Effect of Water on Wear of DLC Coatings in High Temperature and Pressurized Ethanol

In the above-mentioned environment containing water with bioethanol, particularly under simulated high-pressure conditions, no research has been conducted. In this environment, DLC coatings were used as materials to simultaneously suppress corrosion and achieve low friction and wear resistance. Acquisition of the friction and wear characteristics of DLC coatings typically involves coating deposition on the disk side, calculating the specific wear rate from the worn area, and directly obtaining the wear volume from methods such as laser microscopy. However, to conduct high-temperature and high-pressure tests, it is necessary to introduce a friction tester into the pressure chamber. Introducing a rotation mechanism into high-pressure and high-temperature chambers is extremely challenging, thus necessitating the use of a mechanism where the two surfaces in contact inside the chamber slide against each other. Previously, the authors developed a sliding friction tester using an oscillating chamber and conducted experiments. Figure 1a shows the external appearance of the oscillating experimental apparatus, and Fig. 1b shows a schematic diagram of the oscillating mechanism. A slider mechanism fixed with three fixed-balls moves against plate-shaped disk test specimens installed inside the chamber. The reason for having three fixed-balls is to ensure that the slider does not tilt and that the ball tips always contact the disk test specimen. Furthermore, since the slider moves inside the clearance (approximately 0.5 mm) provided between the slider and the slider guide installed on the side, even if DLC coatings are deposited on the disk test specimens, it is not guaranteed that friction will occur at the same location. Therefore, to clarify the friction and wear characteristics of DLC coatings, ball specimens of SUS316 stainless steel with a diameter of 2 mm was press-fitted into a ball holder, were used. Then, DLC coating was deposited on the fixed ball to ensure that a specific spot on the ball was always subjected to friction.

To elucidate the friction and wear characteristics of DLC coatings in high-temperature, high-pressure water-containing ethanol environments, the authors previously prepared test specimens with a-C:H and a-C:H:Si coatings deposited on ball surfaces, as used in water environments. Details are as described in reference [14], prepared using the chemical vapor deposition (CVD) method. To clarify the effect of silicon content on the friction and wear characteristics, coatings were deposited to achieve nearly the same hardness. The mechanical properties of the coatings are shown in Table 1. SUS316 plates were used as the counterpart materials. The plate thickness was approximately 1.0 mm, with a measured hardness of approximately 7.2 GPa using nanoindentation test method (ENT-1100a, Elionix). The Berkovich-type diamond tip applied a force of 200 μN to each coating surface for 1 s holding time at the highest normal load. An arithmetic mean roughness of specimens is approximately 25 nm Ra.

To conduct friction and wear tests on DLC coatings in high-temperature and high-pressure water-containing ethanol environments, we utilized an autoclave-type friction tester. The autoclave, a heat and pressure-resistant container, replicates the conditions of the high-temperature and high-pressure water-containing ethanol environment investigated in this study. The friction testing machine is positioned inside the autoclave chamber, as illustrated in Fig. 1. The fixed ball-type test specimen and disk-type test specimen mentioned earlier make contact with the ball-type test specimen reciprocating over the disk to perform the test. By oscillating the autoclave chamber like a pendulum using the motor and oscillation mechanism located outside the autoclave chamber, the friction testing machine can reciprocate the slider against the disk-type test specimen.

The experimental conditions for assessing the influence of water content and the presence of gases such as oxygen or nitrogen on the wear of DLC coatings in high-temperature and high-pressure water-containing ethanol environments, using an autoclave-type friction tester, are outlined in Table 2. To minimize fluctuations in the dissolved oxygen levels within the solution, the autoclave chamber was pressurized with nitrogen gas up to 7 MPa prior to the friction test, then subsequently depressurized to mitigate the influence of dissolved oxygen from the atmosphere. Following this, gas was introduced using a cylinder to achieve a pressure of 10 MPa at the test temperature. The test was carried out over a sliding distance of 18.6 and 37.2 m for a duration of 60 min. The normal load was set to 0.143 N. To investigate the impact of water content in ethanol solution on DLC coating, the content was set to 0 and 6 vol.%, and different pressurized gases of O₂ and N₂ were bubbled to assess the effect of dissolved oxygen on DLC coatings during the friction and wear test. For clarity in describing the friction test conditions using the autoclave, they were denoted as “a-C:H_W6E94_O₂” (Test specimen_W: Water volume % in the solution, E: Ethanol volume % in the solution_pressure gas).

To elucidate the mechanical properties of DLC coatings following the friction and wear tests, we employed an AFM (atomic force microscopy, Shimadzu: SPM-9700HT). This technique entails scratching the coating surface with an AFM probe coated with diamond (nanoworld, CDT-NCHR-10) featuring a tip radius curvature of 150 nm, enabling us to estimate the coating’s hardness based on the scratch depth [15‐18]. In high-temperature and high-pressure environments, the cause of wear on the outermost surface of DLC coatings is thought to be the softening of the topmost surface. To obtain topmost surface hardness, a normal load of 28 μN is applied, and a scratch of 2 μm in the x-axis direction is made once at a speed of 2 μm/s. Next, a diamond indenter is sent in the y-axis direction by 7.8 nm. These scratches are repeated to create a state where a 2-μm square area is scratched only once. Finally, since a 2-μm square area is formed in a single scratch, the depth is measured by measuring a 6-μm square area including the previously scratched area with a normal load of 0.5 μN to obtain surface morphology information. In this way, the relationship between scratch depth and the number of scratches is obtained.

By considering the relationship between the number of scratches and depth as an abrasive wear model, it is possible to calculate the hardness of the topmost surface. Details are as shown in references [14, 18]. The hardness, H, of the topmost surface can be determined using the following equation, which relates the scratch depth, h, to the contact area, A_c, between the diamond probe and the test piece:here, W represents the applied normal load. The contact area, A_c, is defined as the semicircle attached between the tip of diamond probe and the specimen. The contact projected radius, r′, can be expressed by the following equation:

Furthermore, the geometrically pressed depth, h, can be described by

Assuming that the geometrically pressed depth, h, can be represented as the actual pressed depth, h′, multiplied by the wear coefficient, K, Eq. (3) can be rearranged as

By solving Eq. (4) for the scratch hardness, H, of the as-deposited DLC coating, and considering the measurement conditions of r and W obtained from the scratch test for the as-deposited DLC coating, the wear coefficient, K, can be determined from h′. This allows the scratch hardness, H, of the DLC coating wear mark surface to be determined, as expressed in Eq. (5):

In this study, we utilized a Perkin-Elmer PHI-650 Auger electron spectrometer to assess alterations in the atomic composition of the DLC coating surface. The primary incident electron beam width was maintained at a minimum of approximately 5 µm or less, ensuring accurate analysis of the friction surface.

The intensity of the C KLL peak was standardized. To quantify the oxidation phenomenon of DLC coatings, we employed the O/C ratio, calculated by dividing the oxygen peak intensity by the carbon peak intensity, as an indicator.

Friction tests were conducted between DLC-coated pins and SUS316 disks in a high-temperature (120 °C), high-pressure environment. The experiments utilized gases of oxygen and nitrogen as well as ethanol–water mixtures at 0 and 6 vol.% as dissolved oxygen concentration for pressurization to reach 10 MPa. Two types of a-C:H and a-C:H-Si coatings were tested. Representative optical microscope observations of the pins and disk specimens after the friction tests are shown in Fig. 2a and b. The wear volume of the DLC-coated pins was measured using a laser microscope, and the result is shown in Fig. 3 as the specific wear rate, obtained by dividing the wear volume by the product of the normal load and sliding distance (37.2 m). Both the a-C:H and a-C:H:Si coatings exhibit higher specific wear rates in an O₂ environment compared to an N₂ environment. Additionally, it was observed that the specific wear rate of the a-C:H:Si coating increases proportionally with the increase in water content in ethanol from 0 to 6 vol.%. The results showed that under O₂ pressure, as the water content in ethanol increased from 0 to 6 vol.%, the specific wear rate of a-C:H coatings remained almost constant, changing from 0.54 × 10⁻⁷ to 0.62 × 10⁻⁷ mm³/Nm. However, for a-C:H:Si coatings, it increased from 0.93 × 10⁻⁷ to 1.8 × 10⁻⁷ mm³/Nm. On the other hand, under N₂ pressure, the specific wear rate of a-C:H:Si coatings showed little change, remaining approximately 0.23 × 10⁻⁷ mm³/Nm. In contrast, for a-C:H:Si coatings, it increased from 0.28 × 10⁻⁷ to 0.55 × 10⁻⁷ mm³/Nm, though the variation was small.

The wear amount of DLC coatings varies depending on high-temperature, high-pressure environments, and dissolved oxygen concentrations, suggesting the possibility of chemical reactions between the carbon, hydrogen, and silicon contained in the DLC coatings. Therefore, AES measurements were conducted to determine the composition ratios of carbon, oxygen, and silicon within the wear scars of the DLC-coated pins. Additionally, depth-directional analysis using an Ar ion beam was performed to reveal the depth-directional distribution of elements within the wear scars. The depth analysis results are shown in Fig. 4 for a-C:H:Si under various different friction conditions, and in Fig. 5 for a-C:H.

The a-C:H:Si coating subjected to friction in an O₂ environment with a water content of 6 vol.%, which exhibited the highest wear volume, had the highest oxygen concentration on the coating’s topmost surface (Fig. 4a). Based on the results of Ar sputtering performed from the topmost surface to the depth direction, although the oxygen concentration in the topmost surface range of a few nanometers was around 10 at.%, only approximately 2 at.% of oxygen concentration was detected just underneath of the topmost surface.

Next, under nitrogen pressure conditions with a water content of 6 vol.%, which also resulted in high wear volume for the a-C:H:Si coating, the oxygen concentration on the topmost surface of the a-C:H:Si coating was second highest among the friction conditions. The oxygen concentration in the a-C:H:Si coating from the topmost surface to a few nanometers deep was around 2 at.%, indicating that only the topmost surface was slightly oxidized. Finally, it was found that the oxygen concentration in the a-C:H:Si coating in a pressurized environment with oxygen and nitrogen in ethanol showed almost no change from the topmost surface to the depth direction, remaining approximately 2 at.%, similar to other experimental conditions.

The results of friction test on the a-C:H coating under different environments are shown in Fig. 5. In the oxygen-pressurized environment with 6% water, the topmost surface of the a-C:H coating showed an oxygen concentration of approximately 4 at.%. However, under other experimental conditions, the oxygen concentration was almost constant from the topmost surface to the depth direction of coating, approximately 2 at.%. When Si is contained in the coating, the topmost surface of the coating oxidizes due to the influence of oxygen molecules and water dissolved in the solution. Even when nitrogen gas was pressurized into the solution, if the solution contains water, the surface oxidizes slightly. On the other hand, it was found that in a 100% ethanol environment, pressurizing oxygen gas did not promote the oxidation of the coating.

The wear amount of a-C:H coating increased proportionally to the amount of water present in the solvent, as shown in Fig. 3. However, it is evident that the wear amount significantly increased when the gas under pressure was oxygen. In contrast, the wear amount of a-C:H coating did not increase even when there was water in the ethanol or when oxygen was dissolved in the ethanol, compared to Si-containing coatings. Such differences in wear amount are considered to be due to different oxidation forms on the topmost surface of the coating, resulting in different surface hardness of the coating. Therefore, few nanometer thickness scratch tests using a diamond tip by through using AFM were conducted on each coating surface, and the topmost surface hardness of the coating was evaluated. The representative scratch test result is shown in Fig. 6. When conducting scratch tests on the same spot using AFM, repeated experiments were performed at 11 μN with a 2 μm square area. As shown in Fig. 6, in the regions indicated by arrows from 0 to 5 cycles, the surface height of the a-C:H:Si coating decreases with increasing number of scratch cycles. The average height within this 2 μm area was defined as the scratch depth, measured as the distance by which it decreased compared to the average height outside the wear track. Scratch tests were conducted on wear surfaces prepared under different conditions. Figure 7 shows the relationship between wear depth and number of scratches for a-C:H coating. Similarly, Fig. 8 presents the results for a-C coating.

The wear depth determined from scratch tests is deepest after 1 cycle, indicating the hardness of the topmost surface. In ethanol containing water, the wear depth of a-C:H:Si coating is deeper in both oxygen and nitrogen pressured conditions compared to ethanol alone, showing lower surface hardness. In the ethanol environment containing water, under oxygen pressure, the wear depth remains deeper than under nitrogen pressure even after 2 cycles, decreasing the hardness due to the effects of oxygen and water by approximately 2.0 nm from the surface of the a-C:H coating. In contrast, the influence of the pressured gas species has little effect on the wear depth in ethanol.

Next, the influence of water and gas species on the wear depth of a-C coating is shown in Fig. 8. Regardless of the gas or presence of water, the wear depth after 1 cycle is nearly the same. However, from 2 cycles onwards, the wear scars formed under oxygen gas resulted in slightly lower hardness.

To elucidate how the hardness of each coating changed during repeated friction tests inside the autoclave chamber, and to determine the initial wear coefficient K of each as-deposited coating, separate AFM tests were conducted on the as-deposited coatings. Additionally, using the scratch depth per scratch obtained from AFM scratch tests within wear tracks formed by friction tests in the autoclave chamber, the scratch hardness H of each coating was calculated [14, 18]. The results are presented in Fig. 9. In the figure, the hardness of each as-deposited coating obtained from the aforementioned AFM scratch tests is depicted with dashed lines. The hardness of the a-C:H:Si coating in a solution containing water and ethanol was consistently the lowest among all conditions, ranging approximately from 6.0 to 7.3 GPa under both oxygen and nitrogen pressures. Specifically, in ethanol-only solutions, oxygen pressure resulted in lower hardness compared to nitrogen pressure, measuring 9.6 GPa and 13.3 GPa, respectively. Conversely, for the a-C:H coating, hardness decreased under oxygen pressure compared to nitrogen pressure, measuring 9.7 GPa in the presence of water and 9.9 GPa in ethanol-only environments. These results demonstrated lower hardness under oxygen pressure with water compared to 13.9 GPa and 14.0 GPa under nitrogen pressure and ethanol-only conditions, respectively, revealing that the hardness of the a-C:H coating decreased specifically when water was present in ethanol. This suggests that the presence of Si within the coating compared to a-C:H coating causes a decrease in hardness, particularly under conditions containing water in ethanol.

The authors have previously reported the influence of pressure on the wear of a-C:H coatings in high-temperature and high-pressure environments using water as the solvent [11]. It has been revealed that the effect of pressure on the wear of a-C:H coatings was small, but there was a tendency for wear to increase with higher levels of dissolved oxygen in the water used for pressurization. However, it was unclear which form of oxygen—either oxygen within water molecules or oxygen molecules dissolved in water—has a greater impact on wear. In this study, ethanol was used as the solvent and oxygen and nitrogen were chosen as the pressurizing gases. This choice allows for clarifying the influence of oxygen contained within ethanol molecules on the wear of a-C:H:Si and a-C:H coatings. Furthermore, by using water as the solvent, the effect of water on the wear of each coating can also be clarified. Assuming that material wear follows Archard’s wear law, there exists a proportional relationship between the inverse of material hardness and wear.

Thus, Fig. 10 illustrates the relationship between specific wear rate obtained from friction tests in the autoclave chamber and the inverse of scratch hardness obtained from AFM scratch tests. Closed symbols represent a-C:H:Si coatings, and open symbols represent a-C:H coatings. The result with the highest specific wear rate in Fig. 10 is the a-C:H:Si coating, which contains 6 vol.% water in the ethanol solvent and is pressurized by oxygen. The topmost surface hardness within the wear scar of this a-C:H:Si coating was lower than any other coating. The next point of interest is the effect of oxygen pressurization in 100 vol.% ethanol on the a-C:H:Si coating and the a-C:H coating, as indicated by the triangles in the figure. Despite both coatings having almost the same inverse scratch hardness value, the a-C:H:Si coating has a higher specific wear rate. This implies that the presence of Si accelerates the wear of the coating due to oxygen molecules dissolved in the solvent. In water-containing ethanol, represented by the square symbol, the inclusion of Si tends to reduce the hardness of the topmost surface. When the water in the ethanol is removed, as indicated by the diamond symbol, the hardness of both the a-C:H:Si coating and the a-C:H coating is almost the same, and the amount of wear is also minimal.

As mentioned earlier, both a-C:H:Si coatings and a-C:H coatings follow Archard’s wear law [19], with the wear mode being proportional to the inverse of the scratch hardness. This reduction in hardness is considered to be due to changes in the bonding state of each coating composed of carbon, hydrogen, and Si atoms. Therefore, the relationship between the specific wear rate and the oxygen content on the topmost surface of the coating, obtained through AES analysis, expressed as the O/C ratio by dividing the oxygen content by the carbon content composing the entire coating, is shown in Fig. 11. A proportional relationship was observed between the specific wear rate and the oxygen content on the topmost surface of each coating. For the oxidation states of each coating, in the case of a-C:H:Si coatings, C–O, C=O bonds, Si–O bonds are considered. On the other hand, in the case of a-C:H coatings, only C–O and C=O bonds are considered for the oxidation states. To clarify the factors that promote this oxidation, the relationship between the water content in ethanol and the O/C ratio is summarized, as shown in Fig. 12. Only the O/C ratio of a-C:H:Si coating changes with the presence or absence of water, and it is also clear that it is not affected by the type of gas used for pressurization, although oxygen affected the most to be the highest O/C ratio.

In the oxidation of silicon carbide or silicon nitride in water, two models of oxidation have been proposed: oxidation by H₃O⁺ and OH⁻ ion pairs, and cooperative oxidation by a single water molecule acting as both an acid and a base [20, 21]. If oxidation progresses through cooperative reactions, it does not require the prior breaking of the ceramic’s bonds, which generates dangling bonds. This allows oxidation to occur at a lower energy level than oxidation by O₂, as reported in the reference [21]. This aligns with the above discussion and the schematic image is shown in Fig. 13. Therefore, in a high-temperature, high-pressure environment, a-C:H:Si coatings with 6 vol.% water are oxidized through cooperative reactions with water molecules, leading to oxidation.