3b) and strongly for CrNx single layer coatings(Fig. 3a) compared to the 32 bilayer coating (Fig. 3c), which has in con-trast the largest bulge height of the lines.
TheXRDphase analysis in Fig. 4 reveals both rather similar occurringphases and orientations for coatings on CFC and for Si as well as thelower crystallinity of the coatings on epoxy CFC substrate. The CrN(200) and Cr (110) peaks for coatings on Si have much higher intensitythan on CFC, which is linked to higher phase content (higher crystallin-ity) and/or larger crystal grain size. Explanations could behigher electri-cal conductivity of Si compared to CFC during biased deposition or thecontamination of the deposition atmosphere by water degassing fromthe CFC composite. All peaks are generally broad for both substratetypes. Their sharpness decreases for the multilayer compositions too,revealing smaller grain sizes. Furthermore, peaks for CrNx compounds(face-centered cubic (fcc) CrN, hexagonal (hex) β-Cr2N [17]) are ratherweak for the multilayer coatings, but they dominate the XRD spectra ofthe single layer around labeled (200) CrN/(111) β-Cr2N peaks. Onlyweak and broad diffraction is present for the β-Cr2N phase (revealingnanocrystalline behavior), which is stronger in (111) and (002) orienta-tion for coatings on epoxyCFC. These peaks, decreasing in intensity fromCrNx single layer to the 16 and 32 bilayer Cr-CrNx multilayers, may giveevidence on phase formation in the cones below the domed lines due tosimilar decrease in density of these lines. However, the XRD results donot fully answer the question of distinct CrNx phase formation withinthese cones. Indications for no differences in the phase compositionand orientation are found in the AFM phase image in Fig. 5, revealingsimilar phases inside the domes along the lines and around.
Searching for the reasons of formation of these bulged lines, thecrack formed below the cone in Fig. 1e as well as mechanical consider-ations provide following explanation: Thermal intrinsic stresses occur
in the vacuum coating of epoxy CFC due to the increasing tempera-ture up to maximum ~90 °C during large area deposition of the4 μm films. Much higher thermal expansion coefficients of theepoxy layer at the CFC surface (55 × 10−6 m m−1 K−1) than theceramic CrN, Cr-CrNx, and a-C coatings (values for bulk materials:Cr: 4.9 × 10−6 m m−1 K−1, CrN: 2.3 × 10−6 m m−1 K−1, a-C: 2 -4 × 10−6 m m−1 K−1) [18,19] trigger strong thermal intrinsic ten-sile stresses in the coating, which lead to brittle through-thicknessfracture after exceeding tensile strength in mechanically weakerregions (e.g. starting from defects on the substrate surface).
Bending of the substrate-coating compound, as found in former ownwork for wrinkling of hard films on soft substrates [20], is evident toobecause of the quite different elastic moduli of the coating (Table 2)and epoxy (~2.5 GPa). However, this has much higher importance dur-ing cooling down after coating: The larger thermal shrinking of theepoxy compared to the coating leads to both tensile stress in epoxyand compressive stress in the film. Due to dense film growth, stress re-laxation occurs by common deformation of the coating and substratesurface, as found by a wavy surface topography and described as(nano-)wrinkling for titanium nitride on polyurethane [20]. However,both the thicker coating and the less flexible substrate decrease thewave density (wavelength on themmscale) and increase the amplitude(on micron scale) for the wrinkle effect on coated CFC.
Nevertheless, any spreading of through-thickness cracks along theinterface coating-epoxy is missing thanks to high adhesion by chemi-cal binding between the 50 nm adhesive Cr interface and the polymer,being revealed in scratch tests with critical loads Lc2 for coating adhe-sive failure above 30 N too [9]. The density of the cracks on the surfacedepends both on the thermal stresses and the mechanical toughnessof the coating. Thermal stresses are contributed by the temperaturerise and difference in thermal expansion coefficients between sub-strate and coating. However, the final temperature during depositionis rather similar for all deposits (~90 °C). Differences in the thermalexpansion coefficient between bulk Cr and CrN materials (data seeabove) are small too (factor 2), especially if mentioning the low Crlayer content (~33%, modulation ratio Cr:CrNx = 1:2) and the muchlarger thermal expansion of epoxy. Both do not explain the muchhigher density of domed lines (and cracking during film growth) ofthe single layers (Fig. 3).
Table 2Elastic modulus (E) of 4 μm CrNx-based coatings without 1 μm a-Ctop layers as well as for 1 μm a-C on Si substrates, obtained bynanoindentation.
The explanation can be found in the very different toughness of thecoatings: Toughness has tremendous influence on the fracture under ten-sile stresses, whichwas tested byHV1Vickers indentation (Fig. 6). Gener-ally, the penetration of an indenter results in elasto-plastic deformation ofthematerial. Soft substratematerials like epoxyCFCprovide only low loadsupport and start to deform (visco-)elastically even under low loading.Hence, this results in high surface deformation due to low elastic. Thecoating on the epoxy CFC surface has to follow this deformation and isheavily bent, whereby maximum tensile stresses are found at the inter-face to the substrate below the contact area of the indenter and at thecoating surface outside in the bulged region. This bulge outside the indentis formed by elasto-plastic material flow outwards of the indented zone.Both radial and circular ring cracks form in the coating if stress level ex-ceeds tensile strength in these directions (σr, σΘ) (compare to theshear-lag model in [21,22]). Based on former theoretical and experimen-tal works [12,13], this regionwas also applied in thiswork for qualitative-ly description of coating toughness: SEM surface images reveal for theCrNx single layer coating (Fig. 6a) a much denser crack net in the bulgedregion around the indents than for the tougher 32 bilayer Cr-CrNx multi-layer coating (Fig. 6b). Some circular cracks are found, but radial cracksare missing except along the edges of the Vickers pyramid.
Detailed investigations of the coating fracture were performed byTEM on FIB cut cross sections of indentations, obtained by ball indenta-tion (Fig. 7). Compared to Vickers pyramid indentation with sharpedges and notch effects, ball indentation provides less complex stressconditions, being quite homogenous in circumferential direction [23].While cracks run straight through the coating thickness in CrNx singlelayer coating, toughening mechanisms are well visible in the multi-layers, as shown in Fig. 7 for the 32 bilayer Cr-CrNx coating. The topview of the surface in Fig. 7a shows both higher density of radial cracks
Fig. 6. SEM images of Vickers HV1 indentations in (a) CrNx single layer and (b) 32 bilayerCr-CrNxmultilayer coatingswith 1 μma-C top layer on epoxy CFC. Stresses (σ) and crack-ing in radial (r) and circular (θ) direction are indicated.
Fig. 7. (a) SEM top view of an indent, formed by 2 N indentation load in a 32 bilayer Cr-CrNx multilayer with 1 μm a-C on epoxy CFC. (b) TEM cross section indicating deformation andcracking. The shown cracks propagate from the initiation site at the surface under tensile stresses perpendicular to the surface, but change its direction to a shear crack parallel to the filmsurface after passing about 75% of the film thickness.
and two circular (ring) cracks. Cracks partly follow the domed lines,which are mechanically weaker than the segments in between. Thering cracks, analyzed in the cross section in Fig. 7b, spread from the coat-ing surface at the periphery of the contact area (bulged region) towardsthe substrate similar to single layer coatings. However, crack propaga-tion changes from tensile fracture normal to the substrate surface toshear fracture parallel to the substrate surface before the crack reachesthe coating – epoxy interface. This behavior releases elastic strain ener-gy, toughens thematerial, and finally explains both the lower density ofdomed lines (thermal cracks) and higher domes on these lines. If thenon-fractured segments are larger at similar thermal stress, the separa-tion of these segments is larger due to concentration of straining. Thisgives additional hints on themechanism of preferential growth of crys-tallites at the thermal tensile cracks: In a formerwork [24], we show re-sults of multilayer growth on dust-contaminated surfaces. A large CrNx
cone started to grow above such defect, depicting it on the coating sur-face. In the case of the thermal tensile fracture, the crack edges are sep-arated from another after stress release. During the DC biaseddeposition, such edge provides different electrical potential, attractinga higher degree of charged particles. In combination with film growthat lower compressive stress at and above the crack than found far offthe tensile cracks, this explains the formation of the larger cones(domed lines): Larger separation of segments as found for themultilay-er increases the free space for the formation of cones, leading to finallylarger and higher domes.
4. Conclusions
The currentwork explained the formationmechanismsof elongated,domed lines as micron-scale surface topography feature of magnetron-sputtered Cr-based single and multilayer hard coatings on epoxy CFCpolymers. Formation starts from tensile fracture during film growthdue to thermal stresses (about 10 times higher thermal expansion ofepoxy compared to thin film materials). Micro cones nucleate at thefracture sites, growing preferably in distinct crystallographic structuredue to electrical charging and film strains. Finally, these micro conesform the domed lines, visible in surface imaging. The formation mecha-nismof the coatings in between these lines is quite similar (only slightlyrougher) than found on stiff, crystalline Si substrates. By intruding amultilayered structure of Cr-CrNx instead of the CrNx single layer coat-ing, the tendency to fracture is reduced due to plastic deformabilityand toughness increase in the coating. However, the height of thedomes (bulges) is higher by concentration of larger thermal strainsand, thus, larger separation of segments.
Acknowledgements
The financial support of this work by the Austrian Research Promo-tion Agency (FFG) in the frame of the program “Intelligent Production”,by the Austrian Federal Ministry of Traffic, Innovation and Technologywithin the frame of the Austrian Nanoinitiative programme, by thePolish–Austrian exchange project PL 12/2010 (funded in Austria bythe Oesterreichischer Austauschdienst OeAD), by the Polish NationalScience Centre in the frame of the project No. 3066/B/T02/2011/40, aswell as by the Country of Styria and European Union is highly acknowl-edged. The author’s thank the Department of Physical Metallurgy andMaterials Testing, University of Leoben as well as the Materials CenterLeoben Forschung Ges.m.b.H. for performing SEM imaging and XRDstudies.
References
[1] M. Ivosevic, R. Knight, S.R. Kalidindi, G.R. Palmese, J.K. Sutter, J. Therm. SprayTechnol. 14 (2005) 45.
[2] G. Palumbo, I. Brooks, K. Panagiotopoulos, US patent 7387578.[3] C. Friedrich, R. Gadow, M. Speicher, Surf. Coat. Technol. 151–152 (2002) 405.[4] M. Biron, Thermosets and Composites, Elsevier, Oxford, 2004.[5] S. Zhang, D. Sun, Y. Fu, H. Du, Surf. Coat. Technol. 198 (2005) 2.[6] A. Karimi, Y. Wang, T. Cselle, M. Morstein, Thin Solid Films 420–421 (2002)
275.[7] Z. Chen, B. Cotterell, W. Wang, Eng. Fract. Mech. 69 (2005) 597.[8] T. Kääriäinen, M. Rahamathunnisa, M. Tanttari, D.C. Cameron, Bull. Soc. Vac. Coaters
(2007) 39.[9] J.M. Lackner, Proc. NanoUpDate Helsingborg 2013, Proc, March 27–28 2013.
28.[10] J.M. Lackner, W. Waldhauser, M. Schwarz, L. Mahoney, D. Burtner, 50th Ann. Techn.
Conf. Proc. Society of Vacuum Coaters, 2007. 87.[11] J.M. Lackner, W. Waldhauser, M. Schwarz, L. Mahoney, L. Major, B. Major, Vacuum
83 (2009) 302–307.[12] J.M. Lackner, L. Major, M. Kot, Bull. Pol. Acad. Sci. 59 (2011) 343.[13] J.M. Lackner, W. Waldhauser, B. Major, L. Major, M. Kot, Thin Solid Films (2013),
http://dx.doi.org/10.1016/j.tsf.2013.03.025, (accepted for publication).[14] W.C. Oliver, G.M. Pharr, J. Mater. Res. 7 (1992) 1564.[15] J.A. Thornton, Ann. Rev. Mater. Sci. 1 (1997) 239.[16] D.B. Lewis, Q. Jo, P.E. Hovsepian, W.D. Münz, Surf. Coat. Technol. 184 (2004)
225.[17] JCPDS - International Centre for Diffraction Data, Cards No. 00-027-1402 (Si),
00-035-0803 (β-Cr2N), 03-065-2899 (CrN), 00-027-1402 (Cr) (2006).[18] P.H. Mayrhofer, G. Tischler, C. Mitterer, Surf. Coat. Technol. 142 (2001) 78.[19] F.C. Marques, R.G. Lacerda, Appl. Phys. Lett. 83 (2003) 3099.[20] J.M. Lackner, W. Waldhauser, P. Hartmann, O. Miskovics, F. Schmied, C. Teichert, T.
Schöberl, Thin Solid Films 520 (2012) 2833.[21] J. Andersons, Y. Leterrier, Thin Solid Films 471 (2005) 209.[22] U.A. Handge, J. Mater. Sci. 37 (2002) 4775.[23] D. Tabor, Microindentation techniques inMaterial Science and Engineering, American
Society of Testing of Materials, Ann Arbor, 1986, p. 129.[24] J.M. Lackner, W. Waldhauser, B. Major, J. Morgiel, L. Major, H. Takahashi, T.
