Identification of Mobile Species in Cationic Polymer Lubricant Layer

ARTICLE pubs.acs.org/Langmuir

Identification of Mobile Species in Cationic Polymer Lubricant Layer on Silicon Oxide from AFM and XPS Analyses Erik Hsiao,† Brandon D. Veres,† Gregory J. Tudryn,‡ and Seong H. Kim*,† † ‡

Department of Chemical Engineering Department of Materials Science and Engineering Pennsylvania State University, University Park, Pennsylvania 16802, United States

bS Supporting Information ABSTRACT: The nanoscale spreading of a cationic polymer lubricant (CPL) film consisting of polydimethylsiloxane with quaternary ammonium salt side chains on a SiO2 surface was studied with the disjoining pressure measurements using atomic force microscopy. CPL shows a monotonic decrease in disjoining pressure as the film thickness increases from 1.3 to 4.5 nm, which suggests stable spreading in this thickness range. Comparing the spreading rates calculated from disjoining pressure and the viscosity of CLP to the self-healing time after tribocontacts revealed that the ionic form may not be the main mobile species. The X-ray photoelectron spectroscopy analysis found that the CPL film on SiO2 has about 30% of the quaternary ammonium salts (cationic groups) reduced to tertiary amines (neutral groups). The reduced CPL polymer has much lower viscosity than the original CPL polymer and yields a spreading rate consistent with that measured at the macroscale. Thus, the mobile component in the CPL/SiO2 film responsible for self-healing is concluded to be the reduced tertiary amine components of CPL.

I. INTRODUCTION Cationic polymer lubricant (CPL), a new bound-and-mobile boundary lubricant, consists of a low-molecular-weight poly(dimethylsiloxane) (PDMS) backbone polymer with quaternary ammonium cationic side groups.1,2 The quaternary ammonium cation can electrostatically bind to negatively charged surfaces such as silicon oxides in aqueous solution. Once dried, the deposited polymer film can act as a boundary lubricant layer. An advantage of the CPL films is that the first monolayer polymer is strongly bound to the surface to enhance the wear resistance, and at the same time, the molecules in the multilayer can be laterally mobile for self-healing of the depleted region made by tribological contacts.1,2 This can overcome many issues that current boundary lubricants face. Self-assembled monolayers (SAMs) are one-time coatings for which, once the layer is worn off, wear ensues.3
7 Hard coatings like diamond-like carbon (DLC) films have conformal issues where hidden interfaces cannot be coated.8 Fluorinated coatings such as perfluoropolyethers will degrade when coated onto silicon, releasing HF.9
12 In a tribo-test with atomic force microscopy (AFM), it was observed that, at contact pressures up to ∼5.5 GPa, the friction coefficient remains lower than 0.2, while PDMS of the same molecular weight is unable to withstand such high contact pressures giving a friction coefficient of ∼0.7. In the macroscale pin-ondisk tribo-test, the lateral spreading rate of the 3
4-nm-thick multilayer film of CPL was estimated to be ∼2  10
11 m2/s from the recovery rate of lubrication efficiency.2 These results r 2011 American Chemical Society

implied that CPL can effectively lubricate in the nanoscale as well as in the macroscale. One could predict whether CPL will be an effective lubricant at the microscale using the windshield wiper effect.13,14 In the windshield wiper effect calculations, the logistic function estimates the competition between depletion rate and the replenishment rate and predicts whether the lubricant film can be fully recovered after depletion by contact with the counter-surface. The smaller the logistic function, the higher the steady-state coverage at the contact center will be. For the conditions of the pin-on-disk tribo-test, the logistic function is ∼0.3, suggesting an efficient lubricant. For smaller-scale devices, the logistic function will be even smaller implying faster replenishment of the lubricant film via lateral flow or spreading. The stability and spreading behavior of a boundary lubricant film is governed by the disjoining pressure of the film and its gradient with respect to film thickness.15
20 The disjoining pressure is the interaction energy per area between the lubricant and the solid, which includes van der Waal interactions, electrostatic interactions, and other structural interactions.15,16 If the disjoining pressure of the film is positive, the liquid film wets a solid surface. If the gradient of the disjoining pressure with respect to film thickness is negative, then the liquid film can spread readily. Thus, the disjoining pressure of a liquid film is an Received: January 20, 2011 Revised: April 13, 2011 Published: May 02, 2011 6808

dx.doi.org/10.1021/la2002699 | Langmuir 2011, 27, 6808–6813

Langmuir important property to predict or understand the wetting behavior of the film. Many disjoining pressure studies have been conducted on the spreading and stability of perfluoropolyethers (PFPEs), which are widely used in the magnetic recording industry.12,21
25 Various methods exist to measure the disjoining pressure of films. These include thin film balance,26
31 contact angle studies,32
37 and AFM.15
19,38
41 Among these, AFM is suitable for nanoscale disjoining pressure measurements for boundary lubricant layers which are typically thinner than 5 nm. As the AFM probe contacts a liquid film, the liquid film wets the AFM probe and a meniscus forms bridging the AFM probe and the substrate. As the AFM probe is slowly retracted, the capillary force of the meniscus bends the AFM cantilever downward until the spring action of the cantilever becomes larger than the capillary force.15
18,42
44 In order for the meniscus to stretch further, more liquid film needs to flow into the meniscus that is being stretched. The capillary forces depend on two components: Laplace pressure and surface tension. The surface tension component is negligible compared to the force caused by the Laplace pressure.43,44 The nanoscale curvature of the meniscus results in an extremely large Laplace pressure. The Laplace pressure component measured with AFM can have contributions from the lubricant film deposited on the solid and the condensation of water vapor from the ambient environment. However, for hydrophobic (water contact angles >90°) polymer films such as CPL, the effect of humidity can be ignored because the polymer surface remains dry. When the contact angle of water is 90° on both the substrate and the AFM probe, then the liquid water meniscus cannot grow.44 When the vapor adsorption from the environment is negligible, then the Laplace pressure of the meniscus (PL(h)) is equal to the disjoining pressure of the lubricant film at equilibrium, as given in eq 115,17   1 1 γ þ ð1Þ ΠðhÞ ¼
PL ðhÞ ¼ γ ¼ rm ðhÞ ra ðhÞ reff ðhÞ where Π(h) is the disjoining pressure at a film thickness of h, γ is the liquid
vapor surface tension, rm(h) and ra(h) are the meridional and azimuthal radii of the meniscus curvature, respectively, and reff(h) is the effective radius of the meniscus curvature. In eq 1, rm(h) and ra(h) vary along the meniscus, but reff(h) is constant throughout the entire meniscus surface. Mate et al. demonstrated the use of AFM in calculating reff(h) for a liquid lubricant film.17 Mate measured reff(h) for perfluoropolyethers (PFPEs) films.16
18 In the Mate method, reff(h) is obtained by fitting the force
distance curve data to the following equation:   FðdÞ d ¼ 1þ ð2Þ 4πRγ 2reff ðhÞ where F(d) is the force measured by AFM, R is the radius of the AFM probe tip, and d is the tip
sample separation distance. The reff(h) is extracted from the slope of the linear portion of the retraction curve where the meniscus is stretched from the AFM tip. Bowles and White later proposed a numerical solution to calculate the shape of the retraction curve based on the shape of the meniscus with a given reff(h).15 By slowly retracting an AFM probe in contact with a liquid film, the meniscus is stretched in a quasi-equilibrium state (the stretch rate is slow enough to allow for the change in the meridional and azimuthal radii in a manner

ARTICLE

that reff(h) remains constant throughout the meniscus stretching). The White method analyzes this retraction curve based on a given reff(h) and the cantilever spring constant k.15 The White method fits the experimental data with the following equations: FðdÞ λ ¼ 2 4πRγ 4R

ð3Þ

λ¼

reff ðhÞ R

ð4Þ

R¼

reff ðhÞ ro ðdÞ

ð5Þ

where λ and R are two parameters for the numerical solution of the shape of the meniscus on the AFM probe and ro(d) is a moving boundary where the lower edge of the meniscus meets the unperturbed liquid film of thickness, h. If the surface tension (γ) of the film is known and reff(h) is determined, then the disjoining pressure for a given film thickness can be obtained from eq 1. From measurements with various film thicknesses, a disjoining pressure gradient with respect to film thickness can be obtained and related to the spreading rate of the film using the following equation: SðhÞ ¼


h3 DΠðhÞ 3η Dh

ð6Þ

where S(h) is the spreading rate of the film (m2/s) and η is the viscosity of the polymer film (Pa 3 s). In order to calculate the spreading rate, the following procedure is followed: (i) obtain η and γ for the liquid film, (ii) calculate the approximate reff(h) with the Mate method, (iii) use this reff(h) as an initial guess in the White method and refine the reff(h) value, (iv) calculate Π(h) and δΠ/∂h, and (v) calculate S(h). In this paper, the spreading rate of a cationic polymer lubricant (CPL) film on a SiO2 surface was calculated from the disjoining pressure measured with AFM. The effective radius of curvature and disjoining pressure of two CPL films with 6% and 15% of cationic monomer units were calculated. Both 6% CPL and 15% CPL show a monotonic decrease in disjoining pressure suggesting stable spreading. Spreading rate from the AFM disjoining pressure measurements was compared to the values obtained from the time-delayed macroscopic pin-on-disk measurements. Along with X-ray photoelectron spectroscopy (XPS) analysis, this comparison revealed that the mobile component in the CPL/SiO2 film responsible for self-healing is the reduced polymers containing tertiary amine (neutral group), not the polymer containing quaternary ammonium salt (cationic group).

II. EXPERIMENTAL DETAILS The synthesis of CPL molecules was previously reported.1,2,45 Two CPL molecules used in this study were 6% CPL and 15% CPL (6 mol % and 15 mol % of quaternary ammonium iodide side groups covalently attached to a 2000 g/mol PDMS polymer, respectively). The substrates used were 500-μm-thick silicon (100) wafers (elastic modulus = 160 GPa, Poisson ratio = 0.27) with amorphous native SiO2 layer. The silicon wafers were cleaned with an RCA-1 process.1 The CPL films were deposited on the substrate by spin-coating 0.5
2.5% CPL solution in 9:1 water/ethanol mixtures at a spin speed of 3000 rpm. Film thicknesses in the range 1.3
4.5 nm of both 6% CPL and 15% CPL were used to 6809

dx.doi.org/10.1021/la2002699 |Langmuir 2011, 27, 6808–6813

Langmuir

Figure 1. Representative retraction curve for 1.7-nm-thick 6% CPL film on a silicon oxide. Both the Mate method and the White method to calculate reff(h) are shown. obtain the spreading rates on the nanoscale. An ellipsometer (Stokes LSE, wavelength = 632.8 nm, incidence angle = 70°) was used to measure the average film thickness. The refractive index of PDMS (1.45) was used to calculate the film thickness. AFM (Molecular Imaging Pico-SPM microscope with a RHK SPM 100 controller) was used to obtain force
distance curves. The AFM probe was in contact with the CPL film for at least 30 min to allow the film to fully wet the probe and ensure that all subsequent retractions were at equilibrium. The retraction rates used to obtain a quasiequilibrium stretching of the menisci were 0.5 nm/s to ensure that reff(h) is constant. The tips used were rectangular silicon cantilevers with spring constants ranging 2.0
5.5 N/m with a tip cone angle of 40° (VistaProbe FM25) and resonance frequencies ranging 85
120 kHz. The radius of the AFM tip used for all calculations was estimated to be about 100 nm. All AFM tips used were calibrated with the Sader method for normal force.46 Both the Mate method and the White method were used to calculate reff(h). The reff(h) values of 6 different film thicknesses were measured for 6% CPL and 15% CPL. The reported reff(h) is an average of 9 measurements at each thickness. XPS (Kratos Analytical Axis Ultra spectrometer) was used to investigate the chemical status of N 1s in the CPL film. The photon energy was 1486.7 eV from a monochromatic Al KR source. Highresolution scans were conducted with a pass energy of 40 at 0.3 eV scan step size. Rheologic responses of the CPL bulk polymers were measured in the linear viscoelastic regime using a rheometer (Rheometric Scientific ARES-LS1) with 25.0 mm parallel plates for strain-controlled steady shear and oscillatory shear. Samples were molded in the instrument by heating to 50 °C under nitrogen to remove voids and water and to ensure precise geometry and proper contact. Environmental control was maintained within 0.1 °C of set point values using a nitrogen feed furnace.

III. RESULTS The reff(h) of the meniscus being stretched can be estimated by analyzing the attractive force exerted by the meniscus to the cantilever during the tip retraction from the surface. Figure 1 shows a representative retraction curve for a 1.7-nm-thick 6% CPL film on a silicon wafer at a retraction rate of 0.5 nm/s. The force was normalized with the pull-off force (F(0)) measured at point a where the AFM tip is separated from the silicon wafer surface, but the tip is still pulled by the polymer meniscus. As the cantilever continues retracting from the surface, the meniscus is stretched further. Point b is where the meniscus stretching finishes and the AFM probe tip is released to the free-standing position. In the Mate method, eq 2, the slope of the linear region between points a and b is used to calculate reff(h). For the White

ARTICLE

Figure 2. reff(h) for 6 different film thicknesses of (a) 6% CPL and (b) 15% CPL calculated with both the Mate method and the White method.

Figure 3. Disjoining pressure for 6 different film thicknesses of (a) 6% CPL and (b) 15% CPL calculated from reff(h) and γ. The power law fitting to calculate the disjoining pressure gradient with film thickness is also included.

method, a numerical algorithm15 for the conical tip geometry simulates the complete meniscus stretching profile as a function of tip
sample separation based on an initial guess of reff(h). Figure 1 illustrates the difference between the reff(h) measured with the Mate method (red straight line) and the White method (blue curve). Figure 2 plots the reff(h) values for 6 different film thicknesses of 6% CPL and 15% CPL. As described in ref 15, the White method gives a lower reff(h) value than the Mate method as the film thickness increases. For 15% CPL, the reff(h) calculated from the Mate method is the same as the reff(h) calculated from the White method for films thinner than 2 nm. For 15% CPL films thicker than 2 nm, the reff(h) calculated from the Mate method is smaller than that calculated from the White method. For thicker 6% CPL films (>2 nm), the reff(h) calculated from the Mate method is also smaller than the reff(h) calculated from the White method. The disjoining pressure of the CPL film can be calculated from eq 1 once reff(h) and γ are determined. The water contact angle for CPL films is the same as the water contact angle for PDMS.1 The liquid
vapor surface tension of PDMS is 21 mN/m. Since the water contact angle of CPL film is the same as that for PDMS, the same liquid
vapor surface tension of PDMS could be used to estimate the disjoining pressure of the CPL film. The surface tension of CPL may not be exactly the same as that for PDMS, but a small difference in surface tension will not significantly change the calculations. The surface tensions of other quaternary ammonium halides containing siloxane surfactant compounds similar to CPL were found to range from 20 to 24 mN/m.47 Using the reff(h) values from Figure 2, the disjoining pressures were calculated and are shown in Figure 3. It should be noted that the disjoining pressures are positive, indicating that the film is stable and does not dewet. Although a bimodal thickness distribution of the multilayer of CPL was previously observed 6810

dx.doi.org/10.1021/la2002699 |Langmuir 2011, 27, 6808–6813

Langmuir

ARTICLE

Table 1. Summary of Viscosities and Spreading Rates for 3-nm-thick Films quaternary ammonium (Nþ)

tertiary amine (N0)

S [3 nm] method 6% CPL
Mate

η (Pa-s) 610

S [3 nm]

(m2/s)

η (Pa-s)

(m2/s)

2.3  10
15

0.024

6.0  10
11


15

6% CPL
White 15% CPL
Mate

610 28 000

1.6  10 4.0  10
17

0.024 0.034

4.0  10
11 3.4  10
11

15% CPL
White

28 000

2.8  10
17

0.034

2.4  10
11

Figure 4. Zero shear rate viscosity measurements for bulk 6% CPL, bulk 15% CPL, reduced 6% NCPL, and reduced 15% NCPL.

with noncontact AFM imaging previously,1 it was not the consequence of polymer dewetting from the solid surface. The strongly bound monolayer is always present and covers the solid surface. The disjoining pressures for both 6% CPL and 15% CPL films on SiO2 decrease monotonically as the film thickness increases. This indicates that both CPL films have a stable spreading behavior or flow laterally into the depleted region. In order to obtain the disjoining pressure gradient with respect to film thickness, the measured disjoining pressure data are fitted to a power law15 Π ¼ ahb

ð7Þ

where a and b are fitting parameters. The fit results are also included in Figure 3. Using eqs 6 and 7, the spreading rate for CPL films can be calculated as follows: SðhÞ ¼
ab

h2 þ b 3η

ð8Þ

For S(h) calculation in eq 8, the viscosity of the CPL film is needed. Since the viscosity of a few-nanometer-thick films could not be measured, one can use the viscosity of the bulk liquid as an estimate. The viscosity measured at varying shear rates for 6% CPL and 15% CPL bulk liquids is shown in Figure 4. The viscosity of 6% CPL is 610 Pa 3 s and that of 15% CPL is 28 000 Pa 3 s. At a shear rate higher than ∼10 s
1, the torque due to the polymer viscosity exceeded the instrumental limit. If these values are used in eq 8, the spreading rates for 3-nmthick 6% CPL and 15% CPL films are estimated to be ∼2  10
15 m2/s and ∼3  10
17 m2/s, respectively (Table 1). If the surface
polymer interaction increases, the effective viscosity of the polymer film, then the spreading rate is expected to be even lower. These extremely slow spreading rates calculated from the AFM measurements with the bulk CPL viscosities are not in agreement with those determined on the macroscale at the same film thickness, which are in the range of ∼10
11 m2/s for ∼3 nm thick 6% CPL and 15% CPL films.2 On the macroscale, it was observed that the CPL lubricant film can flow into a ∼50-μmwide depletion zone within ∼32 s.2 If the spreading rate is ∼10
15 m2/s for 6% CPL, it would take approximately 1 week to spread into a 50-μm-wide depleted zone. With a spreading rate of ∼10
17 m2/s for of 15% CPL, it would take roughly 2 years to spread into a 50-μm-wide linear zone. Similarly, the windshield wiper effect model predicts that the logistic function was given as 4.4 and 6.3 for 6% CPL and 15% CPL, respectively, implying that

Figure 5. High-resolution N 1s XPS spectra of 6% CPL and 15% CPL coated onto silicon oxide. There is a 30% reduction from the Nþ state to the N0 state.

the CPL film in the contact depletion region will not be recovered.13,14 Thus, the estimated spreading rate from the AFM measurements is too low to explain the results observed in the macroscopic tribo-test. This discrepancy implies that the viscosity of the spreading layer must be lower than that of the bulk CPL fluid. The lower viscosity would be possible when the quaternary ammonium groups in the CPL film are reduced to tertiary amine groups. The electronic state of nitrogen group in the side chain of CPL was analyzed with XPS. The CPL films were placed in a UHV chamber prior to the XPS chamber to avoid any CPL from contaminating the detectors. For the 6% CPL, the measured film thickness was 1.4 nm, and for the 15% CPL, it was 1.8 nm. The high-resolution N 1s XPS spectra are shown in Figure 5. There are 2 main peaks at 403.5 and 398.1 eV, which correspond to Nþ and N0 states of CPL, respectively. It should be noted that 1H NMR analysis of bulk CPL liquids shows that the conversion of the amine (N0) side group to the quaternary ammonium (Nþ) is 100% (data in Supporting Information).1 No degradation of the ammonium salt is observed upon normal aging in ambient conditions or dissolution in the spin-coating solutions (FTIR and NMR data in the Supporting Information). However, the N 1s XPS analysis of the deposited CPL films reveals that there is an approximate 30% reduction of quaternary ammonium (Nþ) iodide to a neutral tertiary amine (N0). It should be noted that the bound layer only (∼0.8-nm-thick films) also showed about 30% reduction.1 The mechanism for this reduction is unknown; but it must be induced by the interactions with SiO2 surface or a genuine property of the CPL liquid/air interface. Various quaternary ammonium halides containing siloxane surfactant compounds have also shown a small degree of reduction in water.47
49 It was also reported that some halide counterions could be coordinated by multiple nitrogen groups or replaced by water molecules.47,49 Similar reduction processes might take 6811

dx.doi.org/10.1021/la2002699 |Langmuir 2011, 27, 6808–6813

Langmuir

ARTICLE

place during the deposition of CPL (spin-coating from aqueous solution). The neutral polymer liquid, CPL-precursor before ionization with CH3I, is not as viscous as CPL; so, the “reduced” neutral polymer (called NCPL hereafter) chains can spread notably more rapidly than the cationic ones (CPL). The viscosities of the 6% NCPL and 15% NCPL bulk liquids were measured and are also shown in Figure 4. The viscosity of 6% NCPL is 0.024 Pa 3 s and that of 15% NCPL is 0.034 Pa 3 s. Due to low viscosity, the torque upon shearing of the NCPL liquid could be measured at shear rates higher than ∼30 s
1. At lower shear rates, the torque caused by NCPL shearing was lower than the sensitivity of the instrument. Since their viscosities are much lower than those of CPL, it can be assumed that the NCPL chains present in the film are responsible for lateral spreading observed in the macroscopic self-healing tests. When the viscosity of the NCPL polymers is used in eq 8, the spreading rates for 3-nm-thick 6% CPL and 15% CPL films containing NCPL are calculated to be ∼5  10
11 m2/s and ∼3  10
11 m2/s, respectively; data are summarized in Table 1. Once again, by implementing the windshield wiper effect,13,14 the logistic function was calculated to be 0.2 and 0.4 for 6% CPL and 15% CPL, respectively. These values imply that the CPL film would be almost fully recovered as evidenced in the macroscale pin-on-disk tribo-test. These spreading rates are in good agreement with those measured with the macro-scale pinon-disk tribometer. Thus, these results indicate that the reduction of the quaternary ammonium cation, which is ∼30% in the pristine film state, is responsible for the lateral spreading and selfhealing of CPL films.

IV. CONCLUSION The spreading rate calculations from the disjoining pressure measured with AFM and the partial reduction of the quaternary ammonium ions found with XPS suggest that the mobile species in the bound-and-mobile lubricant layer consisting of PDMS with cationic side changes (CPL) are the polymer chains or segments that contain tertiary amine groups rather than quaternary ammonium cations.50,51 The viscosity of the CPL liquid seems to be too high to readily flow into the lubricant depleted region. About ∼30% of the ammonium ions are reduced into tertiary amines during or after the film deposition. The reduction mechanism is not known. When the viscosity of the tertiaryamine-containing PDMS liquid (precursor of CPL) is used, the spreading rate calculated from the disjoining pressure measurements is in agreement with the value estimated from macroscopic tribo-tests. ’ ASSOCIATED CONTENT

bS

1

H NMR and FT-IR of NCPL, CPL - bulk, and CPL in solution. This material is available free of charge via the Internet at http://pubs.acs.org. Supporting Information.

’ AUTHOR INFORMATION Corresponding Author

*Author e-mail: [email protected].

’ ACKNOWLEDGMENT This work was supported by the National Science Foundation (Grant No. CMS-0528141 and CMS-0637028). The authors

appreciate the help with viscosity measurements from Dr. Ralph H. Colby. The authors gratefully acknowledge Dr. Adam Bowles and Dr. Lee R. White for providing Mathematica programs to analyze the shape of the meniscus during the AFM retraction.

’ REFERENCES (1) Hsiao, E.; Kim, D.; Kim, S. H. Langmuir 2009, 25 (17), 9814–9823. (2) Hsiao, E.; Bradley, L. C.; Kim, S. H. Tribol. Lett. 2011, 41 (1), 33–40. (3) Ashurst, W. R.; Yau, C.; Carraro, C.; Maboudian, R.; Dugger, M. T. J. Microelectromech. Syst. 2001, 10 (1), 41–49. (4) Jun, Y.; Zhu, X. Y. J. Adhes. Sci. Technol. 2003, 17 (4), 593–601. (5) Linford, M. R.; Fenter, P.; Eisenberger, P. M.; Chidsey, C. E. D. J. Am. Chem. Soc. 1995, 117 (11), 3145–3155. (6) Romig, A. D.; Dugger, M. T.; McWhorter, P. J. Acta Mater. 2003, 51 (19), 5837–5866. (7) Salmeron, M. Tribol. Lett. 2001, 10 (1
2), 69–79. (8) Kim, S. H.; Asay, D. B.; Dugger, M. T. Nano Today 2007, 2, 22–29. (9) Eapen, K. C.; Patton, S. T.; Smallwood, S. A.; Nainaparampil, J. J.; Zabinski, J. S. Surf. Coat. Technol. 2005, 197 (2
3), 270–277. (10) Kasai, P. H. J. Info. Stor. Proc. Syst. 1999, 1 (1), 23–31. (11) Smallwood, S. A.; Eapen, K. C.; Patton, S. T.; Zabinski, J. S. Wear 2006, 260 (11
12), 1179–1189. (12) Xu, L.; Ogletree, D. F.; Salmeron, M.; Tang, H. A.; Gui, J.; Marchon, B. J. Chem. Phys. 2000, 112 (6), 2952–2957. (13) Abdelmaksoud, M.; Lee, S. M.; Padgett, C. W.; Irving, D. L.; Brenner, D. W.; Krim, J. Langmuir 2006, 22 (23), 9606–9609. (14) Brenner, D. W.; Irving, D. L.; Kingon, A. I.; Krim, J. Langmuir 2007, 23 (18), 9253–9257. (15) Bowles, A. P.; Hsia, Y. T.; Jones, P. M.; Schneider, J. W.; White, L. R. Langmuir 2006, 22 (26), 11436–11446. (16) Mate, C. M. J. Appl. Phys. 1992, 72 (7), 3084–3090. (17) Mate, C. M.; Lorenz, M. R.; Novotny, V. J. J. Chem. Phys. 1989, 90 (12), 7550–7555. (18) Mate, C. M. Phys. Rev. Lett. 1992, 68 (22), 3323–3326. (19) Fukuzawa, K.; Kawamura, J.; Deguchi, T.; Zhang, H. D.; Mitsuya, Y. J. Chem. Phys. 2004, 121 (9), 4358–4363. (20) Ma, X.; Gui, J.; Smoliar, L.; Grannen, K.; Marchon, B.; Jhon, M. S.; Bauer, C. L. J. Chem. Phys. 1999, 110 (6), 3129–3137. (21) Ma, M. L.; Hill, R. M. Curr. Opin. Colloid Interface Sci. 2006, 11 (4), 193–202. (22) Ma, X.; Gui, J.; Smoliar, L.; Grannen, K.; Marchon, B.; Bauer, C. L.; Jhon, M. S. Phys. Rev. E 1999, 59 (1), 722–727. (23) Waltman, R. J.; Khurshudov, A.; Tyndall, G. W. Tribol. Lett. 2002, 12 (3), 163–169. (24) Waltman, R. J.; Tyndall, G. W.; Wang, G. J.; Deng, H. Tribol. Lett. 2004, 16 (3), 215–230. (25) Waltman, R. J.; Zhang, H.; Khurshudov, A.; Pocker, D.; Karplus, M. A.; York, B.; Xiao, Q. F.; Zadoori, H.; Thiele, J. U.; Tyndall, G. W. Tribol. Lett. 2002, 12 (1), 51–60. (26) Claesson, P. M.; Ederth, T.; Bergeron, V.; Rutland, M. W. Adv. Colloid Interface Sci. 1996, 67, 119–183. (27) Nikolov, A. D.; Wasan, D. T. Langmuir 1992, 8 (12), 2985–2994. (28) Stubenrauch, C.; Schlarmann, J.; Strey, R. Phys. Chem. Chem. Phys. 2002, 4 (18), 4504–4513. (29) Bhakta, A.; Ruckenstein, E. Adv. Colloid Interface Sci. 1997, 70, 1–124. (30) Kolaric, B.; Jaeger, W.; von Klitzing, R. J. Phys. Chem. B 2000, 104 (21), 5096–5101. (31) Vaynblat, D.; Lister, J. R.; Witelski, T. P. Phys. Fluids 2001, 13 (5), 1130–1140. (32) Leger, L.; Joanny, J. F. Rep. Prog. Phys. 1992, 55 (4), 431–486. (33) Sharma, A.; Reiter, G. J. Colloid Interface Sci. 1996, 178 (2), 383–399. 6812

dx.doi.org/10.1021/la2002699 |Langmuir 2011, 27, 6808–6813

Langmuir

ARTICLE

(34) Girifalco, L. A.; Good, R. J. J. Phys. Chem. 1957, 61 (7), 904–909. (35) Duncanhewitt, W. C.; Thompson, M. Anal. Chem. 1992, 64 (1), 94–105. (36) Eres, M. H.; Schwartz, L. W.; Roy, R. V. Phys. Fluids 2000, 12 (6), 1278–1295. (37) Schwartz, L. W.; Eley, R. R. J. Colloid Interface Sci. 1998, 202 (1), 173–188. (38) Saramago, B. Curr. Opin. Colloid Interface Sci. 2010, 15 (5), 330–340. (39) Basu, S.; Sharma, M. M. J. Colloid Interface Sci. 1996, 181 (2), 443–455. (40) Chan, D. Y. C.; Dagastine, R. R.; White, L. R. J. Colloid Interface Sci. 2001, 236 (1), 141–154. (41) Neto, C.; Jacobs, K. Physica A 2004, 339 (1
2), 66–71. (42) Asay, D. B.; De Boer, M. P.; Kim, S. H. J. Adhes. Sci. Technol. 2010, 24, 2363–2382. (43) Butt, H. J.; Cappella, B.; Kappl, M. Surf. Sci. Rep 2005, 59 (1
6), 1–152. (44) Hsiao, E.; Marino, M. J.; Kim, S. H. J. Colloid Interface Sci. 2010, 352 (2), 549–557. (45) Sabourault, N.; Mignani, G.; Wagner, A.; Mioskowski, C. Org. Lett. 2002, 4 (13), 2117–2119. (46) Sader, J. E.; Chon, J. W. M.; Mulvaney, P. Rev. Sci. Instrum. 1999, 70 (10), 3967–3969. (47) Schmaucks, G.; Sonnek, G.; Wustneck, R.; Herbst, M.; Ramm, M. Langmuir 1992, 8 (7), 1724–1730. (48) Ramm, M.; Schulz, B.; Sonnek, G.; Schmaucks, G. Cryst. Res. Technol. 1990, 25 (7), 763–769. (49) Vanrysselberghe, P.; McGee, J. M. J. Am. Chem. Soc. 1945, 67 (6), 1039–1040. (50) Colby, R. H.; Zheng, X.; Rafailovich, M. H.; Sokolov, J.; Peiffer, D. G.; Schwarz, S. A.; Strzhemechny, Y.; Nguyen, D. Phys. Rev. Lett. 1998, 81 (18), 3876–3879. (51) Leibler, L.; Rubinstein, M.; Colby, R. H. Macromolecules 1991, 24 (16), 4701–4707.

6813

dx.doi.org/10.1021/la2002699 |Langmuir 2011, 27, 6808–6813