JCT Research, Vol. 1, No. 4, October 2004
Effects of Thickness and Substrate on the Mechanical Properties of Hard Coatings Zenghu Han, Jiawan Tian, Jijun Lao, Geyang Li**—Shanghai Jiao Tong University* Jiawei Dai—Key Lab for High Temperature Materials & Tests of Ministry of Education†
Nanoindentation technique was employed to characterize the mechanical properties of TiN coatings deposited on highspeed steel and stainless steel substrates. Effects of thickness and substrate on the mechanical properties were investigated. The results show that TiN coatings exhibit different mechanical properties corresponding to the variation in thickness and substrate. Keywords: Nanoindentation, TiN coatings, thickness, substrate properties, mechanical properties
H
ard coatings can improve the wear resistance and elongate the service life of tools and dies. TiN coatings have been widely used to enhance the tribological performance of materials because of their high hardness and low friction coefficient. Mechanical properties are the most important properties of hard coatings. Although the chemical compositions and microstructures (crystal structure, crystalline size, orientation, and defects, etc.) are the determining factors of the mechanical properties, the mechanical performance of hard coatings is also strongly influenced by the thickness of coatings and the properties of the substrates on which the coatings are deposited. This influence has been theoretically acknowledged, but accurate and reliable data for quantitative analysis are still lacking because the high hardness and the micron-level thickness of coatings make the measurement of mechanical properties very difficult. For instance, the reported Vickers hardness (HV) of TiN coatings ranges from 21–31 GPa 1 and it is very difficult to distinguish the various influences on the mechanical properties of hard coatings. This presents a great deal of trouble for the evaluation of mechanical properties and the control of the product quality of hard coatings. Many empirical rules and amendatory formulas have been proposed to improve the accuracy of the mechanical property measurement of hard coatings. Some researchers suggested that the ratio of maximum indentation depth to coating thickness be limited to
eliminate the influence of substrate deformation on measurement. For different coating/substrate combinations, this ratio has been proposed at 1/5, 1/10, and even 1/20.2-6 For example, Saha and Nix6 observed that the substrate hardness affected the measured coating hardness in the case of a hard coating hardness on a soft substrate. In such cases, the true hardness of the coatings could be determined from the indentation data only if the indenter displacement was less than 10% of the coating thickness. Yet whether or not the influence of the substrate had been eliminated was not verified. Some researchers7-10 also proposed empirical amendatory formulas for extracting the real coating hardness and elastic modulus from indentation tests of coating/substrate composites under a large testing load. However, because of the diversity among hard coatings and the variation in their properties caused by depositing parameters, coating thickness, microstructure, stress state, and substrate properties, the above empirical rules and amendatory formulas cannot meet the requirements of accurate measurement. In this article, a penetration method which can correctly and reliably measure the mechanical properties of hard coatings using the continuous penetration technique was employed to obtain the hardness and modulus of TiN coatings of different thicknesses on high-speed steel (HSS) and stainless steel (SS) substrates. Then, the effects of coating thickness and substrate properties on the mechanical properties of hard coatings were demonstrated.
*State Key Lab of MMCs, Shanghai 200030, People’s Republic of China. 200030, People’s Republic of China. **Author to whom all correspondence should be addressed. Email:
[email protected]. †Shanghai
www.coatingstech.org
October 2004
337
Z. Han et al. elastic modulus). The effect of substrate hardness on the coating hardness (HV) was negligible only if the substrate did not yield plastically under the applied load. In such cases, the plastic deformation was contained within the coatings and the “true” Vickers hardness could be obtained using the Oliver and Pharr’s method. Compared to hardness, the nanoindentation measurement of the elastic modulus of coatings is more strongly affected by the substrate. This is because the elastic field under the indenter tip is not confined to the coating itself; rather, it is a long-range field that extends into the substrate, especially when the coating thickness is small. To illustrate,
Figure 1—The load/unload indenter displacement with a large load (200 mN).
EXPERIMENTAL PROCEDURES Deposition of TiN Coatings TiN coatings were deposited on HSS and SS substrates by a reactive DC magnetron sputtering at room temperature. The substrates were polished with 0.5-µm diamond paste before being ultrasonically cleaned in acetone and alcohol, and then mounted on the substrate holder in the vacuum chamber. The TiN coatings were deposited from a pure Ti target (99.9%) in an argon and nitrogen mixture atmosphere. The base pressure of the chamber, prior to sputtering, was pumped to 3 × 10–3 Pa. The Ar partial pressure was kept at 3 × 10–1 Pa, while the N2 partial pressure was 2 × 10–2 Pa during deposition. The voltage and current of the DC cathode were kept at 420 V and 0.15 A, respectively, and the deposition time was 60 min and 120 min to prepare TiN coatings of different thicknesses. The TiN coatings had nominal thicknesses of 1.5 and 3.0 µm.
Mechanical Property Measurement The indentation tests were carried out on a Fischerscope H100 nanoindenter. For every different thickness and substrate, there were two samples which were tested five times respectively. The following measurement procedures were followed to accurately obtain the mechanical properties of hard coatings:
in which hpl is the indenter displacement due to the plastic deformation under the indenter tip, while he is due to the elastic deformation. It can be seen that the influence of the substrate on the universal hardness measurement of hard coatings comes from not only the plastic deformation, but also the elastic deformation of the substrate. This suggests that if the universal hardness measurement of hard coatings is not influenced by the substrate deformation, then the influence of both plastic and elastic deformation of substrates on mechanical property measurement of coatings can be neglected, and then “real” hardness and elastic modulus of hard coatings can be obtained.
EXPERIMENTAL RESULTS AND DISCUSSION TiN Coating/HSS Composite A large load (200 mN) was employed for the first step of the test in order to demonstrate the influence of substrate deformation on the mechanical property measurement. Figure 1 shows the continuous loading/unloading curves of the different thicknesses of TiN coating/HSS composites and the HSS substrate.
(1) A large load (200 mN) indentation test was performed to demonstrate the influence of substrate deformation on the universal hardness (HU)11 of a coating/substrate composite. (2) A smaller applied load was then selected according to the result of step one for the penetration test. The hardness and modulus of TiN coatings were determined using Oliver and Pharr’s analysis12 from the load/unload displacement curves. It should be noted that Oliver and Pharr developed their method of analysis for monolithic materials, and now this method has become a standard method of analysis for nanoindentation and is frequently used for coatings’ mechanical property measurement (hardness and 338
October 2004
Figure 2—Universal hardness vs load.
JCT Research
Mechanical Properties of Hard Coatings
Figure 3—Load/unload displacement with a small load (15 mN).
The curves in Figure 1 can be defined into a loading stage and an unloading stage. During loading, the indenter displacement increases with increasing load. The contact area can be calculated from the indenter shape and indenter displacement in order to obtain the universal hardness. For the same indenter tips (in this article, a Vickers indenter was used), the universal and Vickers hardness (HV) have the same calculation formula. The difference is that the contact area for HU was calculated from the indenter displacement during loading, while the HV was calculated after unloading. The correlation between the universal hardness and the load derived from Figure 1 is shown in Figure 2. It can be seen that the universal hardness of HSS substrate versus load variation is almost a horizontal line, which indicates the universal hardness of the homogenous HSS substrate shows little change when the applied load increases. The hardness of TiN coating/HSS composites shows an obvious difference during loading. Below about 3 mN, the hardness values are influenced by the intrinsic interference of the external vibration sources, surface roughness of the coatings, thermal drift, oxidation, etc.11,13 At this stage, the mechanical properties of hard coatings cannot be precisely measured. Between 3 mN and the defined limiting load values determined by the coating thickness and hardness, these curves show horizontal stages, which imply that the universal hardness does not change with increasing applied load. The measurement can avoid not only the factors mentioned above, but also the influence of the substrate deformation (both elastic and plastic deformation). Beyond the defined limiting load values, the measured hardness is determined by the coating and the substrate together, and does not give the real hardness of the coating. At this stage, the deformation zone under the indenter tip has extended to the substrate and the measurement of mechanical properties is influenced by the deformation of the substrate. With increasing indenter displacement, the universal hardness decreases gradually and finally approaches the universal hardness of the substrate. Figure 2 also reveals these results for TiN coatings of different thicknesses. The hardness and length of horizontal stages in hardness versus load curves are quite different,
www.coatingstech.org
Figure 4—Universal hardness vs load.
and the thicker the coating, the higher the hardness and the longer the horizontal stages. The expansion of the horizontal stage indicates that a larger load was essential for the deformation zone under the indenter tip in order to extend to the substrate. The hardness increase is considered to be a result of the different internal stresses in coatings.14 After the load range, within which the mechanical property measurement of TiN coatings was not influenced by the substrate, deformation was determined and a small load of 15 mN was selected for these two kinds of coatings for the penetration test. The load displacement plots are shown in Figure 3. The hardness and elastic modulus are calculated according to Oliver and Pharr’s formula and are shown in Table 1. According to the analysis above, under this load the deformation zone is within the coatings, so the measured hardness and elastic modulus were not influenced by the substrate deformation. It can be seen from the measurement results that TiN coatings of different thicknesses prepared under the same conditions show different hardness and elastic modulus due to the difference of internal stress in hard coatings caused by different coating thicknesses.
TiN Coating/SS Composite Figure 4 shows universal hardness versus load curves of SS substrate and TiN coatings of different thicknesses. It can be seen that the universal hardness of the SS substrate is almost unchanged with increasing load. The slightly higher hardness at the small load is due to the strain hardening at the surface during machining and polishing. Table 1—Mechanical Properties of TiN Coatings with Different Thicknesses on Different Substrates No.
Specimen
1 ............HSS 2 ............1.5 µm TiN coating/HSS 3 ............3.0 µm TiN coating/HSS 4 ............SS 5 ............1.5 µm TiN coating/SS 6 ............3.0 µm TiN coating/SS
Hu (GPa)
Y (GPa)
HV (GPa)
7.28 12.6 13.9 2.10 11.8 12.6
251.2 336.3 375.3 215.7 314.5 340.5
10.1 22.8 24.3 2.09 21.8 23.0
October 2004
339
Z. Han et al. (2) The influence of substrate deformation on hardness and modulus measurement was eliminated as much as possible, so the difference in hardness can be attributed to the difference in the internal stress of TiN coatings.
Figure 5—Load/unload displacement curves with a small load.
Figure 5 shows the load versus indenter displacement curves over the load range selected using the data in Figure 4. Considering the difference in horizontal stages on the HU versus load curves of TiN coatings of 3.0 and 1.5 µm, 15-mN and 5-mN loads were chosen, respectively, for the hardness measurement. The experimental results are listed in Table 1.
Atomic Force Microscopy Observation Atomic Force Microscopy (AFM) was also employed to investigate the indentation shape in order to check the measurement of hardness. Shown in Figure 6a is the AFM morphology of the residual impression of 3.0-µm thick TiN coatings on an HSS substrate after 15-mN penetration. Cellular structure and ± 15 nm surface fluctuations can be seen. The indenter displacement under the small load penetration test was about 200 nm, and was much larger than the fluctuations, so the influence of the surface toughness on the measurement can be neglected. Figure 6b shows the depth variation of the insection diagonal selected to accurately measure the indentation size. The horizontal line represents the average position of the surface. The residual depth and the diagonal length of indentation are 84 and 1100 nm, respectively. Comparing with the maximum indenter displacement of 200 nm and a diagonal of 1400 nm during loading, the recovery of the indentation depth is almost 60%, while the recovery of the diagonal is only 20%. The hardness calculated according to the AFM results is 23.5 GPa, which fits the result determined from the load/unload indenter displacement curves very well.
The mechanical properties of TiN coatings, in the form of coating/substrate composites, are influenced not only by the chemical bonding and microstructure, as in bulk materials, but by the process of deposition and the stress caused by the coating/substrate match. For hard ceramic coatings on metallic substrates, the compressive stress is up to several GPas, and at the same time the compressive stress will increase with increasing coating thickness and will even finally cause the de-adhesion of the coatings from the substrates. The results above show that the hardness of TiN coatings on HSS and SS substrates both increase with increasing thickness. TiN coatings and substrates form coating/substrate composites through fine interfacial adhesion. The compressive stress in TiN coatings can be extended into the substrates, while the internal stress in coatings will cause substrate deformation to release at the interface because the modulus of substrates is lower than that of coatings. The mechanical properties such as modulus, hardness, and plastic deformation of substrates will directly influence the stress state of TiN coatings, and this causes TiN coatings of identical thicknesses on different substrates to show different hardness and modulus. From the results above, the hardness and modulus of TiN coatings can be measured accurately using nanoindentation and the variation in mechanical properties of these coatings caused by different thicknesses and substrates can be demonstrated.
(a)
If we compare the mechanical properties of the coatings of identical thickness shown in Figure 2, Figure 4, and Table 1, it can be seen that TiN coatings of identical thickness on different substrates show obviously different mechanical properties. TiN coatings on HSS substrates have higher hardness and higher moduli than those on SS substrates. It should be noted that this result was obtained under the following conditions:
(b)
(1) The TiN coatings of identical thicknesses were prepared with the same deposition parameters.
Figure 6—(a) AFM morphology and (b) profile of indentation morphology (15 mN).
340
October 2004
JCT Research
Mechanical Properties of Hard Coatings
CONCLUSIONS Effects of thickness and substrate on the mechanical properties of TiN coatings can be demonstrated using nanoindentation technique, and reliable hardness and modulus can be obtained. Hardness and modulus of TiN coatings vary with different substrates and different coating thicknesses. TiN coatings on harder substrates or with higher thicknesses show higher hardness and elastic modulus. These effects can be attributed to the internal stress of TiN coatings.
(5) (6)
(7)
(8)
(9)
ACKNOWLEDGMENT This work was financially supported by the Shanghai Nanomaterials Project, under Grant No. 0352nm084.
References (1) Fischer, K. and Oettel, H., “Microstructure Gradients in Thin Hard Coatings—Tailor-Made,” Surf. Coat. Technol., 97:308-312 (1997). (2) Veprek, S., “The Search for Novel, Superhard Materials,” J. Vac. Sci. Technol., A 17(5):2401-2420 (1999). (3) Shinn, M., Hultman, L., and Barnett, S.A., “Growth, Structure, and Microhardness of Epitaxial TiN/NbN Superlattics,” J. Mat. Res., 7(4):901-911 (1992). (4) Cammarata, R.C. and Schlesinger, T.E. et al. “Nanoindentation Study of the Mechanical Properties of Copper-Nickel
www.coatingstech.org
(10)
(11) (12)
(13)
(14)
Multilayered Thin Films,” Appl. Phys. Lett, 56(19):1862-1864 (1990). Jönsson, B. and Hogmark, S., “Hardness Measurement of Thin Films,” Thin Solid Films, 114:257-269 (1984). Saha, R. and Nix, W.D., “Effects of the Substrate on the Determination of Thin Film Mechanical Properties by Nanoindentation,” Acta Materialia, 50: 23-28 (2002). Wang, H.L., Chiang, M.J., and Hon, M.H., “Determination of Thin Film Hardness for a Film/Substrate System,” Ceram. Int., 27: 385-389 (2001). Fernandes, J.V., Trindade, A.C., Menezes, L.F., and Cavaleiro, A., “A Model for Coated Surface Hardness,” Surf. Coat. Technol., 131: 457-461 (2000). Ahn, J.H. and Kwon, D., “Micromechanical Estimation of Composite Hardness Using Nanoindentation Technique for Thin-Film Coated System,” Mater. Sci. Eng., A 285:172-179 (2000). Hainsworth, S.V. and Soh, W.C., “The Effect of the Substrate on the Mechanical Properties of TiN Coatings,” Surf. Coat. Technol., 163-164: 515-520 (2003). Deutsche Norm DIN 50359-1. “Testing of Metallic Materials— Universal Hardness Test—Part 1: Test Method.” October 1997. Oliver, W.C. and Pharr, G.M., “An Improved Technique for Determining Hardness and Elastic Modulus Using Load and Displacement Sensing Indentation Experiment,” J. Mater. Res., 7(6):1564 -1583 (1992). Mencik, J. and Swain, M.V., “Errors Associated with DepthSensing Microindentation Tests,” J. Mater. Res., 10: 1491-1501 (1995). Veprek, S. and Argon, A.S., “Towards the Understanding of Mechanical Properties of Super- and Ultrahard Nanocomposites,” J. Vac. Sci. Technol B, 20: 650-664 (2002).
October 2004
341