The Effect of Agglomeration Reduction on the Tribological Behavior of WS 2 and MoS 2 Nanoparticle Additives in the Boundary Lubrication Regime

This study investigates the impact of different surfactants and dispersion techniques on the friction and wear behavior of WS2 and MoS2 nanoparticles additives in a Polyalphaolefin (PAO) base oil under boundary lubrication conditions. The nanoparticles were dispersed using Oleic acid (OA) and Polyvinylpyrrolidone (PVP) to investigate their impact on particle agglomeration. The size distribution of the dispersed nanoparticles in PAO was measured by dynamic light scattering. The nanoparticles treated using PVP resulted in the most stable particle size. Friction studies showed that nanoparticle agglomeration reduction and the homogeneity of the suspension did not significantly impact the friction reduction behavior of the lubricant. Reciprocating wear experiments showed that, for our test conditions, both WS2 and MoS2 nano additives exhibited maximum wear depth reduction (45%) when using the PVP surface treatment compared to base oil. The wear results confirmed the significance of minimizing agglomeration and promoting high dispersion in promoting favorable wear resistance under boundary lubricant conditions. Analysis of the wear surfaces showed that a tribofilm formation was the primary wear reduction mechanism for WS2 particles treated by PVP while, in the case of MoS2 treated by PVP, the mechanism was load sharing via particles rolling and/or sliding at the interface.


Introduction
Several studies have shown that tribological behavior of a lubricant can be improved by dispersing a small amount of nanomaterials in base oil.Dispersing nanoparticles in lubricant is a challenging process since the particles tend to agglomerate due to strong Van der Waal forces [1][2][3].The literature shows that dispersion of nanoparticles additives can be improved by using a number of techniques including sonicating, which includes both bath or probe sonication, adding a surfactant, or a combination of the previously mentioned techniques [3][4][5][6][7][8][9][10][11].Specifics of the dispersion method is dictated by the nanoparticles and their surface energy [3,12].
In this study, we selected two dichalcogenides nanoparticles known as WS 2 and MoS 2 that have been shown to improve tribological behavior as additives.The use of MoS 2 nanoparticles has been extensively investigated over the last several years.Multiple studies on MoS 2 nanoparticles did not use a surfactant to stabilize the nanoparticles when added to the lubricant [13][14][15][16][17][18].However, MoS 2 nanoparticles have a strong tendency to agglomerate when added to lubricants [18][19][20].Other studies investigated the dispersion of the MoS 2 nanoparticles and showed that the agglomeration could be reduced by adding surfactant via bath sonication [21][22][23].Similarly, there are studies that have demonstrated improved tribological behavior using WS 2 nanoparticle additives with Refs.[16,[24][25][26][27][28][29][30] and without Refs.[22,23,[31][32][33][34] as well as the use of surfactants.Nevertheless, these results suggest there is a need to determine if reduced agglomeration and dispersion stability directly contribute to improving tribological performance.
The present paper investigates the impact of different surfactants and dispersion techniques on the friction and wear behavior of WS 2 and MoS 2 nanoparticle additives in a Polyalphaolefin (PAO) base oil under boundary lubrication conditions.The substrate chosen was 8620 steel, which is used in drivetrain components in energy generation and agricultural machinery.Oleic acid (OA) and Polyvinylpyrrolidone (PVP) were chosen as surfactants.Both these materials are commonly used in the literature to stabilize nanomaterials additives [7, [35][36][37][38][39].Our initial work showed that, while OA was effective in reducing agglomeration for CuO and WC nanoparticle additives, they were less effective for WS 2 particles.PVP has been shown to be effective for dichalcogenides due to reducing repulsive forces between its hydrophobic chains [22,23,40].

Materials
The neat base oil used in this experiment was Polyalphaolefin (PAO) with the following specifications, as reported by the manufacturer (Exxon Mobile Corporation, Spring, TX, USA): density 826 g/m 3 at 15 • C and viscosity 400 cSt at 40 • C. WS 2 nanoparticles were purchased commercially (Zhengzhou Dongyao Nano Materials Co., Ltd., Zhengzhou, China) and had the following properties: a spherical shape with a nominal diameter of 100 (nm), density of 7.5 (g/m 3 ), and a hardness of 0.75 (mhos).MoS 2 nanoparticles were synthesized using a hydrothermal process and were described as follows.First, separate aqueous solutions of (7 mM) ammonium molybdate tetrahydrate ((NH 4 ) 6 Mo 7 O 24 •4H 2 O; 99.98%) and (35 mM) thiourea (Both purchased from Sigma-Aldrich, St. Louis, Mo, USA) were prepared.Next, the solutions were mixed together to obtain a total volume of 50 mL.The final solution was vigorously stirred for 20 min before it was transferred to a Teflon-lined stainless-steel autoclave.The hydrothermal process was carried out at 180 • C for 15 h after which the autoclave was cooled to room temperature and the black powder of MoS 2 was collected.The MoS 2 powder was centrifuged and washed three times with deionized water and ethanol.The powder was dried at 60 • C for 2 h.The MoS 2 nanoparticles prepared had the following properties: a spherical shape with a nominal diameter of 100 (nm), density of 5.06 (g/m 3 ), and a hardness of 1.5 (mhos).
The surfactants used in this study were Oleic acid and Polyvinylpyrrolidone (both purchased from Sigma-Aldrich).Oleic acid is a surfactant containing a hydrophilic hydroxyl end group and an organophilic alkyl chain [4].PVP is a non-ionic polymer that contains a strong hydrophilic molecule that can attach easily to materials in the solution [36]. Figure 1 shows a schematic for the chemical structure of OA and PVP. and without Refs.[22,23,[31][32][33][34] as well as the use of surfactants.Nevertheless, these results suggest there is a need to determine if reduced agglomeration and dispersion stability directly contribute to improving tribological performance.
The present paper investigates the impact of different surfactants and dispersion techniques on the friction and wear behavior of WS2 and MoS2 nanoparticle additives in a Polyalphaolefin (PAO) base oil under boundary lubrication conditions.The substrate chosen was 8620 steel, which is used in drivetrain components in energy generation and agricultural machinery.Oleic acid (OA) and Polyvinylpyrrolidone (PVP) were chosen as surfactants.Both these materials are commonly used in the literature to stabilize nanomaterials additives [7, [35][36][37][38][39].Our initial work showed that, while OA was effective in reducing agglomeration for CuO and WC nanoparticle additives, they were less effective for WS2 particles.PVP has been shown to be effective for dichalcogenides due to reducing repulsive forces between its hydrophobic chains [22,23,40].

Materials
The neat base oil used in this experiment was Polyalphaolefin (PAO) with the following specifications, as reported by the manufacturer (Exxon Mobile Corporation, Spring, TX, USA): density 826 g/m 3 at 15 °C and viscosity 400 cSt at 40 °C.WS2 nanoparticles were purchased commercially (Zhengzhou Dongyao Nano Materials Co., Ltd., Zhengzhou, China) and had the following properties: a spherical shape with a nominal diameter of 100 (nm), density of 7.5 (g/m 3 ), and a hardness of 0.75 (mhos).MoS2 nanoparticles were synthesized using a hydrothermal process and were described as follows.First, separate aqueous solutions of (7 mM) ammonium molybdate tetrahydrate ((NH4)6Mo7O24•4H2O; 99.98%) and (35 mM) thiourea (Both purchased from Sigma-Aldrich, St. Louis, Mo, USA) were prepared.Next, the solutions were mixed together to obtain a total volume of 50 mL.The final solution was vigorously stirred for 20 min before it was transferred to a Teflon-lined stainless-steel autoclave.The hydrothermal process was carried out at 180 °C for 15 h after which the autoclave was cooled to room temperature and the black powder of MoS2 was collected.The MoS2 powder was centrifuged and washed three times with deionized water and ethanol.The powder was dried at 60 °C for 2 h.The MoS2 nanoparticles prepared had the following properties: a spherical shape with a nominal diameter of 100 (nm), density of 5.06 (g/m 3 ), and a hardness of 1.5 (mhos).
The surfactants used in this study were Oleic acid and Polyvinylpyrrolidone (both purchased from Sigma-Aldrich).Oleic acid is a surfactant containing a hydrophilic hydroxyl end group and an organophilic alkyl chain [4].PVP is a non-ionic polymer that contains a strong hydrophilic molecule that can attach easily to materials in the solution [36]. Figure 1 shows a schematic for the chemical structure of OA and PVP.In this study, the concentration of the nanoparticle additives in oil was chosen to be 1% by weight for each nanoparticle sample based on the recommendation from studies on a variety of material additives [26,30,[41][42][43].The particles added without any surfactant were sonicated for 60 min.For the formulations using Oleic acid as a surfactant, the samples were prepared by mixing 1% of weight Oleic acid to the PAO 1 wt % nanoparticle solutions and sonicating the particles for 120 min.For the formulations using PVP as a surfactant, the samples were prepared as follows.Nanoparticles and PVP were added to the distilled water in the ratio of 3:1 by weight.The solution was subsequently sonicated for 30 min, which is followed by drying in an oven at 60 • C for two hours.The resulting dry nanoparticles were added to the oil and sonicated for 60 min using Cole-Pramer Ultrasonic Bath (Cole-Pramer, Vernon Hills, IL, USA, 115 VAC, 0.8 amps, and High-frequency 42 kHz, at 40 • C).

Dynamic Light Scattering
Extant research demonstrates that the agglomeration of nanoparticles tends to occur when they are added to oil [44].Once agglomeration occurs, the particles size keeps changing.For a better estimate of the nominal particle size in the solution, the size distribution of dispersed nanoparticles in PAO was measured by dynamic light scattering (DLS) (Malvern ZetaSizer Nano ZS, Westborough, MA, USA).The theory underlying the measurements of the Zetasizer are based on effects of Brownian motions of nanoparticles on Rayleigh light scattering data to determine the size of a particle/molecule in the solution [45].The primary result obtained from a DLS measurement represents the intensity-weighted mean diameter derived from the cumulants, which is very sensitive to the existence of nanoparticles agglomeration due to the inherent intensity weighting.Given that the typical friction and wear test time lasts for 60 min, the hydrodynamic diameter distribution of the particles in the solution was measured and collected four times at 15-minute intervals over the span of an hour and the corresponding data were recorded.

Friction and Wear Testing
For the friction and wear experiments, a custom-built reciprocating ball-on-flat micro-tribometer that can produce a microscale (apparent area ~1000 µm 2 ) multi-asperity contact was used [46].A schematic of the micro-tribometer major components is shown in Figure 2. A probe with a specific radius is placed at the end of a crossed I-beam structure, which is lowered using a linear stage to apply a desired normal load to the sample.For this study, a SiC probe (4 mm diameter, hardness = 9 mohs, v = 0.19, E = 415 GPa) and an AISI 8620 steel substrate (4.5 mohs hardness, v = 0.29, E = 200 GPa) were used.The average substrate surface roughness (Ra) was approximately 0.06 µm as measured by a Zygo NewView (Zygo, Middlefield, CT, USA) 7100 non-contact profilometer over a scan area of 0.47 mm × 0.35 mm.
The normal and the friction (lateral) forces in the micro-tribometer are measured using semiconductor strain gages on the cantilevers.Friction forces can be resolved to approximately ±5 µN and normal forces to approximately ±15 µN.The signal from the normal load is monitored and used in a simple proportional-integral (PI) feedback loop to maintain the desired normal force regardless of any slope or waviness in the surface of the sample.The desired sample is affixed to another stage set perpendicular to the beam, which provides a linear motion.An appropriate amount of the formulated oil was dropped to a fresh substrate surface prior to conducting each test.A new SiC probe was used for each friction and wear test.
To evaluate the coefficient of friction, the applied normal load was increased linearly from 0 to 2000 mN for a specific sliding distance of 25 mm at a speed of 1 mm/s.To obtain the coefficient of friction, ramped load tests were performed in which the normal load increased linearly with the sliding distance while the friction force was monitored.The load was increased from 0.2 to 200 mN since the probe was moved across a stroke distance.Based on these parameters, the maximum Hertzian contact pressure was estimated to be 1.27 GPa and the lubrication regime was boundary lubrication.The lateral force was recorded continuously along with the normal load for each test, which provides a coefficient of friction.Five replicates were conducted for each condition and the average coefficient of friction values were reported along with the 90% confidence interval.In order to obtain the wear response, a reciprocating sliding wear test was performed against the SiC probe at a constant load of 5400 mN (maximum Hertzian force of 1.77 GPa) for 200 cycles, a stroke length of 8 mm, and a stroke speed of 10 mm/s.Two replicates were conducted for each sample.After each test, the sample surface was cleaned with isopropyl alcohol wipes and the wear track depth was measured using the Zygo profilometer.An average wear depth was reported from five measurements along each wear track and the data was reported as an average of 10 measurements (from two tests) along with a 90% confidence interval.Note that the test condition consisting of base oil with no additives was used as a control.
An FEI Quanta-250 Scanning Electron Microscope (FESEM, Supra 400VP Gemini, Jena, Germany) was used to obtain high-resolution images of wear tracks (Oxford Aztec energy-dispersive X-ray analysis, Abingdon, UK) was used to perform point analysis of the wear track and adjacent regions for evidence of any tribo-film formation.Backscattered images were analyzed for further tribo-film analysis with an accelerating voltage of 10 kV and a spot size of 4 A.U. for all cases.
Lubricants 2018, 6, x FOR PEER REVIEW 4 of 12 which provides a coefficient of friction.Five replicates were conducted for each condition and the average coefficient of friction values were reported along with the 90% confidence interval.In order to obtain the wear response, a reciprocating sliding wear test was performed against the SiC probe at a constant load of 5400 mN (maximum Hertzian force of 1.77 GPa) for 200 cycles, a stroke length of 8 mm, and a stroke speed of 10 mm/s.Two replicates were conducted for each sample.After each test, the sample surface was cleaned with isopropyl alcohol wipes and the wear track depth was measured using the Zygo profilometer.An average wear depth was reported from five measurements along each wear track and the data was reported as an average of 10 measurements (from two tests) along with a 90% confidence interval.Note that the test condition consisting of base oil with no additives was used as a control.
An FEI Quanta-250 Scanning Electron Microscope (FESEM, Supra 400VP Gemini, Jena, Germany) was used to obtain high-resolution images of wear tracks (Oxford Aztec energy-dispersive X-ray analysis, Abingdon, UK) was used to perform point analysis of the wear track and adjacent regions for evidence of any tribo-film formation.Backscattered images were analyzed for further tribo-film analysis with an accelerating voltage of 10 kV and a spot size of 4 A.U. for all cases.

Results
The results are presented and discussed in the following three sections.First, we describe the results related to reducing agglomeration of nanoparticle additives when dispersed in base oil.Next, we describe the friction behavior of the nano-additives, which is followed by the results of the stable nano-additives on wear behavior.

Dispersion of Nanoparticles in Oil
Figure 3 shows the average particle size as a function of post sonication time for the two nanoparticle additives and surfactants.The use of particles without surfactant resulted in an initially high particles size (350-420 nm) and a gradual decrease over the testing duration with sizes of about 200 nm at 60 min.However, the absence of surfactants did lead to a very high variability in the particle size, which suggests unstable dispersion.For both particles, dispersing with PVP provided the lowest and most stable particle size for the testing duration.The stable particle size was around 150 nm and 100 nm for WS2 and MoS2 nanoparticles during the testing duration used.OA was less effective in reducing agglomeration when compared to PVP and by the 60 min mark, which did not show appreciable improvements when compared to the condition without surfactant.OA as a surfactant was more effective in the case of MoS2 with a particle size of 200 nm but not as effective as PVP.However, the variability in the particle size was very high when compared to PVP.This variability was even more apparent in the case of WS2.

Results
The results are presented and discussed in the following three sections.First, we describe the results related to reducing agglomeration of nanoparticle additives when dispersed in base oil.Next, we describe the friction behavior of the nano-additives, which is followed by the results of the stable nano-additives on wear behavior.

Dispersion of Nanoparticles in Oil
Figure 3 shows the average particle size as a function of post sonication time for the two nanoparticle additives and surfactants.The use of particles without surfactant resulted in an initially high particles size (350-420 nm) and a gradual decrease over the testing duration with sizes of about 200 nm at 60 min.However, the absence of surfactants did lead to a very high variability in the particle size, which suggests unstable dispersion.For both particles, dispersing with PVP provided the lowest and most stable particle size for the testing duration.The stable particle size was around 150 nm and 100 nm for WS 2 and MoS 2 nanoparticles during the testing duration used.OA was less effective in reducing agglomeration when compared to PVP and by the 60 min mark, which did not show appreciable improvements when compared to the condition without surfactant.OA as a surfactant was more effective in the case of MoS 2 with a particle size of 200 nm but not as effective as PVP.However, the variability in the particle size was very high when compared to PVP.This variability was even more apparent in the case of WS 2 .Overall, while the addition of Oleic acid as well as treating the nanoadditives by PVP can form a layer on the nanoparticle surface and contribute to an increased diameter, the data suggest that agglomeration is the greater contributor for the increased diameters observed.This is supported by the fact that the observed diameters for particles with surfactants were lower than that of particles without surfactant.Furthermore, despite the fact that the amount by weight of PVP in the oil formulations for both particles was less than that of oleic acid, the PVP dispersion technique showed a lower particle size, which points to agglomeration effects being the primary phenomenon captured by the DLS data.In addition, a visual impact confirmed that the dispersion did not settle, which eliminates this as a possible reason for observing a reduced particle size.Consequently, the dispersion technique using PVP contribute to effective agglomeration reduction and the most stable dispersion.We speculate that the difference in the chemical structure of the surfactants contributed to the difference in their impact on agglomeration.The hydrophobic chain interactions of PVP combined with its effectiveness in forming surface films appear to lead to lower surface energy when compared to the interactions of the non-polar ends of OA.

Friction Behavior
Figure 4 compares the reduction in the coefficient of friction observed for the formulations of nanoparticle additive samples relative to the coefficient of friction of the neat base oil (PAO).All formulations exhibited a friction reduction with reductions ranging from 12% to 33%.When compared to the base oil, the average reduction using WS2 particles with no surfactants was 25%.Additionally, the addition of oleic acid showed a 29% reduction while there was a 33% reduction using the PVP treatment for the WS2 particles.There were high amounts of variability in friction reduction across the dispersion methods for WS2, which suggests that the observed reductions are Overall, while the addition of Oleic acid as well as treating the nanoadditives by PVP can form a layer on the nanoparticle surface and contribute to an increased diameter, the data suggest that agglomeration is the greater contributor for the increased diameters observed.This is supported by the fact that the observed diameters for particles with surfactants were lower than that of particles without surfactant.Furthermore, despite the fact that the amount by weight of PVP in the oil formulations for both particles was less than that of oleic acid, the PVP dispersion technique showed a lower particle size, which points to agglomeration effects being the primary phenomenon captured by the DLS data.In addition, a visual impact confirmed that the dispersion did not settle, which eliminates this as a possible reason for observing a reduced particle size.Consequently, the dispersion technique using PVP contribute to effective agglomeration reduction and the most stable dispersion.We speculate that the difference in the chemical structure of the surfactants contributed to the difference in their impact on agglomeration.The hydrophobic chain interactions of PVP combined with its effectiveness in forming surface films appear to lead to lower surface energy when compared to the interactions of the non-polar ends of OA.

Friction Behavior
Figure 4 compares the reduction in the coefficient of friction observed for the formulations of nanoparticle additive samples relative to the coefficient of friction of the neat base oil (PAO).All formulations exhibited a friction reduction with reductions ranging from 12% to 33%.
When compared to the base oil, the average reduction using WS 2 particles with no surfactants was 25%.Additionally, the addition of oleic acid showed a 29% reduction while there was a 33% reduction using the PVP treatment for the WS 2 particles.There were high amounts of variability in friction reduction across the dispersion methods for WS 2 , which suggests that the observed reductions are comparable among dispersion conditions even though the average reduction in friction using the PVP treatment was the highest.As such, it appears that agglomeration reduction of the WS 2 nanoparticles can result in improved friction performance under the specified testing conditions.
Friction tests of oil samples containing MoS 2 nano additives in base oil demonstrated lower levels of friction reduction when compared to WS 2 .The measured average reduction percentage using MoS 2 added to the lubricant without surfactant was 25%.In addition, the oleic acid technique showed a 12% reduction on average while there was a 16% reduction using the PVP treatment.MoS 2 nanoparticles dispersed in lubricant without surfactant had the best friction reduction when compared to the other two techniques.The data indicates that agglomeration reduction of MoS 2 nanoparticle additives was not a requirement for effective friction reduction.
Overall, the nano-additives in this study reduced the coefficient of friction when compared to the neat base oil.These results suggest that the WS 2 particles reduce friction by providing avenues for rolling friction rather than sliding friction.Reduced agglomeration (using PVP) enhances the ball bearing effect, which provides lower friction when compared to conditions where agglomeration occurs (no surfactant).In contrast, for the MoS 2 nanoparticles, the primary mechanism is still the traditional shearing of planes, which explains why the occurrence of agglomeration has a better friction reduction when compared to the presence of individual particles [47].The availability of more particles in the contact facilitate the shearing of the particles.It must be noted that intrinsic properties of the nanoparticles including the structure, the number of layers, and the defect may also influence their friction response.
Lubricants 2018, 6, x FOR PEER REVIEW 6 of 12 comparable among dispersion conditions even though the average reduction in friction using the PVP treatment was the highest.As such, it appears that agglomeration reduction of the WS2 nanoparticles can result in improved friction performance under the specified testing conditions.Friction tests of oil samples containing MoS2 nano additives in base oil demonstrated lower levels of friction reduction when compared to WS2.The measured average reduction percentage using MoS2 added to the lubricant without surfactant was 25%.In addition, the oleic acid technique showed a 12% reduction on average while there was a 16% reduction using the PVP treatment.MoS2 nanoparticles dispersed in lubricant without surfactant had the best friction reduction when compared to the other two techniques.The data indicates that agglomeration reduction of MoS2 nanoparticle additives was not a requirement for effective friction reduction.
Overall, the nano-additives in this study reduced the coefficient of friction when compared to the neat base oil.These results suggest that the WS2 particles reduce friction by providing avenues for rolling friction rather than sliding friction.Reduced agglomeration (using PVP) enhances the ball bearing effect, which provides lower friction when compared to conditions where agglomeration occurs (no surfactant).In contrast, for the MoS2 nanoparticles, the primary mechanism is still the traditional shearing of planes, which explains why the occurrence of agglomeration has a better friction reduction when compared to the presence of individual particles [47].The availability of more particles in the contact facilitate the shearing of the particles.It must be noted that intrinsic properties of the nanoparticles including the structure, the number of layers, and the defect may also influence their friction response.

Wear Behavior
Figure 5 shows the results of the wear tests using WS2 and MoS2 nano additives and different dispersion techniques.The results shown are reductions in wear scar depths for each formulation compared to the wear depth observed when using base oil.Positive values indicate improved wear performance while negative values indicate worse performance when compared to the wear depth observed with the base oil only.The changes on the wear behavior varied from −22% to +45%, which indicates that the nano additives could lead to an increase or decrease in the wear scar depth when compared to the base oil depending on the nanomaterial and the dispersion method used.Figure 6 shows representative wear scar particles for the various nanomaterial additives and dispersion conditions.
WS2 nanoparticles exhibited a relative wear reduction in all of the tested dispersion techniques.For example, when compared to the base oil, the average reduction percentage observed using no surfactants was 9%.Treatment with oleic acid yielded a 15% wear reduction while treatment with PVP resulted in a 43% reduction.When taken together with the DLS data in Figure 3a, these results

Wear Behavior
Figure 5 shows the results of the wear tests using WS 2 and MoS 2 nano additives and different dispersion techniques.The results shown are reductions in wear scar depths for each formulation compared to the wear depth observed when using base oil.Positive values indicate improved wear performance while negative values indicate worse performance when compared to the wear depth observed with the base oil only.The changes on the wear behavior varied from −22% to +45%, which indicates that the nano additives could lead to an increase or decrease in the wear scar depth when compared to the base oil depending on the nanomaterial and the dispersion method used.Figure 6 shows representative wear scar particles for the various nanomaterial additives and dispersion conditions.
WS 2 nanoparticles exhibited a relative wear reduction in all of the tested dispersion techniques.For example, when compared to the base oil, the average reduction percentage observed using no surfactants was 9%.Treatment with oleic acid yielded a 15% wear reduction while treatment with PVP resulted in a 43% reduction.When taken together with the DLS data in Figure 3a, these results show that reducing agglomeration and enhancing dispersion stability promoted increased wear resistance in terms of WS 2 nanoparticles.show that reducing agglomeration and enhancing dispersion stability promoted increased wear resistance in terms of WS2 nanoparticles.The MoS2 nano-additives in base oil demonstrated different wear behavior than that of WS2.The average wear reduction for MoS2 when added to lubricant without surfactant was approximately −5%, which means that the wear scar depth increased by 5% when MoS2 was added to the base oil.In addition, treatment with oleic acid increased wear depth by 20% when compared with the base oil.In contrast, the treatment with PVP, which is shown in Figure 3b, had the lowest and most stable particle size, which resulted in a 45% wear scar depth reduction when compared with base oil.This result suggests that agglomeration reduction of the nano-additives alleviate the wear resistance of the lubricant.
A primary interpretation of the nano additives behavior in boundary lubrication is that the small particles fill the valleys of the surface asperities and increases the real contact area.The nano additives then help share the load subjected to the surface asperities and reduce wear.We investigated substrate surfaces lubricated with nanoparticles treated by PVP and without surfactant to better identify the prevailing mechanism of wear reduction.Figure 7 shows representative SEM images and EDS spectra of the wear tracks from the tests using the nanoparticles with and without PVP.Figure show that reducing agglomeration and enhancing dispersion stability promoted increased wear resistance in terms of WS2 nanoparticles.The MoS2 nano-additives in base oil demonstrated different wear behavior than that of WS2.The average wear reduction for MoS2 when added to lubricant without surfactant was approximately −5%, which means that the wear scar depth increased by 5% when MoS2 was added to the base oil.In addition, treatment with oleic acid increased wear depth by 20% when compared with the base oil.In contrast, the treatment with PVP, which is shown in Figure 3b, had the lowest and most stable particle size, which resulted in a 45% wear scar depth reduction when compared with base oil.This result suggests that agglomeration reduction of the nano-additives alleviate the wear resistance of the lubricant.
A primary interpretation of the nano additives behavior in boundary lubrication is that the small particles fill the valleys of the surface asperities and increases the real contact area.The nano additives then help share the load subjected to the surface asperities and reduce wear.We investigated substrate surfaces lubricated with nanoparticles treated by PVP and without surfactant to better identify the prevailing mechanism of wear reduction.Figure 7 shows representative SEM images and EDS spectra of the wear tracks from the tests using the nanoparticles with and without PVP.Figure 7a shows the SEM image of the wear track of the surface lubricated with 1% WS2 dispersed without The MoS 2 nano-additives in base oil demonstrated different wear behavior than that of WS 2 .The average wear reduction for MoS 2 when added to lubricant without surfactant was approximately −5%, which means that the wear scar depth increased by 5% when MoS 2 was added to the base oil.In addition, treatment with oleic acid increased wear depth by 20% when compared with the base oil.In contrast, the treatment with PVP, which is shown in Figure 3b, had the lowest and most stable particle size, which resulted in a 45% wear scar depth reduction when compared with base oil.This result suggests that agglomeration reduction of the nano-additives alleviate the wear resistance of the lubricant.
A primary interpretation of the nano additives behavior in boundary lubrication is that the small particles fill the valleys of the surface asperities and increases the real contact area.The nano additives then help share the load subjected to the surface asperities and reduce wear.We investigated substrate surfaces lubricated with nanoparticles treated by PVP and without surfactant to better identify the prevailing mechanism of wear reduction.Figure 7 shows representative SEM images and EDS spectra of the wear tracks from the tests using the nanoparticles with and without PVP.Figure 7a shows the SEM image of the wear track of the surface lubricated with 1% WS 2 dispersed without surfactant.It demonstrates an existence of abrasive wear via plastic deformation that changed the surface topography of the wear track region.In contrast, Figure 7b demonstrates fewer changes in the surface topography inside the wear track when WS 2 nanoparticles were treated by PVP. Figure 7b also shows a lighter color inside the wear track, which suggests the existence of material layers other than steel.This layer could be debris from the contact surfaces or nano additives embedded in the surface.Figure 7c shows the chemical analysis of the unworn and worn surfaces for WS 2 additives with and without PVP.EDS data identified a higher peak intensity of W particles present inside the wear track when lubricated using WS 2 treated by PVP.However, no trace of sulfur was detected because the intensity line of sulfur is weaker.This supports nanoparticle-based film formation on the wear track surface that could be the cause of the higher wear reduction observed for this dispersion condition.It is also possible that PVP acts as a film promoter, but that mechanism needs further investigation.the surface topography inside the wear track when WS2 nanoparticles were treated by PVP. Figure 7b also shows a lighter color inside the wear track, which suggests the existence of material layers other than steel.This layer could be debris from the contact surfaces or nano additives embedded in the surface.Figure 7c shows the chemical analysis of the unworn and worn surfaces for WS2 additives with and without PVP.EDS data identified a higher peak intensity of W particles present inside the wear track when lubricated using WS2 treated by PVP.However, no trace of sulfur was detected because the intensity line of sulfur is weaker.This supports nanoparticle-based film formation on the wear track surface that could be the cause of the higher wear reduction observed for this dispersion condition.It is also possible that PVP acts as a film promoter, but that mechanism needs further investigation.Figure 7e and Figure 7f compares MoS2 nanoparticles dispersed with and without PVP.Wear tracks of the substrates show evidence of abrasive wear via ploughing, which is similar to that observed using WS2.PVP treated nanoparticles resulted in less surface deformation on the worn surface and fewer changes on the surface topography between the outside and inside wear track regions.Analyses of SEM images and EDS data for the substrates reveal no evidence of any MoS2 nanoparticle-based material on any of the wear tracks.This suggests the mechanism of wear reduction for well-dispersed MoS2 in the boundary lubrication condition, which is via load support of rolling nanoparticles rather than the formation of a tribo-film.The particles rolling mechanisms Figure 7e,f compares MoS 2 nanoparticles dispersed with and without PVP.Wear tracks of the substrates show evidence of abrasive wear via ploughing, which is similar to that observed using WS 2 .PVP treated nanoparticles resulted in less surface deformation on the worn surface and fewer changes on the surface topography between the outside and inside wear track regions.Analyses of SEM images and EDS data for the substrates reveal no evidence of any MoS 2 nanoparticle-based material on any of the wear tracks.This suggests the mechanism of wear reduction for well-dispersed MoS 2 in the boundary lubrication condition, which is via load support of rolling nanoparticles rather than the formation of a tribo-film.The particles rolling mechanisms was also found in the literature as the dominant wear reduction [48][49][50].The agglomeration reduction and homogeneous particle dispersion help to make particles roll more and contribute to the wear reduction.
We were expecting that WS 2 nanoparticles treated by PVP would provide a higher level of relative wear reduction compared to MoS 2 because of the tribo-film formed on the substrate.However, as demonstrated in Figure 4, the wear reduction in WS 2 was equivalent to MoS 2 nano additives treated by PVP, which we believe reduced wear via a nanoscale rolling effect.This equivalency may be due to the lack of durability of the WS 2 , which is a tribo-film, or MoS 2 , which may have formed a film as well but was removed before the 200 cycles were used in the previous analyses (Figure 5).In order to investigate these possibilities, we performed wear depth measurements and EDS analyses at 100, 200, 400, and 800 cycles at the same load.Figure 8a shows variations of the average wear scar depth as a function of sliding cycles.The wear scar depth increased sharply at first, but then gradually increased between 200 and 800 cycles.WS 2 showed a slightly lower wear rate increase when compared to MoS 2 .At 800 cycles, WS 2 exhibited a slightly lower wear scar depth.However, this difference is minimal.Figure 8b shows EDS data for the wear tracks of various sliding cycles for WS 2 with PVP.A decrease in intensity of the adhered W materials to the surface is observed as the number of cycles increase.The data suggest that tribo-film formation is the dominant wear reduction mechanism when WS 2 with PVP is used.This tribo-film is susceptible to wear.This explains why it did not show higher relative wear reduction when compared with MoS 2 PVP.Despite the tribo-film wear susceptibility, it is quite durable under the tested conditions.Figure 8c shows the intensity level of Mo particles inside the wear track regions on the substrate surfaces when MoS 2 with PVP lubricated the particles for different wear track cycles.The data shows a lack of any particle or particle-based film formation when MoS 2 with PVP nano additives were used.The data reinforces the load sharing via rolling and that there was not any tribo-film formed under this condition.
Lubricants 2018, 6, x FOR PEER REVIEW 9 of 12 and homogeneous particle dispersion help to make particles roll more and contribute to the wear reduction.
We were expecting that WS2 nanoparticles treated by PVP would provide a higher level of relative wear reduction compared to MoS2 because of the tribo-film formed on the substrate.However, as demonstrated in Figure 4, the wear reduction in WS2 was equivalent to MoS2 nano additives treated by PVP, which we believe reduced wear via a nanoscale rolling effect.This equivalency may be due to the lack of durability of the WS2, which is a tribo-film, or MoS2, which may have formed a film as well but was removed before the 200 cycles were used in the previous analyses (Figure 5).In order to investigate these possibilities, we performed wear depth measurements and EDS analyses at 100, 200, 400, and 800 cycles at the same load.Figure 8a shows variations of the average wear scar depth as a function of sliding cycles.The wear scar depth increased sharply at first, but then gradually increased between 200 and 800 cycles.WS2 showed a slightly lower wear rate increase when compared to MoS2.At 800 cycles, WS2 exhibited a slightly lower wear scar depth.However, this difference is minimal.Figure 8b shows EDS data for the wear tracks of various sliding cycles for WS2 with PVP.A decrease in intensity of the adhered W materials to the surface is observed as the number of cycles increase.The data suggest that tribo-film formation is the dominant wear reduction mechanism when WS2 with PVP is used.This tribo-film is susceptible to wear.This explains why it did not show higher relative wear reduction when compared with MoS2 PVP.Despite the tribo-film wear susceptibility, it is quite durable under the tested conditions.Figure 8c shows the intensity level of Mo particles inside the wear track regions on the substrate surfaces when MoS2 with PVP lubricated the particles for different wear track cycles.The data shows a lack of any particle or particle-based film formation when MoS2 with PVP nano additives were used.The data reinforces the load sharing via rolling and that there was not any tribo-film formed under this condition.

Conclusions
This study tested different suspension techniques for WS2 and MoS2 nanoparticles to enhance agglomeration reduction and improve the particle dispersion in PAO.Subsequently, the tribological behavior exhibited by the various dispersion techniques of the nanoparticles in a boundary lubricant

Conclusions
This study tested different suspension techniques for WS 2 and MoS 2 nanoparticles to enhance agglomeration reduction and improve the particle dispersion in PAO.Subsequently, the tribological behavior exhibited by the various dispersion techniques of the nanoparticles in a boundary lubricant regime was investigated.The nanoparticles were dispersed using the following techniques: (1) 60-minute sonication without using a stabilizing agent, (2) 60-minute sonication with 1% of weight oleic acid (OA), and (3) treating the nanoparticles using PVP (3:1 ratio of nanoparticles to PVP).The dispersion conditions using PVP showed the most stable particle size and homogeneous mixture dispersion for both nanoparticles.
Friction studies showed that nanoparticle agglomeration reduction and the homogeneity of the suspension did not significantly impact the friction reduction behavior of the lubricant.Reciprocating wear experiments showed that, for our test conditions, both WS 2 and MoS 2 nano additives showed maximum wear depth reduction (45%) when using the PVP surface treatment when compared to base oil.The wear results confirmed the significance of minimizing agglomeration, promoting high dispersion, and enabling favorable wear resistance under boundary lubricant conditions.Further analysis using SEM of the wear surfaces showed that a tribo-film formation was the primary wear reduction mechanism for WS 2 particles treated by PVP while, in the case of MoS 2 , the mechanism was load sharing through particle rolling and/or sliding.Overall, the study demonstrates that stabilizing the dispersion of WS 2 and MoS 2 nanoparticles and minimizing agglomeration will improve wear resistance under boundary lubricant conditions.

Figure 1 .
Figure 1.Schematic of the chemical structures of Oleic acid (OA) and Polyvinylpyrrolidone (PVP) that were chosen as surfactants.

Figure 1 .
Figure 1.Schematic of the chemical structures of Oleic acid (OA) and Polyvinylpyrrolidone (PVP) that were chosen as surfactants.

Figure 2 .
Figure 2. (a) Schematic of micro-tribometer used for friction and wear tests.(b) A photograph of the micro-tribometer.

Figure 2 .
Figure 2. (a) Schematic of micro-tribometer used for friction and wear tests.(b) A photograph of the micro-tribometer.

Figure 3 .
Figure 3. Dynamic light scattering (DLS) data show the average hydrodynamic diameter of (a) WS 2 , (b) MoS 2 , nanoparticles in PAO dispersed using different surfactants.

Figure 4 .
Figure 4.A comparison of the reduction in the coefficient of friction exhibited by the various dispersion techniques for WS2 and MoS2 nanoparticles.Reductions are expressed as percentages with respect to the coefficient of friction observed in a base oil condition.Average values are shown along with 90% confidence intervals.

Figure 4 .
Figure 4.A comparison of the reduction in the coefficient of friction exhibited by the various dispersion techniques for WS 2 and MoS 2 nanoparticles.Reductions are expressed as percentages with respect to the coefficient of friction observed in a base oil condition.Average values are shown along with 90% confidence intervals.

Figure 5 .
Figure 5.A comparison of the reduction in wear exhibited by the various dispersion techniques for WS2 and MoS2 nanoparticles.Reductions are expressed as percentages with respect to the wear observed with only the base oil.Positive values indicate improved wear performance while negative values indicate worse performance.Average values are shown along with 90% confidence intervals.

Figure 6 .
Figure 6.Representative wear scar profiles for (a) WS2 nanoparticles and (b) MoS2 nanoparticles with different surfactants, compared to the results for the base oil.

Figure 5 .
Figure 5.A comparison of the reduction in wear exhibited by the various dispersion techniques for WS 2 and MoS 2 nanoparticles.Reductions are expressed as percentages with respect to the wear observed with only the base oil.Positive values indicate improved wear performance while negative values indicate worse performance.Average values are shown along with 90% confidence intervals.

Figure 5 .
Figure 5.A comparison of the reduction in wear exhibited by the various dispersion techniques for WS2 and MoS2 nanoparticles.Reductions are expressed as percentages with respect to the wear observed with only the base oil.Positive values indicate improved wear performance while negative values indicate worse performance.Average values are shown along with 90% confidence intervals.

Figure 6 .
Figure 6.Representative wear scar profiles for (a) WS2 nanoparticles and (b) MoS2 nanoparticles with different surfactants, compared to the results for the base oil.

Figure 6 .
Figure 6.Representative wear scar profiles for (a) WS 2 nanoparticles and (b) MoS 2 nanoparticles with different surfactants, compared to the results for the base oil.

Figure 7 .
Figure 7. Scanning Electron Microscope (SEM) images of wear tracks on a steel substrate from sliding tests against a diamond probe of the wear track.(a) WS2 as additives and (b) WS2 treated by PVP.Dotted lines indicate the edges of the wear track.(c) EDS data for WS2 nano additives from regions inside and outside the wear tracks.The figure demonstrates a lack of any particles or particle-based film formation when no surfactant used and adhered W materials to the surface forming a layer of particles or particle-based film when the additives treated by PVP.(d) MoS2 as additives and (e) MoS2 treated by PVP.(f) EDS data for the different MoS2 nano additives from regions inside and outside the wear track show a lack of any particles or particle-based film formation.

Figure 7 .
Figure 7. Scanning Electron Microscope (SEM) images of wear tracks on a steel substrate from sliding tests against a diamond probe of the wear track.(a) WS 2 as additives and (b) WS 2 treated by PVP.Dotted lines indicate the edges of the wear track.(c) EDS data for WS 2 nano additives from regions inside and outside the wear tracks.The figure demonstrates a lack of any particles or particle-based film formation when no surfactant used and adhered W materials to the surface forming a layer of particles or particle-based film when the additives treated by PVP.(d) MoS 2 as additives and (e) MoS 2 treated by PVP.(f) EDS data for the different MoS 2 nano additives from regions inside and outside the wear track show a lack of any particles or particle-based film formation.

Figure 8 .
Figure 8.(a) Wear scar depth data in µm for different cycles when WS2 and MoS2 nanoparticles treated by PVP.(b) EDS data for WS2 nano additives treated by PVP from regions inside the wear tracks for different cycles, which shows a decrease in the adhered W materials to the surface formed a layer of particles when the number of cycles increase.(c) EDS data for the MoS2 nano additives treated by PVP from regions inside the wear tracks for different cycles shows a lack of any particle or particle-based film formation.

Figure 8 .
Figure 8.(a) Wear scar depth data in µm for different cycles when WS 2 and MoS 2 nanoparticles treated by PVP.(b) EDS data for WS 2 nano additives treated by PVP from regions inside the wear tracks for different cycles, which shows a decrease in the adhered W materials to the surface formed a layer of particles when the number of cycles increase.(c) EDS data for the MoS 2 nano additives treated by PVP from regions inside the wear tracks for different cycles shows a lack of any particle or particle-based film formation.