3.1 Characterization of the Ti- DLC films
The use of a metal buffer layer between the DLC film and the substrate significantly reduces the internal stress and increases the adhesion of the DLC film [14]. The Ti buffer layer (~ 178 nm thick) was grown by sputtering, using a DC power of Ti target, a DC power of 180 Watt and a deposition time of 6 min. Figure 1 shows that the surface AFM and cross-sectional SEM morphology of the Ti-DLC/Ti buffer bi-layer film are relative to Rm flow ratio. The DLC films exhibit a smooth particular structure with a relatively small surface roughness (Ra = 6.99 ~ 8.80 nm). The film is dense and uniform in structure, with no delamination or peeling after coating, and it is perfectly bonded to the substrate. The growth rate is calculated by dividing the DLC film thickness by the 25 min deposition time. The growth rate for Ti-DLC films decreases from 23.24 nm min–1 (581 nm thick) to 16.52 nm min–1 (413 nm thick) when the Rm flow ratio is increased from 4–12%.
Ti atoms are sputtered and incorporated into the film, which contains carbon. If the methane flow rate is increased, the surface of the Ti target becomes covered with carbonaceous material due to target poisoning [15] so the Ti concentration and the growth rate for the DLC film decreases. In the right parts of Fig. 1, the cross-sectional SEM morphology shows correlation between the thickness of the film and the Rm flow ratio. As the Rm flow ratio is 4%, 6%, 8%, 10% and 12%, the thickness of the films is 581, 569, 514, 484 and 413 nm, respectively. This result show that the thickness of the film decreases while the Rm flow ratio increases. This depicts Ti atoms become more difficult to be sputtered because the carbon coating has been coated on the Ti target during sputtering. This result was also demonstrated by Wang et al. [16].
The C, Ti and O content in the Ti-DLC films is shown in Table 2 as a function of the Rm flow ratio. As Rm increases, the carbon concentration increases from 48.62 at.% to 58.63 at.%, the Ti concentration decreases from 37.23 at.% to 28.21 at.% and the O concentration ranges from 13.16 at.% to 14.15 at.%. The doped Ti element bonds with carbon to form a Ti-C structure. As the Rm flow ratio increases, the proportion of carbide bonds decreases and the proportion of carbon bonds increases. The Ti content of the DLC film is controlled by adjusting the Rm flow ratio. This is also demonstrated in the study by Cao et al. [17].
Table 2
Element concentration of a Ti-DLC film coating as a function of Rm flow ratio.
Rm
|
Element composition (atomic %)
|
C
|
Ti
|
O
|
Totals
|
4%
|
48.62
|
37.23
|
14.15
|
100
|
6%
|
51.83
|
34.45
|
13.72
|
100
|
8%
|
53.46
|
32.47
|
14.07
|
100
|
10%
|
56.73
|
30.07
|
13.20
|
100
|
12%
|
58.63
|
28.21
|
13.16
|
100
|
Figure 2 shows the GIXRD diffraction spectrum for a Ti-DLC film coating for various Rm flow ratios. There are no diffraction peaks for the carbon and diamond phase, so the Ti-DLC film is amorphous. There are peaks for nano-crystalline TiC at values for 2θ of 35.88o, 41.52o and 60.14o, which respectively correspond to the (111), (200) and (220) planes [JCPDS No.89-3828] of the face centered cubic (FCC) structure. The peaks for the (200) and (220) planes are relatively weak. This is in agreement with the results of the study by Gayathri et al. [18]. The preference for orientation in the (111) plane is attributed to the surface energy [19]. As the Rm flow ratio is increased from 4 to 12%, the position of the (111) peak shifts towards a lower diffraction angle of 35.88o to 35.12°. The microstructure of Ti-DLC using TEM is shown in Fig. 3. Typical amorphous/nanocrystalline composite carbon structure embedded in DLC matrix. The electron diffraction pattern of a selected area (Fig. 3a) shows crystalline diffraction rings that correspond to (111), (200) and (220) crystal planes of TiC, indicating the presence of a polycrystalline structure. This is in agreement with the GIXRD diffraction results. The interplanar distance calculated from the inverse fast Fourier transform (FFT) resulted in 0.223 nm (Fig. 3b inset 1) and 0.246 nm (Fig. 3b inset 2), which corresponds to (200) and (113) planes.
The grains size in the films is calculated using the full width at half maximum (FWHM) and the Scherrer equation [20]. The 2θ, FWHM and grain size for Ti-DLC (111) plane that are grown using various Rm flow ratios, corresponding to Fig. 2 are shown in Table 3. The grain size of the Ti-DLC (111) plane is among 13.63 nm to 15.66 nm, 2θ is 35.120 ~ 35.882 degree, and FWHM is 0.533 to 0.612. The incorporation of carbon into the interstitial spaces of the FCC TiC lattice results in lattice distortion and compressive stress inside the film, as demonstrated by the shift in the TiC (111) XRD peak to a lower angle. Figure 4 shows the SIMS depth profile for a Ti-DLC film coating for a Rm flow ratio of 4%. The distribution of Ti and C elements is uniform throughout the depth of the film.
Table 3
The 2θ, FWHM and grain size for Ti-DLC (111) films that are grown using various Rm flow ratios, corresponding to Fig. 2.
Rm
|
2θ (degree)
|
FWHM
|
Grain size (nm)
|
4%
|
35.882
|
0.545
|
15.32
|
6%
|
35.676
|
0.610
|
13.68
|
8%
|
35.696
|
0.533
|
15.66
|
10%
|
35.541
|
0.612
|
13.63
|
12%
|
35.120
|
0.547
|
15.23
|
High-resolution XPS spectroscopy is used to determine the bond state structure and the elemental composition information of Ti-DLC films. The XPS spectra for characteristic binding energies for C 1s, Ti 2p and O 1s for Ti-DLC film coatings with a Rm flow ratio of 4%, are shown in Fig. 5. The O 1s signal is observed in the DLC films, probably there is a residual atmosphere in the chamber or because the sample is oxidized in air [21]. The C 1s spectra in Fig. 5 (a) has a main peak und at ~ 283.4 eV (C-C/C-H bonds, amorphous carbon), which is typical of DLC film structure. There is a major peak ~ 280.4 eV and a weak peak at ~ 287.3 eV, which respectively correspond to C-Ti (Ti carbide components) and C = O (carbide bonds). The peaks at around 280.4 eV are due to C-Ti components. This shows that C atoms interact with doped Ti atoms to form a nanocrystalline TiC structure [22]. The Ti 2p spectrum in Fig. 5 (b) has peaks at ~ 453.7 eV and ~ 459.8 eV, which respectively correspond to the Ti 2p3/2 and Ti 2p1/2 peaks for the Ti-C bond structures (TiC state) [23]. The binding energies at around 457.2 eV and 462.9 eV are respectively attributed to Ti 2p3/2 and Ti 2p1/2 for the Ti-O (TiO2 structure). Moreover, in the interval between the Ti-C peak and the Ti-O peak, there are two weak peaks near 454.8 eV (Ti 2p3/2) and 461.6 eV (Ti 2p1/2), which are denoted as Ti-X. These peaks consist of TiCx (x < 1) and TiOx (or TiCxOy) [24]. The O 1s spectrum in Fig. 5 (c) has binding energies at around 527.9 eV, 528.6 eV, 530.2 eV and 531.6 eV that are respectively associated with O-Ti, O-Ti*, O = C and O-C bonding. This result is consistent with the study by Zhang et al. [25].
Raman spectroscopy is used to determine the DLC carbon bonding state in the C-C phase structure [26]. Figure 6 shows the Raman spectra for a Ti-DLC that is deposited using various Rm flow ratios. The Raman spectrum has an asymmetrically dispersed, curved spectrum with peaks between 1000 and 1800 cm− 1, which is typical for a DLC structure. The D peak has a smaller shoulder (disordered, indicated C-C sp2 bonding), which is associated with carbon disorder, due to breathing patterns for those sp2 carbon bonds that are only in aromatic rings. The G peak at 1560 ~ 1680 cm− 1 (suggested C-C sp3 bonding) is attributed to the C-C stretching patterns of sp2 atoms in aromatic rings and carbon chains [27]. Figure 6 shows that increasing the Rm flow ratio (reducing Ti metal doping) increases the intensity of the Raman spectrum, because there is an increase in sp2-C (sp2/sp3 ratio) site fraction of DLC components [28]. These results are in agreement with those of the study by Guo et al. [29], which showed that increasing the metal doping decreases the intensity of the Raman spectrum for a DLC film, because the proportion of centrosymmetric structures (such as TiC) increases, which hinders Raman excitation.
The position of the G peak and the ID/IG ratio are used to characterize the sp2/sp3 bonding ratio for the DLC structure [1, 17, 30]. The G peak position and the ID/IG ratios are shown in Table 4. As the Rm flow ratio is increased from 4–12%, the ID/IG ratio increases from 0.757 to 0.919 and the G peak position ranges from 1562.56 to 1576.23 cm− 1. Similar results were obtained by Dai et al. [28], which showed that as the C2H2 fraction increases, the G peak position shifts to higher wavenumbers and the value of the ID/IG ratio also increases, so there is in an increase in the sp2/sp3 ratio. If the Rm flow ratio is relatively low (the doping concentrations of Ti increases), the sp2/sp3 ratio for the DLC film decreases and the structural phase of the coating is carbide nanocomposite Ti-DLC. Further increasing the Rm flow ratio causes the structure to transformed into an amorphous carbon-based Ti-DLC coating [28].
Table 4 The ratio of ID/IG and the position of the G peak for Ti-DLC films as a function of the Rm flow ratio.
Rm
|
ID
|
IG
|
ID/IG
|
G peak position
|
4%
|
1747.24
|
1946.97
|
0.757
|
1562.56
|
6%
|
2712.98
|
3521.65
|
0.770
|
1576.08
|
8%
|
3636.29
|
4208.89
|
0.864
|
1571.53
|
10%
|
4494.26
|
5188.46
|
0.866
|
1576.23
|
12%
|
5317.22
|
5779.88
|
0.919
|
1567.72
|
3.2 Mechanical properties and cutting tool inserts features of Ti-DLC films
To determine the mechanical properties of the Ti-DLC films, the load-displacement curve is plotted using the results of nano-indentation tests, and the elastic modulus (E, GPa), hardness (H, GPa), and elastic recovery (%Re) values are determined [31]. The %Re represents the ability of a film to return to its original shape after indentation [32] and is defined as:
$$\% {R_e}=\frac{{{h_{\hbox{max} }} - {h_r}}}{{{h_{\hbox{max} }}}} \times 100\%$$
1
where hmax is the maximum indentation depth and hr is the residual indentation depth. The indentation depth is less than 10% of the film's thickness to eliminate substrate effects. Four identical nano-indentation tests were performed on each sample. Table 5 lists the hardness (H), elastic modulus (E), H/E ratio H3/E2 ratio and %Re value for Ti-DLC films that are grown using various Rm flow ratios. The hardness increases from 11.07 GPa to 14.76 GPa as the Rm flow ratio is increased from 4–12%. The hardness of DLC coatings is directly related to the sp3-C fraction in the carbon matrix [15]. A higher H/E ratio and H3/E2 ratio are respectively related to good resistance to elastic strain failure and resistance to plastic deformation. The results in Table 5 show that if the Rm flow ratio is increased to 12%, the maximum respective values for the H/E ratio, the H3/E2 ratio and %Re are 0.12, 0.22 and 62.47%. The formation of TiC carbide bonds is thought to be responsible for an increase in the elastic modulus so adjusting the doping content by adjusting the Rm flow ratio controls the H/E ratio, the H3/E2 ratio and the %Re value for the coating. These results are in agreement with those for the study by Dai et al. [6]. A high H/E, H3/E2 ratio and Re value are characteristic of good tribological features for a film coating [15].
Table 5 The H, E, H/E ratio H3/E2 ratio, %Re value for Ti-DLC films that are grown using various Rm flow ratios.
Rm
|
Hardness
(H, GPa)
|
Elastic
modulus
(E, GPa)
|
H/E
|
H3/E2
|
hmax
|
hr
|
Elastic recovery
(%Re)
|
4%
|
11.07 ± 0.32
|
125.3 ± 8.93
|
0.09
|
0.09
|
177.5
|
98.1
|
44.73
|
6%
|
13.30 ± 0.59
|
117.9 ± 7.72
|
0.11
|
0.17
|
175.7
|
96.9
|
44.85
|
8%
|
13.46 ± 0.58
|
118.0 ± 6.79
|
0.11
|
0.18
|
181.1
|
91.8
|
49.31
|
10%
|
14.63 ± 0.55
|
119.9 ± 6.82
|
0.12
|
0.22
|
154.1
|
58.6
|
61.97
|
12%
|
14.76 ± 0.61
|
120.6 ± 7.91
|
0.12
|
0.22
|
146.3
|
54.9
|
62.47
|
The adhesive strength of the film coating onto cermet cutting tool inserts is measured using a scratch test [33]. Typical crack modes quantify the critical load for failure [34]. To determine the failure characteristics for a DLC film coating, scratch tracking is used to test the adhesion of the film to the substrate. Figure 7 shows the scratch tracking, normal load and acoustic emission (AE) recordings for a Ti-DLC/Ti buffer bi‑layer film coating. The normal load on the diamond stylus is gradually increased from 500 mN (begin load) to 50 N (end load) at a loading rate of 49.5 N/min, using a scratch speed of 5 mm/min, with a total scratch length of 5 mm. The scratch adhesion critical normal loads LC1 (first crack) and LC2 (first peel) are respectively defined as the load that causes incipient cracking and initial adhesive failure [35]. The respective normal loads for LC1 and LC2 are about 37.5 N and 47.5 N. The corresponding AE signals are shown at about 3.7 mm (incipient crack) and 4.8 mm (initial peeling), so this film adheres well to the substrate.
Dry milling without a coolant reduces the environmental impact [36]. This study uses a cermet cutting insert that is coated with a Ti-DLC/Ti buffer bi‑layer film for the dry machining of an Inconel 718 superalloy workpiece (hard-to-cut material), as shown in Fig. 8. For uncoated tool inserts, the surface roughness and flank wear are Ra = 3.81 µm and 56.73 µm, respectively. For a Ti-DLC coated cutter insert, the Rm flow ratio increases from 4–12%, the surface roughness decreases from Ra = 3.31 to Ra = 1.96 µm and the flank wear is reduced from 47.55 to 30.81 µm. These experimental results are consistent with the results for the nano-indentation test. A Ti-DLC/Ti buffer bi‑layer coating on a cutter insert reduces surface roughness and flank wear. All coated tools have a significantly longer life than uncoated tools.