Dynamical Shear Studies of Nanoscale Lubricant Films Confined Between Iron Oxide Surfaces Covered with DTP Wear Inhibitors

Yanhua Zhou 1, Tahir Cagin1, Elaine S. Yamaguchi2, Andrew Ho2, Rawls Frazier2, Yongchun Tang2, and William A. Goddard III1
1Materials and Process Simulation Center, California Institute of Technology
2 Oronite Global Technology, Chevron Chemical Company


       Behaviors and properties of fluids in ultra-thin films are very different from their bulk counterparts, in general. We are particularly interested in the behavior of a thin layer of lubricant oil sandwiched between metal and/or metal oxide surfaces.

       Here we present the results from shear-dynamics simulations of hexadecane lubricant (n-C16H34) confined between two Fe2O3 surfaces, each of which is covered with a self-assembled monolayer (SAM) of wear inhibitors [dithiophosphate molecules DTP = S2P(OR)2, where R symbolizes an organic group]. We believe that this is the first study of realistic systems (realistic lubricants, realistic solid surfaces, and realistic adsorbed molecules) under the influence of a shear flow. A typical structure and setup of the system is shown in Figure 1. In the simulation, the top surface was sheared with respect to the bottom one at a rate of 100~m/s (0.1~nm/ps) and temperature of 500~K. The surface layers of the iron-oxide, the DTP molecules, and the lubricant were allowed to respond to the imposed shear forces without constraint.

       An objective of the study is to investigate the effects of the wear inhibitors on the behavior of the lubricant. For this purpose, we did simulations for three different organic groups in dithiophosphates: R= iPr (isopropyl), iBu (isobutyl), and Ph (phenyl).

       We found significant density oscillations near the walls (within 1.5 to 2.0~nm) (Figure 2). In the central region of the fluid, the layering decreases, with the density approaching that of the bulk liquid. For a separation of 4.5 nm we can define 10 layers. For a thinner system with just 2.0 nm separation, we found 5 layers and the density oscillations are even more evident (Figure 2d).

       In Figure 3 we trace each molecule and see how its distribution contributes to the overall density oscillation. It clearly shows layering taking place in the lubricant. Most molecules are confined within 1 to 2 layers during the entire time span of the shear simulations.

       Figure 4 shows the displacement of the center-of-mass of each lubricant molecule in the shearing direction (x) as a function of time. For the iBu (b) we see clearly a stick-slip motion of the lubricant molecules near the bottom SAM, characterized by a sequence of move (slip), pause (stick), and move (slip) again, with a periodicity of ~40 ps. On the other hand, after 100~ps of dynamics the iPr case shows little shear at the boundaries (a) while the Ph case shows almost continuous shear at the boundaries (c).

       The average velocity profiles of the lubricant are also quite different for the three cases (Figure 5). Here we see for iPr (a) all the shear occuring in the middle 2.0 nm region, whereas for Ph (c) the shear distributed uniformly from bottom to top boundary. For iBu (b) its behavior falls in intermediate. Thus, for iPr the velocity gradients near the interfaces are essentially zero, creating a boundary layer inside the lubricate. Intuitively, this existence of an additional lubricating layer at the interface should aid in protecting the iron-oxide surface from wearing. Indeed, iPr does lead to the lowest engine wear among the three wear inhibitors in actual engine test.

Figure 1: The structure and setup of the system for the shearing dynamics simulations. Shown here is a snapshot of the iPr DTP case after 200~ps simulation. The middle region contains 32 C16 molecules and is ~4.5 nm thick. These molecules are partitioned into 8 layers according to their average value of zcm; each layer is plotted with a different color.

Figure 2: The density distribution of the lubricant as a function of z. Only the average results of the last 100~ps in the 200~ps run are plotted.

Figure 3: The layering in the lubricant for the case of R = iPr. The background is the average density over all the molecules (Figure 2a). The colored curves show the density distribution for each molecule.

Figure 4: The displacement of lubricant molecules in the shearing direction as a function of time. The molecules are grouped according to their final x positions. Each group is indicated with a different color. The upward inclined line shows the motion of the top SAM layer, which reads the velocity of 0.1 nm/ps. The bottom horizontal line shows the motion of the bottom SAM layer.

Figure 5: The average velocity profiles. The figures denote the velocities of the top and bottom lubricant layers.


    This research was supported by the Chevron Chemical Company (Oronite Global Technology) and grants from the DOE-ASCI-ASAP and NSF (CHE 95-12279 and ASC 92-17368).

Yanhua Zhou<yanhua@wag.caltech.edu>
Last modified: Wed Mar 15 16:02:37 PST 2000