Article pubs.acs.org/Langmuir
Tribochemistry of Carbon Films in Oxygen and Humid Environments: Oxidative Wear and Galvanic Corrosion Ala Alazizi,† Andrew Draskovics,† Giovanni Ramirez,‡ Ali Erdemir,‡ and Seong H. Kim*,† †
Department of Chemical Engineering and Materials Research Institute, The Pennsylvania State University, University Park, Pennsylvania 16802, United States ‡ Energy Systems Division, Argonne National Laboratory, Argonne, Illinois 60439, United States S Supporting Information *
ABSTRACT: The effects of oxidation on wear of carbon/steel tribological interfaces were studied. When mechanical wear was small, the oxidation behavior of hydrogenated diamond-like carbon (H-DLC) and stainless steel (SS) sliding interface varied depending on the nature of the oxidizing environment. In dry air or oxygen, both H-DLC and SS wore readily. The wear debris of SS did not form iron oxide in dry air and oxygen. In humid nitrogen, however, the wear of H-DLC diminished with increasing humidity, and the SS surface showed mild wear and iron oxide debris accumulated around the sliding contact region. These results revealed that different tribochemical reactions occur in dry oxygen and humid environments. In the absence of water, oxygen oxidizes the H-DLC surface, making it susceptible to wear, creating debris, and inducing wear on both H-DLC and SS. In contrast, adsorbed water molecules at less than 40% RH act as a molecular lubricant of the oxidized DLC surface, while multiwater layers adsorbed at near-saturation act as electrolyte inducing electrochemical galvanic corrosion reactions on the SS surface. When hydrogen-free amorphous carbon (a-C) was used in tribo-tests, severe wear of the SS surface occurs, in addition to the tribochemical wear observed for H-DLC, due to the high hardness of the a-C film.
1. INTRODUCTION Most carbon allotropes including diamond, graphite, and amorphous carbons experience oxidation of their outer surfaces, edge sites, or defects when they are exposed to water vapor or oxygen in ambient air. Surface oxidation of these carbon materials is a complex process.1 Decades long research on diamond and graphite showed that their oxidation depends on various factors such as oxidation temperature, oxidizing environment, crystal facets, and concentration of defects.2−4 Despite the long-held belief that amorphous carbons, especially hydrogenated ones, are inert, their surfaces are readily oxidized at ambient conditions, and their oxidation behavior is even more complex than that of crystalline forms of carbon.5−8 Several studies reported that amorphous carbon oxidation is dependent on its microstructure and the oxidizing environment as well.7,8 In tribological contacts, the shearing action by countersurfaces adds another layer of intricacy to the interfacial oxidation behaviors.9 The presence of water vapor inside the sliding contact area further complicates the oxidation process © 2016 American Chemical Society
since it can change chemical and mechanical properties of the sliding surfaces during the course of sliding.9 For example, as stainless steel slides on copper in dry atmosphere, catastrophic adhesive wear of the soft copper takes place at intermediate humidity.10 In contrast, when humidity is increased to nearsaturated conditions, galvanic corrosion takes place causing the much harder stainless steel to wear, instead wearing the soft copper surface.10 The presence of water vapor can also trigger tribochemical reactions of silicon materials.11 During friction tests of boron oxide and boron carbide in humid air, chemical reactions between water vapor and these surfaces result in the formation of lubricious boric acid which prevents wear of the substrate materials.12 Water can also change the bulk properties of materials. Adsorbed water on glass, for instance, can facilitate the hydrolysis of the strained Si−O−Si bonds and dissociate Received: November 16, 2015 Revised: January 24, 2016 Published: February 4, 2016 1996
DOI: 10.1021/acs.langmuir.5b04207 Langmuir 2016, 32, 1996−2004
Article
Langmuir
thickness increases upon increasing RH toward saturation, the SS counter-surface becomes subject to tribo-assisted electrochemical galvanic corrosion and wear. Similar tribochemical behaviors are also observed for much harder a-C films, although the environmental effects are less prominent due to the mechanical wear induced on SS counter-surface by the much harder a-C.
them, causing detrimental changes to the bulk properties of glass such as fatigue and fracture resistance.13 Numerous studies reported the effect of water vapor and oxygen in the testing environment on the frictional behavior of amorphous carbon coatings. It was reported that the formation of the carbon-rich film on the counter-surface after dry sliding (referred to as transfer film) is affected drastically by the presence of oxygen and water vapor in the surrounding environment.14 It was also suggested that water vapor and oxygen oxidize and passivate carbon surfaces.5,15 Several mechanisms have been proposed. The question of which mechanism is dominant seems to depend on the properties of the carbon film and the pressure or concentration of the oxidizing molecules in the surrounding environment. On hydrogenated diamond-like carbon (H-DLC) coatings, the oxidation by ambient air forms a surface layer that is rich in oxygenated functional groups.6 When friction of H-DLC is tested in dry nitrogen or vacuum environments, friction coefficient starts high but quickly drops during the initial conditioning period, which is referred to as run-in period. After this period, the friction coefficient drops to an ultralow value of 0.01 or lower. Momentary exposure to water vapor disrupts the ultralow friction observed in inert environment and causes another run-in behavior.5,7 Friction coefficients of H-DLC increase with increasing water vapor pressure.16 This increase was attributed to the high shear force required to slide on adsorbed water layer.14 The higher shear force could be related to the increase in adhesion between H-DLC surface and the counter-surface due to the adsorbed water in the contact.17 However, more recent experimental evidence suggested that oxidation causes chemical as well as structural changes of the surface layer of H-DLC.7 Surface analysis of native and thermally oxidized H-DLC showed that oxidation of H-DLC results in clustering of the sp2 aromatic structure, altering the mechanical properties of oxide layers which makes them more prone to wear and causes run-in behavior.7,18 The presence of oxygen was also found to disrupt the ultralow friction behavior of H-DLC.5,19 An in situ gas analysis study in vacuum conditions showed that the tribo-chemical reaction products produced during sliding were different for water and oxygen adsorption cases.8 Polycrystalline and hydrogen-free amorphous carbon films (a-C) show water vapor effects opposite to the H-DLC case; water vapor reduces friction of a-C significantly compared to dry environments.20,21 This was attributed to the ability of water to dissociate at carbon surfaces, forming a passivated layer at the sliding interface.20,21 The presence of oxygen, however, increased friction and hampered the desired effect of water vapor when they were present together.22,23 In this paper, we elucidated the difference in tribochemical reactions induced by water vapor (H2O) and oxygen (O2) environments when H-DLC and a-C surfaces are rubbed with steel counter-surface. The results in this paper show that the mechanical properties of the sliding materials become insignificant when the sliding materials have significantly difference electrochemical potential and are sliding in a highhumidity environment. Oxygen oxidizes the H-DLC surface, which facilitates wear of both H-DLC and SS surfaces. In contrast, the wear behavior in humid environment depends on relative humidity (RH). In low RH, the adsorbed water layers show molecular lubrication behavior, reducing wear of both the oxidized H-DLC surface and SS surface and giving a friction coefficient around 0.2. However, as the adsorbed water layer
2. EXPERIMENTAL TECHNIQUES Hydrogenated diamond-like carbon (H-DLC) and non-hydrogenated amorphous carbon (a-C) films were deposited on silicon substrates. HDLC was deposited in a plasma-enhanced chemical vapor deposition (PECVD) chamber, with a gas feed of 75% hydrogen and 25% methane.24 The silicon substrate was first cleaned and a silicon bond layer was deposited to improve the adhesion of H-DLC to silicon. The final coating consisted of ∼40 at. % hydrogen. a-C films were deposited using magnetron sputtering applying 2000 W (4.5 W/cm2) on a graphite target and using bias voltage of −35 V. Argon etching was used to clean the silicon substrate and high-power impulse magnetron sputtering was used to deposit the chromium bonding layer to enhance the adhesion between the Si substrate and the a-C coating. The thicknesses of both H-DLC and a-C films on silicon were measured with cross-sectional scanning electron microscopy imaging and found to be about 1.5 μm. The RMS surface roughness of H-DLC and a-C, as measured with optical profilometry, is about 4 nm for both films. The hardness (H) and elastic moduli (E) of H-DLC, a-C, and SS were measured using nanoindentation. Figure 1 shows the results of
Figure 1. Force−distance curves and the calculated hardness and elastic modulus values for H-DLC, a-C, and SS. the mechanical property tests. The measurements were performed using a Hysitron TI 900 nanoindenter, equipped with a three-sided diamond Berkovich tip with a radius of curvature of ∼150 nm and 65.35° half-angle. Figure 1 shows the displacement-controlled indentations performed to a depth of 150 nm on H-DLC, a-C, and 440C. The hardness and elastic modulus values were calculated using the Oliver−Pharr method (inset to Figure 1). The hardness of 440C SS is comparable to that of H-DLC, but much lower than that of a-C. As it pertains to this study, the main difference between DLC and a-C are their mechanical properties and their hydrogen content. Other than that there is only a slight difference in the D/G ratio of their Raman spectra. Friction tests were performed using a custom-made reciprocating ball-on-flat tribometer with a continuous-flow environmental control capability. The continuous flow of the test gas ensured that the small amounts of gaseous reaction products are purged out of the system and therefore have negligible effect on the sliding surfaces. All surfaces were rinsed with ethanol and water and blow-dried with nitrogen before testing. AISI 440 C stainless steel (SS) balls of 3 mm diameter and RMS surface roughness of ∼10 nm were reciprocally slid at a lateral speed of 4 mm/s over a 2.5 mm long sliding track (a frequency of 1.6 Hz) on the carbon-coated substrate under a load of 1 N. These 1997
DOI: 10.1021/acs.langmuir.5b04207 Langmuir 2016, 32, 1996−2004
Article
Langmuir conditions created an ∼50 μm wide contact area with a Hertzian contact pressure of ∼0.5 GPa. The depth of the wear found on the flat carbon films was uniform along the length of sliding with the exception of some pile-ups at the ends of the wear tracks. Friction tests were performed in dry oxygen, dry air, and humid nitrogen with RH ranging from 20 to 90% RH (Figure S1 in the Supporting Information). Among all RH conditions, results for only 20, 40, and 90% RH were investigated in depth. The choice is based on the wear behavior that was observed. 20% and 90% RH represent two distinct wear regimes, and 40% RH shows the transition between these two regimes. The friction coefficient showed no clear dependence on relative humidity. This is because at 20% RH and higher, there is a monolayer or more of adsorbed vapor,25 which always gives a friction coefficient ranging from 0.12 to 0.2 as long as wear is prevented.9,26−29 Previous studies revealed a clear friction dependence of DLC and a-C on RH at low partial pressure of water.30,31 It was shown that friction coefficient plateaus after 20% RH at around 0.15 for both films25,30,31the molecular lubrication friction value. After each friction test, the transfer film and wear debris were analyzed. Chemical analysis of the transfer film and wear debris was done using a WITec Raman instrument using a laser with a wavelength of 488 nm. Raman spectra of various points within and around the contact area were collected, and a representative spectrum was chosen. Energy dispersive X-ray (EDX) spectroscopy was used to perform elemental analysis of the debris on the wear tracks of H-DLC surfaces. A FEI Nova NanoESEM 630 field emission scanning electron microscope was used for EDX analysis. Optical profilometry was used to investigate the wear of carbon surfaces and steel balls after sliding. Three-dimensional images and line profiles of the wear tracks and wear marks were constructed from profilometry measurements taken with a Zygo NewView 7300 instrument utilizing a white light.
Table 1. Wear Rates of the Flat H-DLC/a-C and Stainless Steel Balls Measured after Friction Tests in Different Environments
3. RESULTS AND DISCUSSION The experimental results will be discussed in two parts. The first is the friction and wear behaviors of the H-DLC/SS interface. In this case, friction was relatively low, and there was a mild or negligible wear in all environments. The absence of mechanical wear made it easy to observe differences in oxidation behaviors of different environments (oxygen vs water vapor). The second part shows the friction and the wear behaviors of a-C/SS interface. The high hardness of a-C resulted in significant mechanical wear on the SS balls in all environments. However, careful analyses of wear volume and debris suggested that there are contributions from the same tribochemical reactions observed for H-DLC by oxygen and water vapor. The friction tests were repeated at least four times for each system, and the wear rates were compared after every experiment. All experiments showed similar wear patterns, and data from one representative experiment were included. Table 1 summarizes the wear rates of the H-DLC/SS and a-C/SS interfaces. For the flat carbon surfaces, the wear rate was expressed as the volume of materials removed per cycle per applied load (N). For the ball surface, the wear was measured as the volume of material removed per sliding distance (mm) per applied load (N). These wear rates will be discussed in each section. 3.1. Oxidation Effects on Wear of H-DLC/SS Interface. Figure 2 shows the friction coefficient measured as a function of reciprocating cycle when the SS ball was slid on H-DLC in various environments. In dry nitrogen, typical H-DLC friction behaviors are observedafter a short run-in period, the friction dropped to an ultralow value. During the run-in period, the surface oxide layer formed by exposure to air during the sample storage is removed.7 The introduction of oxygen and water vapor to the test environment showed two different friction
Figure 2. Friction coefficient of 440C SS sliding on H-DLC in various environments.
behaviors. In dry oxygen and dry air, the friction of H-DLC/SS interface showed the run-in period initially, but after this period, friction coefficients were noisy and varied randomly within 0.05−0.2. In humid nitrogen, the H-DLC/SS friction did not show the run-in period observed in dry environments; instead, it varied gradually and smoothly between 0.15 and 0.3. These differences in the friction coefficient suggested that the chemical reactions occurring at the H-DLC/SS interface due to oxygen and water vapor must be different. Optical images of the SS ball wear marks are shown in Figure 3a. In the case of dry nitrogen, there was a thick carbon transfer film covering the whole contact area, but in all other environments, wear debris were outside and/or on the periphery of the contact area. It was noted that the amount of visible wear debris was always larger in the dry oxygen and air cases compared to humid environment tests. Raman spectra of transfer films and wear debris were collected at multiple locations, and representative ones are shown in Figure 3b. As Figure 3b shows, the transfer films formed during friction tests in nitrogen, oxygen, and air show the typical spectral features of H-DLC with D and G bands; the D band of the transfer film is slightly more intense than that of pristine H-DLC.25 Oxidation of the H-DLC was found to contribute to the higher intensity of the D band due to the sp2clustering effect that takes place when H-DLC oxidizes.7,18 In the case of dry nitrogen, the enhanced D band is due to oxidation of the transfer film during the sample mounting to the Raman spectrometer.5,7 The D band intensity increased as the test environment was changed from dry N2 to dry air and 1998
DOI: 10.1021/acs.langmuir.5b04207 Langmuir 2016, 32, 1996−2004
Article
Langmuir
Figure 3. (a) Optical microscope images of the 440C SS ball after friction tests in Figure 2. (b) Raman spectra of the transfer films in (a). Iron oxide peaks were detected on the periphery of contact areas (see arrows in part a). Scale bar is 50 μm.
dry O2 due to oxidation during the tribo-test.5,7 As the oxygen partial pressure increases during the friction test, the D band enhancement of the transfer film becomes more prominent due to oxidation. The Raman spectra of wear debris formed in humid environments were quite different from those of dry test environments. At 20% RH, there was a small amount of wear debris around the periphery of the contact area. Raman spectrum of the wear debris showed a D/G band ratio lower than that of dry O2 and air cases. These observations indicated that the H-DLC transfer film formation was substantially reduced, and the oxidation of the transfer film was much lower. As RH increased further, the characteristic Raman features of H-DLC became less noticeable within the transfer film, and multiple sharp peaks in the low wavenumber region were dominant. At 90% RH, sharp peaks were observed at 235, 305, 415, 505, 625, 670, and 1320 cm−1. These peaks are similar to iron oxide peaks, particularly hematite.32,33 The transfer film of the 40% RH test showed a main peak around 670 cm−1, and the rest of the peaks had much lower intensity. The peak ratios in both 40% and 90% RH transfer film spectra do not exactly resemble the reference peaks for hematite or magnetite only. This suggests that the wear debris contains a mixture of hematite and magnetite after friction tests in intermediate and high humidity environments.32,34,35 The fact that iron oxide peaks in the 90% RH transfer film spectrum resemble hematite more than those in 40% RH spectrum is an indication that the tribochemical reaction is dominated by amount of water present. At near saturation, stoichiometric oxidation of the steel counter surface occurs forming more of hematite (Fe2O3) than other types of iron oxide. However, at 40% RH steel is less oxidized forming magnetite (Fe3O4). Overall, the Raman analysis results indicated that the adsorbed water layers in the sliding H-DLC/SS interface suppress the transfer of H-DLC onto the SS counter-surface and induces oxidation of the SS wear debris. The identification and relative abundance of polymorphic iron oxide species were difficult and beyond the scope of this paper.36−39 We have checked if iron oxide debris from the SS countersurface were deposited or left on the H-DLC substrate surface. The Raman spectra of the ends and edges of the wear tracks on the H-DLC substrate did not show any iron oxide peaks above the noise level of the background (Figure 4). Here again, the
Figure 4. Raman spectra collected from the wear debris on the sliding tracks on DLC after friction tests shown in Figure 2.
higher D/G ratio in the wear debris around the wear tracks of air and oxygen environments tests compared to humid environments tests indicate oxidation of DLC was prominent in the presence of oxygen but less in the presence of water vapor. EDX analysis was also performed on the sliding tracks to verify the Raman findings. The FESEM images and the EDS spectra and analyses can be found in the Supporting Information. The amount of iron detected was very close to the detection limits of EDX. This suggests that a negligible amount of iron oxide was left on the H-DLC surface and that the majority of wear debris were adhered onto the SS countersurface surface. The results shown in Figures 3 and 4 reveal that the effects of oxygen and water vapor on transfer films or wear debris accumulated on the counter-surface are quite different. Oxidation of the SS counter-surface sliding on H-DLC does not occur in dry oxygen and air; but when RH increases, the SS surface is oxidized. Figure 2 also indicate that the effects of oxygen and humidity on friction are quite different. In order to fully understand oxidation mechanisms, the wear of both HDLC and SS surfaces were measured using optical profilometry. Figures 5 and 6 show optical images and line profiles of the wear tracks on H-DLC and the wear mark on the SS ball, respectively, after friction tests in the aforementioned environments. The steel balls were vigorously cleaned by sonication 1999
DOI: 10.1021/acs.langmuir.5b04207 Langmuir 2016, 32, 1996−2004
Article
Langmuir
The effects of humidity on wear of H-DLC and SS were quite different from those of oxygen. In the presence of water vapor, the wear of H-DLC did not increase at all. As Figure 5 shows, at 90% RH, the wear of H-DLC was almost completely suppressed; the wear depth was very small or negligible, and it was extremely difficult to find the wear track. The wear track could be identified only by the small amount of debris left at the both ends of the slide track (which could be seen as dark spots in the 90% RH optical microscope image in Figure 5). Whereas the H-DLC wear decreased, the SS counter-surface wear increased slightly up to 10−15 nm deep as RH was augmented (Figure 6). This is also quantitatively shown in the data within the top red rectangle of Table 1. It is important to note that the small wear of SS in high RH conditions resulted in formation of iron oxide species that could be easily detected with Raman, while the larger wear of SS in dry oxygen did not produce iron oxide phases. The results described here raise an interesting and important question: why does wear occur dif ferently in dif ferent oxidizing gas environments? Archard’s relationship predicts that the wear of a material would be inversely proportional to its yield stress.40 Therefore, in the absence of tribo-chemical oxidative reactions or structural changes during sliding, mechanical properties of the sliding materials would dictate wear processes. H-DLC and 440C SS have comparable hardness and elastic modulus values. Although we could not measure the hardness of the oxidized surface layer of H-DLC (too thin to measure with nanoindentation), this layer has a lower density than the pristine HDLC due to partial removal of carbons through reactions with oxygen producing CO2 and H2O.7,41−43 Thus, the mechanical properties of H-DLC oxide layer are expected to be inferior compared to the bulk film and therefore more susceptible to wear during the run-in period. In dry nitrogen, the oxide layers of H-DLC are removed, and the removed portion of H-DLC is transferred to the SS counter-surface, protecting the SS surface from wear. Then, the interfacial shear occurs between H-DLC and the transferred H-DLC, which results in ultralow friction. In the presence of oxygen, continuous formation of oxide layers and their removal by the counter surface result in significant H-DLC wear.44 This is evident by the significantly higher wear rate in dry oxygen (PO2 = 760 Torr) compared to dry air (PO2 = 152 Torr), indicating that the amount of oxygen in the environment plays a critical role. Removal of the oxidized
Figure 5. Sliding tracks on DLC surface and their profile.
and rinsing with ethanol to remove the transfer film and wear debris accumulated on the ball surface during the friction test. Since the radius of curvature of the ball is much larger than the wear depth in all cases, the curvature of the ball was mathematically removed to highlight the wear formed in different test environments. In a dry N2 environment, H-DLC wear is very mild (Figure 5); typical wear depth was about 10 nm. The SS ball wear was negligible (Figure 6) since the friction coefficient of the HDLC/SS interface dropped quickly to the ultralow friction state (Figure 1). In dry oxygen and dry air, much deeper wear tracks (60−90 nm) were created on H-DLC (Figure 5). Based on the previous findings about the easy oxidation of the H-DLC surface and the wear susceptibility of the oxidized surface layer,5,7,8 the increase in H-DLC wear in O2 and air environments can be attributed to the combined actions of continuous surface oxidation by oxygen and repeated removal of the oxidation product by interfacial shear during sliding. The wear of H-DLC substrate in dry O2 and dry air was accompanied by the SS counter-surface wear. The optical profilometry images and line profiles of these wear marks on the SS ball surface (Figure 6) showed that the wear of the ball in dry air and oxygen was the largest among all environmental conditions tested in this study (Table 1). It is intriguing that this wear of the SS ball surface occurred without formation of iron oxide species detectable with Raman spectroscopy (Figure 2b).
Figure 6. (a) Optical profilometry 3D images of the 440C SS balls with curvature removed after friction tests on DLC. (b) Line profiles across the wear marks of the 440C SS ball after friction tests in 0 to 90% RH. Inset to (b) shows line profiles in (b) along with the line profiles of the wear mark after air and oxygen tests for comparison. 2000
DOI: 10.1021/acs.langmuir.5b04207 Langmuir 2016, 32, 1996−2004
Article
Langmuir
Figure 7. (a) A diagram showing the electrochemical galvanic potential ranges for graphite and different kinds of steels. Gold and copper are added for comparison. For passivated steels, dark red and light red show galvanic potential range with and without the passivation layer, respectively. (b) A schematic showing how an electrochemical reaction can take place within the contact area of steel/DLC in the presence of sp2-rich transfer film and adsorbed water.
the H-DLC structure were negligible during the tribo-test in high RH conditions. This also supported the molecular lubrication effect of the adsorbed water layer, protecting the H-DLC surface that was preoxidized by the sample storage in ambient air. As RH approaches the saturation point, multilayers of water can be adsorbed on the H-DLC surface.25 The presence of thick multilayers of water can introduce another mechanism beside lubrication by physisorbed molecules. When the adsorbed water layer becomes thick enough, they can act as an electrolyte,10 allowing an electrochemical reaction, or corrosion, between two dissimilar materials with different corrosion potentials. Measurements of corrosion rate and electrical resistivity within the vicinity of corroding steel showed that as RH increases, the electrical resistivity decreases and corrosion rate increases, indicating that adsorbed water indeed acts as an electrolyte that facilitates corrosion.49,50 It was also previously found that adsorbed water can trigger electrochemical reactions when multilayer of adsorbed water is present at the sliding interface of steel and copper.10 This electrochemical reaction could explain the formation and accumulation of iron oxide particulates around the sliding contact region after friction tests in 90% RH and, to a lesser extent, 40% RH. Galvanic corrosion is an electrochemical reaction that occurs when two materials with different electrochemical potential are in contact in the presence of an electrolyte. This leads to a preferential corrosion of the more anodic material. The galvanic corrosion potentials of 440C SS is more anodic than that of graphite (Figure 7a). The electrochemical potential of H-DLC is unknown, but it may not be as cathodic as graphite. However, oxidized H-DLC layers are rich in sp2 aromatic rings,7,41 and the interfacial shear of H-DLC during friction test can transform H-DLC to graphitic carbons.7,51,52 When a graphitic layer is present, then the steel counter-surface could be galvanically corroded by the sp2-rich carbon in the presence of sufficient amounts of water adsorbed from the gas phase. Figure 7b schematically illustrates of how electrochemical reaction can take place at the sliding interface of sp2-rich layers/steel in the presence of thick multilayer of adsorbed water,25 protecting HDLC and corroding the steel surface. The stoichiometry of the electrochemical reaction suggests the formation of hydrogen gas. However, in our environmental cell, the production of small amount of hydrogen could not be detected. Generally, metal oxides can play an important role in friction behavior when wear is prevented. This is why at the beginning
layers from the DLC surface produces wear debris that can have abrasive effect on the counter surface resulting in mutual wear of both surfaces.45 The abrasive effect of wear particles cause the third body contacts, resulting in noisy friction coefficients in dry oxygen and air environments (Figure 2). Wear debris are present within and around the contact area on the steel balls (Figure 3a) after friction tests in dry oxygen and dry air environments. The absence of iron oxide after friction tests in dry air and oxygen suggests that chemical oxidation of the worn steel ball surface or wear debris does not occur prominently. Studies on corrosion of steel showed that the oxidation rate of steel by oxygen in the absence of water vapor is very low or negligible.46−48 The oxidation rate would have been even lower in the presence of carbons (DLC transfer films) which can act as a reducing agent. In humid environments, water molecules readily adsorb on the oxidized DLC surface providing molecular lubrication and reducing wear.25,44 Our previous measurement of water adsorption isotherm on DLC showed that as RH increases, the thickness of adsorbed water increases quickly to a monolayer of water (∼0.3 nm thick) at around 20% RH. The adsorbed water layer keeps on growing slowly as RH increases from 20% to 90% RH. It reaches ∼0.5 nm at 40% RH and increases to a maximum thickness of ∼1 nm near saturation RH.42 In equilibrium conditions, adsorbed vapor molecules cannot be squeezed out of the contact area.26 It was found that physisorbed molecules adsorbed from vapor prevent wear and give a friction of 0.15−0.2 as long as no mechanical failure takes place.9,26 At least a monolayer of adsorbed water is expected to form on H-DLC at RH higher than 20%.25 Lubrication by physisorbed molecules typically gives a friction coefficient of about 0.15−0.2, and as the size of the physisorbed molecules decreases, this value could increase slightly to ∼0.25.26 The friction coefficient values observed for humid conditions in Figure 2 falls into this range. While water vapor can induce oxidation, which can be a competing mechanism beside molecular lubrication, a previous study showed that tribooxidation effect of water vapor is very small especially when compared to that of liquid water.25 When liquid water is present, continuous oxidation of DLC and removal of oxidation products takes place, causing significant oxidative wear.25 This behavior is not observed in the presence of water vapor. Moreover, the Raman spectra of H-DLC wear tracks tested at 40% and 90% RH conditions (as shown in Figure 4) were very similar to the pristine H-DLC substrate with relatively weak D band shoulders. This indicated that shear-induced oxidations in 2001
DOI: 10.1021/acs.langmuir.5b04207 Langmuir 2016, 32, 1996−2004
Article
Langmuir
experiment a-C contains no hydrogen; therefore, it gives very high friction in dry nitrogen. The friction is also high and noisy in dry air and oxygen environments. When water vapor is present in the surrounding environment, friction drops due to water adsorption.59,60 The friction coefficient value in humid environments (∼0.1) is lower than that observed in the presence of adsorbed monolayer (0.2).42,49 This might be due to the reduced actual contact area created by the presence of wear debris. The optical microscope images of the balls taken after the friction tests in Figure 8 indicate that the SS counter-surface had severe mechanical wear (Figure 9). The hardness of the 440C SS ball is significantly lower than that of a-C (Figure 1). As predicted by the Archard’s relationship, the mechanical wear of the softer SS surface occurred readily by the harder a-C surface. The wear of a-C in dry nitrogen is too small to be measured (Table 1). Even though severe mechanical wear of the SS countersurface occurred, the electrochemical corrosion effects by water vapor could be observed. For example, the wear volume of the SS ball at 90% RH was significantly higher than the wear in all other environments (Table 1). The formation of iron oxide species was observed even at 20% RH tests (Figure 10). The
of metal-against-metal sliding in dry nitrogen, friction is somewhat low due to the presence of oxide layers but increases as soon as the oxide layers wear away.10,27 So what role does the native iron oxide layer on SS and the iron oxide particles formed during sliding in high RH environment play in SS/DLC sliding? It is important to note that in this experiment iron oxide does not significantly affect the friction and wear behavior. First, the effect of metal oxide is suppressed because the steel countersurface gets covered with carbon film transformed during the first few cycles of sliding. This is evident by the Raman spectra collected from the surface of the steel balls (Figure 3). The D and G bands appear in the spectra after all tests. Consequently, the friction is effectively arising from carbon film-covered steel ball sliding on DLC. However, this layer of carbon film at the interface is amorphous and dynamic due to sliding action and affects reactions between the adsorbed vapor molecules and the metal surface. Second, in the presence of adsorbed vapor molecules, the chemistry of the sliding solid does not dominate the behavior of sliding surfaces.9,26,28,53 Rather, it is the chemistry of the adsorbate, which is water in this case, not iron oxide or DLC, that dictates friction and wear.9,26,54 3.2. Oxidation Effects on Wear of a-C/SS Interface. Figure 8 shows friction coefficients as a function of
Figure 8. Friction coefficient of 440 C SS/a-C in different environments.
Figure 10. Raman spectra of the transfer films collected from to the surface of 440C SS balls after sliding on a-C.
reciprocating cycle measured with SS balls sliding on a-C surfaces in different environments. Unlike H-DLC, a-C does not give ultralow friction in dry or inert environment.55−58 The absence of hydrogen makes friction high in inert environments.55 Hydrogen reduces friction because it passivates the dangling bonds that form during sliding, reducing adhesion, friction and wear in inert and dry environments.55−58 In our
friction coefficients in humid environments were low and smooth, consistent with the molecular lubrication effect by adsorbed molecules. There was no measurable wear of a-C slid on with SS ball (Figure 9c) in humid conditions.
Figure 9. (a) Optical microscope images of the 51 200 S balls after 300 cycles of sliding on a-C in different environments. Scale bar is 100 μm. (b) Line profiles across the wear marks. (c) Optical microscope images of the wear tracks on a-C (scale bar is 100 μm) with line profile overlaid. 2002
DOI: 10.1021/acs.langmuir.5b04207 Langmuir 2016, 32, 1996−2004
Langmuir
■
In dry O2 and air, the wear of SS surface was significantly reduced (Figure 9b) and the wear of a-C was increased, leaving 35 and 40 nm deep tracks after 1000 cycles of rubbing with SS balls (Figure 9c). Similar to the H-DLC case, this must be attributed to the oxidation of a-C surface by oxygen and continuous removal of the oxidized a-C surface layer by shearing action of the SS counter-surface. As Figure 10 shows, in Raman analyses, very small amounts of iron oxide signals were detected. However, when the signal intensity was normalized with the wear volume of the SS surface, the iron oxide phases detected with Raman were very small compared to the H-DLC/SS case.
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.5b04207. Friction coefficients in a RH range from 0 to 80%, FESEM and EDS analysis of wear debris, tribo-test with a sapphire ball at 90% RH (PDF)
■
REFERENCES
(1) Joshi, A.; Nimmagadda, R.; Herrington, J. Oxidation kinetics of diamond, graphite, and chemical vapor deposited diamond films by thermal gravimetry. J. Vac. Sci. Technol., A 1990, 8 (3), 2137−2142. (2) Sun, Q.; Alam, M. Relative Oxidation Behavior of Chemical Vapor-Deposited and Type-Iia Natural Diamonds. J. Electrochem. Soc. 1992, 139 (3), 933−936. (3) de Theije, F. K.; van der Laag, N. J.; Plomp, M.; van Enckevort, W. J. P. A surface topographic investigation of {001} diamond surfaces etched in oxygen. Philos. Mag. A 2000, 80 (3), 725−745. (4) Hahn, J. R.; Kang, H.; Lee, S. M.; Lee, Y. H. Mechanistic study of defect-induced oxidation of graphite. J. Phys. Chem. B 1999, 103 (45), 9944−9951. (5) Marino, M. Understanding Run-In Behavior of Diamond-Like Carbon Friction. Langmuir 2011, 27, 12702−12708. (6) Yang, M.; Marino, M. J.; Bojan, V. J.; Eryilmaz, O. L.; Erdemir, A.; Kim, S. H. Quantification of oxygenated species on a diamond-like carbon (DLC) surface. Appl. Surf. Sci. 2011, 257 (17), 7633−7638. (7) Al-Azizi, A. A.; Eryilmaz, O.; Erdemir, A.; Kim, S. H. Surface Structure of Hydrogenated Diamond-like Carbon: Origin of Run-In Behavior Prior to Superlubricious Interfacial Shear. Langmuir 2015, 31 (5), 1711−1721. (8) Wu, X.; Ohana, T.; Nakamura, T.; Tanaka, A. Gaseous Tribochemical Products of Hydrogenated DLC Film and Stainless Steel Pair in Air Detected by Mass Spectrometry. Tribol. Lett. 2015, 57 (2). (9) Alazizi, A.; Barthel, A.; Surdyka, N.; Luo, J.; Kim, S. Vapors in the ambientA complication in tribological studies or an engineering solution of tribological problems? Friction 2015, 3 (2), 85−114. (10) Barthel, A.; Gregory, M.; Kim, S. Humidity Effects on Friction and Wear Between Dissimilar Metals. Tribol. Lett. 2012, 48 (3), 305− 313. (11) Yu, J.; Kim, S. H.; Yu, B.; Qian, L.; Zhou, Z. Role of Tribochemistry in Nanowear of Single-Crystalline Silicon. ACS Appl. Mater. Interfaces 2012, 4 (3), 1585−1593. (12) Erdemir, A.; Fenske, G. R.; Erck, R. A. A Study of the formation and self-lubrication mechanism of boric acid films on boric oxide coatings. Surf. Coat. Technol. 1990, 43−44 (1−3), 588−596. (13) Wiederhorn, S. Influence of Water Vapor on Crack Propagation in Soda-Lime Glass. J. Am. Ceram. Soc. 1967, 50 (8), 407−414. (14) Scharf, T. W.; Singer, I. L. Role of the Transfer Film on the Friction and Wear of Metal Carbide Reinforced Amorphous Carbon Coatings During Run-in. Tribol. Lett. 2009, 36 (1), 43−53. (15) Konicek, A. R.; Grierson, D. S.; Gilbert, P. U. P. A.; Sawyer, W. G.; Sumant, A. V.; Carpick, R. W. Origin of Ultralow Friction and Wear in Ultrananocrystalline Diamond. Phys. Rev. Lett. 2008, 100 (23), 235502. (16) Andersson, J.; Erck, R. A.; Erdemir, A. Frictional behavior of diamondlike carbon films in vacuum and under varying water vapor pressure. Surf. Coat. Technol. 2003, 163−164 (0), 535−540. (17) Tagawa, M.; Ikemura, M.; Nakayama, Y.; Ohmae, N. Effect of water adsorption on microtribological properties of hydrogenated diamond-like carbon films. Tribol. Lett. 2004, 17 (3), 575−580. (18) Wang, D.-Y.; Chang, C.-L.; Ho, W.-Y. Oxidation behavior of diamond-like carbon films. Surf. Coat. Technol. 1999, 120−121, 138− 144. (19) Fontaine, J.; Le Mogne, T.; Loubet, J. L.; Belin, M. Achieving superlow friction with hydrogenated amorphous carbon: some key requirements. Thin Solid Films 2005, 482 (1−2), 99−108. (20) Konicek, A. R.; Grierson, D. S.; Gilbert, P. U. P. A.; Sawyer, W. G.; Sumant, A. V.; Carpick, R. W. Origin of ultralow friction and wear in ultrananocrystalline diamond. Phys. Rev. Lett. 2008, 100 (23), 235502. (21) Andersson, J.; Erck, R. A.; Erdemir, A. Frictional behavior of diamondlike carbon films in vacuum and under varying water vapor pressure. Surf. Coat. Technol. 2003, 163, 535−540. (22) Abou Gharam, A.; Lukitsch, M. J.; Qi, Y.; Alpas, A. T. Role of oxygen and humidity on the tribo-chemical behaviour of non-
4. CONCLUSIONS This paper investigated the effects of different oxidizing environments on two amorphous carbon coatings with different mechanical properties: hydrogenated diamond-like carbon (HDLC) and hydrogen-free amorphous carbon (a-C). Oxygen acts as a mere oxidant for the carbon surfaces, but water adsorbed at low humidity acts as a molecular lubricant. In contrast, thick adsorbed water layer formed at high humidity serves as an electrolyte facilitating electrochemical reactions. The sp2-rich surface layers of amorphous carbon coatings can act like a cathode material when in contact with steel anode. Sliding under these conditions triggers galvanic corrosion of steel surfaces and results in the formation of iron oxide during sliding. This can easily be observed when sliding takes place on H-DLC where no severe mechanical wear occurs. Sliding a SS ball on hydrogen free a-C, conversely, causes mechanical wear of the steel balls due to differences in hardness. However, analysis of the wear pattern and rate of a-C and SS in different environments shows that transition from oxidation to galvanic corrosion also occurs as the environment changes from oxygen to high humidity.
■
Article
AUTHOR INFORMATION
Corresponding Author
*Phone + 1 814 863 4809, e-mail
[email protected] (S.H.K.). Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the National Science Foundation (Grant CMMI-1131128). The authors acknowledge Dr. Osman L. Eryilmaz for preparing H-DLC samples for this study. G.R. and A.E. were supported by the U.S. Department of Energy, Basic Energy Sciences, Office of Energy Efficiency and Renewable Energy, under Contract #DE-AC02-06CH11357. 2003
DOI: 10.1021/acs.langmuir.5b04207 Langmuir 2016, 32, 1996−2004
Article
Langmuir hydrogenated diamond-like carbon coatings. Wear 2011, 271 (9−10), 2157−2163. (23) Yen, B. K. Influence of water vapor and oxygen on the tribology of carbon materials with sp(2) valence configuration. Wear 1996, 192 (1−2), 208−215. (24) Erdemir, A. Synthesis of superlow-friction carbon films from highly hydrogenated methane plasmas. Surf. Coat. Technol. 2000, 133− 134, 448−454. (25) Al-Azizi, A. A.; Eryilmaz, O.; Erdemir, A.; Kim, S. H. Nanotexture for a wear-resistant and near-frictionless diamond-like carbon. Carbon 2014, 73 (0), 403−412. (26) Barthel, A. J.; Kim, S. H. Lubrication by Physisorbed Molecules in Equilibrium with Vapor at Ambient Condition: Effects of Molecular Structure and Substrate Chemistry. Langmuir 2014, 30 (22), 6469− 6478. (27) Barthel, A. J.; Luo, J.; Hwang, K. S.; Lee, J.-Y.; Kim, S. H. Boundary lubrication effect of organic residue left on surface after evaporation of organic cleaning solvent. Wear 2016, 350−351, 21−26. (28) Barnette, A. L.; Ohlhausen, J. A.; Dugger, M. T.; Kim, S. H. Humidity Effects on In Situ Vapor Phase Lubrication with n-Pentanol. Tribol. Lett. 2014, 55 (1), 177−186. (29) Al-Azizi, A. A.; Eryilmaz, O.; Erdemir, A.; Kim, S. H. Effects of Nanoscale Surface Texture and Lubricant Molecular Structure on Boundary Lubrication in Liquid. Langmuir 2013, 29 (44), 13419− 13426. (30) Li, H.; Xu, T.; Wang, C.; Chen, J.; Zhou, H.; Liu, H. Humidity dependence on the friction and wear behavior of diamond-like carbon film in air and nitrogen environments. Diamond Relat. Mater. 2006, 15 (10), 1585−1592. (31) Kim, H. I.; Lince, J. R.; Eryilmaz, O. L.; Erdemir, A. Environmental effects on the friction of hydrogenated DLC films. Tribol. Lett. 2006, 21 (1), 51−56. (32) Chourpa, I.; Douziech-Eyrolles, L.; Ngaboni-Okassa, L.; Fouquenet, J.-F.; Cohen-Jonathan, S.; Souce, M.; Marchais, H.; Dubois, P. Molecular composition of iron oxide nanoparticles, precursors for magnetic drug targeting, as characterized by confocal Raman microspectroscopy. Analyst 2005, 130 (10), 1395−1403. (33) Pol, V. G.; Motiei, M.; Gedanken, A.; Calderon-Moreno, J.; Mastai, Y. Sonochemical Deposition of Air-Stable Iron Nanoparticles on Monodispersed Carbon Spherules. Chem. Mater. 2003, 15 (6), 1378−1384. (34) Oh, S.; Cook, D. C.; Townsend, H. E. Characterization of Iron Oxides Commonly Formed as Corrosion Products on Steel. Hyperfine Interact. 1998, 112 (1−4), 59−66. (35) Yuan, Y.-X.; Ling, L.; Wang, X.-Y.; Wang, M.; Gu, R.-A.; Yao, J.L. Surface enhanced Raman spectroscopic readout on heavy metal ions based on surface self assembly. J. Raman Spectrosc. 2007, 38 (10), 1280−1287. (36) Bauer, M.; Davydovskaya, P.; Janko, M.; Kaliwoda, M.; Petersen, N.; Gilder, S.; Stark, R. W. Raman spectroscopy of laser-induced oxidation of titanomagnetites. J. Raman Spectrosc. 2011, 42 (6), 1413− 1418. (37) Shebanova, O. N.; Lazor, P. Raman study of magnetite (Fe3O4): laser-induced thermal effects and oxidation. J. Raman Spectrosc. 2003, 34 (11), 845−852. (38) de Faria, D. L. A.; Venâncio Silva, S.; de Oliveira, M. T. Raman microspectroscopy of some iron oxides and oxyhydroxides. J. Raman Spectrosc. 1997, 28 (11), 873−878. (39) Varadwaj, K. S. K.; Panigrahi, M. K.; Ghose, J. Effect of capping and particle size on Raman laser-induced degradation of γ-Fe2O3 nanoparticles. J. Solid State Chem. 2004, 177 (11), 4286−4292. (40) Agelet de Saracibar, C.; Chiumenti, M. On the numerical modeling of frictional wear phenomena. Computer Methods in Applied Mechanics and Engineering 1999, 177 (3−4), 401−426. (41) Shpilman, Z.; Gouzman, I.; Grossman, E.; Shen, L.; Minton, T. K.; Paci, J. T.; Schatz, G. C.; Akhvlediani, R.; Hoffman, A. Oxidation and Etching of CVD Diamond by Thermal and Hyperthermal Atomic Oxygen. J. Phys. Chem. C 2010, 114 (44), 18996−19003.
(42) Haasz, A. A.; Chiu, S.; Pierre, J. E.; Gudimenko, Y. I. Thermooxidative erosion of amorphous hydrogenated carbon films. J. Vac. Sci. Technol., A 1996, 14 (1), 184−193. (43) Wang, W.; Jacob, W.; Roth, J. Oxidation and hydrogen isotope exchange in amorphous, deuterated carbon films. J. Nucl. Mater. 1997, 245 (1), 66−71. (44) Wu, X.; Ohana, T.; Nakamura, T.; Tanaka, A. Gaseous Tribochemical Products of Hydrogenated DLC Film and Stainless Steel Pair in Air Detected by Mass Spectrometry. Tribol. Lett. 2015, 57 (1), 1−7. (45) Denape, J.; Lamon, J. Sliding friction of ceramics: Mechanical action of the wear debris. J. Mater. Sci. 1990, 25 (8), 3592−3604. (46) Asteman, H.; Svensson, J. E.; Norell, M.; Johansson, L. G. Influence of Water Vapor and Flow Rate on the High-Temperature Oxidation of 304L; Effect of Chromium Oxide Hydroxide Evaporation. Oxid. Met. 2000, 54 (1−2), 11−26. (47) Ehlers, J.; Young, D. J.; Smaardijk, E. J.; Tyagi, A. K.; Penkalla, H. J.; Singheiser, L.; Quadakkers, W. J. Enhanced oxidation of the 9% Cr steel P91 in water vapour containing environments. Corros. Sci. 2006, 48 (11), 3428−3454. (48) Halvarsson, M.; Tang, J. E.; Asteman, H.; Svensson, J. E.; Johansson, L. G. Microstructural investigation of the breakdown of the protective oxide scale on a 304 steel in the presence of oxygen and water vapour at 600 °C. Corros. Sci. 2006, 48 (8), 2014−2035. (49) Enevoldsen, J. N.; Hansson, C. M.; Hope, B. B. The influence of internal relative humidity on the rate of corrosion of steel embedded in concrete and mortar. Cem. Concr. Res. 1994, 24 (7), 1373−1382. (50) Tuck, C. W.; Odgers, M.; Sachs, K. The oxidation of iron at 950°C in oxygen/water vapour mixtures. Corros. Sci. 1969, 9 (4), 271− 285. (51) Sánchez-López, J. C.; Erdemir, A.; Donnet, C.; Rojas, T. C. Friction-induced structural transformations of diamondlike carbon coatings under various atmospheres. Surf. Coat. Technol. 2003, 163− 164 (0), 444−450. (52) Ma, T.-B.; Wang, L.-F.; Hu, Y.-Z.; Li, X.; Wang, H. A shear localization mechanism for lubricity of amorphous carbon materials. Sci. Rep. 2014, 4, 3662. (53) McFadden, C. F.; Gellman, A. J. Metallic friction: the effect of molecular adsorbates. Surf. Sci. 1998, 409 (2), 171−182. (54) Marino, M. Is Ultra-Low Friction Needed to Prevent Wear of Diamond-Like Carbon (DLC)? An Alcohol Vapor Lubrication Study for Stainless Steel/DLC Interface. Tribol. Lett. 2011, 42, 285−291. (55) Eryilmaz, O. L.; Erdemir, A. On the hydrogen lubrication mechanism(s) of DLC films: An imaging TOF-SIMS study. Surf. Coat. Technol. 2008, 203 (5−7), 750−755. (56) Erdemir, A. Genesis of superlow friction and wear in diamondlike carbon films. Tribol. Int. 2004, 37 (11−12), 1005−1012. (57) Fontaine, J.; Belin, M.; Le Mogne, T.; Grill, A. How to restore superlow friction of DLC: the healing effect of hydrogen gas. Tribol. Int. 2004, 37 (11−12), 869−877. (58) Donnet, C.; Fontaine, J.; Grill, A.; Le Mogne, T. The role of hydrogen on the friction mechanism of diamond-like carbon films. Tribol. Lett. 2001, 9 (3), 137−142. (59) Andersson, J.; Erck, R. A.; Erdemir, A. Friction of diamond-like carbon films in different atmospheres. Wear 2003, 254 (11), 1070− 1075. (60) Konicek, A. R.; Grierson, D. S.; Sumant, A. V.; Friedmann, T. A.; Sullivan, J. P.; Gilbert, P. U. P. A.; Sawyer, W. G.; Carpick, R. W. Influence of surface passivation on the friction and wear behavior of ultrananocrystalline diamond and tetrahedral amorphous carbon thin films. Phys. Rev. B: Condens. Matter Mater. Phys. 2012, 85 (15), 155448.
2004
DOI: 10.1021/acs.langmuir.5b04207 Langmuir 2016, 32, 1996−2004
