*3.4. Frictional Properties Measured by the Progressive Load Scratch Test and Drawing Friction Test*

Table 5 summarizes the frictional forces determined by the progressive-load scratch test. Both DLC-coated disk specimens had significantly lower initial and total frictional forces than the non-coated disk specimen. There was no significant difference between the two DLC-coated specimens in terms of the distance for detachment.


**Table 5.** Frictional forces obtained by the progressive-load scratch test (N).

Notes: Values are presented as the mean ± SD; Identical letters indicate that mean values were not significantly different (*p* < 0.05) by one-way ANOVA followed by the Tukey multiple test; † There was no significant difference between the two DLC-coated specimens in terms of the distance for detachment by Student's *t*-test.

Table 6 summarizes the static and kinetic frictional forces determined from drawing-friction testing of the non-coated and DLC-coated wire specimens under dry and wet conditions. Two-way ANOVA showed that the coating procedure (non-coating, DLC-1, DLC-2) and test environment (dry, wet) were statistically-significant factors affecting both the static and kinetic frictional forces. One-way ANOVA and Tukey's tests showed that the DLC-2 had a significantly lower frictional force than the non-coated specimen, with the exception of the static frictional force under the dry condition. On the other hand, the DLC-1 showed frictional force values that were similar to those of the non-coated specimen, with the exception of the kinetic frictional force under the dry and wet condition. According to the drawing-friction testing, the DLC layers were partially ruptured for the DLC-1 case, while no rupture was observed for the DLC-2 condition (Figure 6).

**Table 6.** Static and kinetic frictional forces for the non-coated and DLC-coated wires in dry and wet conditions (N).


Notes: Values are presented as the mean ± SD; Identical letters indicate that mean values were not significantly different (*p* < 0.05) by one-way ANOVA followed by the Games–Howell test.

**Figure 6.** Stereomicroscope images of DLC-coated wires after the drawing-friction test. A portion of the DLC layer had ruptured from the interface. (**a**) DLC-1 and (**b**) DLC-2. Original magnification: 50×.

#### **4. Discussion**

In this study, ca. 300 nm-thick DLC layers were deposited on orthodontic stainless steels. The coatings were amorphous, which was consistent with previous findings [24]. The type of DLC can be identified using a ternary phase diagram [16]. This diagram shows the fraction of carbon sites that have *sp*<sup>2</sup> (graphite-like) bonding, *sp*<sup>3</sup> (diamond-like) bonding or bonding with hydrogen. Quantitative analysis of *sp*<sup>2</sup> and *sp*<sup>3</sup> bonding in a DLC can be performed by XPS analysis [25,26]. In the present study, the DLC-1 had a higher *sp*2/*sp*<sup>3</sup> ratio (0.343) at the external surface region (ca. 13 nm deep), while the DLC-2 had a lower *sp*2/*sp*<sup>3</sup> ratio (0.181) at the external surface region. This indicated that the external surface of the DLC-1 had a more diamond-rich structure than the DLC-2. After four more layers had been sputtered, the *sp*2/*sp*<sup>3</sup> ratio (measured at a depth of ca. 65 nm) was similar for DLC-1 (0.235) and DLC-2 (0.283). Furthermore, this trend changed after 10 layers were sputtered (measured at a depth of ca. 130 nm) when the DLC-1 displayed a lower *sp*2/*sp*<sup>3</sup> ratio (0.201), although the DLC-2 had a higher *sp*2/*sp*<sup>3</sup> ratio (0.343). This indicated that the inner surface of the DLC-2 had a more diamond-rich structure than the DLC-1. Nanoindentation testing suggested that the DLC-1 had better mechanical properties than the DLC-2 at the external surface region, while the DLC-2 seemed to have better mechanical properties than the DLC-1 at the inner surface region. These findings are supported by the *sp*2/*sp*<sup>3</sup> ratios measured at the different depths in this study, because the diamond structure is harder than the graphite structure [16]. Quantitative analysis of hydrogen in a DLC can be performed by elastic recoil measurements [27]. Using this technique, the average hydrogen content of DLC-2 (27%) was slightly higher than that of DLC-1 (23%). A higher hydrogen content of a DLC coating layer can lead to a higher hardness and elastic modulus [28,29], which may influence wear rate and frictional properties.

Most DLC films are harder than metallic materials. DLC coatings using PBIID methods provide hardnesses ranging from 6 to 20 GPa, depending on the deposition conditions [16,18,19]. The hardness of the DLC layers determined by nanoindentation testing in this study ranged from 9.18 to 9.69 GPa (when measured at a depth of ca. 70 nm), which is much higher than the 6.4 GPa measured by nanoindentation testing under the 20-mN load of the as-received stainless steel orthodontic wire. Additionally, the DLC layers showed a much higher elastic modulus compared with non-coated stainless steel orthodontic wires [30], which should influence the elastic modulus of whole archwires. This is supported by the three-point bending results of the present study. The DLC-coated wire exhibited a significantly higher elastic modulus (by 6%–11% as measured by the three-point bending test) than the non-coated wire. Fortunately, variation of this level may not influence clinical orthodontic tooth movement because a wide range of initial orthodontic forces (18–1500 gf) has been proposed as the optimum force for orthodontic tooth movement, and evidence is lacking regarding the optimal force level [31]. Three-point bending at a span of 1.0 mm caused the coating layer to detach from the inner core for only the DLC-1 wire. None of the coatings of the DLC-2 wires were damaged, probably because the DLC-2 coating had better mechanical properties and adhesion.

Several recent studies of DLC coating reported excellent frictional properties [9–12,18–20], fine cell growth with non-cytotoxicity [18], less bacterial adhesion [32] and inhibited biofilm formation on the metal with DLC coatings [33]. Similarly, the progressive-load scratch test in the present study revealed that both DLC-coated disk specimens (DLC-1, DLC-2) displayed significantly lower frictional forces than the non-coated disk specimens. One explanation for this behavior is that the DLC layer, with higher hardness due to the diamond-rich structure, produced lower frictional forces because of a lower wear rate [16]. Additionally, the hydrogen content might have contributed to lower friction under the dry condition because of the elimination of free σ-bonds on the surface [12]. However, only DLC-2 produced significantly lower frictional force than the non-coated case in the drawing-friction test with a 10◦ positioning of the bracket under the wet condition. This was attributed to partial rupture of the coating of DLC-1, causing increasing wire-binding at the edge of the bracket [34], thereby increasing the frictional force. Crack initiation and ruptured coating regions were not observed for DLC-2, which suggested that the DLC-2 coating had good flexibility as a functionally-graded material with outstanding adhesion to the orthodontic stainless steel substrate. Additionally, the hydrogen content of the DLC layers might be important under the wet condition. Water molecules might react with a hydrogenated DLC coating to form oxygen-containing hydrophilic groups on the surface that could provide lubrication for the sliding counter surface [21,22]. Another possibility is that hydrogen-terminated surfaces of a hydrogenated DLC coating may interact through weak van der Waals forces [16,22].

The improved frictional properties demonstrated in this work for the DLC-coated samples suggest that tooth movement by sliding mechanics using DLC-coated stainless steel wire may be superior to that using conventional stainless steel wire. However, further randomized controlled trials are required to assess the clinical efficacy.

#### **5. Conclusions**

Two types of DLC coatings (DLC-1, DLC-2), differing in hydrogen content, *sp*2/*sp*<sup>3</sup> ratio and mechanical properties, were deposited on orthodontic stainless steel substrates. These coatings affected in vitro bending and frictional properties. DLC-2 showed superior frictional properties, good flexibility and adhesion to the stainless steel. A DLC coating with a higher hydrogen content may provide a better orthodontic wire.

**Author Contributions:** M.I. conceived of and designed the experiments. T.M., M.I. and M.K. performed the experiments. M.I., T.M. and I.M. wrote the paper.

**Funding:** This study was partially supported by a Grant-in-Aid Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology, Japan (No. 18K09864).

**Acknowledgments:** The authors thank Masahiko Sugihara and Yoshimi Nishimura at Kurita Seisakusho for their expert technical assistance with the DLC coating procedure.

**Conflicts of Interest:** The authors declare no conflict of interest.
