Atom Probe Tomography Unveils Formation Mechanisms of Wear

Jun 20, 2017 - ... Sang† , Donovan N. Leonard‡, Jun Qu‡ , and Jonathan D. Poplawsky† ... Here, the tribofilms on a gray cast iron surface form...
3 downloads 0 Views 4MB Size
Research Article www.acsami.org

Atom Probe Tomography Unveils Formation Mechanisms of WearProtective Tribofilms by ZDDP, Ionic Liquid, and Their Combination Wei Guo,*,† Yan Zhou,‡ Xiahan Sang,† Donovan N. Leonard,‡ Jun Qu,*,‡ and Jonathan D. Poplawsky† †

Center for Nanophase Materials Sciences and ‡Materials Science and Technology Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, United States S Supporting Information *

ABSTRACT: The development of advanced lubricant additives has been a critical component in paving the way for increasing energy efficiency and durability for numerous industry applications. However, the formation mechanisms of additive-induced protective tribofilms are not yet fully understood because of the complex chemomechanical interactions at the contact interface and the limited spatial resolution of many characterizing techniques currently used. Here, the tribofilms on a gray cast iron surface formed by three antiwear additives are systematically studied; a phosphonium-phosphate ionic liquid (IL), a zinc dialkyldithiophosphate (ZDDP), and an IL+ZDDP combination. All three additives provide excellent wear protection, with the IL+ZDDP combination exhibiting a synergetic effect, resulting in further reduced friction and wear. Atom probe tomography (APT) and scanning transmission electron microscopy (STEM) imaging and electron energy loss spectroscopy (EELS) were used to interrogate the subnm chemistry and bonding states for each of the tribofilms of interest. The IL tribofilm appeared amorphous and was Fe, P, and O rich. Wear debris particles having an Fe-rich core and an oxide shell were present in this tribofilm and a transitional oxide (Fe2O3)-containing layer was identified at the interface between the tribofilm and the cast iron substrate. The ZDDP+IL tribofilm shared some of the characteristics found in the IL and ZDDP tribofilms. Tribofilm formation mechanisms are proposed on the basis of the observations made at the atomic level. KEYWORDS: tribofilm, atom probe tomography, ionic liquid, ZDDP, lubricant, STEM-EELS

1. INTRODUCTION

A tribofilm refers to any surface layer grown upon rubbing, which would simply be an oxide layer that often is seen on a mental surface in dry sliding or lubricated by a base oil. On the other hand, when discussing a lubricated contact with antiwear additive(s), e.g., in this study, a tribofilm often specifically represents the surface protective layer produced by the antiwear additive and does not include the oxide interlayer if there is one. It has widely been accepted that the wear protection of an antiwear additive primarily relies on forming a sacrificial and/or self-healing tribofilm on the contact surfaces.3,4 Surface characterization has repeatedly confirmed the existence of ZDDP or IL tribofilms on rubbed metal surfaces and revealed their morphology, thickness, nanostructure, chemical composition, and mechanical properties.3,4 However, the mechanisms

Friction and wear reduction during service is important to maintain component mechanical integrity, mobility, and efficiency, where lubrication playsan essential role.1 Additives2 are crucial part of any commercial lubricant used in internal combustion engines, turbines, gears, and other bearing systems, which are specifically designed for protecting the contact surfaces from excessive wear, where Zinc Dialkyldithioposphates (ZDDP)3,4 have been the dominant additive used for engine lubrication since its development in the 1940s. Oilmiscible ionic liquids (ILs)5 are recently developed antiwear additives for next-generation lubricants, and have been shown to have better thermal stability, wear protection,6 and a less adverse impact on exhaust emission catalysts.7 An interesting synergistic effect was later observed between phosphoniumphosphate ILs and ZDDP additives that showed better wear properties than a lubricant containing either additive alone.8 © XXXX American Chemical Society

Received: April 3, 2017 Accepted: June 20, 2017 Published: June 20, 2017 A

DOI: 10.1021/acsami.7b04719 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 1. (a) Schematic showing the reciprocating sliding test lubricated by a base oil containing an IL, ZDDP, or a combination of both. (b) Tribofilm sample sectioned by FIB and then lifted-out using a micromanipulator in the FIB instrument. (c) SEM image showing the mounted sample on the Si micropost. (d) Final sharpened APT needle. (e) Schematic of the APT experiment.

Other techniques, such as SIMS and AFM, can acquire a larger region of interest (ROI), but they lack a sufficient lateral resolution. Three-dimensional (3D) information can be obtained from several techniques, such as XPS, AES, and SIMS, in the form of composition depth-profiling combined with elemental mapping; however, such profiles are significantly affected by shadowing effects, which limits the lateral resolution with spatial ambiguity at the nm-scale. In contrast, atom probe tomography (APT) has no such ambiguities and has been applied into a wide range of materials science research. APT is a 3D compositional mapping technique based on field evaporation of individual atoms and/or compounds as positive ions from the tip of a cryogenically cooled needle-shaped specimen. The chemical identity of each ion is determined from time-offlight mass spectrometry, whereas the x,y positions are determined from the position where the ion strikes the detector. A 3D atom-by-atom reconstruction can be created from the known shape of the specimen before evaporation, the xy positional data from the detector, and the order of evaporation events, which is used to calculate the z-position of each ion. The 3D reconstruction can thus reveal the atomic level composition and impurity distributions with ten parts-permillion sensitivity and subnm spatial resolution.22 APT has been used for studying surface related phenomena such as metal oxidation;23−26 however, little work has been done in the field of tribology, where there is a great need in understanding the formation and growth mechanisms of tribofilms induced by tribochemical reactions between the metal contact surface and lubricant additives.27 In this work, three different tribofilms formed on a cast iron worn surfaces produced by a lubricant containing IL, ZDDP, and IL and ZDDP additives, respectively, were characterized by correlative APT and STEM-EELS. The APT and STEM-EELS

governing the tribofilm formation are not well understood. The dominant hypothesis for ZDDP tribofilm formation is that ZDDP molecules decompose and self-react to produce a network of cross-linked organometallic phosphates that deposit onto the contact area.9,10 Another mechanism for ZDDP tribofilm growth was proposed by Martin et al.,11,12 whereby iron oxide wear particles are embedded in the ZDDP film by mechanical mixing and then chemically digested (react with zinc phosphates) to form iron phosphates. The mechanisms proposed for ZDDPs are not applicable to the oil-miscible ILs because the metal-free ILs cannot self-supply metal cations. There is little consensus on the IL tribofilm formation process despite several literature reports.6,8,13 The ideal characterization of tribofilms would involve the identification of the roles played by the metal surface elements, oxygen, and lubricant active agents, and their interactions. Chemical characterization of IL and ZDDP tribofilms in the literature have primarily been performed using energy dispersive spectroscopy (EDS) in a scanning electron microscope (SEM),14 X-ray photoemission spectroscopy (XPS),8,15 X-ray absorption near edge structure spectroscopy (XANES),16,17 Auger electron spectroscopy (AES),18 secondary ion mass spectroscopy (SIMS),19 atom force microscopy (AFM),10,20 and transmission electron microscopy (TEM).21 Because spatial resolution is a critical factor for understanding the formation process of tribofilms, TEM or scanning TEM (STEM) have the advantage of extracting structural information with resolutions less than 2 nm; but STEM-based EDS and electron energy loss spectroscopy (EELS) typically have limited sensitivities (∼0.1%) and generally lack three dimension information unless challenging electron tomography experiments are performed, which are time-consuming and do not necessarily work for beam-sensitive samples such as tribofilms. B

DOI: 10.1021/acsami.7b04719 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

fabrication of multiple APT needles from one wedge lift-out.29 APT analysis was performed using a local electrode atom probe LEAP 4000X HR equipped with a 10 pico-second 355 nm UV laser from CAMECA Instruments (Figure 1e). The vacuum pressure in the analysis chamber was 10 nm within the oxide-rich interlayer reflects the alternations of Fe and O and is not due to APT aberrations. Such composition change in the oxide interlayer is also confirmed by the 5 at. % P proximity analysis shown in Figure 4e, where O increases from 4.23 ± 0.74 to 29.69 ± 0.61 at. % within a width of 12 nm. The Fe content gradually decreases to D

DOI: 10.1021/acsami.7b04719 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 3. STEM-EDS-EELS characterization of the tribofilm on the cast iron flat lubricated by the base oil containing IL+ZDDP. (a) Cross-sectional BF-STEM image and (b) EDS elemental maps. (c) EEL spectra of the cast iron substrate, iron-tribofilm interface, and tribofilm. (d) High-resolution HAADF-STEM image of the cast iron-tribofilm interface showing the structure. (e) FFT of a typical amorphous region of the tribofilm; (f) schematic of a unit cell of magnetite observed at the cast iron−tribofilm interface.

3.2.2. ZDDP Tribofilm. Figure 6 shows the tribofilm grown on the cast iron surface lubricated by the base oil with ZDDP. The S+ peaks overlap with O2+ peaks in the mass spectrum, and therefore O/S ionic species were used to consider the contribution of both elements in Figure 6b. Distinct features are evident in the IL tribofilm, as shown in Figures 4 and 5: first, the phosphate related ionic species exist as both FePOx and ZnPOx (ZnPO2+,2+, ZnPO3+,2+, ZnPO4+,2+), indicating selfpolymerization of ZDDP during tribofilm formation; second, the iron oxide ionic species together with phosphorus-related ionic species, i.e., Fe2O, FeOx, FePOx, and ZnPOx, are distributed homogeneously within the tribofilm. No gradient oxide interlayer was found on this particular APT sample, which differs from the IL sample APT results. Although this leads to the conclusion that the antioxidant properties of ZDDP can eliminate the oxide layer, the nonexistence of an oxide layer in the APT results shown in Figure 7 may be simply because the oxide interlayer is noncontinuous, and the APT sample was extracted from a location without an oxide interlayer. The proximity histograms (using an iron isosurface of 65 at. %) revealed the chemical composition at the tribofilm−iron interface (Figure 7). The average composition within the tribofilm was 39.33 ± 0.29 at. % Fe, 40.10 ± 0.29 at. % O, 8.23 ± 0.16 at. % P, and 3.98 ± 0.12 at. % Zn. The tribofilm contains ten times the amount of Fe as much as Zn, demonstrating significant involvement of the cast iron substrate and/or wear debris in the ZDDP tribofilm formation. 3.2.3. IL+ZDDP Tribofilm. Multiple attempts were made to delineate the mechanism of tribofilm growth in the base oil containing IL+ZDDP. Two APT samples selected from two

Table 2. Fe-L2,3, O-K Edge, Peak Energy (eV) and Fe-L2,3 Edge White-Line Ratio of the Tribofilm on the Cast Iron Flat Lubricated by the Base Oil Containing IL+ZDDP O-K peak (eV) cast iron substrate iron−tribofilm interface tribofilm

Fe-L3 peak (eV)

Fe-L2 peak (eV)

Fe-L32 ratio

N/A 537.8

709.4 709.4

722.6 722.6

1.75 2.18

537.8

709.7/710.9

723.2

2.50

47.31 ± 0.64 at. % as the O increases, while P increases rapidly to 13.87 ± 0.46 at. %, defining the beginning of the bulk tribofilm amorphous layer. The concentrations of these three elements remain nearly constant within the bulk amorphous tribofilm, indicating a relatively uniform chemical composition. Another surface region of the IL tribofilm was analyzed and reconstructed by APT (Figure 5). As shown in Figure 5a, a layer of iron wear debris of thickness greater than 10 nm is present above the tribofilm, with an oxygen-rich shell observed on top. The proximity histogram analysis of a selected region of the interface (marked as arrow 1 in Figure 5a) indicates that the wear debris has a composition close to that of the cast iron (Figure 5c). Also, the chemical composition of the tribofilm at this localized region is different from the first sample analyzed, which possessed more oxygen (∼44.4 at. %) than Fe (∼38.7 at. %), and a smaller P concentration. The proximity histograms analysis of the second selected interface (marked as arrow 2 in Figure 5a) confirms the existence of an oxygen-rich interlayer, which is a general characteristic for the growth of an IL tribofilm (Figure 5d). E

DOI: 10.1021/acsami.7b04719 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 4. (a) 3D APT reconstruction of the tribofilm produced by IL showing a graded-layer structure. (b) Spatial distributions of Fe, Fe2O, O, FePOx, FeOx, and Ni ionic species. (c) Graded-layer tribofilm structure revealed by APT. (d) Proximity histogram analysis for a 80 at. % Fe isoconcentration surface. (e) Proximity histogram analysis for a 5 at. % P isoconcentration surface.

Although the results indicatea graded oxide structure along the tribofilm depth direction (normal to the cast iron surface), the thickness of the phosphate-rich amorphous layer is limited to ∼9 nm, which may not fully represent the overall landscape of the IL+ZDDP tribofilm formation. To more fully interrogate the uniformity of the IL+ZDDP tribofilm, we sampled another surface region for APT analysis (sample #2) and the APT reconstruction is shown in Figure 9. A thick oxide interlayer was not observed in sample #2, but rather exhibited a greater than 50 nm thick phosphate-rich layer. Zn, O, ZnPOx, FePOx, and other iron-oxide related ionic species are detected within the tribofilm, as shown by the atom maps in Figure 9a. Figure 10 shows the composition profile across the thickness of the IL +ZDDP tribofilm for the sample shown in Figure 8, which shows a relatively uniform concentration of zinc (12.85 ± 0.42 at. %), phosphorus (7.25 ± 0.33 at. %), iron (27.30 ± 0.56 at. %), and oxygen (48.30 ± 0.63 at. %). Note that the content of Zn in the IL+ZDDP tribofilm (12.85 at. %) is significantly higher than that found in the ZDDP tribofilm (3.98 at. % as shown in Figure 6), which is consistent with previous X-ray photoelectron (XPS) compositional analyses for these two systems.8 Even though the Zn content in the oil containing 0.52%IL+0.4%ZDDP is only a half of that for the oil containing 0.8%ZDDP, the IL+ZDDP enables the participation of more Zn atoms toward forming the tribofilm due to a synergistic effect between IL and ZDDP. The discrepancy in the Zn contents between the two IL+ZDDP tribofilm samples (Figures 8 and 9) is believed to be due to the spatial heterogeneity

separate surface locations of the IL+ZDDP tribofilm are presented (sample #1 and #2). Figure 8 shows the atom maps of the main ionic species in sample #1, which exhibits similar characteristics observed for both theZDDP and the IL tribofilms. A significant amount of iron oxide-related ionic species, i.e., Fe2O, Fe2O3, and FeO, appear adjacent to the cast iron substrate, which is similar to that of the IL tribofilm (Figure 4). The iron-oxide related ionic species tend to evaporate as Fe2O during atom probe experiments. Beyond the Fe2O-rich zone, other iron oxide ionic species, such as Fe2O3 and FeO are identified within the oxide interlayer. Also, the phosphate related ions, i.e., FePOx and ZnPOx, coexist inside the IL+ZDDP tribofilm, similar to the observations of the ZDDP tribofilm. To distinguish any sublayers within the IL+ZDDP tribofilm, we plotted 80 at. % Fe and 60 at. % Fe isoconcentration surfaces in Figure 8c. A 10 × 10 × 120 nm3 analysis volume (white rectangle shown in Figure 8c) was positioned at the center of the 3D volume to include the different layers identified at the surface. Figure 8e shows the 1D composition profile acquired along the film growth direction (yellow arrow in Figure 8c). In the oxide-rich interlayer, a gradual increase in Fe and decrease in O are observed, with limited Zn and P, compared with the overall composition of the tribofilm. The gradual increase of Zn and P concentrations from the oxide interlayer to the tribofilm, as shown in Figure 8e, suggests that the additive molecules (IL+ZDDP) likely reacted with the oxide layer surface during the tribofilm formation. F

DOI: 10.1021/acsami.7b04719 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 5. (a) 3D APT reconstruction of the tribofilm produced by IL that contains an iron/iron-oxide wear debris particle. (b) Spatial distributions of Fe, Fe2O, O, FePOx, FeOx, and Ni ionic species. (c) Proximity histogram analysis for a 60 at. % Fe isoconcentration surface (marked by arrow 1 in Figure 5a). (d) Proximity histogram analysis for a 75 at. % Fe isoconcentration surface (marked by arrow 2 in panel a).

Figure 6. (a) 3D APT reconstruction of the tribofilm produced by ZDDP. (b) Spatial distributions of Fe, Fe2O, FeOx, FePOx, Zn, ZnPOx, O/S, and Ni ionic species are shown.

4. DISCUSSION

because APT samples are so small. The XPS analysis data

4.1. Iron Oxidation States in the Tribofilms. The APT results presented show that the oxide−iron interfaces exhibited different Fe ionic species, such as Fe2O, FeO, and Fe2O3, which appear as a graded layer structure along the tribofilm growth direction normal to the cast iron surface. An open question is whether these ionic species yielded during the atom probe field evaporation process can be directly related to a change in

reported previously were based on signals from a much larger area, with a 400 μm diameter, and the averaged XPS composition results clearly showed a higher Zn content in the IL+ZDDP tribofilm than that in the ZDDP tribofilm.8 G

DOI: 10.1021/acsami.7b04719 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 7. APT proxigram analysis of selected 65 at. % Fe isoconcentration surfaces at the iron/tribofilm interface. (a) APT reconstructed volume with 65 at. % Fe isosurface. (b) Proximity histograms showing the compositional profile across the cast iron−tribofilm interface.

Figure 8. (a) 3D APT reconstruction of the IL+ZDDP produced tribofilm (sample #1). (b) Spatial distributions of Fe, Fe2O, FeOx, FePOx, Zn, O, ZnPOx, and Ni ionic species are shown. (c) 3D APT volumes decorated by 60 at. % Fe and 80 at. % Fe isoconcentration surfaces. The white rectangle designates the region used for 1D composition analysis. (d) Magnified image showing the ionic maps within the rectangle. (e) 1D compositional profile along the direction of yellow arrow marked in panel a.

oxidation state change for the iron atoms. In general, the measured Fe/O ratio should be directly related to the number of oxygen atoms that are associated with Fe atoms, which would be a qualitative indicator of the oxidation state of the Fe, i.e., more oxygen atoms around Fe atoms indicates a higher Fe

oxidation state if all the Fe and O are bound. Dong et al. studied the oxidation states of corroded Zr alloys and found that the oxygen evaporated as O+, O2+, ZrO2+, ZrO3+, ZrO2+, and ZrO3+ ions, with instances of Zr2O23+ and Zr2O33+. They attributed these ionic species and their relative concentrations H

DOI: 10.1021/acsami.7b04719 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 9. (a) 3D APT reconstruction of the IL+ZDDP-produced tribofilm (sample #2). (b) Spatial distributions of Fe, Fe2O, FeOx, FePOx, Zn, O, ZnPOx, and Ni ionic species are shown. (c) Atom maps with 40 at. % Fe (pink) and 75 at. % Fe (green) isoconcentration surfaces, the dashed rectangular region indicated on volume (black dotted lines) is selected for proxigram analysis. (d) Proximity histogram analysis with respect to the 40 at. % Fe isosurface. (e) Proximity histogram analysis with respect to 75 at. % Fe isosurface.

together with smaller amounts of Fe2O and Fe2O3 ions. When the O/Fe ratio decreases to below 0.8, the FeO/Fe2O and FeO/Fe2O3 ratios increase significantly and FeO becomes the dominant evaporated oxide species. When the O/Fe ratio is reduced to below 0.4 at the interface, the iron oxide ionic species evaporate primarily as Fe2O and metallic Fe ionic species. Thus, the evaporated ionic species depends on the relative ratio of the constituting elements. A large-area STEM structural scan was conducted at the cast iron−tribofilm interface to identify the specific structure of the oxides formed at the interface. By comparing measured lattice spacing from the FFT with known iron oxide crystallographic structures, it is concluded that Fe3O4 Magnetite and Fe2O3 Hematite structures are the primary oxide products (Figures 2a, 2c, and 3d); no other oxide structures, i.e., FeO and Fe2O, were found at the interface. This experimental evidence further demonstrates that the observed Fe−O ionic species obtained from APT do not necessarily reflect the oxidation state, but rather support a composition-dependent evaporation preference during atom probe field evaporation. Therefore, the main chemical reactions at the surface of the as-cast iron are believed to proceed as

Figure 10. 1D composition profiles for O and Fe atoms, together with 1D concentration profile of iron oxide-related ionic species. The O/Fe ratio is shown for clarity.

as indirect evidence of a graded Zr-oxide structure in the form of a ZrO2 front, a thin layer of equiatomic ZrO and a saturated solid-solution of ZrOsat.34 If this postulation holds true, the oxide layer structure identified in the present work will possess Fe2O, FeO, and Fe2O3 suboxides in the oxide rich interlayer. The concentration profiles for the different ionic species acquired along the IL tribofilm growth direction (normal to cast iron substrate) are plotted in Figure 10. The O/Fe ratio is also plotted for clarity. When the atomic ratio, O/Fe, is close to 1, the evaporated iron oxide is in the form of FeO ions,

3Fe + 2O2 → Fe3O4 (magnetite)

2Fe + 1.5O2 → Fe2O3(hematite)

ZDDP is known to act as both an antiwear and an antioxidant additive.35 Only Fe2O3 is present at the interface between the I

DOI: 10.1021/acsami.7b04719 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 11. Schematics of tribofilm structures produced by IL, ZDDP, and IL+ZDDP.

the interface to grow the tribofilm during the dynamic wear process. However, a significant amount of iron was observed in the tribofilms formed by IL, ZDDP, and IL+ZDDP, indicating the important role of iron in the formation of each of the tribofilms. ZDDP is capable of exchanging zinc with ions that originates from the iron oxides of the digested debris particles and the cast iron substrate, a process driven by the chemical hardness principle.11,12,36 Without such a driving force, a large energy input is required to transform iron oxide to iron phosphate in the sole presence of phosphate ions, which is the case for IL. There are two iron sources for tribofilm formation in the present study: the cast iron substrate below the tribofilm and iron-based wear debris in the lubricant above the tribofilm. It would be difficult for iron atoms from the cast iron substrate to diffuse to the top of the tribofilm. Moreover, Fe−Fe bonds are strong and difficult to break and generate free iron atoms. For the second iron source, the iron-based wear debris in the lubricant exists in different chemical states and physical sizes. Some relatively large (from a few to tens of nanometers) particles of wear debris were observed within the tribofilm formed by IL lubrication (e.g., Figure 5a). On the other hand, chemical fluctuations were observed at different locations of the tribofilm (see Figures 4 and 5), indicating the role of mechanical mixing on the formation of the tribofilm, even though a relatively constant composition across the depth of the tribofilms was observed. Fe in the form of small clusters could react with oxygen and the lubricant additives, and subsequently deposit on the surface of the tribofilm, and thereby contribute to the thickness increase in the tribofilm. Although the APT-STEM-based tribofilm characterization presented here is interesting and informative, we understand that in situ characterization of the tribofilm growth is even more desired. However, there currently is no available technique providing in situ high-resolution nanostructure examination or chemical composition analysis of the tribofilm. Carpick’s group [10] developed a new approach using AFM single-point contact to grow a ZDDP tribofilm and provide in situ observation of the tribofilm morphology and thickness. Future work will be conducted to use a similar AFM-based approach to gain further understanding of the tribofilm growth for the IL and IL+ZDDP combination.

cast iron and the tribofilm on the sample lubricated by IL, but Fe3O4 is found on the samples lubricated by either ZDDP or ZDDP + IL, possibly because the antioxidant functionality of ZDDP hindered the iron oxidation process. 4.2. Formation of Tribofilm. The processes for the tribofilms formed by the IL, ZDDP, and IL+ZDDP are illustrated in Figure 11. All three tribofilms are formed as a result of the additive and oxygen, triggered by the thermomechanical stresses in rubbing, reacting with the iron surface to grow bottom-up as well as reacting with the wear debris to deposit the reaction products from top-down, which are mechanically mixed and at the same time competing with the material removal (wear) in rubbing against the counterface. Although an oxide interlayer was observed on the APT samples of the IL and IL+ZDDP tribofilms but not on APT sample of the ZDDP tribofilm, we believe that such an oxide interlayer is inevitable because of the good availability and high reactivity of oxygen. The iron oxide interlayer is likely in a discrete manner due to the heterogeneous nature of the surface condition upon rubbing and wear. As a result, whether such an oxide interlayer is observed or not by APT or STEM largely depends on the APT or FIB sampling location. Because the thickness of the native oxide film on an Fe-based alloy23 (a few hundred nanometers) is much less than the wear scar depth on the cast iron in this work (5−15 μm), the oxide interlayer underneath the tribofilm definitely is not the residual native oxide film. It must be formed during rubbing through the following two possible processes. (i) Fresh surface oxidation: on a freshly exposed cast iron surface as a result a scratch (wear), oxygen transporting through the iron lattice interstitial sites creates iron oxides and the oxidation layer grows inward. The organophosphate molecules, from the IL or ZDDP, may compete with oxygen to react with the freshly exposed iron surface; however, it is almost impossible for the much larger organophosphate molecules to transport through the iron lattice interstitial sites, at the same rate or to the same distance as far as the smaller oxygen ions. This is supported by the APT analysis in Figure 4, where the oxide interlayer contains a negligible amount of phosphorus. (ii) Permeation through the tribofilm. The oxide interlayer could also form by oxygen permeation through the tribofilm that then reacts with the iron substrate. A major difference between IL and ZDDP is that IL does not contain metal ions and therefore needs metal ions supplied to J

DOI: 10.1021/acsami.7b04719 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

5. CONCLUSIONS In this work, APT and STEM were used to investigate the tribofilms formed by three different lubricant antiwear additives, a phosphonium-phosphate IL, ZDDP, and a combination of both IL+ZDDP. The main findings are summarized as follows: (1) All three tribofilms exhibit a mixture of amorphous− crystalline hybrid structures; however, the structure and compositions vary depending on the additive chemistry. The ionic state of iron ions gradually changes from +2 to +3 along the direction of tribofilm growth perpendicular to the cast iron substrate. (2) For the IL tribofilm, wear debris particles with an ironrich core and an iron oxide shell were observed by APT. The chemical composition of the tribofilm measured by APT varies from one location to another, demonstrating the role of wear debris particles in contributing to the inhomogeneity of the tribofilm. (3) The growth of the ZDDP tribofilm is beyond simple selfreaction, because a noticeable amount of Fe ions were detected within the tribofilm. (4) The IL+ZDDP tribofilm showed similar features observed in the IL tribofilm with an oxide interlayer adjacent to the tribofilm layer. The well-developed tribofilm contains more Zn and less S than the ZDDP tribofilm, indicating a synergistic effect between IL and ZDDP in the growth of tribofilm. (5) Tribofilms are formed as a result of the additive and oxygen, triggered by thermomechanical stresses in rubbing, reacting with the iron surface to grow bottomup as well as reacting with the wear debris to deposit the reaction products from top-down, which are mechanically mixed and at the same time competing with material removal (wear).



Energy (DOE). Atom probe tomography and electron microscopy characterization was performed at Oak Ridge National Laboratory’s Center for Nanophase Materials Sciences (CNMS), which is a U.S. DOE Office of Science User Facility.



(1) Carnes, K. The Ten Greatest Events in Tribology History. Tribol. Lubr. Technol. 2005, 61, 38−47. (2) Rudnick, L. R. Lubricant Additives Chemistry and Applications; CRC Press: Boca Raton, FL, 2003. (3) Spikes, H. The History and Mechanisms of Zddp. Tribol. Lett. 2004, 17, 469−489. (4) Nicholls, M. A.; Do, T.; Norton, P. R.; Kasrai, M.; Bancroft, G. M. Review of the Lubrication of Metallic Surfaces by Zinc DialkylDithlophosphates. Tribol. Int. 2005, 38, 15−39. (5) Barnhill, W. C.; Qu, J.; Luo, H.; Meyer, H. M., 3rd; Ma, C.; Chi, M.; Papke, B. L. Phosphonium-Organophosphate Ionic Liquids as Lubricant Additives: Effects of Cation Structure on Physicochemical and Tribological Characteristics. ACS Appl. Mater. Interfaces 2014, 6, 22585−22593. (6) Qu, J.; Luo, H.; Chi, M.; Ma, C.; Blau, P. J.; Dai, S.; Viola, M. B. Comparison of an Oil-Miscible Ionic Liquid and Zddp as a Lubricant Anti-Wear Additive. Tribol. Int. 2014, 71, 88−97. (7) Xie, C.; Toops, T. J.; Lance, M. J.; Qu, J.; Viola, M. B.; Lewis, S. A.; Leonard, D. N.; Hagaman, E. W. Impact of Lubricant Additives on the Physicochemical Properties and Activity of Three-Way Catalysts. Catalysts 2016, 6, 54. (8) Qu, J.; Barnhill, W. C.; Luo, H.; Meyer, H. M., 3rd; Leonard, D. N.; Landauer, A. K.; Kheireddin, B.; Gao, H.; Papke, B. L.; Dai, S. Synergistic Effects between Phosphonium-Alkylphosphate Ionic Liquids and Zinc Dialkyldithiophosphate (ZDDP) as Lubricant Additives. Adv. Mater. 2015, 27, 4767−4774. (9) Mosey, N. J.; Muser, M. H.; Woo, T. K. Molecular Mechanisms for the Functionality of Lubricant Additives. Science 2005, 307, 1612− 1615. (10) Gosvami, N. N.; Bares, J. A.; Mangolini, F.; Konicek, A. R.; Yablon, D. G.; Carpick, R. W. Mechanisms of Antiwear Tribofilm Growth Revealed in Situ by Single-Asperity Sliding Contacts. Science 2015, 348, 102−106. (11) Martin, J. M. Antiwear Mechanisms of Zinc Dithiophosphate: A Chemical Hardness Approach. Tribol. Lett. 1999, 6, 1−8. (12) Minfray, C.; Le Mogne, T.; Martin, J. M.; Onodera, T.; Nara, S.; Takahashi, S.; Tsuboi, H.; Koyama, M.; Endou, A.; Takaba, H.; Kubo, M.; Del Carpio, C. A.; Miyamoto, A. Experimental and Molecular Dynamics Simulations of Tribochemical Reactions with ZDDP: Zinc Phosphate-Iron Oxide Reaction. Tribol. Trans. 2008, 51, 589−601. (13) Sharma, V.; Doerr, N.; Erdemir, A.; Aswath, P. B. Interaction of Phosphonium Ionic Liquids with Borate Esters at Tribological Interfaces. RSC Adv. 2016, 6, 53148−53161. (14) Cai, Z.-b.; Meyer, H. M.; Ma, C.; Chi, M.; Luo, H.; Qu, J. Comparison of the Tribological Behavior of Steel−Steel and Si3N4− Steel Contacts in Lubricants with ZDDP or Ionic Liquid. Wear 2014, 319, 172−183. (15) Zhou, Y.; Dyck, J.; Graham, T. W.; Luo, H. M.; Leonard, D. N.; Qu, J. Ionic Liquids Composed of Phosphonium Cations and Organophosphate, Carboxylate, and Sulfonate Anions as Lubricant Antiwear Additives. Langmuir 2014, 30, 13301−13311. (16) Pereira, G.; Munoz-Paniagua, D.; Lachenwitzer, A.; Kasrai, M.; Norton, P. R.; Capehart, T. W.; Perry, T. A.; Cheng, Y.-T. A Variable Temperature Mechanical Analysis of Zddp-Derived Antiwear Films Formed on 52100 Steel. Wear 2007, 262, 461−470. (17) Sharma, V.; Gabler, C.; Doerr, N.; Aswath, P. B. Mechanism of Tribofilm Formation with P and S Containing Ionic Liquids. Tribol. Int. 2015, 92, 353−364. (18) Minfray, C.; Martin, J. M.; Esnouf, C.; Le Mogne, T.; Kersting, R.; Hagenhoff, B. A Multi-Technique Approach of Tribofilm Characterisation. Thin Solid Films 2004, 447−448, 272−277.

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.7b04719. Friction coefficient traces of the three lubricants over the 1000 m sliding; 2D profile of crossing the wear scars on the cast iron flats; mass spectrum and the main associated ranges for the IL+ZDDP samples displayed in Figure 9 (PDF)



REFERENCES

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (W.G). *E-mail: [email protected] (J.Q.). ORCID

Wei Guo: 0000-0002-9534-1902 Xiahan Sang: 0000-0002-2861-6814 Jun Qu: 0000-0001-9466-3179 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We thank William C. Barnhill for performing the tribological tests and Dr. Huimin Luo for synthesizing the ionic liquid. Research supported by Vehicle Technologies Office, Office of Energy Efficiency and Renewable Energy, US Department of K

DOI: 10.1021/acsami.7b04719 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces (19) Ito, K.; Martin, J.-M.; Minfray, C.; Kato, K. Low-Friction Tribofilm Formed by the Reaction of Zddp on Iron Oxide. Tribol. Int. 2006, 39, 1538−1544. (20) Li, H.; Cooper, P. K.; Somers, A. E.; Rutland, M. W.; Howlett, P. C.; Forsyth, M.; Atkin, R. Ionic Liquid Adsorption and Nanotribology at the Silica-Oil Interface: Hundred-Fold Dilution in Oil Lubricates as Effectively as the Pure Ionic Liquid. J. Phys. Chem. Lett. 2014, 5, 4095−4099. (21) Zhou, Y.; Leonard, D. N.; Meyer, H. M.; Luo, H.; Qu, J. Does the Use of Diamond-Like Carbon Coating and Organophosphate Lubricant Additive Together Cause Excessive Tribochemical Material Removal? Adv. Mater. Interfaces 2015, 2, 1500213. (22) Guo, W.; Sneed, B. T.; Zhou, L.; Tang, W.; Kramer, M. J.; Cullen, D. A.; Poplawsky, J. D. Correlative Energy-Dispersive X-Ray Spectroscopic Tomography and Atom Probe Tomography of the Phase Separation in an Alnico 8 Alloy. Microsc. Microanal. 2016, 22, 1251−1260. (23) Lozano-Perez, S.; Saxey, D. W.; Yamada, T.; Terachi, T. AtomProbe Tomography Characterization of the Oxidation of Stainless Steel. Scr. Mater. 2010, 62, 855−858. (24) Parvizi, R.; Hughes, A. E.; Tan, M. Y.; Marceau, R. K. W.; Forsyth, M.; Cizek, P.; Glenn, A. M. Probing Corrosion Initiation at Interfacial Nanostructures of AA2024-T3. Corros. Sci. 2017, 116, 98− 109. (25) Kruska, K.; Lozano-Perez, S.; Saxey, D. W.; Terachi, T.; Yamada, T.; Smith, G. D. W. Nanoscale Characterisation of Grain Boundary Oxidation in Cold-Worked Stainless Steels. Corros. Sci. 2012, 63, 225− 233. (26) Langelier, B.; Persaud, S. Y.; Newman, R. C.; Botton, G. A. An Atom Probe Tomography Study of Internal Oxidation Processes in Alloy 600. Acta Mater. 2016, 109, 55−68. (27) Kim, Y.-J.; Baik, S.-I.; Bertolucci-Coelho, L.; Mazzaferro, L.; Ramirez, G.; Erdermir, A.; Seidman, D. N. Atom-Probe Tomography of Tribological Boundary Films Resulting from Boron-Based Oil Additives. Scr. Mater. 2016, 111, 64−67. (28) Barnhill, W. C.; Qu, J.; Luo, H. M.; Meyer, H. M.; Ma, C.; Chi, M. F.; Papke, B. L. Phosphonium-Organophosphate Ionic Liquids as Lubricant Additives: Effects of Cation Structure on Physicochemical and Tribological Characteristics. ACS Appl. Mater. Interfaces 2014, 6, 22585−22593. (29) Thompson, K.; Lawrence, D.; Larson, D. J.; Olson, J. D.; Kelly, T. F.; Gorman, B. In Situ Site-Specific Specimen Preparation for Atom Probe Tomography. Ultramicroscopy 2007, 107, 131−139. (30) Guo, W.; Garfinkel, D. A.; Tucker, J. D.; Haley, D.; Young, G. A.; Poplawsky, J. D. An Atom Probe Perspective on Phase Separation and Precipitation in Duplex Stainless Steels. Nanotechnology 2016, 27, 254004. (31) Varela, M.; Oxley, M. P.; Luo, W.; Tao, J.; Watanabe, M.; Lupini, A. R.; Pantelides, S. T.; Pennycook, S. J. Atomic-Resolution Imaging of Oxidation States in Manganites. Phys. Rev. B: Condens. Matter Mater. Phys. 2009, 79, 254004. (32) Schmid, H. K.; Mader, W. Oxidation States of Mn and Fe in Various Compound Oxide Systems. Micron 2006, 37, 426−432. (33) Tan, H.; Verbeeck, J.; Abakumov, A.; Van Tendeloo, G. Oxidation State and Chemical Shift Investigation in Transition Metal Oxides by Eels. Ultramicroscopy 2012, 116, 24−33. (34) Dong, Y.; Motta, A. T.; Marquis, E. A. Atom Probe Tomography Study of Alloying Element Distributions in Zr Alloys and Their Oxides. J. Nucl. Mater. 2013, 442, 270−281. (35) Yamaguchi, E. S.; Onopchenko, A.; Francisco, M. M.; Chan, C. Y. The Relative Oxidation Inhibition Performance of Some Neutral and Basic Zinc Dithiophosphate Salts. Tribol. Trans. 1999, 42, 895− 901. (36) Martin, J. M.; Onodera, T.; Minfray, C.; Dassenoy, F.; Miyamoto, A. The Origin of Anti-Wear Chemistry of Zddp. Faraday Discuss. 2012, 156, 311−323.

L

DOI: 10.1021/acsami.7b04719 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX