Enhancing Lubrication of Electrified Interfaces by Inert Gas Atmosphere

With the proliferation of electric vehicles (EVs) around the world, the interest in further enhancement of their performance, efficiency, and reliability has also increased significantly in recent years. In particular, the moving mechanical assemblies in EV drivetrains were found to present some reliability challenges due to stray electricity or shaft currents passing through the contact interfaces of bearings and gears and thus causing surface damage [1,2]. The shaft currents primarily arise from the magnetic flux asymmetry but inverter-induced voltage fluctuations and electrostatic effects may also contribute to them and thus compromise the reliability of the driveline systems of EVs [1].

The main problem with shaft currents is that when they flow through the contact interfaces of bearings or gears, they can create micro-craters on their rolling/sliding surfaces which may lead to larger-scale surface damages and hence premature failures [3–5]. Further, it was shown that passing or discharging electricity at such interfaces can generate high heat (due to Joule heating) which can, in turn, reduce the viscosity of lubricating oils and increase oxidation. Collectively, these can lead to accelerated wear and micro-pitting in the drivetrain components [6–12]. In fact, some of the most notable wear losses in lubricated contacts under electrified conditions have lately been ascribed to accelerated oxidational wear at the sliding contact interfaces [6]. Mechanistically, corrosion and/or oxidation of metals involves some electrochemistry, hence the presence of stray electricity can only further exacerbate such degradation process. To diminish or eliminate the adverse effect of stray electricity or shaft currents, a variety of corrective actions have been developed in recent years. These include the use of shaft grounding brushes [13], conductive lubricants [14], and/or highly dielectric materials/coatings [15], etc.

Reducing or eliminating the detrimental effects of oxygen and water molecules in air on friction and wear has been studied extensively for many decades [16,17]. For example, under dry N₂ (a relatively inert/nonreactive gas), hydrogenated DLC coatings exhibit some of the lowest friction and wear coefficients [18]. In CH₄, MoN-Cu coating can cut down friction by half and reduce wear losses by factors of 3−4 orders of magnitude by forming a carbon-rich tribolayer on sliding surfaces [19]. Likewise, in other inert gasses, steel−steel test pairs can also provide much superior wear performance by limiting the degree of oxidation and promoting the formation of a carbon-based tribolayer as in the case of MoN-Cu tested in methane gas or under oil lubrication [20–22]. So, the critical impact of test environments on friction and wear of sliding surfaces has long been recognized.

From these studies, it has become clear that oxygen together with water vapor in air can cause tribo-oxidation and hence oxidational wear regardless of being tested under dry or lubricated conditions. As studied extensively by Terence Quinn back in the 1970s, oxygen in the air (potentially from water molecules as well) can chemically react with nascent iron in steel during sliding and thus form an oxide-rich tribolayer. Initially, such films may limit wear by protecting the sliding surface, but, as the oxide film grows thicker and becomes vulnerable to breakage and/or delamination, a transition to severe oxidational wear takes place through a third-body abrasive wear mechanism [23–25]. Recent experimental studies have further confirmed that tribo-oxidation is indeed very detrimental to the wear performance of sliding contact interfaces. The creation of an inert gas environment prevented oxidation and hence abrasive third-body wear [22,26–28]. Instead, the nascent iron surfaces created during sliding were able to catalyze the hydrocarbon molecules of lubricating oils and thus result in a protective carbon-based tribolayer as in the case of the catalytically active MoN-Cu coatings which provided extreme resistance to wear [26,29]. From all these studies, it became evident that the chemical nature of the surrounding atmosphere plays an important role in the friction of wear behaviors of sliding contact interfaces regardless of being tested under dry or lubricated conditions.

Considering the fact that sliding contact interfaces suffer severe oxidation and hence oxidational wear under electrified conditions [6], in this paper, we explored the possible enhancement of lubrication of electrified interfaces by using an inert gas environment (i.e., N₂). Our results confirm that under electrified conditions, sliding steel surfaces suffer much more severe oxidational wear than unelectrified sliding conditions in the open air, whereas little wear occurs upon testing in dry N₂ regardless of electrification or the lubricant being used. Such an impressive performance results from the more effective or favorable formation of a carbon-rich tribolayer under electrification.

Tribological tests were performed using steel (AISI 52100) test pairs in a pin-on-disk tester (Anton Paar, TRB3) equipped with an air-tight plexiglass enclosure and polymeric dynamic seals for maintaining the inert gas atmosphere (N₂ gas) through the tribological tests. It was modified to allow testing under unelectrified and electrified sliding conditions via a two-electrode cell, as shown in Fig. 1. Different lubricants such as two synthetic base oils (PAO 2 and PAO 10), two transmission oils with different specifications (a Synthetic ATF and commercial Dexron VI), and a gear oil were tested under boundary lubrication conditions (ʌ < 0.7) to elucidate the extent and nature of oxidation and/or oxidational wear in different oils and under electrified/unelectrified test conditions. The kinematic viscosities of these oils are given in Table 1. Both AISI 52100 steel balls (12.7 mm diameter) and disks (80 mm diameter) had a hardness value of 64 HRC, and a surface roughness of standard ball and disk specimens, Sa, of 0.02 ± 0.01 μm, as measured by the optical profilometer for all the tested specimens.

The sliding contact was insulated by using a ball holder made of PEEK and an insulating disk holder which was also made of PEEK. The test conditions are listed in Table 2. An electrical DC power supply device (Keithley, 2230-30-1) was used to electrify the sliding contact by connecting it to the ball and disk, respectively. DC power was applied to keep a continuous electrical current passing through the sliding contact interface thus allowing to study the effects of electricity on friction and wear during lubricated tests. The magnitude of the applied current was 3 A, which was selected to reproduce a harsh electrical environment and generate significant wear alterations by running short tribological experiments. It is noteworthy that short experiments were wanted since very large amounts of gas were required for running longer experiments. Due to the application of a constant DC current, i.e., 3 A, the voltage was variable (in the range of 0.2–1.5 V for the lubricants and contact conditions tested) and dependent on the contact resistance of each sliding contact interface according to Ohm´s law. A fixed wire was used for connecting to the ball by inserting the wire into a cavity of the holder and establishing electrical contact with the stationary ball sample. A carbon brush was used for connecting the power supply to the rotating disk sample. The disk was slid against the carbon brush to allow continuous electrical contact. In the experiments, a small amount of oil (10 μL) was spread over to cover the sliding wear track area in the disk and found to be more than enough to keep the wear track lubricated during the entire test. The coefficient of friction (CoF) was measured and logged by the tribometer's data acquisition software for post analyses. Both unelectrified and electrified tests were run in ambient air and in N₂ atmosphere (for which the chamber was continuously fed with N₂ gas through a supply inlet to create and maintain a positive pressure inside the test chamber and prevent air from backfilling the chamber). Three repeat tests were run for the ambient air cases while two repeats were done for the N₂ atmosphere cases. In the last, only two repeat tests were done due to the large amount of gas required.

Figure 2 compares the lubricated wear performance for both ball and disk samples under unelectrified and electrified conditions in ambient air. As is clear, most of the ball and disk test pairs suffer significant wear under electrification, especially in PAO2. This is most likely due to the very low viscosity and the absence of anti-wear additives in base PAO2 collectively giving rise to more severe metal-to-metal contact conditions, and hence much higher wear. In fact, the wear of the balls tested in PAO2 increases by more than sixfold under electrification in air. In the case of higher viscosity PAO10, the amount of wear was much lower even under electrified sliding conditions. Marked increases in the wear of test pairs under electrification in PAO oils can be attributed to much faster and more extensive oxidation of the rubbing surfaces (especially in the case of PAO2 oil) as will be further elaborated later in light of Raman and EDS analyses.

As for formulated oils, Fig. 2 reveals that formulated oils (ATF, Dexron, and gear oil) with all those functional additives afford reasonable wear resistance to all test pairs regardless of electrification and ambient air environment. Obviously, functional additives in such formulated oils enable them to form protective boundary films and thus limit wear [30,31]. As highlighted in the inset in the top right corner of Fig. 2(a), except for ATF in air, all of the other formulated lubricants (especially Dexron) caused somewhat higher wear upon siding under electrification; but the extent of wear is not as severe as those observed in pure PAO oils. Overall, contact electrification increased the wear of the disk samples as shown in Fig. 2(b). However, the magnitude of their wear losses is still small compared to the disks tested in pure PAO oils under electrification.

In contrast to open-air test results presented in Fig. 2; tests in dry N₂ resulted in dramatic reductions in wear volumes of steel balls and disks in all lubricants tested (see Fig. 3). The greatest reduction (i.e., by factors of 8−10) was seen on samples tested in PAO2 and PAO10 oils in dry N₂. Again, such a dramatic reduction in wear in N₂ can be attributed to the absence of oxygen and/or water molecules in test environments. In formulated oils, the decrease in wear was not as dramatic upon switching from air to N₂ but was still noticeable. Overall, dry N₂ was beneficial to the wear performance of ball and disk samples.

Much higher wear of test pairs in the air can be attributed to severe oxidation of their rubbing surfaces leading third body wear. Conversely, the suppression of tribo-oxidation and hence oxidational wear leads to much lower wear losses in dry N₂. As reported earlier, on catalytically active surfaces and/or dry N₂, a protective carbon tribolayer forms under lubrication and thus reduces the wear [32]. Note that iron in steel is a well-known catalyst metal but when in an oxidizing environment (such as air), its extreme affinity to react with oxygen to form an oxide layer overrides its catalytic reactivity, as a result, very little or no such carbon layers form on sliding surfaces in open-air environments.

Concerning the tests under electrified conditions, catastrophic wear losses occur on rubbing surfaces of balls and disks during tests in air as shown in Fig. 4, especially in base PAO2 and PAO10 oils. This is most likely due to accelerated tribo-oxidation leading to the generation of highly abrasive iron oxide wear debris triggering third body wear of contact interface. Conversely, when tests were run in dry N₂, the amount of wear on ball and disk samples was reduced dramatically (i.e., by a factor of two orders of magnitude reduction when tested in PAO2,) as shown in Fig. 4. Again, the lack of oxidation and hence the absence of abrasive third body oxide particles at sliding interface lead to much-reduced wear in dry N₂. As will be discussed, the formation of a carbon-rich tribolayer was also an important factor in reduced wear. Note that reductions in the wear of test pairs in formulated oils were also significant but were not as dramatic as in the base PAO oils.

Contact electrification can also affect the CoF of the test pairs in both air and N₂ environments (see Fig. 5). In particular, the CoF of base oils (PAO 2 and PAO 10) increased upon electrification regardless of the test environment. The change in the CoF in formulated oils was less evident (e.g., within margins) regardless of the test environment and electrification. From these results, it can be deduced that additives in formulated oils are highly effective in ensuring a relatively stable CoF, regardless of electrification and/or test environment. With regards to the increase in CoF for PAO2 and PAO10 under electrification, it can be attributed to the severe oxidation of rubbing surfaces combined with the lack of a protective tribolayer as in the formulated oils. Moreover, it is noteworthy that electrical discharge machining (EDM) may also have a certain influence on the wear under all the electrified tests. Nevertheless, we did not find significant evidence on the wear tracks to support this. The lack of EDM evidence in these experiments is due to (1) the tests were run under boundary lubrication, which means that there are always contact spots at the interface, and thus EDM is limited; (2) the abrasion caused by the ball sliding removes material from most of the possible eroded regions.

Figures 6 and 7 present the microscopic images and surface profiles of the wear scars and tracks formed during tests in ATF under unelectrified conditions in the open air. It is noteworthy to mention that similar wear scar patterns were also observed after tests in other commercial oils. So, an ATF was selected to get a deeper understanding of the wear mechanisms involved. It is clear that electrification caused a significant increase in wear damage on both the ball and disk sides, as shown in Fig. 7. Also, in contrast to the unelectrified tests, electrified tests generated much more wear debris (identified visually due to their russet color) which were accumulated in and around the wear scar and track, as shown in Fig 7. According to the Raman analyses conducted on such debris particles accumulated around the edges of wear scars and tracks (see Fig. 8), the formation of a thick iron oxide layer and debris particles (hematite (α-Fe₂O₃)) was confirmed for both the unelectrified and electrified cases. The Raman bands corresponding to peaks at 227, 293, 409, 498, 611, 658, and 1320 cm⁻¹ are all indicative of the formation of hematite [33]. Interestingly, carbon D and G bands also appeared on wear tracks/scars formed under electrification. The formation of such carbonaceous products can be attributed to the increased polymerization and/or limited decomposition of oil molecules under electrification. A further increase in flash heating (due to resistive or Joule heating) of interacting asperities can accelerate this decomposition process. Specifically, electrons passing through the contact interface can potentially break down the carbon−carbon bonds of the long-chain hydrocarbon molecules of oils and deposit them as solid carbon products on rubbing surfaces. Increased contact resistance due to tribo-oxidation can also give rise to much higher ohmic heating (or resistive heating in addition to frictional heating) and hence accelerate this decomposition process. Conversely, without such events under unelectrified conditions, very little or no carbon deposit forms during the test in open air (see Fig. 8(a)).

Figures 9 and 10 show the images and profiles of the wear scar and track formed in the ball and disk samples, respectively, after testing in the N₂ atmosphere under unelectrified and electrified conditions using the same ATF oil. As is clear, the extent of wear on both sides is insignificant regardless of being electrified or not compared to the open-air test results. Simply, the N₂ atmosphere effectively diminished the oxidation of contact interfaces, and hence the severity of the oxidational wear. However, some black/dark-contrasting tribolayers were observed across the ball's wear scar (Fig. 11(b)) formed under electrification. The Raman spectra obtained from such black/dark-contrasting regions revealed a carbon-rich tribolayer as denoted by arrows in Fig. 11(b). There were also some weak Raman bands belonging to iron oxide (hematite) (though at low intensity) but by and large, surface tribolayer was mostly made of carbon-based products. Under unelectrified conditions (Fig. 11(a)), Raman spectroscopy of a dark contrasting spot also revealed the existence of some oxides and perhaps some carbonaceous products as manifested by a strong 2D band. All of the bands belonging to oxides may have been due to the native oxide layers that were present on both the ball and disk surfaces before the test. The presence of some ppm level oxygen in the test chamber continuously flooded by N₂ gas cannot also be ruled out and hence may have been the cause of such minor oxide peaks.

In general, these Raman spectra revealed that the most prominent peaks belong to amorphous or disordered carbons with distinct D and G bands, as in Fig. 11(b), suggesting that under an N₂ environment, passing of electricity through the contact interfaces results in the formation of a carbon-rich tribolayer and abrasive nano-carbon particles as a consequence, as that found recently by Deshpande et al. [34]. The absence of oxygen and/or water molecules in the test chamber ceases oxidation, and hence enhances the catalytic reactivity of iron and thus the formation of a carbon-rich tribolayer. However, much deeper and more extensive systematic studies on tribochemistry occurring under electrified inert-gas environments must be carried out, which is a topic of ongoing research of our group.

The findings of this work confirm that creating a dry N₂ atmosphere can dramatically enhance the lubrication of steel surfaces under electrified conditions. In air, severe oxidational wear occurs and compromises the overall tribological performance of such surfaces. Base and formulated oils tested in our study exhibited similar wear trends with respect to the presence or absence of electrification and a dry N₂ atmosphere. Raman spectroscopy of the worn surfaces confirmed the formation of more extensive iron oxide (hematite) and hence abrasive debris particles on rubbing surfaces under electrified sliding conditions in air. In dry N₂, Raman bands belonging to D and G bands of a carbon-rich tribolayer that is similar to the ones reported in Ref. [32] emerged. With the formation of such tribolayers, the lubrication performance of all fluids was improved. Overall, from the findings of our study, it is clear that in the presence of dry N₂, the extent of oxidation and hence the oxidational wear diminish, while the formation of a carbon-rich tribofilm takes precedence and minimizes the extent of wear. Accordingly, the creation of an inert environment around the driveline systems of future EVs could be beneficial to their long-term performance and hence reliability.

The authors acknowledge the Texas A&M Engineering Experiment Station startup funds and the Governor's University Research Initiative funds for supporting this work.

There are no conflicts of interest.

The authors attest that all data for this study are included in the paper.