Nanotribology Fundamentals: Predicting the viscous coefficient of friction

Show full item record

Title: Nanotribology Fundamentals: Predicting the viscous coefficient of friction
Author: Coffey, Tonya Shea
Advisors: Jacqueline Krim, Committee Chair
Christine Grant, Committee Member
Marco Buongiorno-Nardelli, Committee Member
Dale Sayers, Committee Member
Abstract: In this work, I have used the Quartz Crystal Microbalance (QCM) to study nanoscale friction of monolayer adsorbates on (111) metals. The friction of these systems is viscous friction, defined as F = hv = (m/t)v. Here, h is the viscous coefficient of friction, v is the velocity of the adsorbate, m is adsorbate mass, and t is the slip time, which is the time required for the film's speed to fall to 1/e of its original value. The main focus of this dissertation is to determine the factors that control h, the viscous coefficient of friction. I have examined three different parameters in order to determine their effect on h. An equation for predicting the viscous coefficient of friction has been proposed: h=hsubs + aUo2. Here, hsubs is the damping of adsorbate sliding energy within the substrate, a is a constant depending on mainly temperature and adsorbate film coverage, and Uo is the atomic-scale surface corrugation. I have closely examined the effect of varying Uo while holding the lattice spacing relatively constant by studying the slippage of xenon films on Cu(111), Ni(111), graphene, and C60 substrates at 77.4 K. I have also examined the effect of varying hsubs while controlling other parameters by studying the slippage of n-octane films on Cu(111) vs. Pb(111) surfaces at room temperature. It was found that the slippage of xenon on Cu(111), Ni(111), and graphene/Ni(111) was very well fit by the proposed equation. These three systems had very similar nearest neighbor lattice spacings (0.255 nm, 0.249 nm, and 0.249 nm, respectively) but varying atomic scale surface corrugations (1.9 meV, 14 meV, and 5.3 meV, respectively.) The xenon monolayer slip times (t) of 15.5 ns, 0.41 ns, and 1.7 ns, respectively, were well fit by the relation t ~ Uo-2. Specifically, when plotted on a ln t vs. ln Uo plot, the data were fit by a slope of –1.82 +/- 0.20. It has been proposed that hsubs should be linearly proportional to the damping of frustrated translational (FT) phonon modes (g) of an adsorbate-substrate system via g = hh. The parallel FT modes are believed to be directly linked to the sliding friction, but it is not clear how damping of the perpendicular FT modes (FTz) affect sliding friction. To explore this question, I have examined the sliding friction of n-octane on Cu(111) vs. Pb(111) surfaces, which have g = 0.45 meV and g = 0.26 meV, respectively. I have observed that the slip time for a monolayer of n-octane/Cu(111) is 0.94 ns +/- 0.36 ns, and the slip time of n-octane/Pb(111) is 0.59 ns +/- 0.13 ns. I therefore observe no direct evidence of a link between the damping of perpendicular FT modes and sliding friction. It is still possible, however, that the damping of the parallel FT phonon mode affects sliding friction. Finally, I studied the slippage of monolayer methanol films at room temperature on rotating, rigid, and slowly ratcheting C60 substrates, to examine the effect that the molecular rotation of the substrate surface has on the sliding friction of an adsorbate. It had been hypothesized that the rotation of the C60 molecules might reduce friction, because the round, rotating C60 could act as a nanoscale ball bearing. I found that the slippage and mass uptake for methanol on rigid and slowly ratcheting C60 was indistinguishable. I found that at all coverages, the slip time for methanol on rigid and slowly ratcheting C60 was longer (hence lower friction) than the slip time for methanol on rotating C60, defying the ball bearing analogy.
Date: 2004-06-29
Degree: PhD
Discipline: Physics

Files in this item

Files Size Format View
etd.pdf 3.461Mb PDF View/Open

This item appears in the following Collection(s)

Show full item record