Article pubs.acs.org/Langmuir
Nanotribological Characterization of Lubricants between Smooth Iron Surfaces Motohiro Kasuya,† Kazuhito Tomita,† Masaya Hino,† Masashi Mizukami,† Hiroyuki Mori,§ Seiji Kajita,§ Toshihide Ohmori,§ Atsushi Suzuki,∥ and Kazue Kurihara*,†,‡ †
Institute of Multidisciplinary Research for Advanced Materials and ‡WPI-Advanced Institute for Material Research, Tohoku University, Sendai 980-8577, Japan § Toyota Central R & D Laboratories, Inc., Nagakute, Aichi 480-1192, Japan ∥ Toyota Motor Co., Toyota, Aichi 471-8572, Japan S Supporting Information *
ABSTRACT: We performed the resonance shear measurement (RSM) for evaluating the nanorheological and tribological properties of model lubricants, hexadecane and poly(α-olefin) (PAO), confined between iron surfaces. The twin-path surface forces apparatus (SFA) was used for determining the distance between the surfaces. The obtained resonance curves for the confined lubricants showed that the viscosity of the confined hexadecane and PAO increased due to liquid structuring when the surface separation (D) decreased to a value less than 24 and 20 nm, respectively. It was also determined that the iron surfaces were lubricated by the hexadecane when normal load (L) was less than 1.1 mN, while the confined hexadecane behaved almost solid-like and showed poor lubricity when L was greater than 1.1 mN. In contrast, PAO between the iron surfaces showed high lubricity even under the high load (L > 2 mN). The surface separation of hexadecane and PAO at a hard wall contact between the iron surfaces was determined to be 4.6 ± 0.5 and 5.0 ± 0.4 nm by applying the fringes of equal chromatic order (FECO) for half-transparent iron films deposited on mica surfaces as substrates. We also characterized hexadecane and PAO confined between mica surfaces for studying the effect of substrates on the confined lubricants.
1. INTRODUCTION The reduction of friction is essential for energy savings, which is one of the most important challenges for current and future engineering. For example, one-third of the fuel energy is used to overcome friction in the engine, transmission, tires, and brakes in passenger cars.1 Lubricants have been used for many years in order to reduce friction between solid surfaces for many engineering applications such as machines and drawn steel. For improving the performance of the engine in the fluid lubrication regime, the development of low viscosity lubricants has been attempted. It is crucial for such lubricants to sufficiently reduce the friction and the wear in the boundary lubrication regime, which appears during low shear speed and high normal load. For this lubrication regime, it is believed that the thickness of the lubricants between surfaces is similar to the roughness or less and the surfaces are partly in solid−solid contact on the bump.2 However, recent reports from several research groups have demonstrated that the molecules of the lubricants remain between the surfaces even under a high normal load.3−8 Therefore, understanding the detailed mechanism of the boundary lubrication is necessary. The surface forces measurement can be a unique tool for solving this problem because of its ability of varying the thickness (surface separation, D) of the lubricants at a 0.1 nm resolution.8,9 Various shear measurement techniques employing © 2017 American Chemical Society
a surface forces apparatus (SFA) have also been developed for characterizing the lubricity of a liquid film in a molecular level of thickness.10−19 It was reported that the properties of the bulk liquids such as viscosity could not be directly correlated to the properties of the liquid under confinement.6 Thus, it is important to directly evaluate the properties of the confined liquids. We developed the resonance shear measurement (RSM) based on the SFA and studied the nanorheological and tribological properties of the confined liquids.17−19 The resonance shear responses are sensitive to changes in the properties of confined liquids and insensitive to noises which are advantages of the RSM. In the case of shear measurements based on a conventional SFA, only transparent substrates, generally mica, have been practically studied because the FECO (fringes of equal chromatic order)8 method used for the distance determination of the conventional SFA requires transparent surfaces. However, mica is not a substrate used for conventional tribology. Therefore, it is important to study substrates other than mica. Received: January 14, 2017 Revised: March 27, 2017 Published: April 10, 2017 3941
DOI: 10.1021/acs.langmuir.7b00148 Langmuir 2017, 33, 3941−3948
Article
Langmuir Recently, we developed a new surface forces apparatus using two-beam (twin-path) interferometry, i.e., twin-path SFA, for measuring the interactions between opaque substrates,20 and used it for measuring the interaction between two gold surfaces.21 Using RSM with the twin-path SFA, it is now possible to study the nanorheological and tribological properties of lubricants between nontransparent substrates, such as iron and aluminum, which are important in tribology. In order to study various substrates, it is also necessary to develop methods for preparing smooth surfaces for the measurement. The “mica stripping method”22 uses an atomically smooth mica surface as a template for metal deposition. Metal-deposited templates are glued on the second substrate such that the deposited metal side is down, and then the mica template is removed by stripping, affording the smooth surfaces. This method was used for the preparation of gold substrates. However, we found this method was not suitable for deposited iron and aluminum because mica remained on their surfaces after stripping due to their strong adhesion to the mica. In this study, we performed the RSM based on the twin-path SFA for characterizing the nanorheological and tribological properties of lubricants, hexadecane and poly(α-olefin) (PAO), between the iron surfaces. Smooth iron surfaces for the measurement were prepared by sputtering deposition. We also characterized these lubricants confined between mica surfaces, popular substrates in most studies of confined liquids by the SFA-based shear measurements. The results demonstrated that properties of the confined liquids were different in this case depending on the kinds of lubricants and substrates. RSM employing the twin-path SFA is a powerful method for characterizing such differences.
2. EXPERIMENTS 2.1. Materials. Hexadecane (Wako) was dried over molecular sieves and distilled just prior to use to avoid the influence of water. The water content of the distilled hexadecane evaluated by Karl Fischer titration was less than 10 ppm. Poly(α-olefin), PAO (hydrogenated oligomers of 1-decene, industrial grade), was used as received. The bulk viscosity of PAO was 76 mPa·s at 25 °C. Iron (99.99%, Kojundo Chemical Laboratory) was used as the target material for the sputtering deposition. 2.2. Preparation of Iron Film by Sputtering Deposition. For preparing the iron film for the twin-path SFA, a radio frequency (RF; 13.56 MHz) magnetron sputtering apparatus (Toei Scientific Industrial) was used. Freshly cleaved mica sheets were glued onto cylindrical silica disks (radius of curvature (R), 20 mm; diameter of the disk, 10 mm) with epoxy resin (Epikote1004, Shell) and used as the substrates for the RSM. The sputtering gas was Ar (99.9995%). The back-pressure in the deposition chamber was 0.04 mN. Dt0 in hexadecane between the mica surfaces was 1.7 ± 0.4 nm for L > 0.02 mN, which was similar to reported values28,29 and was lower than that between the iron surfaces. We attempted to perform the RSM for hexadecane between the thin iron film (7 nm), however, shearing the surfaces, the solid− solid contact in air, or repeatedly applying a high load damaged the film during the measurement. Therefore, we used this method only for normal forces measurement and performed the RSM using the twinpath SFA with the thicker iron film (70 nm). 2.5. Resonance Shear Measurement (RSM). The experimental setup of the RSM17−19 based on the twin-path SFA20 is shown in
Figure 3. (a) Schematic illustration of surface separations between the surfaces, D, Dt0, Dt, and FECO images at solid−solid contact between two mica surfaces in air before (b) and after deposition (c) of iron and hard wall contact between iron surfaces in hexadecane (d). Figure 2. A pair of stiff leaf springs supported the upper surface. The lower surface was supported by double cantilever springs. The upper surface was laterally moved by a four-sectored piezo tube, which was driven by applying a sinusoidal voltage to the two opposite electrodes. The amplitude of the applied voltage was Uin, and the angular frequency was ω. The deflection of the leaf spring was detected using a capacitance probe as its output voltage (Uout). The amplitude ratio, Uout/Uin, vs ω was plotted as the resonance curve. At the beginning of each measurement, the resonance curves for the surfaces separated in air (AS) and those for a solid−solid contact (SC) were measured as the reference states. In the former curves, the peak frequency, ωAS, was characterized by the mass and the spring constant of the upper surface unit. On the other hand, the upper and lower surfaces moved identically in a strong adhesion contact when we measured the latter curves. Thus, the resonance frequency of the solid−solid contact in air, ωSC, shifted to a higher frequency due to contribution of the mass and spring constant of the lower surface units. After measuring the SC reference peak, we repeated the measurement and confirmed no change of the peak amplitude and position, indicating no surface damage under shearing. After these measurements, hexadecane was injected between the iron surfaces, and the resonance curves were measured at various surface separations (D) and the applied normal load (L). The D value was determined using the same method for the surface forces measurement. L was obtained from the deflection of the double cantilever springs.8 Twin-path SFA does not enable us to evaluate the contact area, which could be evaluated by a conventional SFA using FECO. Therefore, we plotted the RSM data vs “normal load”, not “contact pressure”, in this study. The Uout/Uin and frequency of the resonance peak ωres reflected the sample properties, such as the viscosity and friction. All measurements were done at room temperature (22 ± 2 °C).
3. RESULTS AND DISCUSSION 3.1. Surface Forces for Hexadecane between Iron Surfaces. Figure 4 shows a profile of the interactions between 3943
DOI: 10.1021/acs.langmuir.7b00148 Langmuir 2017, 33, 3941−3948
Article
Langmuir
Figure 4. Profiles of forces between the iron (▲) or mica (○) surfaces in hexadecane obtained using twin-path SFA. D was determined as the sum of the separation determined from twin-path interferometer (Dt) and hard wall thickness (Dt0) determined from FECO-SFA using the half-transparent thin iron (see Figure 3a).
the iron surfaces across hexadecane measured using the twinpath SFA. Here, the surface separation D was obtained by adding the distance Dt determined by the twin-path SFA to the surface separation Dt0 at the hard wall contact, D = Dt + Dt0, as explained in Figure 3a. Repulsion between the iron surfaces in hexadecane was observed at D < 22 nm, which could be attributed to steric forces from the structured hexadecane under confinement. It was reported that hexadecane confined between mica showed structuring repulsion at D < 4 nm.28,29 This structuring repulsion was observed at same D region in the case of hexadecane between mica surfaces in our study. On the other hand, structuring repulsion was observed at D < 22 nm in the case of hexadecane between iron surfaces in our study. This D value for the iron surfaces was larger than that for the mica surfaces (4 nm), indicating hexadecane between the iron surfaces could be more structured than that between the mica surfaces. It was reported that hexadecane confined between mica oriented to horizontal direction against substrate and showed layered structures.28 The thickness of the layered structure was 0.5 nm, suggesting that the observed structuring repulsion for hexadecane between iron at D < 22 nm corresponded to thickness of 44 layers of hexadecane. The repulsion increased with the decreasing D and the hard wall contact was found at D = 4.6 nm and F/R > 0.8 mN/m. A profile of forces between mica surfaces across hexadecane was also measured using the twin-path SFA and plotted in Figure 4. Repulsion between the mica surfaces in hexadecane was observed at D < 4 nm, and the hard wall contact was found at D = 1.7 nm and F/R > 0.3 mN/m. D at the onset of the repulsion between the iron surfaces (22 nm) was much higher than twice the PV roughness (4.8 nm), indicating that this repulsion force was not caused by a partial solid−solid contact due to the roughness but should be due to liquid structuring. Therefore, this D value for the iron surfaces was larger than that for the mica surfaces (4 nm), indicating hexadecane between the iron surfaces could be more structured than that between the mica surfaces. 3.2. Resonance Shear Measurement for Hexadecane between the Iron Surfaces. Figure 5a shows the resonance curves for the hexadecane between the iron surfaces at various separations D and applied normal loads L. D was obtained as the sum of Dt, determined using the twin-path SFA, and Dt0,
Figure 5. Resonance curves obtained for hexadecane confined between the iron (a) or mica (b) surfaces at various surface separations D and the normal loads L using twin-path SFA. D was determined as the sum of the separation determined from twin-path interferometer (Dt) and hard wall thickness (Dt0) determined from FECO-SFA using the half-transparent thin iron (see Figure 3a).
i.e., D = Dt + Dt0 (see Figure 3a). Two reference states, the curves for the separation in air (AS) and the solid−solid contact (SC), are also plotted for convenience. The resonance curve at D = 104 nm showed the peak at the resonance frequency of ωres = ca. 180 rad/s, the same frequency as the AS peak. It did not change with the decreasing D from 100 to 27 nm. The resonance peak amplitude gradually decreased at D < 24 nm, and the peak almost disappeared when D reached the hard wall contact (at D = 4.6 nm and L = 0.06 mN). This indicated that the viscosity of the confined hexadecane increased due to liquid structuring under confinement at D < 24 nm. This D region was identical to that for the onset of the steric repulsion. The resonance peaks shifted toward higher frequencies with the increasing L when D was at the hard wall contact, while the peak amplitude remained low at 0.06 mN < L < 0.3 mN. This resonance frequency shift indicated that the viscosity of the confined hexadecane further increased. However, the low resonance amplitude indicated that the friction between the surfaces was low; i.e., the hexadecane molecules were mobile under lateral shear. At L > 0.44 mN, the resonance peak amplitude increased with the increasing L. The amplitude reached a plateau at L > 1.1 mN and was ca. 20% lower than the SC peak amplitude. This indicated that the hexadecane still existed between the iron surfaces and dissipated the energy of 3944
DOI: 10.1021/acs.langmuir.7b00148 Langmuir 2017, 33, 3941−3948
Article
Langmuir shearing. This indication about the remaining hexadecane between the iron surfaces was confirmed by the thickness of the hexadecane at the hard wall contact determined using FECOSFA, 4.6 ± 0.5 nm, as described in section 2.4. Only a 20% lower amplitude of the peak at the hard wall contact than the SC peak amplitude indicated that the remaining hexadecane showed poor lubricity because it behaved solid-like. Lubricity of confined liquids was generally determined by viscosity relating with structuring and the slipping plane of liquid molecules between solid surfaces or solid−liquid interfaces. The results of FECO observation showed that hexadecane and PAO remained between the surfaces at the hard wall contact, indicating that interactions between liquid molecules and solid surfaces were stronger than interactions between liquid molecules. Thus, the large apparent viscosity of the confined liquid directly affected to the lubricity. We performed physical model analysis19 of the resonance curves of the hexadecane between iron surfaces for evaluating viscosity and the friction coefficient. The viscosity of the confined hexadecane was plotted at various D as shown in Figure S4a. It increased at D < 20 nm and reached to 1.4 × 104 mPa·s at the hard wall contact. The friction coefficient for the confined hexadecane were 0.75, which was obtained from a friction force vs load plot shown in Figure S4b. 3.3. Resonance Shear Measurement for Hexadecane between the Mica Surfaces. We performed RSM for the hexadecane confined between the mica surfaces as a reference and compared the results to that for the iron surfaces because the mica surfaces were used as substrates in most studies of confined liquids by the SFA-based shear measurements. Figure 5b shows the resonance curves at various D and L values for hexadecane between the mica surfaces obtained using the twinpath SFA. Reference peak of mica−mica contact in the air (MC) was shown. The resonance curve at D = 43 nm showed a peak at the resonance frequency ωres = ca. 170 rad/s, the same frequency as the AS peak. It did not change with the decreasing D from 43 to 15 nm. The resonance peak amplitude gradually decreased at D < 4 nm, and the peak almost disappeared when D was at the hard wall contact (L = 0.04 mN). This indicated that the viscosity of the hexadecane confined between the mica increased due to liquid structuring under confinement at D < 4 nm. This D region was identical to that for the onset of the steric repulsion. The resonance curves shifted toward higher frequencies with the decreasing L, while the peak amplitude remained low at 0.04 mN < L < 0.17 mN. This frequency shift and the low amplitude of the peak indicated that the viscosity of the confined hexadecane further increased and friction was low. At L > 0.48 mN, the resonance peak amplitude increased with the increasing L. It reached a plateau of ca. 60% of the SC peak amplitude at L > 2.1 mN, indicating that hexadecane still existed between the surfaces and showed low lubricity. In order to compare the distance at which the viscosity of the confined hexadecane started to increase, we plotted the resonance peak amplitude normalized to that of the AS peak, Uout/Uin, as a function of D for both of the mica and iron surfaces as shown in Figure 6. A sharp decrease in the amplitude of the resonance peak at the separation (AS) side was observed when the viscosity of a liquid increased by confinement.10−12 The resonance peak amplitude for hexadecane confined between the iron surfaces gradually decreased below 24 nm with the decreasing D. This indicated that the viscosity of the confined hexadecane increased due to liquid structuring under confinement below 24 nm. On the other
Figure 6. Plots of normalized resonance amplitude of the confined hexadecane as a function of surface separations, D. Plots for hexadecane between the iron or mica surfaces are indicated by triangles and circles, respectively. D was determined as the sum of the separation determined from twin-path interferometer (Dt) and hard wall thickness (Dt0) determined from FECO-SFA using the halftransparent thin iron (see Figure 3a).
hand, the resonance peak amplitude for hexadecane confined between the mica surfaces sharply decreased below 3 nm with the decreasing D. This result indicated that the viscosity of the hexadecane between the mica surfaces sharply increased below D = 4 nm, which was shorter than the value for the iron, 24 nm. The difference in the distance at which the viscosity started to increase, ca. 20 nm, was greater than twice the PV roughness, ca. 5 nm. Therefore, the observed difference should be due to different increase in the viscosity under confinement. In order to compare the lubricity of the hexadecane between those of the iron and of the mica surfaces, the resonance peak amplitude normalized to the SC peak amplitude (MC in the case of mica) with the increasing L is shown in Figure 7. An
Figure 7. Resonance peak amplitude normalized to the SC peak amplitude under various normal applied load (L). Plots for hexadecane between the iron or mica surfaces were drawn using triangles and circles, respectively.
increase in the amplitude of the resonance peak at the SC side indicated an increase in the friction. The normalized amplitude of the hexadecane between the iron surfaces sharply increased and reached a plateau of ca. 0.8 at L > 1.1 mN. On the other hand, for the hexadecane between the mica surfaces, the normalized amplitude remained lower than that for the iron surfaces at L > 0.4 mN and reached a plateau of ca. 0.6 at L > 2.1 mN. The higher normalized amplitude for the iron surfaces than that for the mica surfaces indicated that the friction between the iron surfaces was higher than that between the 3945
DOI: 10.1021/acs.langmuir.7b00148 Langmuir 2017, 33, 3941−3948
Article
Langmuir
hard wall contact, while the peak amplitude remained low at 0.03 mN < L < 0.3 mN. This resonance frequency shift and the low resonance amplitude indicated that the viscosity of the confined PAO further increased and friction was low. The amplitude reached a plateau at L > 0.3 mN and was ca. 10% compared with the SC peak amplitude. This indicated that the PAO between the iron surfaces had high lubricity. In order to compare the distance at which the viscosity of the confined PAO started to increase, we plotted the resonance peak amplitude normalized to that of the AS peak, Uout/Uin, as a function of D for both of the iron and mica surfaces as shown in Figure 9. The resonance peak amplitude for PAO confined
mica surfaces. The viscosity increase by confinement seemed more pronounced in the case of the iron surfaces (see Figure 6); thus, the higher friction for the iron surfaces well agreed with this tendency of more structuring of the hexadecane between iron surfaces. Another factor that should be taken into account was the surface roughness. The thickness of hexadecane at the hard wall contact between the iron surfaces was similar to twice the PV roughness (4.8 nm) of the iron surface, indicating that some part of the two surfaces could be in the solid−solid contact which caused some friction. It was reported that liquid molecules between rough surfaces were less structured at molecular level of thickness than those between flat surfaces.30 This suggested that solid−solid contact due to larger roughness could significantly contribute to lower lubricity between iron than that between mica. As already mentioned, hexadecane between the iron surfaces showed a viscosity increase at longer D regions, lower lubricity, and longer Dt0 than that between the mica surfaces. All of these results indicated that hexadecane between the iron surfaces was more structured than that between the mica surfaces. The plot of viscosity vs distance for hexadecane between mica surfaces is shown in Figure S4a. It increased at D < 3 nm and reached to 1.2 × 104 mPa·s at the hard wall contact. The friction coefficient for the hexadecane confined between mica was 0.42, which was obtained from a friction force vs load plot shown in Figure S4b. This value was smaller than the value from the hexadecane between iron surfaces (0.75). 3.4. Resonance Shear Measurement for PAO between the Iron Surfaces. Figure 8 shows the resonance curves for
Figure 9. (a) Plots of normalized resonance amplitude of the confined PAO as a function of surface separations, D. Plots for PAO between the iron or mica surfaces are indicated by triangles and circles, respectively. (b) Resonance peak amplitude normalized to the SC peak amplitude under various normal applied load. Plots for PAO between the iron or mica surfaces obtained using twin-path SFA were drawn using triangles and circles, respectively.
between the iron surfaces gradually decreased below 20 nm with the decreasing D. This indicated that the viscosity of the confined PAO increased due to liquid structuring under confinement below 20 nm. On the other hand, the resonance peak amplitude for PAO confined between the mica surfaces sharply decreased below 8 nm with the decreasing D. This result indicated that the viscosity of the PAO between the mica surfaces sharply increased below D = 8 nm, which was shorter than the value for the iron, 20 nm. In order to compare the lubricity of the PAO between those of the iron and of the mica surfaces, the resonance peak amplitude normalized to the SC peak amplitude with the increasing L is shown in Figure 9b. The normalized amplitude of the PAO between the iron surfaces sharply increased and reached a plateau of ca. 0.1 at L > 0.4 mN. On the other hand, for the PAO between the mica surfaces, the normalized amplitude remained higher than that for the iron surfaces at L > 0.4 mN and reached a plateau of ca. 0.3 at L > 0.3 mN. The
Figure 8. Resonance curves obtained for PAO confined between the iron surfaces at various surface separations D and the normal loads L using twin-path SFA.
the PAO between the iron surfaces at various separations D and applied normal loads L. The resonance curve at D = 216 nm showed the peak at the resonance frequency of ωres = ca. 190 rad/s, the same frequency as the AS peak. It did not change with the decreasing D from 216 to 26 nm. The resonance peak amplitude gradually decreased at D < 20 nm, and the peak almost disappeared when D reached the hard wall contact (at D = 5.0 nm and L = 0.03 mN). This indicated that the viscosity of the confined PAO increased due to liquid structuring under confinement at D < 20 nm. The resonance peaks shifted toward higher frequencies with the increasing L when D was at the 3946
DOI: 10.1021/acs.langmuir.7b00148 Langmuir 2017, 33, 3941−3948
Langmuir
■
lower normalized amplitude for the iron surfaces than that for the mica surfaces indicated that the friction between the iron surfaces was lower than that between the mica surfaces. In addition, PAO between iron surfaces showed a higher RSM peak intensity at the hard wall contact than that for the confined hexadecane. This indicated that PAO had higher lubricity than hexadecane. PAO is branched alkane, which exhibits generally less structuring under confinement due to worse packing than normal alkanes such as hexadecane. This could be the reason for higher lubricity of PAO. Hexadecane and PAO show similar trends that those between iron surfaces showed longer D regions for the viscosity increase than those between mica. In contrast, lubricity of PAO between iron surfaces was higher than that between mica surfaces, which was opposite to the case of hexadecane. These results suggested it was important to characterize the confined liquid for each substrate because substrates could show different effects on nanorheological and tribological properties of each liquid under confinement.
Article
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.7b00148. XPS results for the iron surfaces prepared by the mica template method and sputtering deposition, surface force profiles obtained using FECO-SFA, surface forces profiles between mica surfaces in hexadecane at various approaching rate, and viscosity and friction forces obtained from the physical model analysis of the resonance curves (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*Phone +81-(0)22-217-6153; Fax +81-(0)22-217-6154; e-mail
[email protected] (K.K.). ORCID
Motohiro Kasuya: 0000-0002-2324-6121 Notes
4. CONCLUSION We characterized hexadecane confined between the iron surfaces using the RSM based on the twin-path SFA, which could evaluate the nanorheological and tribological properties of a liquid confined between nontransparent surfaces. For this measurement, we used smooth iron surfaces (rms and PV roughness were 0.29 ± 0.06 and 2.4 ± 0.4, respectively) prepared by magnetron sputtering deposition. The obtained resonance curves from the hexadecane confined between the iron surfaces indicated: (1) Liquid structuring of the hexadecane confined between the iron surfaces occurred at D < 24 nm. (2) The confined hexadecane between the iron surfaces at hard wall contact showed lubricity at L < 1.1 mN. One of the problems from using the twin-path SFA for this type of study is the determination of the hard wall thickness. We used the FECO-SFA for the determination of the surface separation at hard wall contact using half-transparency thin iron films to be Dt0 = 4.6 ± 0.5 nm. This thin iron surfaces were easily damaged when the surfaces were under shear and/or under high applied loads. We characterized the hexadecane between mica surfaces which are used in most studies for studying the effect of substrates on the properties of the confined liquid. The results indicated that hexadecane between the iron surfaces showed a viscosity increase in the longer D regions, lower lubricity, and larger Dt0 than that between the mica surfaces. We also characterized a practical synthetic lubricant, poly(αolefin) (PAO), confined between iron or mica surfaces using the same methods. The results demonstrated that PAO confined between the iron surfaces showed a viscosity increase in the longer D regions (20 nm) than that between the mica surfaces (8 nm), which was same trend as the case of hexadecane. In contrast, lubricity of PAO between the iron surfaces was higher than that between mica surfaces, which was opposite to the case of hexadecane. Our study showed that the properties of confined liquids were varied by the substrates and also other factors, and the RSM based on the twin-path SFA was important for characterizing such properties which are significant factors important in many applied sciences and engineering fields such as tribology.
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the CREST program of the Japan Science and Technology Agency (JST), “Green Tribology Innovation Network” Advanced Environmental Materials Area, Green Network of Excellence (GRENE) Program and “Tohoku Innovative Materials Technology Initiatives for Reconstruction (TIMT)” sponsored by the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan.
■
REFERENCES
(1) Holmberg, K.; Andersson, P.; Erdemir, A. Global energy consumption due to friction in passenger cars. Tribol. Int. 2012, 47, 221−234. (2) Bowden, F. P.; Tabor, D. The Friction and Lubrication of Solids; Clarendon Press: Oxford, 1986. (3) Mate, M. C. Tribology on the Small Scale; Oxford University Press: Oxford, 2007. (4) Zhu, Y.; Ohtani, H.; Greenfield, M. L.; Ruths, M.; Granick, S. Modification of boundary lubrication by oil-soluble friction modifier additives. Tribol. Lett. 2003, 15, 127−134. (5) Drummond, C.; Israelachvili, J. N. Dynamic Behavior of Confined Branched Hydrocarbon Lubricant Fluids under Shear. Macromolecules 2000, 33, 4910−4920. (6) Watanabe, J.; Mizukami, M.; Kurihara, K. Resonance Shear Measurement of Confined Alkylphenyl Ether Lubricants. Tribol. Lett. 2014, 56, 501−508. (7) Xu, L.; Ogletree, D. F.; Salmeron, M.; Tang, H.; Ma, X.; Gui, J. Thickness and drainage of perfluoropolyethers under compression. J. Chem. Phys. 2001, 114, 10504−10509. (8) Israelachvili, J. N. Intermolecular and Surface Forces, 3rd ed.; Academic Press, Ltd.: New York, 2010. (9) Horn, R. G.; Israelachvili, J. N. Direct measurement of structural forces between two surfaces in a nonpolar liquid. J. Chem. Phys. 1981, 75, 1400−1411. (10) Israelachvili, J. N.; McGuiggan, P. M.; Homola, A. M. Dynamic Properties of Molecularly Thin Liquid Films. Science 1988, 240, 189− 191. (11) Gee, M. L.; McGuiggan, P. M.; Israelachvili, J. N.; Homola, A. M. J. Chem. Phys. 1990, 93, 1895−1906. (12) Yoshizawa, H.; Chen, Y. L.; Israelachvili, J. N. Fundamental Mechanisms of Interfacial Friction. 1. Relation between Adhesion and Friction. J. Phys. Chem. 1993, 97, 4128−4140. 3947
DOI: 10.1021/acs.langmuir.7b00148 Langmuir 2017, 33, 3941−3948
Article
Langmuir (13) Van Alsten, J.; Granick, S. Molecular Tribometry of Ultrathin Liquid Films. Phys. Rev. Lett. 1988, 61, 2570−2573. (14) Peachey, J.; Van Alsten, J.; Granick, S. Design of an Apparatus to Measure the Shear Response of Ultrathin Liquid Films. Rev. Sci. Instrum. 1991, 62, 463−473. (15) Klein, J.; Kumacheva, E. Confinement-Induced Phase Transitions in Simple Liquids. Science 1995, 269, 816−818. (16) Raviv, U.; Tadmor, R.; Klein, J. Shear and Frictional Interactions between Adsorbed Polymer Layers in a Good Solvent. J. Phys. Chem. B 2001, 105, 8125−8134. (17) Dushkin, C. D.; Kurihara, K. A Nanotribology of thin liquidcrystal films studied by the shear force resonance method. Colloids Surf., A 1997, 129−130, 131−139. (18) Sakuma, H.; Otsuki, K.; Kurihara, K. Viscosity and Lubricity of Aqueous NaCl Solution Confined between Mica Surfaces Studied by Shear Resonance Measurement. Phys. Rev. Lett. 2006, 96, 046104. (19) Mizukami, M.; Kurihara, K. A new physical model for resonance shear measurement of confined liquids between solid surfaces. Rev. Sci. Instrum. 2008, 79, 113705. (20) Kawai, H.; Sakuma, H.; Mizukami, M.; Abe, T.; Fukao, Y.; Tajima, H.; Kurihara, K. New surface forces apparatus using two-beam interferometry. Rev. Sci. Instrum. 2008, 79, 043701. (21) Kamijo, T.; Kasuya, M.; Mizukami, M.; Kurihara, K. Direct Observation of Double Layer Interactions between the Potentialcontrolled Gold Electrode Surfaces Using the Electrochemical Surface Forces Apparatus. Chem. Lett. 2011, 40, 674−675. (22) Hegner, M.; Wagner, P.; Semenza, G. Ultralarge atomically flat template-stripped Au surfaces for scanning probe microscopy. Surf. Sci. 1993, 291, 39−46. (23) Israelachvili, J. N. Thin Film Studies Using Multiple-Beam Interferometry. J. Colloid Interface Sci. 1973, 44, 259−272. (24) Smith, C. P.; Maeda, M.; Atanasoska, L.; White, H. S.; McClure, D. J. Ultrathin Platinum Films on Mica and the Measurement of Forces at the Platinum Water Interface. J. Phys. Chem. 1988, 92, 199− 205. (25) Ruths, M.; Heuberger, M.; Scheumann, V.; Hu, J.; Knoll, W. Confinement-Induced Film Thickness Transitions in Liquid Crystals between Two Alkanethiol Monolayers on Gold. Langmuir 2001, 17, 6213−6219. (26) Mizuno, H.; Haraszti, T.; Mizukami, M.; Kurihara, K. Nanorheology and Nanotribology of Two-Component Liquid Crystal. SAE Int. J. Fuels Lubr. 2009, 1, 1517−1523. (27) Zeng, H.; Tian, Y.; Anderson, T. H.; Tirrell, M.; Israelachvili, J. N. New SFA Techniques for Studying Surface Forces and Thin Film Patterns Induced by Electric Fields. Langmuir 2008, 24, 1173−1182. (28) Christenson, H. K.; Gruen, D. W. R.; Horn, R. G.; Israelachvili, J. N. Structuring in liquid alkanes between solid surfaces: Force measurements and mean-field theory. J. Chem. Phys. 1987, 87, 1834− 1841. (29) Bureau, L. Rate Effects on Layering of a Confined Linear Alkane. Phys. Rev. Lett. 2007, 99, 225503. (30) Capozza, R.; Fasolino, A.; Ferrario, M.; Vanossi, A. Lubricated friction on nanopatterned surfaces via molecular dynamics simulations. Phys. Rev. B: Condens. Matter Mater. Phys. 2008, 77, 235432.
3948
DOI: 10.1021/acs.langmuir.7b00148 Langmuir 2017, 33, 3941−3948
