Interaction of phosphonium ionic liquids with borate esters at tribological interfaces

Abstract

Synergistic interaction between ionic liquids (ILs) and soluble borate (SB) esters was examined in the context of tribofilm formation for antiwear applications. ILs composed of phosphonium cation and phosphate anion (P_DEHP) and dithiophosphate anion (P_DEPDT), respectively, were mixed with borate esters in group I base oil at 1000 P ppm and 200 B ppm and their tribological properties were evaluated using a cylinder on reciprocating flat and ball on rotating disc contact tribometers. Friction and wear results were compared with those of a reference oil containing zinc dialkyl dithiophosphate (ZDDP). Results indicate that the addition of SB to ILs results in a shortened incubation time for tribofilm formation and significantly better wear protection in comparison to ZDDP. The chemical compositions of the interfacial tribofilms examined with X-ray absorption near edge structure (XANES) spectroscopy suggest that interaction between IL + SB results in incorporation of boron as boron oxides and boron phosphate in the bulk of the tribofilms while the surface is largely composed of short chain iron polyphosphates. On the other hand, tribofilms derived from only ILs are composed of short chain iron polyphosphate and tribofilms derived from SB1 and SB2 exhibit trigonal B chemistry.

---

1. Introduction

The discipline of tribology primarily deals with mechanical systems involving interactions between surfaces that are in contact or/and in relative motion to each other. Thus tribology is defined as the science and technology of interacting surfaces in relative motion and deals with force transference between surfaces moving relative to each other.¹ This force is known as friction which is a resistance to the relative motion between the interacting surfaces that results in loss of energy as well as loss of materials (i.e. wear) to the mechanical system. The losses due to friction and wear can be overcome by the application of a lubrication system which is applied between the interacting surfaces to perform primary tasks i.e. reducing friction resistance and minimizing material deterioration from wear by promoting a protective layer or film between the contact surfaces. These protective films are commonly known as tribofilms. Tribofilms are thin patchy films, with typical thickness ranges between 100 nm and 400 nm,^2–13 basically a byproduct of the decomposition of lubricant additives which cover the surfaces of the moving parts in order to avoid the direct contact of interacting surfaces to ease the relative motion as well as minimize material removal due to wear.^6,14–19 Lubricant additives primarily an anti-wear additives (e.g. ZDDP etc.) play a significant role in the formation of tribofilms which either get physically or chemically absorbed to the surfaces of moving parts or decompose to form new ions to activate the substrate surfaces and subsequently react to form chemically active species which adhere onto the wear prone surfaces to enhance protection.^11,15,20,21

However, increasingly stringent government regulation on emissions (EPA Emissions Standard Reference Guide²²) and latest CAFE (Corporate Average Fuel Economy) standards requiring an average fuel economy of 54.5 mpg (combined cars and trucks)²³ by 2025 impose significant challenges to the automotive and lubricant industries calling for the development and implementation of lower viscosity ILSAC GF-5&6 (International Lubricants Standardization and Approval Committee for Gasoline Fueled vehicles) and API-CJ4&5 (American Petroleum Institute Compression Ignition) oils which further limit the amount of SAPS (sulfated ash, phosphorous and sulfur) and deposits in engines.²⁴ Development of additives that result in lower ash content, volatility and anti-wear property plays a crucial role in being able to reach these standards. In addition, new engine technologies such as turbo-charged gasoline direct ignition (TGDI) and complex engine designs result in much higher combustion temperature and severe oxidative conditions, which limit the use of traditional lubricant additives. To overcome these challenges, a number of new approaches have been used in the development of new additives. Recent studies^25–42 have shown that ionic liquids (ILs) that are ashless additives, offer great potential as high performance environmentally friendly additives for the next generation of lubricants.

Several IL structures have been examined for their tribological properties, which include imidazolium, ammonium and phosphonium as cationic moieties and fluorinated anions such as tetrafluoroborate (BF₄), hexafluorophosphate (PF₆), bis(trifluoromethylsulfonyl)amide (NTF₂).^29,43–50 These ILs demonstrated promising lubricating properties. However, toxicity and corrosion due to fluorinated structures^29,51 have restricted their applications. Moreover, the high polarity of these compounds limits their solubility in commonly used non-polar base oils. Recent progress in IL lubrication has led to the development of a new class of phosphonium alkylphosphate ILs that are oil miscible by designing 3D structures with relatively longer alkyl chains.²⁸ Somers et al.²⁶ found that phosphonium ILs with long alkyl chains on both anions and cations show better miscibility in non-polar base oil. Qu et al.³⁴ studied the tribological properties of phosphonium phosphate IL (PP-IL) and reported comparable or even superior antiwear and anti-scuffing properties than zinc dialkyl dithiophosphate (ZDDP) at 0.1 wt% phosphorus. Phosphonium alkylphosphate ILs have also been shown to synergistically interact with ZDDP⁵² and result in a 30% reduction of friction and >70% reduction in wear. In this context, a synergistic interaction of organo borate esters with phosphonium class IL's has been explored in this study. Organo borate ester chemistry was chosen to improve the solubility of boron based additive chemistry to the hydrocarbon based base oil. Previously, solid lubrication using boron oxide/boric acid either as solid lubricants or coated on mating surfaces have been studied.^53–59 Erdemir et al.⁵⁷ in 1990 proposed a lubrication mechanism of super lubricity of boron oxide films coated on metallic and ceramic surfaces. Using Raman and scanning electron microscopy (SEM) surface characterization techniques, they reported that boron oxide coating in an open air tribological experiment under steady state sliding action produces boric acid film which has a layered triclinic structure. They hypothesized that super lubrication is achieved once these layers aligned themselves to the direction of relative motion to provide relative ease to slide over one another.

In this study, beneficial synergistic interaction of phosphonium ILs with soluble borate esters (SB) has been examined with the intent to determine the mechanism, chemistry and structure of tribofilms and its impact on wear and frictional properties. Tribological outcomes indicate a unique synergism between the boron and IL additives that together resulted in better outcomes compared to ZDDP. The coordination chemistry of tribofilms formed at the interface of the contacting surfaces were studied using X-ray absorption near edge structure (XANES) spectroscopy to elucidate the underlying mechanism of interfacial tribofilm formation responsible for improved tribological outcomes.

2. Experimental details

2.1 Description of additive chemistry and thermogravimetric analysis

Table 1 details the chemistry of the additives used. ZDDP used in this study, is a secondary alcohol derived ZDDP with approximately 70% neutral and 30% basic characteristics, details of which are available in Parekh et al.⁶⁰ Phosphonium ionic liquids, tetrabutylphosphonium O,O-diethyl-phosphoro-dithionate (P_DEPDT) and trihexyltetradecylphosphonium bis(2-ethylhexyl)phosphate (P_DEHP) and borate ester 2-methoxy-4,4,6-trimethyl-1,3,2-dioxaborinane (SB1) were provided from AC2T research GmbH, Austria. Trimethoxyboroxine (SB2) was provided from Argonne National Laboratory. Base oil (BO) blends were prepared in a group 1 base oil (mixture of 60 wt% solvent neutral 150 W and 40 wt% bright stock 90 W).

Table 1 Description of antiwear additive chemistry

Coded name	Chemical name	Chemical structure
ZDDP	Zinc dialkyl dithiophosphate
P_DEPDT	Tetrabutylphosphonium O,O-diethyl-phosphoro-dithionate
P_DEHP	Trihexyltetradecylphosphonium bis(2-ethylhexyl)phosphate
SB1	2-Methoxy-4,4,6-trimethyl-1,3,2-dioxaborinane
SB2	Trimethoxyboroxine

---

Thermogravimetric analysis (TGA) of ZDDP and ionic liquids was performed using Shimadzu TGA-51 thermogravimetric analyzer. TGA was performed in nitrogen (N₂) atmosphere at a heating rate of 10 °C min⁻¹ from room temperature to 600 °C with a N₂ flow rate of 20 mL min⁻¹. Each test was repeated twice for the consistency of the data.

2.2 Tribological experiments

Oil samples were formulated by blending ZDDP, ILs and borate esters individually or as mixtures of IL and SB in base oil. All blends were prepared by keeping the phosphorus (P) and boron (B) concentration at 1000 ppm and 200 ppm respectively as applicable. Tribological experiments were conducted at Argonne National Laboratory in ambient air (relative humidity ≈50%). The first group of tribological experiments employed a cylinder-on-reciprocating flat contact test machine⁶¹ (test setup 1) to determine the performance characteristics of the additives. The best performing blends were then further examined using a ball on rotating disc contact test configuration using a Nanovea tribometer (test setup 2). Table 2 provides schematic of the test configurations and highlights the test parameters. Test setup 1 was also equipped with electric contact resistance (ECR) measurement setup. ECR data was collected in situ during the test in order to gain insight into the incubation time for film formation at the tribological contacts. Specimens were cleaned before each test using Stoddard solution followed by isopropanol and acetone to completely remove any oil and dust present on the surfaces. The rubbed surfaces after the tests were cleaned with n-heptane and isopropanol and then saved by submerging them in sulfur free base oil.

Table 2 Schematic of the test configurations and detail of tribometrical parameters for test setup 1 and test setup 2

Test configuration	Cylinder on reciprocating flat contact	Ball on rotating disc contact
Base oil	Group I mineral oil (60 wt% SN 150 W + 40 wt% BS 90 W)	Group I mineral oil (60 wt% SN 150 W + 40 wt% BS 90 W)
Viscosity	10.1 mm^2 s^−1 at 100 °C	10.1 mm^2 s^−1 at 100 °C
Treat rate	1000 ppm phosphorus and 200 ppm boron	1000 ppm phosphorus and 200 ppm boron
Applied load	82 N	2 N
Initial Hertzian contact pressure	500 MPa	510 MPa
Temperature	100 °C	100 °C
Speed	0.06 m s^−1, 5 Hz	0.06 m s^−1
Stroke length	6 mm	—
Cylinder/Ball	Φ 4 mm × 6 mm, 52 100 steel (60–61 HRc)	Φ 12.7 mm, 52 100 alloy steel (58–60 HRc)
Flat/Disc	12 mm × 12 mm × 8 mm, 52 100 alloy steel (60–61 HRc)	Φ 20 mm, 52 100 alloy steel (60–61 HRc)
Duration	1 hour	1 hour

---

2.3 Surface characterization

Wear scar width generated on cylinder (from test setup 1) and wear scar diameter (WSD) generated on ball (from test setup 2) after rubbing were measured using optical images obtained from an Olympus brand metallographic microscope. Wear width was measured at nine locations on each cylinder (from test setup 1) and each oil sample was tested twice. Thus, wear width for each oil sample is an average of 18 measurements. WSD on the ball (from test setup 2) is an average of two measurements from two repeated tests for each oil sample. The surface topography of the wear track was examined by obtaining 3-dimensional (3D) images of the rubbed surfaces using Hysitron Triboscope™ in scanning probe microscopy (SPM) imaging mode.

Chemical nature of the tribofilms formed in situ during tribological tests was characterized using XANES spectroscopy. XANES spectra were obtained at The Canadian Light Source synchrotron facility at Saskatoon Canada. The phosphorus L-edge (P L-edge) and boron K edge (B K-edge) spectra were collected at VLS-PGM (variable line spacing plane grating monochromator) beam station that operates at the energy range of 5.5–250 eV with a photon resolution of more than 10 000 E/ΔE. All the spectra were collected using a 100 μm × 100 μm photon beam spot size. In this study, P L-edge and B K-edge spectra are analyzed in both total electron yield (TEY) and fluorescent yield (FY) mode. TEY spectra offer more surface sensitive information whereas FY spectra give information from the bulk of the sample. Exemplarily, Si L-edge the sampling depth of TEY and FY mode is about 5 nm and 70 nm respectively.⁶²

3. Results and discussion

3.1 Coefficient of friction and wear volume

Fig. 1 shows the coefficient of friction (CoF) and wear volume obtained from test setup 1 for all the blends. CoF curve acquired from each oil blend is plotted in Fig. 1(a) and the average coefficient of friction for the last 15 minutes of the test is shown in the bar chart embedded in Fig. 1(a). Friction curve for group 1 BO without additives (labeled as A) exhibits large variance and high CoF value for the duration of the test in comparison with oil blends containing ZDDP (B), ILs and binary mixtures of IL + SB. Oil blends containing only SB1 (C) and SB2 (D) show similar frictional behaviours as that of group 1 BO characterized by large variance and spikes in the friction curve with high CoF value. This result indicates that boron concentration (i.e. 200 ppm) was too small for any improvement in friction when using only boron based additives in base oil (i.e. SB1 and SB2). However, the 200 ppm B concentration was chosen to meet the current industrial standard of boron in engine oils.⁶³ Comparing the oil with ZDDP alone with blends containing ILs and binary mixture of IL + SB, ZDDP exhibits a stable friction curve with consistent CoF value while P_DEHP (H) and P_DEHP + SB2 (J) exhibit higher CoF at the beginning of the test and the CoF decreases as the test progresses and the test ends with a CoF smaller than that of ZDDP. In the case of P_DEHP + SB1 (I), large fluctuations were observed at the beginning and higher CoF value than P_DEHP and P_DEHP + SB2. On the other hand, oils with P_DEPDT (E) exhibit a significantly lower CoF for the first 500 seconds of the test and then friction increases and remains consistent for the remainder of the test. Similar behavior was seen for P_DEPDT + SB1 (F) and P_DEPDT + SB2 (G), where CoF is lower in the initial stage of the test and then friction increases and reaches a plateau, however, the duration of the lower friction was reduced from 500 seconds for P_DEPDT to 100 seconds with P_DEPDT + SB2 and almost negligible for P_DEPDT + SB1. These results indicate that addition of borate esters to phosphonium ILs (P_DEHP and P_DEPDT) does not result in significant improvement in their stable friction behavior at the end of test but plays a significant role in determining the frictional performance at the early stages of the test. The lowest average CoF at the end of the test was achieved with a binary mixture of P_DEHP + SB2 as shown in the bar chart in Fig. 1(a).

Fig. 1 Coefficient of friction (a) and wear volume (b) obtained from test setup 1 for group 1 BO, ZDDP, IL, SB and IL + SB in group 1 BO at 0.1 wt% P and 0.02 wt% B.

Wear volume calculated from the cylinders (from test setup 1) are plotted in Fig. 1(b). Group 1 BO (A) shows the largest wear on the cylinder while the addition of additive chemistry results in significant reduction in wear. Comparing the wear volume of ZDDP (B) with ashless blends containing only SB or ILs and blends P_DEPDT + SB, both SB1 (C) and SB2 (D) as well as P_DEHP (H) and P_DEPDT (E) including its blends P_DEPDT + SB1 (F) and P_DEPDT + SB2 (G) show poorer wear protection than ZDDP. It is noticeable that all samples containing P_DEPDT (E to G) have similar wear indicating that SBs have little effect on overall wear in these combinations. On the other hand, P_DEHP + SB1 (I) and P_DEHP + SB2 (J) show comparable or better antiwear performance than ZDDP at equal phosphorus levels (i.e. 1000 ppm). The error bar on the bar chart denotes the consistency of the test results.

Fig. 2 illustrates the coefficient of friction and WSD obtained from test setup 2 for selected oils evaluated in test setup 1. An average CoF for the different oil blends obtained are plotted in Fig. 2(a). None of the blends exhibited a significant improvement in friction compared to ZDDP with the lowest and smoothest friction recorded for the oil with P_DEHP + SB1. The WSD measured are plotted in Fig. 2(b) and their respective optical images of wear scar on ball are also presented. Error bar on the friction and WSD bar diagram denotes the consistency and reproducibility of the data. Tribological tests from group 1 BO results in the highest WSD as expected. Addition of antiwear additives to the base oil significantly improves the wear protection. Moreover, in comparison to ZDDP, both P_DEHP and binary mixture of P_DEHP + SB1 exhibited smaller WSD. The new ashless additive chemistry of P_DEHP and P_DEHP + SB1 showed remarkable improvement in wear protection by 19% and 36% respectively compared to ZDDP at 0.1 wt% P treat rate.

Fig. 2 Coefficient of friction (a) and wear scar diameter (b) obtained from test setup 2 for group1 BO, ZDDP, P_DEHP and P_DEHP + SB1 in group 1 BO at 0.1 wt% P and 0.02 wt% B.

3.2 Incubation time for the tribofilm formation and thermogravimetric analysis (TGA) of additive chemistry

The glassy phosphate films formed on the tribological surfaces are known to have very low electric conductivity.⁶¹ Hence, an ECR measurement can be used as a way of determining the effectiveness of tribofilm formation by applying an electric circuit between counter surfaces. As the protective tribofilm starts to form, the resistance increases and voltage drop builds up between the counter surfaces. Fig. 3 summarizes the ECR data (i.e. voltage drop) collected from test setup 1 for the selected additive chemistries and a TGA plot acquired for ZDDP and ILs (P_DEHP and P_DEPDT). TGA graph is positioned on the bottom right corner in Fig. 3 and remaining graphs are the ECR data (bottom) plotted together with CoF (top) as a function of test duration. The voltage applied between the counter surfaces is 100 mV. The voltage drop readings at the interface are shown as black dots in ECR plot obtained from each test and values of each voltage drop reading were integrated and plotted as the red curve superimposed on the voltage drop data. The integrated values of voltage drop have been normalized to the maximum value obtained in the case of P_DEHP + SB2 to the remaining plots to illustrate the overall effectiveness of tribofilm formation over time. Voltage drop close to 100 mV correspond to more net coverage of tribofilm while values closer to 0 mV correspond to no tribofilm coverage. In addition, the slope of the red curve indicates the effectiveness of tribofilm formation with larger slopes having progressively larger tribofilm formation. From Fig. 3, the ECR plot of the test with ZDDP indicates that voltage drop reaches maximum in the very early stages of the test, indicating very short incubation time for the tribofilms formation using ZDDP. The ECR plot exhibits low voltage drop for about 1500 seconds in the case of P_DEHP and 600 seconds for P_DEPDT at the start of the test, suggesting a prolonged incubation time for the tribofilm formation. On the other hand, the addition of SB to P_DEHP reduced the incubation time significantly and the tribofilm begins to form as early as the start of the tests. ECR data indicates that the binary mixtures of P_DEHP and SB result in a quick and stable film formation that exhibit better wear protection as seen in Fig. 1(b) and 2(b). These results indicate a synergistic interaction of P_DEHP with oil soluble boron additives.

Fig. 3 ECR plot of oil blends from test setup 1 and thermogravimetric analysis (bottom right) of additives of ZDDP and ILs (P_DEHP and P_DEPDT).

The friction response of each additive blend can also be correlated with the tribofilm formation. In the case of ZDDP, tribofilms form at an early stage of the test and its friction response remains relatively consistent, however, the ECR data indicate partial removal of the tribofilm periodically followed by new film formation. In the case of P_DEPDT, a clear correlation of friction and tribofilm formation can be seen. P_DEPDT shows poor film formation at earlier stages of the test till about 600 seconds and then voltages drop starts to gain to an intermediate level (60–70 mV) for the test duration from 600–1200 seconds and finally the voltage drops reaches maximum. Correlation between the frictional response and the observed ECR data of P_DEPDT indicates that the lower friction probably results with the formation of easily removable tribofilms that are not very protective for the surfaces. However, once stable tribofilms are formed after the initial incubation period, the ECR and frictional data indicate more stability. Analogous results were found with P_DEHP. One of the striking benefits of the addition of SB additives to DEHP can be deduced from the ECR graphs where we can see that in the presence of SB2, DEHP exhibits the best tribofilm formation.

When analyzing the thermal stability of these additive chemistries along with the time scale formation of the tribofilms during the test, a correlation can be concluded between the thermal stability of the additive chemistry and their incubation time for tribofilms formation. In Fig. 3, TGA graph of the additives shows that ZDDP starts to decompose at a lower temperature (about 200 °C) compared to both P_DEHP (about 325 °C) and P_DEPDT (about 300 °C). The ECR plots show that ZDDP results in the shortest incubation time for the tribofilms formation (about at 100 seconds) than P_DEPDT (about 600 seconds) and P_DEHP (about 1200 seconds). Hence, it can be hypothesized that higher thermal stability of the additive chemistry results in delayed decomposition of the additives to form protective tribofilms by reaction with the rubbing surfaces. This is significantly mitigated by the addition of SB into the mix where the incubation time is significantly reduced.

3.3 Topography of the wear surfaces

Topography of the wear track generated under ZDDP, IL and IL + SB lubrication was studied by 3D images obtained from SPM imaging technique. Fig. 4 illustrates the SPM images of the wear track for the flat samples (test setup 1). The arrow next to the images delineates the sliding direction. SPM images are produced for 60 μm × 60 μm area of the wear track while keeping the z-axis same i.e. ±200 nm. SPM image of ZDDP lubricated wear track demonstrates island type features that are typical of ZDDP tribofilms.^61,64 These islands of patchy tribofilms are unevenly distributed on the wear track. Deep scratches are also observed on the wear track, which are oriented in the sliding direction. SEM image covering a larger region also illustrates non-uniform patches of films along with the deep scratches on ZDDP wear track. In an earlier study, Taylor et al.⁶⁵ suggested that uneven distribution of ZDDP antiwear films and the roughness oriented in the direction of rubbing results in high friction under ZDDP lubrication. Recently, Gosvami et al.⁶⁶ studied the mechanism of ZDDP antiwear film growth under single-asperity sliding contact on Fe-coated Si wafer using AFM. They reported that patchy tribofilm formation is a result of heterogeneous nucleation dependent on the atomic-scale surface roughness. The surface roughness leads to varying contact stress, which means that the energy barrier for tribochemical reaction is overcome easily where contact stress is higher.

Fig. 4 3D SPM images of rubbed surfaces generated under ZDDP, P_DEPDT, P_DEHP, P_DEHP + SB1 and P_DEHP + SB2 lubrication.

3D image of P_DEPDT reveals similar topography as seen for ZDDP consisting of non-uniform distribution of patchy tribofilms owing to similar anion chemistry both based on dithiophosphate. This is also supported by the tribological behavior of P_DEPDT in the last half hour of the test which is consistent with ZDDP: P_DEPDT shows CoF values very close to ZDDP (Fig. 2(a)) and ECR data (Fig. 3) indicate that P_DEPDT films start to form and remain intact. Combining these results proposes a correlation between friction and tribofilm topography as P_DEPDT tribofilms exhibit topographical features similar to ZDDP once formed show similar frictional behavior. In contrast, the first half of the P_DEPDT test reveals much smaller CoF (in Fig. 1(a)). Since, there was no protective film on the surface for that duration as learnt from ECR data (Fig. 3), the lower CoF arises from large polishing wear (abrasive type) leading to increase in the contact area (from line contact to area contact), which consequently reduces maximum contact pressure.

Wear track of P_DEHP also consists of many island type features along with few scratches. The features on the wear tracks are much smaller than ZDDP and P_DEPDT and the surface looks smoother and more uniform. It can be speculated that a polishing type wear had occurred resulting in a smooth surface during the prolonged incubation time of P_DEHP films (as evident in Fig. 3). The polishing mechanism produces uniform asperities on the surface, which consequently results in a more uniform nucleation and growth of the patchy tribofilms.

The wear surfaces derived from the lubrication of P_DEHP + SB1 and P_DEHP + SB2 exhibit a significantly different topography. The wear tracks are much smoother and no island type features are observed. The SPM images reveal few minor scratches on the surface, which are attributed to be present from the sample preparation process. It is evident that with the presence of SB1 or SB2 in addition to P_DEHP, there is minimal wear that occurs and the larger island like patchy tribofilms are not formed. Moreover, the ECR data indicate that larger areas are covered with tribofilms and this is matched by the SPM images that indicate that the whole region of interest is covered by a uniform tribofilm.

3.4 XANES characterization of tribofilms

The antiwear properties of additives are derived from their ability to form protective film on the rubbing surfaces under high contact temperature and shear forces in a tribological system. The knowledge of the nature of chemical film formed and the local coordination of the elements present on the surface and in the bulk of the film is critical to advance the mechanistic understanding of surface protection with tribofilm formation. For this purpose, XANES technique has been extensively used to examine the chemical nature of the tribofilms formed using various antiwear additives.^{2,14,41,61,67–79}

3.4.1. Phosphorous L_2,3-edge. P L-edge TEY and FY spectra of tribofilms are plotted and compared with model compounds (Zn₃(PO₄)₂, FePO₄ and BPO₄) and IL standards (P_DEPDT std and P_DEHP std) in Fig. 5(a) and (b). P L-edge spectra of the tribofilms were acquired from three different locations on the wear scar to examine the variance in the chemical composition on the wear track. The pre-edge shoulders and primary absorption edge from model compounds Zn₃(PO₄)₂ and FePO₄ are identified as a′, b′, c′ and a, b, c respectively. Peaks labeled as a′, a and b′, b correspond to transition from spin–orbit split of phosphorus 2p electrons (2p_3/2 and 2p_1/2 levels, which are usually referred to as the L₃-and L₂-edges, respectively) into the first unoccupied 3s-like antibonding state⁸⁰ in Zn₃(PO₄)₂ and FePO₄. Peaks labeled as c′ and c are attributed to the transition of the phosphorus 2p electrons to the 3p-like antibonding state⁸⁰ in Zn₃(PO₄)₂ and FePO₄ respectively. Peak d in Fig. 5(a) and (b) is commonly present in both Zn₃(PO₄)₂ and FePO₄ as well as in tribofilms spectra, and is a shape resonance peak owing to 2p to 3d transition.⁸¹ Comparing the P L-edge TEY and FY spectra of tribofilms with model compounds based on fingerprint match analysis, peak d is commonly found in all tribofilms spectra as well as in model compounds and IL standards. This indicates that all the tribofilms derived from either ZDDP or ILs and IL + SB form phosphate antiwear films. ZDDP tribofilms showed main absorption edge at peak position c′ which corresponds to the main absorption edge of Zn₃(PO₄)₂, which suggests that ZDDP tribofilms are primarily compounds of phosphate of zinc. However, since Fe cations were also available from the substrate during the test, possibility of FePO₄ formation cannot be ruled out completely. Both P L-edge TEY and FY spectra of the tribofilms derived from P_DEPDT, P_DEHP and mixture of P_DEHP + SB exhibit main absorption edge at peak position c which aligns with the main absorption edge of FePO₄ that strongly suggests that tribofilms in these cases are composed of phosphates of iron. In addition, both P_DEHP + SB1 and P_DEHP + SB2 tribofilms exhibited the main absorption edge at relatively higher energy (at peak c) than the main absorption edge from BPO₄. This indicates no significant contribution of BPO4 in the P L-edge spectra of the P_DEHP + SB1 and P_DEHP + SB2 tribofilms. However, in the next section using the B K-edge spectra, presence of absorption edge peak for BPO₄ model compound is found in the P_DEHP + SB1 and P_DEHP + SB2 tribofilms. Here, it can be speculated that, since availability of iron is more redundant at the actual contact than boron, phosphate chemistry is dominated by FePO₄ formation than BPO₄. All tribofilms exhibited similar peak intensity for spectra collected at multiple locations on each wear scar except ZDDP tribofilms. In the case of P L-edge FY spectra of ZDDP tribofilms, one spectra exhibited significantly higher counts than the other two, indicating a chemical variance in the film.

Fig. 5 (a). P L-edge TEY spectra of tribofilms derived from ZDDP, ILs and IL + SB lubrication (black lines) with reference compounds (green lines). (b). P L-edge FY spectra of tribofilms derived from ZDDP, ILs and IL + SB lubrication (black lines) with reference compounds (green lines).

P L-edge spectra can be further characterized to determine the chain length of polyphosphates formed on the wear track. The degree of polymerization of the phosphate films is determined by analyzing the peak intensity ratio of pre-edge peak a or a′ to main absorption edge c or c′, which is called as a/c ratio. The a/c ratio below 0.3 represents short-chain polyphosphates and a/c above 0.6 represents long-chain polyphosphate.^68,71,82 The a/c ratio of P L-edge TEY and FY spectra of tribofilms were calculated and shown in Fig. 5(a) and (b). The a/c ratio of TEY and FY spectra for all the tribofilms is less than 0.3, which suggest that the surface of the all the tribofilms are composed of short-chain polyphosphates of zinc (in the case of ZDDP) and iron (in the case of both ILs and binary mixture of IL + SB). The a/c ratio of P L-edge FY spectra which represents more bulk chemistry of tribofilms also showed formation of short-chain polyphosphates for P_DEPDT, P_DEHP, P_DEHP + SB1 and P_DEHP + SB2 while ZDDP films exhibited region where medium chain phosphates are found. Tribofilms derived from ILs and IL + SB showed similar a/c ratio irrespective of the location where spectra were acquired. On the other hand, in the case of P L-edge FY spectra of ZDDP films, a significant variance in the a/c ratio was observed depending on location of measurement. This again confirms the inhomogeneity in the chemical make-up of ZDDP tribofilms. IL P_DEPDT exhibited shorter chain polyphosphate formation in the bulk of the film in comparison to the surface as the a/c ratio becomes smaller from TEY to FY spectra. On the other hand, P_DEHP showed higher degree of polymerization of phosphate films in the bulk than the surface. Similar results were observed in the previous study,^25,41 where the sulfur containing ionic liquid (IL-TP) exhibited longer chain polyphosphates formation at the surface than the bulk and only phosphorus containing ionic liquid (IL-P) yielded an increase in phosphate chain length from the surface towards the bulk of the tribofilms. It can be hypothesized that sulfur from sulfur containing IL i.e. P_DEPDT dominates the reaction with the metal substrate first to form sulfates/sulfides of iron thus hinders the polymerization of phosphate films near the film–substrate interface, while in the case of P_DEHP, more iron cations are available near the interface than the surface of the films, higher polymerization occurs near the interface than surface. However, in the case of ZDDP, since the availability of zinc cation is similar at the surface of the tribofilm and at the interface of the tribofilm and the metal surface, similar a/c ratio observed in TEY and FY spectra except one location where a high a/c ratio observed.

3.4.2. Boron K-edge. This study is focused on the synergistic interaction of borate esters with phosphonium ionic liquids. Thus, boron chemistry in the tribofilms have also been investigated using XANES spectroscopy. Fig. 6(a) is a plot of B K-edge TEY (bottom) and FY (top) spectra of tribofilms. For the purpose of the fingerprint match type analysis of tribofilms spectra, B K-edge spectra of several model compounds were also acquired and presented in Fig. 6(b). As reported by Kasrai et al.,⁸³ B K-edge TEY probes at near the surface (∼6 nm) and FY obtain information from the bulk (∼100 nm). Kasrai et al.⁸³ also reported that TEY spectra get influenced by the surface modification as TEY measurement only probes up to 6 nm from the surface. In order to present a clean spectral feature of the model compounds, only B K-edge FY spectra (bulk) are shown in Fig. 6(b). From Fig. 6(b), three peaks are identified for B K-edge spectra of model compounds and are labelled as peak a (194.0 eV), b (198.4 eV) and c (202.9 eV, broad). Peaks a and c are attributed to the boron in trigonal coordination and peak b is assigned to tetrahedral coordination.⁸³ Peaks a and c are commonly present in B₂O₃ and H₃BO₃. BPO₄ exhibits all three peaks, where peak b originates from tetrahedral B and is a dominant peak, since the compound in its pure form has boron in tetrahedral coordination. Relatively weak peaks a and c in BPO₄ spectra are believed to originate from surface modification.⁸³ The main absorption edge for FeB is located at 191.9 eV. FeB spectra also exhibit peak a, which could originate due to surface modification. The peak a for h-BN is shifted to 192.4 eV, which is about 1.6 eV lower energy than that of B₂O₃. The borate esters SB1 standard and SB2 standard spectra also exhibits peak a owing to the trigonal coordination of boron, in addition SB2 standard spectra exhibits a post edge peak at 194.7 eV. The model compounds spectra can now be used to characterize the tribofilms spectra in Fig. 6(a). The B K-edge TEY and FY spectra of SB1 and SB2 tribofilms look very similar and exhibits a dominant peak a indicating that boron is primarily present as trigonal B. In addition, the width of peak a in SB1 tribofilms TEY and FY spectra (FWHM is ∼1.2 eV) further suggests a greater likelihood of the formation of B₂O₃/H₃BO₃ (FWHM 0.9 eV/1 eV) on the surfaces since the width of peak a for SB1 standard spectra is too sharp (FWHM ∼0.5 eV). In the case of SB2 tribofilms TEY and FY spectra, a post edge peak after peak a from SB2 standard is missing, which also eliminates the possibility of residual SB2 standard chemistry within the SB2 tribofilms. A mere presence of peak b suggests that a partial transformation had occurred to produce tetrahedral boron species in SB1 and SB2 tribofilms under the thermo-mechanical action. Zhang et al.⁸⁴ also reported transformation of trigonal coordination to tetrahedral coordination in boron in the tribofilms upon rubbing. Interestingly, the blending of ILs P_DEHP and P_DEPDT with the borate esters SB1 and SB2 show a noticeable effect in the spectra. B K-edge FY spectra of P_DEHP + SB1 tribofilms exhibit dominant presence of peak a, which is again attributed to the formation of B₂O₃/H₃BO₃. In addition, a resolved peak b can be seen that also suggests the formation of BPO₄ in the tribofilms due to the interaction of borate esters with P_DEHP. Furthermore, P_DEHP + SB1 FY spectrum exhibits a weak peak at peak a′ (195.4 eV). In an earlier study, Zhang et al.⁸⁴ reported that peak a′ originates from the p_2s from phosphate structure. Hence, the presence of peak a′ in P_DEHP + SB1 can be assigned to the FePO₄. P L-edge (in earlier Section) also showed FePO₄ in P_DEHP + SB1 tribofilms. On the other hand, P_DEHP + SB1 TEY spectra exhibit a broad peak at peak a′, suggesting that surface of P_DEHP + SB1 tribofilms is dominated by the FePO_4. However, a pre-edge shoulder at peak a is seen, indicating that to some extent boron is also present on the surface in the form of B₂O₃/H₃BO₃. This indicates that the chemistry of P_DEHP + SB1 tribofilm changes from the surface towards the bulk as boron was mainly found in the bulk (FY spectra) as both trigonal B (B₂O₃/H₃BO₃) and tetrahedral B (BPO₄) and a small proportion of boron found at the surface (TEY spectra) as trigonal B only. Similar results were observed when SB1 added to P_DEPDT. Here again, boron is mostly present in the bulk (FY) both as trigonal B (B₂O₃/H₃BO₃) and tetrahedral B (BPO₄) and in less extent on the surface (TEY) as trigonal B (B₂O₃/H₃BO₃). In cases of P_DEHP + SB1, the chemistry of the tribofilms at the surface is largely dominated by the phosphorus in the form of FePO₄. B K-edge TEY and FY spectra of P_DEHP + SB2 tribofilms exhibit peak a and peak b where the intensity of peak a is stronger than peak b indicating that both in bulk and at surface, boron is primarily present as trigonal B (B₂O₃/H₃BO₃) while the presence of tetrahedral B (BPO₄) can be also noticed clearly. Peak a′ is not observed in FY spectra whereas merely noticed in TEY spectra attributing to FePO₄ at the surface.

Fig. 6 B K-edge TEY and FY of tribofilms derived from ZDDP, ILs and IL + SB lubrication (a) and B K-edge FY of reference compounds (b).

B K-edge FY spectra of P_DEPDT + SB2 again indicates that boron is present in the bulk as trigonal B (B₂O₃/H₃BO₃) and tetrahedral B (BPO₄) however, in this case, intensity of peak b is stronger than peak a. This result suggests that SB2 largely interacts with P_DEPDT to form BPO₄ film. B K-edge TEY spectra of P_DEPDT + SB2 also exhibit similar chemistry as in FY, in addition, peak a′ is also observed. This indicates that the boron chemistry in P_DEPDT + SB2 is similar in bulk and at surface however, FePO₄ exhibits a noticeable presence at the surface (TEY). In a summary, all tribofilms derived from borate ester indicate that boron is primarily present as trigonal B. The addition of phosphonium ILs (P_DEHP and P_DEPDT) results in addition BPO₄ films in the tribofilms. In the case of SB1 mixture (i.e. P_DEHP + SB1 and P_DEPDT + SB1) boron chemistry changes significantly from surface towards the bulk, borates and boron phosphates are mainly found in the bulk (FY) while the surface (TEY) of the tribofilms are dominated by phosphate of iron (peak a′). On the other hand, mixtures of SB2 (i.e. P_DEHP + SB2 and P_DEPDT + SB2) exhibit borates and boron phosphates in bulk as well as at the surface. However, the peak a′ becomes noticeable in TEY spectra attributing to the FePO₄ at surface. In a similar study using XANES, Zhang⁸⁴ reported that when borated dispersant are mixed with the ZDDP, composition of boron changes in the surface and the bulk as the film grows. From B K-edge spectra, we can also eliminate the possibility of the formation of FeB in the P_DEHP + SB and P_DEPDT + SB tribofilms, as the B K-edge spectra of the tribofilms do not show features matching with the B K-edge spectra of FeB model compound.

These results suggests borate chemistry mostly present in the bulk whereas iron phosphate film dominates boron chemistry at the surface of the tribofilms as the spectra for P L-edge and B K-edge are acquired at the same time at the same location of tribofilms and from similar depths as the energy range of acquisition are very similar. Combining these findings with ECR results, we can speculate that when boron additive is mixed with phosphorus ILs, borate films start to deposit on the surface in the earliest stages of the test leading to very short incubation time (Fig. 3 for P_DEHP + SB1 and P_DEHP + SB2) but are subsequently consumed during the wear process thus less boron is found at the surface which is then compensated by the formation of phosphate films which start to form after relatively longer incubation time (Fig. 3 for P_DEHP and P_DEPDT). Hence, the binary additive mixture of IL with SB showed improvement in the wear protection when compared with only ILs blends as seen in Fig. 1(b).

4. Conclusions

The synergistic effect of borate esters with phosphonium ILs was studied for antiwear application. Results indicate that addition of borate ester to phosphonium IL results in significant improvement in wear protection compared to ZDDP as well as ILs and SBs alone at 0.1 wt% P and 0.02 wt% B treat rate both in cylinder on reciprocating flat contact and ball on rotating contact disc tribotest setup.

ECR and TGA of additive chemistry indicate that incubation time for tribofilms formation depends on the thermal stability of the additives, ZDDP with lowest decomposition temperature results in shorter incubation time compared to P_DEPDT and P_DEHP. However, addition of boron chemistry to P_DEHP and P_DEHP results in significantly shorter incubation time for tribofilm formation which subsequently resulted in lower wear.

XANES analysis of P L-edge indicated that ZDDP tribofilms are primarily composed of zinc polyphosphate films that are inhomogeneous in nature with variation in phosphate chain length at different locations and shorter chain phosphates at the surface compared to the bulk. On the other hand, tribofilms derived from P_DEHP and P_DEPDT lubrication are composed of short chain iron phosphate films and exhibited similar phosphate chain length at different locations. P containing IL (P_DEHP) exhibited relatively longer phosphate chain length in the bulk compared to the surface while both P and S containing IL (P_DEPDT) had shorter phosphate chain length in the bulk compared to the surface. Tribofilms formed from P_DEHP + SB blends had short chain iron polyphosphates both at surface and in the bulk.

Boron in SB1 and SB2 derived tribofilms is primarily found as trigonal B mainly as B₂O₃/H₃BO₃ and to some extent tetrahedral B both at surface and in the bulk. Tribofilms derived from IL + SB blends showed the interaction of borate ester with phosphonium IL as BPO₄ is formed in the bulk tribofilms along with B₂O₃/H₃BO₃.

Acknowledgements

XANES experiments were conducted at the Canadian Light Source, Saskatoon, Saskatchewan, Canada that is supported by NSERC, NRC, CIHR and the University of Saskatchewan. Tribological tests were performed at Argonne National Laboratory. Scanning probe microscopy experiments were conducted at Center for Characterization for Materials and Biology at The University of Texas at Arlington. This work was also supported by the "Austrian COMET-Program" in the frame of K2 XTribology (project no. 849109).

References

1. K. Miyoshi, Solid lubrication fundamentals and applications, NASA, Report Number 19980218790 1998.
2. B. Kim, R. Mourhatch and P. B. Aswath, Wear, 2010, 268, 579–591.
3. R. Kapadia, R. Glyde and Y. Wu, Tribol. Int., 2007, 40, 1667–1679.
4. K. Topolovec-Miklozic, T. R. Forbus and H. A. Spikes, Tribol. Lett., 2007, 26, 161–171.
5. H. Fujita and H. A. Spikes, Proc. Inst. Mech. Eng., Part J, 2004, 218, 265–277.
6. J. M. Martin, C. Grossiord, T. Le Mogne, S. Bec and A. Tonck, Tribol. Int., 2001, 34, 523–530,  DOI:10.1016/S0301-679X(01)00029-9.
7. C. Minfray, J. M. Martin, C. Esnouf, T. Le Mogne, R. Kersting and B. Hagenhoff, A multi-technique approach of tribofilm characterisation, Elsevier, San Diego, CA, United States, 2004.
8. M. S. Fuller, L. R. Fernandez, G. Massoumi, W. Lennard, M. Kasrai and G. Bancroft, Tribol. Lett., 2000, 8, 187–192.
9. A. Somayaji, R. Mourhatch and P. B. Aswath, J. Nanosci. Nanotechnol., 2007, 7, 4378–4390.
10. R. Mourhatch and P. B. Aswath, Tribol. Int., 2011, 44, 201–210.
11. K. Ito, J. Martin, C. Minfray and K. Kato, Tribol. Int., 2006, 39, 1538–1544.
12. G. Nehme, R. Mourhatch and P. B. Aswath, Wear, 2010, 268(9–10), 1129–1147.
13. Z. Zhang, E. S. Yamaguchi, M. Kasrai and G. M. Bancroft, Tribofilms generated from ZDDP and DDP on steel surfaces: part 1. Growth, wear and morphology, American Society of Mechanical Engineers, New York, NY 10016-5990, Washington, D.C., United States, 2005.
14. R. Mourhatch and P. B. Aswath, Tribol. Int., 2011, 44, 187–200.
15. H. A. Spikes, Tribol. Lett., 2004, 17, 469–489.
16. A. M. Barnes, K. D. Bartle and V. R. A. Thibon, Tribol. Int., 2001, 34, 389–395.
17. R. Mourhatch and P. B. Aswath, J. Nanosci. Nanotechnol., 2009, 9, 2682–2691.
18. Z. Yin, M. Kasrai, G. M. Bancroft, K. F. Laycock and K. H. Tan, Tribol. Int., 1993, 26, 383–388.
19. J. M. Martin, J. L. Mansot, I. Berbezier and H. Dexpert, Wear, 1984, 16(93), 117–126.
20. S. M. Hsu, Tribol. Int., 2004, 7(37), 553–559.
21. L. L. Cao, Y. M. Sun and L. Q. Zheng, Wear, 1990, 140, 345–357.
22. Federal Emission Standards Reference Guide, United States Environment Protection Agency, March 31st 2016. http://www3.epa.gov/otaq/standards/allstandards.htm#1.
23. Federal Register, ed. Environmental Protection Agency and Department of Transportation National Highway Traffic Safety Administration, National Archive and Records Administration, 2012, p. 62623.
24. F. Kevin, Technical Bulletin 1 API 1509, Engine Oil Licensing and Certification System, June 17th 2010. http://www.api.org/%7E/media/Files/Certification/Engine-Oil-Diesel/Forms/Whats-New/1509-Technical-Bulletin-1.pdf?la=en.
25. V. Sharma, N. Dörr and P. Aswath, RSC Adv., 2016, 6, 22341.
26. A. E. Somers, B. Khemchandani, P. C. Howlett, J. Sun, D. R. MacFarlane and M. Forsyth, ACS Appl. Mater. Interfaces, 2013, 5, 11544–11553.
27. M. Anand, M. Hadfield, J. Viesca, B. Thomas, A. H. Battez and S. Austen, Wear, 2015, 334, 67–74.
28. W. C. Barnhill, J. Qu, H. Luo, H. M. Meyer III, C. Ma, M. Chi and B. L. Papke, ACS Appl. Mater. Interfaces, 2014, 6, 22585–22593.
29. L. Pisarova, C. Gabler, N. Dörr, E. Pittenauer and G. Allmaier, Tribol. Int., 2012, 46, 73–83.
30. M. Cai, Y. Liang, M. Yao, Y. Xia, F. Zhou and W. Liu, ACS Appl. Mater. Interfaces, 2010, 2, 870–876.
31. I. Minami, Molecules, 2009, 14, 2286–2305.
32. B. Yu, D. G. Bansal, J. Qu, X. Sun, H. Luo, S. Dai, P. J. Blau, B. G. Bunting, G. Mordukhovich and D. J. Smolenski, Wear, 2012, 289, 58–64.
33. R. Gusain, P. Gupta, S. Saran and O. P. Khatri, ACS Appl. Mater. Interfaces, 2014, 6(17), 15318–15328.
34. J. Qu, H. Luo, M. Chi, C. Ma, P. J. Blau, S. Dai and M. B. Viola, Tribol. Int., 2014, 71, 88–97.
35. V. Totolin, I. Minami, C. Gabler, J. Brenner and N. Dörr, Tribol. Lett., 2014, 53, 421–432.
36. J. Qu, P. J. Blau, S. Dai, H. Luo and H. M. Meyer III, Tribol. Lett., 2009, 35, 181–189.
37. A. E. Jiménez and M. Bermúdez, Tribol. Lett., 2009, 33, 111–126.
38. A. Jimenez and M. Bermudez, Tribol. Lett., 2010, 37, 431–443.
39. M. Kronberger, V. Pejaković, C. Gabler and M. Kalin, Proc. Inst. Mech. Eng., Part J, 2012, 226, 933–951.
40. M. Bermúdez, A. Jiménez, J. Sanes and F. Carrión, Molecules, 2009, 14, 2888–2908.
41. V. Sharma, C. Gabler, N. Doerr and P. B. Aswath, Tribol. Int., 2015, 92, 353–364.
42. R. González, M. Bartolomé, D. Blanco, J. Viesca, A. Fernández-González and A. H. Battez, Tribol. Int., 2016, 98, 82–93.
43. A. E. Jiménez, M. D. Bermúdez, P. Iglesias, F. J. Carrión and G. Martínez-Nicolás, Wear, 2006, 260, 766–782.
44. H. Wang, Q. Lu, C. Ye, W. Liu and Z. Cui, Wear, 2004, 256, 44–48.
45. M. Yao, M. Fan, Y. Liang, F. Zhou and Y. Xia, Wear, 2010, 268, 67–71.
46. L. Pisarova, V. Totolin, C. Gabler, N. Dörr, E. Pittenauer, G. Allmaier and I. Minami, Tribol. Int., 2013, 65, 13–27.
47. V. Totolin, M. Conte, E. Berriozábal, F. Pagano, I. Minami, N. Dörr, J. Brenner and A. Igartua, Lubrication Science, Special Issue: LUBMAT 2012-Lubrication, Maintenance and Tribotechnology, 2014, vol. 26, (7–8), pp. 514–524.
48. X. Liu, F. Zhou, Y. Liang and W. Liu, Wear, 2006, 261, 1174–1179.
49. I. Minami, T. Inada, R. Sasaki and H. Nanao, Tribol. Lett., 2010, 40, 225–235.
50. C. Gabler, N. Dörr and G. Allmaier, Tribol. Int., 2014, 80, 90–97.
51. R. P. Swatloski, J. D. Holbrey and R. D. Rogers, Green Chem., 2003, 5, 361–363.
52. J. Qu, W. C. Barnhill, H. Luo, H. M. Meyer, D. N. Leonard, A. K. Landauer, B. Kheireddin, H. Gao, B. L. Papke and S. Dai, Adv. Mater., 2015, 27(32), 4767–4774.
53. A. Erdemir, Nanolubricants, John Wiley and Sons, NY, USA, 2008, pp. 203–224.
54. A. Erdemir, Lubrication with boric acid additives, US Patent US 6025306, February 15th 2000.
55. A. Erdemir, H. Sliney and M. Peterson, Lubr. Eng., 1991, 47, 168–173.
56. A. Erdemir, G. Fenske, R. Erck, F. Nichols and D. Busch, Tribological properties of boric acid and boric-acid-forming surfaces: Part 2, Formation and self-lubrication mechanisms of boric acid films on boron-and boric-oxide-containing surfaces, 1990.
57. A. Erdemir, G. Fenske and R. Erck, Surf. Coat. Technol., 1990, 43, 588–596.
58. M. R. Lovell, M. A. Kabir, P. L. Menezes and C. F. Higgs III, Philos. Trans. R. Soc., A, 2010, 368, 4851–4868,  DOI:10.1098/rsta.2010.0183 [doi].
59. P. Deshmukh, M. Lovell, W. G. Sawyer and A. Mobley, Wear, 2006, 260, 1295–1304.
60. K. Parekh, X. Chen and P. B. Aswath, Tribol. Lett., 2009, 34, 141–153,  DOI:10.1007/s11249-008-9373-7.
61. V. Sharma, A. Erdemir and P. B. Aswath, Tribol. Int., 2015, 82, 43–57.
62. M. Kasrai, W. N. Lennard, R. W. Brunner, G. M. Bancroft, J. A. Bardwell and K. H. Tan, Appl. Surf. Sci., 1996, 99, 303–312.
63. J. C. Dodd, Gas engine lubricating oil composition, US Patent Application 20150159107 A1, U. S. A, June 11th 2015.
64. O. L. Warren, J. F. Graham, P. R. Norton, J. E. Houston and T. A. Michalske, Tribol. Lett., 1998, 4, 189–198,  DOI:10.1023/A:1019194903262.
65. L. Taylor, A. Dratva and H. A. Spikes, Friction and wear behavior of zinc dialkyldithiophosphate additive,Society of Tribologists and Lubrication Engineers, 2000.
66. N. N. Gosvami, J. A. Bares, F. Mangolini, A. R. Konicek, D. G. Yablon and R. W. Carpick, Science, 2015, 348, 102–106,  DOI:10.1126/science.1258788.
67. M. Nicholls, M. N. Najman, Z. Zhang, M. Kasrai, P. R. Norton and P. U. P. A. Gilbert, Can. J. Chem., 2007, 85, 816–830.
68. Y. Li, G. Pereira, M. Kasrai and P. R. Norton, Tribol. Lett., 2007, 28, 319–328.
69. Y. Li, G. Pereira, A. Lachenwitzer, M. Kasrai and P. R. Norton, Tribol. Lett., 2007, 27, 245–253.
70. M. A. Nicholls, T. Do, G. M. Bancroft, P. R. Norton, M. Kasrai, T. W. Capehart, Y.- Cheng and T. Perry, Tribol. Lett., 2003, 15(3), 241–248.
71. M. N. Najman, M. Kasrai, G. M. Bancroft, B. H. Frazer and G. DeStatio, Tribol. Lett., 2004, 17(4), 811–822.
72. X. Chen, R. L. Elsenbaumer and P. B. Aswath, Tribol. Int., 2013, 69, 114–124.
73. X. Chen, B. Kim, R. Elsenbaumer, P. B. Aswath and P. B. Aswath, Tribol.-Mater., Surf. Interfaces, 2012, 6(3), 121–133.
74. G. Pereira, D. Munoz-Paniagua, A. Lachenwitzer, M. Kasrai, P. R. Norton, T. W. Capehart, T. A. Perry and Y. Cheng, Wear, 2007, 262, 461–470.
75. M. N. Najman, M. Kasrai and G. M. Bancroft, Tribol. Lett., 2004, 17(2), 217–229.
76. M. Najman, M. Kasrai, G. Michael Bancroft and R. Davidson, Tribol. Int., 2006, 39, 342–355.
77. L. G. Yu, E. S. Yamaguchi, M. Kasrai and G. M. Bancroft, Can. J. Chem., 2007, 85, 675–684.
78. V. N. Bakunin, M. Kasrai, G. N. Kuzmina, G. M. Bancroft and O. P. Parenago, Tribol. Lett., 2007, 26, 33–43.
79. M. T. Costello and M. Kasrai, Tribol. Lett., 2006, 24, 163–169.
80. G. Harp, Z. Han and B. Tonner, J. Vac. Sci. Technol., A, 1990, 8, 2566–2569.
81. D. Li, G. M. Bancroft, M. Kasrai, M. E. Fleet, X. H. Feng and K. H. Tan, Am. Mineral., 1994, 79, 785–788.
82. Z. Yin, M. Kasrai, G. M. Bancroft, K. H. Tan and X. Feng, Phys. Rev. B: Condens. Matter Mater. Phys., 1995, 51, 742–750.
83. M. Kasrai, M. E. Fleet, S. Muthupari, D. Li and G. Bancroft, Phys. Chem. Miner., 1998, 25, 268–272.
84. Z. Zhang, E. S. Yamaguchi, M. Kasrai and G. M. Bancroft, Tribol. Trans., 2004, 47, 527–536.

---