1484
Langmuir 2008, 24, 1484-1488
Nanotribology of Surface-Grafted PEG Layers in an Aqueous Environment† Tanja Drobek and Nicholas D. Spencer* Laboratory for Surface Science and Technology, Department of Materials, ETH Zurich, Wolfgang-Pauli-Strasse 10, CH-8093 Zurich, Switzerland ReceiVed July 28, 2007. In Final Form: September 5, 2007 The lubrication properties of adsorbed poly(L-lysine)-graft-poly(ethylene glycol) in aqueous buffer solution were studied with the surface forces apparatus. In general, the polymer brushes revealed extremely low friction forces. Two distinct regimes could be identified. In response to lateral shear, the friction forces of intact polymer films at moderate loads were below the detection limit. At high loads, when the films were compressed to about 10% of the original equilibrium film thickness, the friction showed a reversible increase with load. Under certain conditions, film destruction was observed, immediately followed by a dramatic increase in the frictional force and an expansion of the adsorbed brush layer. By the addition of free polymer to the buffer solution, the resistance of the polymer brushes to abrasion was dramatically increased by readsorption of the polymer following friction-induced desorption. This self-healing capacity and the extremely low friction of the adsorbed copolymer films contribute to their excellent properties as lubricant additives for water-based lubrication under boundary conditions.
Introduction Surface-grafted, brushlike polymers can dramatically modify the lubricious properties of surfaces.1-3 The ability to bind a significant amount of solvent in a surface layer is thought to be one of the key mechanisms for low-friction, polymer-brush films. A brush composed of water-soluble, biocompatible polymers, such as poly(ethylene glycol), in an aqueous environment can provide an oil-free, environmentally friendly, food-compatible lubricious surface. Poly(L-lysine)-graft-poly(ethylene glycol) (PLL-g-PEG) is a water-soluble copolymer consisting of a poly(L-lysine) backbone and poly(ethylene glycol) side chains.4 The PLL chain carries multiple positive charges and spontaneously adsorbs onto negatively charged surfaces, such as many metal oxides or mica. This leads to a straightforward, reliable method for coating surfaces with a dense PEG brush simply by immersing them in an aqueous PLL-g-PEG solution.5,6 The adsorption of PLL-gPEG onto a substrate is primarily controlled by the balance between the attractive electrostatic interaction of the backbone and the steric repulsion of the PEG side chains.5 In macroscopic tribological contacts, layers of PLL-g-PEG can reduce the coefficient of friction under sliding conditions and even more effectively under rolling conditions.7,8 As long as the polymer is present in the solution, damage to the film is easily repaired
by readsorption.9,10 This self-healing capacity makes the polyelectrolyte-anchoring approach attractive for tribological applications in which the contacting surfaces are immersed in a reservoir of lubricant solution. In previous studies, we have investigated the properties of PLL-g-PEG films with the surface forces apparatus (SFA) under compression.11-13 The molecules form brushlike homogeneous films on the surface, exhibiting predominantly repulsive, nearly elastic interaction forces upon compression. The equilibrium film thickness is dependent on the polymer architecture, adsorption conditions, and temperature. A comparison of brushbrush and brush-hard wall experiments revealed a significant overlap of the two opposing films. In this article, we report an investigation of the tribological properties of PLL-g-PEG films carried out with a surface forces apparatus. Although the conditions applied in the experiments (moderate surface pressures, low velocities) are far removed from those of many practical macrotribological applications, they help to elucidate the underlying mechanisms of film structure, lubrication, and repair. Experimental Section
Part of the Molecular and Surface Forces special issue. * Corresponding author. Tel: +41 44 63 25850. Fax: +41 44 63 31027. E-mail:
[email protected].
Mica Samples. Thin mica sheets (1.5-5 µm) were prepared by the manual cleavage of optical-quality ruby mica blocks (Spruce Pine Mica Company, Spruce Pine, NC). To avoid the presence of nanoparticles,14-16 the individual pieces were cut with surgical scissors. The mica samples were covered on one side with a 30-40 nm silver film by thermal evaporation. To glue the mica to cylindrical
(1) Klein, J. Ann. ReV. Mater. Sci. 1996, 26, 581-612. (2) Raviv, U.; Frey, J.; Sak, R.; Laurat, P.; Tadmor, R.; Klein, J. Langmuir 2002, 18, 7482-7495. (3) Lee, S.; Spencer, N. D. Achieving Ultralow Friction by Aqueous, BrushAssisted Lubrication. In Superlubricity; Erdemir A., Jean-Michel Martin, J.-M., Eds.; Elsevier: Amsterdam, 2007; Chapter 21. (4) Sawhney, A. S.; Hubbell, J. A. Biomaterials 1992, 13, 863-870. (5) Kenausis, G. L.; Vo¨ro¨s, J.; Elbert, D. L.; Huang, N. P.; Hofer, R.; RuizTaylor, L.; Textor, M.; Hubbell, J. A.; Spencer, N. D. J. Phys. Chem B 2000, 104, 3298-3309. (6) Huang, N. P.; Michel, R.; Voros, J.; Textor, M.; Hofer, R.; Rossi, A.; Elbert, D. L.; Hubbell, J.; Spencer, N. D. Langmuir 2001, 17, 489-498. (7) Lee, S.; Mu¨ller, M.; Ratoi-Salagean, M.; Voros, J.; Pasche, S.; De Paul, S. M.; Spikes, H. A.; Textor, M.; Spencer, N. D. Tribol. Lett. 2003, 15, 231-239. (8) Mu¨ller, M.; Lee, S.; Spikes, H. A.; Spencer, N. D. Tribol. Lett. 2003, 15, 395-405.
(9) Mu¨ller, M. Aqueous Lubrication by Means of Surface-Bound Brush-Like Copolymers. Doctoral Thesis No. 16030, Swiss Federal Institute of Technology (ETH) Zu¨rich, Zu¨rich, Switzerland, 2005. (10) Lee, S.; Mu¨ller, M.; Heeb, R.; Zu¨rcher, S.; Tosatti, S.; Heinrich, M.; Amstad, F.; Pechmann, S.; Spencer, N. D. Tribol. Lett. 2006, 24, 217-223. (11) Heuberger, M.; Drobek, T.; Spencer, N. D. Biophys. J. 2005, 88, 495504. (12) Heuberger, M.; Drobek, T.; Vo¨ro¨s, J. Langmuir 2004, 20, 9445-9448. (13) Drobek, T.; Spencer, N. D.; Heuberger, M. Macromolecules 2005, 38, 5254-5259. (14) Ohnishi, S.; Hato, M.; Tamada, K.; Christenson, H. K. Langmuir 1999, 15, 3312-3316. (15) Kohonen, M. M.; Meldrum, F. C.; Christenson, H. K. Langmuir 2003, 19, 975-976. (16) Heuberger, M.; Za¨ch, M. Langmuir 2003, 19, 1943-1947.
†
10.1021/la702289n CCC: $40.75 © 2008 American Chemical Society Published on Web 10/16/2007
Nanotribology on Surface-Grafted PEG Layers glass lenses of 20 mm radius, the glass was spin coated with a film of pure epoxy resin (EPON 1004, Shell Chemicals) and heated to 140 °C. The mica sheets were glued with the silver-coated side on the glass lenses. After mounting into the SFA and purging for at least 1 h with dry nitrogen, the thickness of the mica samples was determined in a dry nitrogen atmosphere in unloaded adhesive contact prior to the adsorption of the polymer. Polymer Film. The molecules used in this study were synthesized and characterized according to standardized procedures, which have been previously described in great detail.6,17,18 In the experiments described here, PLL(20)-g[3.5]-PEG(2) and PLL(20)-g[2.9]-PEG(5) were used. The molecular weight of the backbone as a hydrobromide salt and the molecular weight of the side chains are in parentheses, and the grafting (Lys to PEG) ratio is in square brackets. A chain of 96 lysine units was used as a backbone. Yielding grafting ratios of 2.9 and 3.5, 32 grafted PEG 44-mers (PEG(2)) or 27 grafted PEG 114-mers (PEG(5)) were used as side chains, respectively. All PEGs were terminated with methoxy groups. The polymer was dissolved in 10 mM 4-[2-hydroxyethyl]piperazine-1-[2-ethanesulfonic acid] (HEPES) buffer (Fluka, Switzerland), adjusted to pH 7.4 with NaOH. After determining the thickness of the mica sheets in the SFA, the samples were demounted and completely immersed in solutions containing 0.5 mg/mL PLLg-PEG at room temperature for more than 30 min. After removal from the polymer solution, the samples were thoroughly rinsed with a jet of ultrapure water (Fluka, Switzerland) and mounted in the SFA with a drop of either HEPES buffer or polymer solution placed between the mica surfaces. In one experiment, the surfaces were rinsed with 150 mM KCl, followed by rinsing with 10 mM HEPES buffer. The surfaces were mounted with a drop of polymer solution between them prior to the friction experiments. Surface Forces Apparatus. The experiments were carried out with a modified version of the Mk 3 SFA (Surforce, Santa Barbara, CA).19-21 The instrument is equipped with a fully automated data acquisition and evaluation program based on the fast spectral correlation algorithm and is mounted in a box with precise temperature control.22 All experiments were carried out at 25 °C. For the friction experiments, the lower surface was moved by means of a piezo bimorph with an amplitude of 50 µm and velocities in the range of 0.5 to 5 µm/s. The bimorph was mounted on a spring with a spring constant of 551 N/m. The shear forces acting on the upper surface were measured with a strain gauge mounted on a measuring spring (for technical details see Luengo et al.23) with a measuring rate of 300 Hz. The approach control mechanism of the SFA contains a stepper motor under computer control and a manual micrometer screw. A schematic view of the experimental setup is shown in Figure 1. An experimental setup with only a drop of liquid was necessary to avoid immersing the piezo and the strain gauge in the liquid. Low-load approach and retraction measurements (up to 4 mN) were carried out by continuous motor travel, in part simultaneously with the friction measurements. For the high-load experiments described below, a manual approach with the micrometer screw was employed. The average contact pressure was calculated by dividing the load by the contact area. This latter parameter changed considerably during loading because of surface deformation, which varied from experiment to experiment as a result of differences in the thickness of mica and glue. Because of this deformation, the (17) Pasche, S.; De Paul, S. M.; Vo¨ro¨s, J.; Spencer, N. D.; Textor, M. Langmuir 2003, 19, 9216-9225. (18) Pasche, S. Mechanisms of Protein Resistance of Adsorbed PEG-Graft Copolymers. Doctoral Thesis No. 15712, Swiss Federal Institute of Technology (ETH) Zu¨rich, Zu¨rich, Switzerland, 2004. (19) Israelachvili, J. N.; McGuiggan, P. M. J. Mater. Res. 1990, 5, 22232231. (20) Heuberger, M. ReV. Sci. Instrum. 2001, 72, 1700-1707. (21) Za¨ch, M.; Heuberger, M. ReV. Sci. Instrum. 2003, 74, 260-266. (22) Heuberger, M.; Vanicek, J.; Za¨ch, M. ReV. Sci. Instrum. 2001, 72, 35563560. (23) Luengo, G.; Schmitt, F. J.; Hill, R.; Israelachvili, J. Macromolecules 1997, 30, 2482-2494.
Langmuir, Vol. 24, No. 4, 2008 1485
Figure 1. Schematic drawing of the SFA setup for friction measurements. The two mica surfaces with the grafted polymer films are separated by a drop of aqueous polymer solution. The distance between the two surfaces is measured by means of thinfilm interferometry.
Figure 2. Compression isotherms measured upon approach of two opposing layers of PLL(20)-g[2.9]-PEG(5) in HEPES buffer (10 mM, pH 7.4). One data set (filled symbols) was taken without lateral movement of the surfaces, whereas a shear motion of 5 µm/s and an amplitude of 50 µm were applied during the measurement of the other data set (open symbols). pressure does not increase linearly with load but shows a substantial lower-order dependency. The drop configuration used in the experiments led to additional mechanical drifts of the surfaces as a result of the evaporation of the solvent. For this reason, approach and retract measurements had to be carried out at higher velocities (typically 10 nm/s) than in previously reported experiments, in which the surfaces were immersed in a bath of conductivity water.11-13
Results In Figure 2, compression isotherms measured on two surfaces with adsorbed PLL(20)-g[2.9]-PEG(5) are shown. Data were taken upon approaching the surfaces with 10 nm/s, once without shearing and once with a shear rate of 5 µm/s. The difference between the two data sets lies within the scatter of this kind of experiment, which is caused by changes in the mechanical drift due to the evaporation of the solvent drop. In this set of experiments as well as in the others reported in this article, no change in the film thickness caused by the shear motion was observed. The equilibrium film thickness is slightly higher than previously reported in a publication for a comparable copolymer.13 This difference is caused by a small difference in the grafting ratio and the use of 10 mM HEPES buffer instead of conductivity water. Both of these parameters affect the amount of polymer adsorbing onto the surface. The amount is determined by the balance between the electrostatic interaction of the polyelectrolyte backbone with the oppositely charged surface of the mica and the steric repulsion of the PEG side chains, which are forced into a brushlike configuration. During the compression isotherms, refractive index data were collected. The refractive index, n(D), is modeled by2
1486 Langmuir, Vol. 24, No. 4, 2008
n(D) ) nH2O +
mζ(nPEG + nH2O) D
Drobek and Spencer
for D > ζ
(1)
Here, nH2O ) 1.337 is the refractive index of the aqueous solution, and nPEG ) 1.51 is the refractive index of the dry PEG polymer. The prefactor m ) 2 accounts for the brush-brush configuration. The equivalent dry thickness ζ ) Γ/F of a dried polymer layer is the ratio of the adsorbed polymer mass, Γ, and the dry polymer density, F ≈ 1.12 g/cm3. Fitting the data shown in Figure 3 reveals a dry polymer thickness of 2.6 nm, which corresponds to an adsorbed mass of 290 ( 100 ng/cm2. This implies an average distance of grafting points of about dg ) 3.5 nm and an average PEG monomer volume fraction of φ ) 0.17 of the uncompressed, solvated polymer layer. In the friction experiments, two distinct regimes could be identified. An intact polymer film at moderate loads (several tens of micronewtons corresponding to average pressures on the order of 10-20 MPa) always showed extremely low frictional forces, below the detection limit of 20 µN, in response to lateral shear. At high loads of about 200 mN, when the film was compressed to about 10% of the original equilibrium film thickness, the friction showed a reversible increase with load (i.e., upon reducing the load, the friction decreased again). Under some experimental conditions, film destruction was observed, immediately followed by a dramatic increase in the frictional force. Intact Polymer Films. Under a variety of experimental conditions, the friction force between the two polymer-coated surfaces remained below the detection and cross-talk limit of 20 µN. An example of the data measured between brushes of PLL(20)-g[2.9]-PEG(5) in HEPES buffer with a sliding velocity of 0.5 µm/s is shown in Figure 4. In this experiment, the load was gradually increased up to 30 mN without a significant increase in the friction signal being detected. This corresponds to a friction coefficient of
