4.1 Simulation results
The results of the MD experiments of DLC/Ni-DLC with different process parameters are shown in Tables 2 and 3.
Table 2
MD simulation results of orthogonal experimental DLC films
Number | Sputtering temperature/K | Sputtering voltage/V | Deposition air pressure /Pa | Sp2 | Sp3 | Sp2/Sp3 | Film thickness/Å | Residual stress/Gpa | Film base bonding/N |
1 | 348.15 | 300 | 0.2 | 0.3865 | 0.0387 | 0.4235 | 17.96 | 0.108 | 33.72 |
2 | 348.15 | 400 | 0.7 | 0.3284 | 0.06661 | 0.4438 | 21.17 | 1.363 | 33.03 |
3 | 348.15 | 500 | 1.2 | 0.2674 | 0.09553 | 0.47499 | 22.66 | 1.788 | 34.41 |
4 | 348.15 | 600 | 1.7 | 0.2736 | 0.09240 | 0.45462 | 20.33 | 1.749 | 34.22 |
5 | 423.15 | 300 | 0.7 | 0.3228 | 0.05201 | 0.46066 | 18.44 | 1.274 | 31.63 |
6 | 423.15 | 400 | 0.2 | 0.2561 | 0.10488 | 0.46802 | 21.30 | 5.003 | 34.35 |
7 | 423.15 | 500 | 1.7 | 0.3288 | 0.06087 | 0.45378 | 20.54 | -0.504 | 32.17 |
8 | 423.15 | 600 | 1.2 | 0.2861 | 0.08849 | 0.47773 | 19.32 | 1.658 | 32.91 |
9 | 498.15 | 300 | 1.2 | 0.3627 | 0.04616 | 0.45631 | 18.18 | 0.612 | 33.23 |
10 | 498.15 | 400 | 1.7 | 0.2829 | 0.09403 | 0.46329 | 19.82 | 1.628 | 33.93 |
11 | 498.15 | 500 | 0.2 | 0.2387 | 0.11108 | 0.48876 | 18.84 | 2.079 | 33.97 |
12 | 498.15 | 600 | 0.7 | 0.2633 | 0.10500 | 0.48405 | 21.28 | 2.667 | 32.36 |
13 | 573.15 | 300 | 1.7 | 0.2811 | 0.09330 | 0.46412 | 20.64 | 1.645 | 32.98 |
14 | 573.15 | 400 | 1.2 | 0.3935 | 0.03508 | 0.41681 | 19.01 | 0.195 | 34.30 |
15 | 573.15 | 500 | 0.7 | 0.2993 | 0.09104 | 0.44658 | 18.82 | 2.576 | 33.45 |
16 | 573.15 | 600 | 0.2 | 0.2918 | 0.08433 | 0.44896 | 17.88 | 2.178 | 34.09 |
Table 3
MD simulation results of orthogonal experimental Ni-DLC films
Number | Sputtering temperature/K | Sputtering voltage/V | Deposition air pressure /Pa | Sp2 | Sp3 | Sp2 + Sp3 | Film thickness/Å | Residual stress/Gpa | Film base bonding/N |
1 | 348.15 | 300 | 0.2 | 0.2837 | 0.1465 | 0.5170 | 36.87 | -0.969 | 38.54 |
2 | 348.15 | 400 | 0.7 | 0.2736 | 0.1995 | 0.3460 | 34.03 | 0.065 | 44.80 |
3 | 348.15 | 500 | 1.2 | 0.2263 | 0.2266 | 0.3791 | 34.69 | -0.244 | 44.75 |
4 | 348.15 | 600 | 1.7 | 0.2331 | 0.2167 | 0.3849 | 34.47 | -0.103 | 44.79 |
5 | 423.15 | 300 | 0.7 | 0.2491 | 0.1937 | 0.3927 | 37.09 | -0.010 | 44.82 |
6 | 423.15 | 400 | 0.2 | 0.2362 | 0.2001 | 0.3936 | 36.33 | -1.002 | 44.55 |
7 | 423.15 | 500 | 1.7 | 0.2297 | 0.2265 | 0.3765 | 37.89 | -0.133 | 44.63 |
8 | 423.15 | 600 | 1.2 | 0.2327 | 0.2104 | 0.3861 | 35.70 | -0.322 | 44.87 |
9 | 498.15 | 300 | 1.2 | 0.2365 | 0.2066 | 0.3896 | 35.43 | -0.383 | 44.67 |
10 | 498.15 | 400 | 1.7 | 0.2328 | 0.2157 | 0.3830 | 36.41 | -0.073 | 44.82 |
11 | 498.15 | 500 | 0.2 | 0.2495 | 0.1833 | 0.3858 | 37.48 | -0.488 | 45.36 |
12 | 498.15 | 600 | 0.7 | 0.2667 | 0.1899 | 0.4201 | 35.32 | 0.095 | 44.24 |
13 | 573.15 | 300 | 1.7 | 0.2378 | 0.1965 | 0.3816 | 37.49 | -0.784 | 45.19 |
14 | 573.15 | 400 | 1.2 | 0.2385 | 0.2056 | 0.3659 | 34.75 | -2.110 | 43.96 |
15 | 573.15 | 500 | 0.7 | 0.2320 | 0.2103 | 0.3773 | 37.00 | -0.558 | 44.88 |
16 | 573.15 | 600 | 0.2 | 0.2590 | 0.1968 | 0.3870 | 36.77 | -0.428 | 45.15 |
After the magnetron sputtering MD simulation, the simulation results were visualized and post-processed by OVITO, and the final stable DLC/Ni-DLC-42CrMo structural model was obtained as shown in Fig. 2, and the thickness of the Ni-doped film was about 4 nm.
As shown in Fig. 3, during the growth of DLC/Ni-DLC films, the formation of a mixed film layer between DLC/Ni-DLC and 42CrMo substrate surface layer was randomly observed at 3ps, 81ps, 196ps, 282ps, and the thickness of such a mixed layer reaches about 2 nm before generating a DLC/Ni-DLC film layer without 42CrMo alloy, and the existence of such a mixed layer is the key to the good adhesion of the films to the substrate, the nature of its residual stress, the distribution of its mechanical properties such as bonding force, and the distribution of sp2/sp3 hybrid structure. The existence of the hybrid layer is the key to the good adhesion of the film on the substrate, and the size, nature, and distribution of its residual stress, the mechanical properties such as the size of the bonding force and the distribution of the sp2/sp3 hybridization structure have an important influence on the wear resistance and self-lubrication of the diamond-like film.
The residual stress distribution of each atom in the atomic-scale DLC/Ni-DLC films for a certain set of parameters in the MD simulation is shown in Fig. 4. It is observed that the outermost layer of carbon atoms is less stressed, while the subsurface layer is a mixture of carbon atoms and 42CrMo with a non-uniform stress profile of the carbon atoms. As can be determined from Fig. 5, the overall compressive stresses in the mixed layer at 0-1.5 nm are predominant for the non-Ni-doped DLC films. With the increase of thickness, the tensile stress is exhibited after 1.5 nm the stress value increases rapidly, and the residual compressive stress of the film increases and then decreases along the thickness. Statistical calculations of the residual stresses in the film layers yielded an average residual stress of 1.516 Gpa for the films. While the thickness of Ni-DLC film is about 3.678 nm, the internal tensile and compressive stresses of the film are lower and more uniformly distributed, and the average residual stress value is -0.969 Gpa.
The proportion of sp2, sp3, and sp2/sp3 heterostructures of the films was counted by the OVITO module in Python, and the results are shown in Fig. 6, the proportion of sp3, sp2/sp3 heterostructures in the DLC after Ni doping increased from (3.87%; 42.35) to (14.65%; 51.70%), and the proportion of sp2 heterostructures reduced from 38.65–28.37%, and the proportion of sp3 hybridized structures in the film organization is about 2.56 times higher after Ni doping than without Ni doping.
As shown in Fig. 7, fix-box/relax aniso calculations show that there are 4405 atoms of C and Fe, Cr, Mo, etc. in the DLC film, with an average binding energy of about 11.81ev, and the film-based binding force per unit cm length of the mixed layer is calculated as 33.72N through the relationship between the distance and the energy; in the Ni-DLC film, there are 6421 atoms of C and Fe, Cr, Mo, Ni, etc., with an average binding energy of about 9.26ev, and the film-based binding force per unit cm length of the mixed layer is 38.54N through the distance and energy relationship. In the Ni-DLC film, there are 6421 atoms of C and Fe, Cr, Mo, Ni, etc., and the average binding energy is about 9.26ev, and the film-based binding force of the mixed layer per unit cm length is 38.54 N calculated from the distance-energy relationship. The number of deposited atoms increases within the same deposition time after Ni doping and the inter-atomic binding energy decreases in the thin film organization, which may be an important factor in the reduction of the residual stress of the film.
4.2 Verification of molecular dynamics simulation results
To verify the reliability of simulated experimental data, this study used a PCVD8060 high vacuum multifunctional coating machine to deposit Ni-DLC films under the same experimental parameters and measured the residual stress of 16 sets of Ni-DLC films using TD-3500 X-ray polycrystalline diffractometer. The sedimentary system is shown in Fig. 8, The 42CrMo pad after coating is shown in Fig. 9, and the X-ray polycrystalline diffractometer is shown in Fig. 10.
The error between the sedimentary experiment and the MD experiment results is shown in Fig. 11. In the case of doping Ni element, the residual stress of Ni-DLC film exists in the form of residual compressive stress or lower tensile stress as a whole. The maximum error between experimental results and simulated stress values is 29.65%, the minimum is 1.28%, and the average error is about 8.8%. The small error proves the feasibility of using the molecular dynamics simulation method to analyze the residual stress of Ni-DLC film prepared by magnetron sputtering.
4.3 Effect of process parameters and Ni doping on the structure and mechanical properties of Thin Films
By analyzing the results of MD simulation, the effects of sputtering temperature, sputtering voltage, and deposition air pressure on the proportion of sp2, sp3, and sp2/sp3 heterostructures in the organization of DLC/Ni-DLC thin films as well as the residual stresses in their thin films were obtained, and the magnitude of the film-base bonding and the film thicknesses were compared between Ni-doped and non-Ni doped diamond-like thin films.
As shown in Fig. 12, keeping the temperature constant and under different deposition air pressures, the sp2 hybridized structure in the Ni-doped film gradually decreases with the increase of air pressure and voltage; while without Ni doping, the sp2 hybridized structure in the film shows an overall increasing trend with the increase of air pressure, and gradually less with the increase of voltage.
As shown in Fig. 13, keeping the deposition air pressure unchanged and under different sputtering temperatures, the sp2 hybridized structure in the Ni-doped film increases with the increase of temperature; when the film is not doped with Ni, the sp2 hybridized structure in the film increases slowly with the increase of temperature, and combined with Fig. 12, it can be seen that the overall decrease of sp2 hybridized structure in the film is observed after doping with Ni.
As shown in Fig. 14, keeping the sputtering temperature unchanged, under different deposition air pressures, the sp3 hybridized structure in the Ni-doped film increases less and then increases with the increase of air pressure, and increases and then decreases with the increase of voltage; when the film is not doped with Ni, the sp3 hybridized structure in the film increases slowly with the increase of air pressure, and gradually increases with the increase of voltage.
As shown in Fig. 15, keeping the deposition air pressure constant, at different sputtering temperatures, the sp3 hybridized structure in the Ni-doped film increases and then decreases with the temperature, while the sp3 hybridized structure in the film increases slowly with the increase of the temperature when the film is not doped with Ni. Combined with Fig. 14, it can be seen that the overall increase of sp3 hybridized structure in Ni-doped films. It has been shown that the sp3 hybridized structure will be transformed into an sp2 graphite structure when the temperature exceeds 500°C [16–17], therefore, the sp3 hybridized structure can effectively ensure the self-lubricating performance of the DLC film on the surface of the rolling body of wind turbine bearings.
As shown in Fig. 16, keeping the deposition air pressure unchanged, under different sputtering temperatures, the residual stress of the Ni-doped film decreases slowly with the increase of temperature, and decreases and then increases with the increase of voltage; when the film is not doped with Ni, the residual stress of the film increases and then decreases with the increase of temperature and voltage.
As shown in Fig. 17, keeping the sputtering temperature unchanged, the residual stresses of the Ni-doped thin films under different deposition air pressures do not change much with the increase of air pressure, while the residual stresses of the thin films without Ni-doping fluctuate greatly with the increase of air pressure. Combined with Fig. 16, it can be seen that after Ni doping, the residual stress of the films is dominated by compressive stress, and the average stress value is lower than that of the non-Ni doped diamond-like films, which may be due to the formation of antibonds between Ni and C as well as the relaxation of the distorted bond angles and bond lengths leading to the reduction of the residual stress.
As shown in Figs. 18–19, for the same deposition conditions and time, the thickness of the Ni-doped film increased by about 82% and the film base bonding force increased by about 32.78% compared to the non-Ni-doped film, with the thickness of the Ni-DLC film ranging from 3.4 nm-3.75 nm, and the film base bonding force ranging from 38.54 N-45.36 N. The thickness of the Ni-DLC film was about 1.5 mm, and the film base bonding force was about 1.5 mm, and the film base bonding force was about 1.5 mm.