We arc deposit Cr-rich Cr-N coatings and show that these coatings are a promising alternative to electrodeposited hard chrome. We find that the substrate bias is of importance for controlling the N content in the grown coatings as it determines the degree of preferential resputtering of N. The substrate bias also affects the substrate temperature and film growth rate. Higher bias results in higher temperatures due to higher energy transfer to the substrate, while the growth rate decreases due to an increased re-sputtering. The N content affects the morphology, microstructure, hardness, and resistivity of the coatings. The hardness increases from 10 GPa with 0.5 at. % N to 17 GPa with 7.5 at. % N, after which no further increase in hardness is seen. At the same time, the grain structure changes from columnar to more featureless and the resistivity rises from 15 to 45 μΩ cm.
Almost 100 years have passed since Colin G. Fink was granted a patent1 for an industrially viable method of electrodepositing Cr on metal components. Since then, Cr coatings have been widely applied for mechanical and decorative purposes on metal- and nonmetal components alike. For wear resistance applications, so-called electrodeposited hard chrome (EDHC) is the dominating coating, finding its use in, for example, hydraulic piston rods,2 piston rings for internal combustion engines,3 gun barrel bores,4 and woodcutting equipment.5 The high hardness of EDHC, recently reported to be ∼10 GPa, as determined by nanoindentation,6 is an important feature in many of the applications. This hardness, much higher than the 2–3 GPa of bulk Cr7 and 5–6 GPa of coatings deposited by physical vapor deposition (PVD),6 has been attributed to the inclusion of oxygen in the EDHC during the deposition process.8
While the high hardness of EDHC makes it attractive to use in many applications, a major drawback is the toxic compounds used in the deposition process. The most common source of Cr in the process is CrO3, which is found in solution in the electrolyte of the plating bath. In CrO3, Cr is hexavalent, a form in which Cr is strongly carcinogenic and consequently is included in the European Chemicals Agency’s list of substances of very high concern,9 meaning that its use is strongly regulated within the European Union. Thus, it is important to find replacement coatings for applications where EDHC has historically been employed.
Several studies on using PVD as a replacement have been carried out,3,10–13 with common replacement candidates being CrN and TiN, materials which have been shown to exhibit hardness above that of EDHC. In the present work, we investigate the path of PVD as a replacement of EDHC coatings in the application of woodcutting. Previous studies have shown that it is possible to increase the hardness of Cr above the level of EDHC with 5 at. % N in the coatings for the case of high power impulse magnetron sputtering (HiPIMS),14 and as little as 2 at. % for direct current magnetron sputtering (DCMS).15 We explore the area of low-level N-alloying of Cr grown by cathodic arc deposition (CAD), investigating what process parameters influence the coating composition and structure and how these, in turn, influence the hardness of the coatings.
The Cr-rich Cr-N coatings were arc deposited onto Si(100) substrates without external substrate heating in a high vacuum (HV) chamber at a base pressure <10−2 Pa using two high purity (99.8%) Cr cathodes supplied by Plansee Composite Materials GmbH, Lechbruck am See, Germany. The cathode-to-substrate distance was 17 cm and the cathodes were positioned parallel above the substrate table, see Fig. 1. The arc discharge was sustained by means of two SM66-AR-100 power supplies from Delta Elektronika B.V., Zierikzee, Netherlands, one for each cathode. The current was kept at a constant 65 A for both cathodes in all depositions, with the voltage varying depending on the process, typically 21 ± 1 V. The N2 partial pressure was varied, with steps at 0.05, 0.11, 0.16, 0.21, 0.27, 0.32, 0.37, and 0.43 Pa with Ar to balance for a total pressure of 0.67 Pa. Because the pressure changes upon ignition of the arcs, the pressure during operation differs from 0.67 Pa. For the sake of clarity, all pressure values reported in this work refer to the preignition pressure. Some depositions were carried out with the substrate at the floating potential, which was measured to be ∼−4 V. Otherwise, an intentionally applied DC substrate bias was used with the amplitude of −30, −50, and −70 V. The deposition time was 10 min for all coatings. Prior to deposition, the substrates were ultrasonically cleaned in isopropanol for 5 min.
The elemental composition of the coatings was determined by ToF-E ERDA. The measurements were carried out with a 36 MeV 127I ion beam at the Tandem Laboratory of Uppsala University. The incident angle of primary ions and exit angle of recoils were 67.5° to the sample surface normal, resulting in a recoil angle of 45°. The measured ToF-E ERDA spectra were converted into relative atomic concentration depth profiles using the POTKU code.16 The chemical bonding structure of the film surfaces was determined using x-ray photoelectron spectroscopy (XPS), carried out with a Kratos AXIS Ultra DLD spectrometer, employing monochromatic Al Kα (1486.6 eV) radiation. All spectra were measured at the normal emission angle with no charge compensation. The analyzer pass energy was set to 20 eV, resulting in the full width at half maximum (FWHM) of 0.55 eV for the Ag 3d5/2 peak from the calibration samples. The Cr 2p and N 1 s spectra were recorded before and after sputter etching with Ar+-ions. For the latter treatment, 4 keV Ar+ beam incident at 70° from the surface normal was first used during 150 s After that, the ion energy was reduced to 0.5 keV for the additional 10 min, with the aim to minimize the effects of sputter damage.17 The sample area etched by the ion beam was 3 × 3 mm2 while the analyzed area was 0.3 × 0.7 mm2. The spectrometer binding energy scale was calibrated according to the ISO standards for monochromatic Al Kα sources that place Au 4f7/2, Ag 3d5/2, and Cu 2p3/2 peaks at 83.96, 368.21, and 932.62 eV, respectively.18 All core-level spectra were charge-referenced to the Fermi edge cut-off which is a more reliable approach than the C 1s method.19 The CasaXPS software (version 2.3.19)20 was used to analyze the XPS data.
X-ray diffraction (XRD) was used to determine the phase distribution of the coatings. This was characterized by θ/2θ scans in a PANalytical X’Pert PRO diffractometer, using Cu Kα radiation (Cu Kα, λ = 1.54 Å) at 40 kV and 40 mA.
The microstructure and thicknesses of the coatings were determined from cross-sectional scanning electron microscopy (SEM) images captured using a Zeiss Sigma 300, operated at an acceleration voltage of 5 kV. The macroparticle coverage of the sample surfaces was calculated from top-view SEM images using the image processing software imagej.21 This was done for the coatings deposited at the floating potential and at −70 V.
A Hysitron TI 950 nanoindenter was used to measure the hardness of the coatings via the Oliver–Pharr method.22 The indenter was equipped with a Berkovich tip and operated at a load that resulted in an indentation depth <10% of the coating thickness, to avoid contributions to the measured hardness from the Si substrate.
Room temperature four-point probe measurements were performed on 20 × 20 mm samples with a probe spacing of 1 mm (geometrical correction factor 0.98) using a Jandel RM3000 system. The in-plane resistivity was then calculated by multiplying the obtained sheet resistance by the film thickness obtained by SEM imaging of cross sections of films deposited on Si(100) substrates.
III. RESULTS AND DISCUSSION
A. Temperature evolution
As seen in Fig. 2, the substrate temperature increases with deposition time for all amplitudes of the substrate bias, including the floating potential. This shows that the plasma heating23 is significant, even at the floating potential. When substrate bias is applied, the energy of the ions impinging on the film surface is
where is the energy of the ion before acceleration through the plasma sheath, n is the charge state of the ion, is the substrate bias, and is the plasma potential. This increase in energy results in the substrate temperature rising further, as can be seen in Fig. 2 where, for example, the final temperature is ∼40% higher in the case of −70 V than at the floating potential (310 vs 220 °C).
B. Chemical composition and bonding structure
The atomic concentration of N as determined by ToF-E ERDA as a function of N2 partial pressure for different bias amplitudes is shown in Fig. 3 with (a) floating potential, (b) −30 V, (c) −50 V, and (d) −70 V bias. As a general trend, the higher the substrate bias, the lower the amount of N in the coatings for each applied N2 partial pressure. The exceptions are the coatings deposited at 0.05 and 0.11 Pa, where there is virtually no N present, and 0.43 Pa, where the amount of N is around 26 at. % independent of substrate bias. Thus, the incorporation of N in the coatings is affected by preferential resputtering upon ion bombardment, the energy of which scales with increasing the bias amplitude, see Eq. (1). Given the fact that the species evaporated by a cathodic arc are typically ionized to a high degree (+2 has been shown to be the average charge state for arc evaporated Cr at similar ambience24), the energy of the bombarding Cr ions will be, on average, at least 60, 100, and 120 eV for −30, −50, and −70 V bias, respectively. At these energy levels, preferential sputtering of N has previously been reported for sputter deposition of Cr-rich Cr-N films,15 where an increase in substrate bias from −30 to −100 V reduced the N content from 18 to 2 at. %.
The fact that preferential sputtering stops at 0.43 Pa can be understood by considering the phases in which N is incorporated. For N2 pressures lower than 0.43 Pa, there is a mix between incorporation of N in the bcc-Cr phase and formation of Cr2N. When substrate bias is applied, the N incorporated in bcc-Cr is preferentially removed. The higher the substrate bias, the more significant this removal becomes. At 0.43 Pa, however, the N atoms are predominantly found in the Cr2N-phase, from which it has been previously shown that N is not preferentially sputtered, even at significantly higher ion bombardment energies.25 The shift in bonding environment can be seen from the XPS spectra in Fig. 4, where the coating containing ∼26 at. % N exhibits characteristics of Cr2N,14 with a reduction of peak intensities and shift toward lower binding energies for Cr 2p and a shift toward higher binding energy for the N 1s peak. The existence of, and transition between, bcc-Cr and Cr2N is further discussed in Sec. III D.
C. Microstructure, morphology, and thickness
For low N2 partial pressures, the coatings grow in a columnar fashion, as can be seen in Fig. 5(a). As the N2 partial pressure is increased, the N content in the films increases and coatings become more featureless. An example of this is seen in Fig. 5(b) for a coating which contains ∼14 at. % N. This change in microstructure is likely due to the presence of nano-dispersed Cr2N, effectively hindering grain coalescence during growth, resulting in smaller grains. A similar transition has been shown for Cr-rich Cr-N films grown by HiPIMS.14 Grain size reduction is further corroborated by the XRD results, especially noticeable for low substrate bias, see Sec. III D
Figure 6 shows the thickness of the deposited coatings as a function of N2 partial pressure, averaged over all substrate bias levels, as well as the thickness as a function of substrate bias averaged over all N2 partial pressures. No effect on the growth can be seen when changing the N2 partial pressures. The growth rate is, however, affected by the substrate bias. This can be seen when comparing the coatings deposited at −70 V where the average thickness is 20% lower than in the case of coatings grown at the floating potential. This is due to either re-sputtering effects, film densification, or a combination thereof.
Macroparticles are present for all samples, as is typical for unfiltered CAD. In Fig. 7, the surface area covered by macroparticles is presented as a function of N content for the case of the floating potential and −70 V. The area coverage can be seen to increase with increasing N content up until ∼15 at. %, to then decrease for the two samples containing ∼25 at. % N. This coincides with the transition from bcc-Cr to Cr2N, as discussed in Sec. III B and which will be described further in Sec. III D.
From Fig. 7, it can also be observed that the coatings deposited at −70 V contain less macroparticles than those deposited at the floating potential. This is because macroparticles attain negative charge while traversing the plasma between the cathodes and the substrate, resulting in a repulsive force that scales with the substrate bias. For macroparticles of small mass, the force at −70 V is strong enough to prevent them from being incorporated into the sample surface.28
D. Crystalline phase content
The crystalline phase content of the coatings as determined by XRD in a Bragg–Brentano geometry is shown in Fig. 8. The Cr 110, 200, 211, 220, and 310 peaks are visible at 44.4, 64.6, 81.7, 98.5, and 310°, respectively. The first sign of the Cr2N 111 peak, with a powder reference peak position of 42.6°, appears at 0.16 Pa for the sample deposited at the floating potential, containing ∼4 at. % N, where the Cr 110 peak is asymmetrically broadened toward the lower angle side. At 0.21 Pa, this broadening can be seen also for −30 V (8 at. % N) and −50 V (2 at. % N). For N2 pressures equal to or higher than 0.32 Pa, all coatings contain at least 7 at. % N and broadening or defined shoulder peaks from Cr2N 111 are present for all coatings. For the coatings deposited at 0.43 Pa, two additional Cr2N peaks are discernible, namely, 110 at 37.3° and 002 at 40.2°, while the intensity from Cr 110 is reduced and showing up as a shoulder peak on the high angle side of the Cr2N 111 peak. Also, because the Cr2N 302- and the Cr 211 peak positions are so close, it is difficult to discern them from each other, but it is likely the case that the Cr 211 peak is reduced in intensity as the N content increases, as can be seen when comparing the coatings deposited different substrate bias at 0.16 Pa. The peak reappearing at its place at higher N contents is consequently the Cr2N 302 peak. For all N2 partial pressures, the peak widths are reduced with increasing substrate bias, improving the discernability of the individual peaks. This is likely an effect of increased grain size, brought on by the increased temperature at higher bias, see Fig. 2. Additionally, as the N content increases, the apparent peak width increases for the same bias, corroborating the decrease in grain size as discussed in Sec. III C. The evolution of the width of the Cr 110 peak with N content is shown in Fig. 9. Because this peak is convoluted with the Cr2N 111 peak on the low angle side for all coatings with N content > ∼1 at. %, a slight overestimation of FWHM could be expected for those samples. The samples deposited at 0.43 Pa is excluded because the Cr 110 peak only appears as a weak shoulder on the Cr2N 111 peak. The width increase with increasing N content is largest for the coatings deposited at the floating potential, as shown in Fig. 9(a). In Figs. 9(b)–9(d), it can be seen that the increase in FWHM is not as large for the coatings deposited with an applied substrate bias. This is likely because of the aforementioned temperature driven grain coarsening counterbalancing the grain size reduction brought on by the presence of Cr2N.
The hardness as determined by the Oliver–Pharr method using nanoindentation is shown as a function of N content in Fig. 10. For little or no N, the hardness is ∼9 GPa and increases to 16 GPa when the N content increases up to 8 at. %. This is close to the reported values for bulk Cr2N28 and higher than the 10 GPa of EDHC.6 No hardness increase is seen with further increasing N content. This is likely a result of the nano dispersed structure discussed in Sec. III C. This gives rise to a hardness increase because the Cr2N nanograins effectively hinder dislocations movement in the coatings. The plateauing of hardness after ∼8 at. % N is in line with the work by Greczynski et al.,14 but it occurs at lower N concentrations than what was reported in that study. The fact that the increase in hardness takes place at an N content lower than 10 at. % is important information for future development of EDHC replacement coatings for woodcutting applications. The grain size difference between coatings deposited at the same N2 partial pressure with different substrate bias, as reported in Sec. III D, seems to have no influence on the coating hardness.
The coating resistivity as a function of N content is shown for the different substrate bias levels in Fig. 11. The resistivity values for the coatings containing the least N are around 15 μΩ cm, close to the bulk Cr value of 12.9 μΩ cm.29 Addition of N raises the resistivity value to ∼75 μΩ cm at around 15 at. % N, after which additional addition of N has a smaller effect on the resistivity. This trend is observed for all bias levels, but the absolute values differ, with higher bias yielding lower resistivity. At ∼26 at. % N, the values are 87, 82, 79, and 76 μΩ cm for floating, −30, −50, and −70 V, respectively. The lower resistivity at higher bias can be attributed to larger grains and thus fewer grain boundaries, a factor known to affect resistivity.30
The general trend of the resistivity increase slowing down with increasing N content is in line with previous work by Greczynski et al.,14 where a plateau was reached for the resistivity, albeit at a N content of ∼5 at. %. The mechanism behind the evolution of resistivity with N content is likely a product of simultaneous loss of bcc-Cr in favor of the more resistive Cr2N and an increase in the number of grain boundaries. The more bcc-Cr there is in the coatings the lower the resistivity is, but also the larger the effect of grain refinement is since more grains are affected. As the amount of N increases in the samples and Cr2N becomes the majority phase, the impact of bcc-Cr grain refinement on the resistivity diminishes and the resistivity approaches a pure Cr2N resistivity.
IV. SUMMARY AND CONCLUSIONS
We apply CAD to grow Cr-rich Cr-N coatings as a replacement alternative for EDHC in woodcutting applications. We show that for small N2 partial pressures, the substrate bias is of importance for controlling the N content. For floating substrate bias, the N content increases linearly with N2 partial pressure while also yielding the highest growth rates, lowest substrate temperature, and smallest grains. The N content, in turn, controls the morphology, crystal structure, hardness, and resistivity of the coatings. For example, by changing the N content from 0.5 to 7.5 at. % the coating hardness can be increased from 10 to 17 GPa while the resistivity increases from 17 to 45 μΩ cm. Our results are promising for the prospect of applying CAD to grow Cr-rich Cr-N coatings as a replacement for EDHC.
The research leading to these results has received funding from the Swedish Foundation for Strategic Research (SSF) and Contract No. ID17-0055. Accelerator operation was supported by the Swedish Research Council VR-RFI (Contract No. 2019-00191). H.H. acknowledges financial support from the Swedish Government Strategic Research Area in Materials Science on Advanced Functional Materials at Linköping University (Faculty Grant SFO-Mat-LiU No. 2009-00971).
Conflict of Interest
J.N., M.J., N.S., and S.K. report a relationship with Husqvarna AB that includes employment.
Johan Nyman: Conceptualization (equal); Formal analysis (equal); Investigation (lead); Writing – original draft (lead). Grzegorz Greczynski: Formal analysis (supporting); Investigation (supporting); Writing – original draft (supporting); Writing – review & editing (equal). Muhammad Junaid: Conceptualization (equal); Project administration (supporting); Supervision (supporting); Writing – review & editing (supporting). Niklas Sarius: Conceptualization (equal); Funding acquisition (equal); Project administration (supporting); Supervision (supporting); Writing – review & editing (supporting). Sören Kahl: Conceptualization (equal); Funding acquisition (equal); Project administration (supporting). Jens Birch: Funding acquisition (supporting); Project administration (supporting); Supervision (supporting). Hans Högberg: Conceptualization (equal); Funding acquisition (equal); Project administration (supporting); Supervision (supporting); Writing – review & editing (supporting).
The data that support the findings of this study are available from the corresponding author upon reasonable request.