www.acsprf.org

Many lubricants are made of petroleum-derived components and a common class of components is hydrocarbons, or molecules made completely of carbon and hydrogen. Experimental studies have shown that hydrocarbon liquids tend to form solid-like layers next to solid surfaces. This behavior has been observed in several simulation studies as well and is not limited to hydrocarbon liquids. The goal of this work is to examine how the presence of a solid substrate influences the structural properties of a liquid film in the vicinity of a surface.

While most of the previous simulation work has focused on a single monolayer of molecules at the surface or on one-component liquids, this work focuses on multicomponent mixtures of hydrocarbons and other organic molecules. First, we are examining the effect of molecular architecture on adsorption behavior and liquid structure. A series of simulations of binary and ternary mixtures of three isomers of decane (n-decane, 4-propylheptane, and 2,2-dimethyloctane) have been performed using Monte Carlo simulation techniques. Initial results show a strong preference for the more linear molecule to adsorb at the surface. For example, in an equimolar mixture of all three isomers at 350 K, the mole fraction of n-decane within 9 Å of the substrate is 32% (±6%) greater than the total mole fraction of n-decane in the mixture. In binary mixtures of the branched isomers, 2,2-dimethyloctane is preferentially adsorbed over 4-propylheptane by 10% (±4%). Previous simulations of neat normal alkanes show very distinct layering arrangements within 25 Å of the substrate, meaning that regions of high molecular density alternate with those of low density. The inclusion of branched isomers reduces the range of these layers to within 10 Å of the substrate, corresponding to approximately two layers of molecules. Examination of the simulations shows that the branched isomers have a tendency to orient themselves to span the molecular layers, whereas normal alkanes are much more likely to align parallel to the substrate. Further analysis of these simulations is necessary to quantify this effect.

Additionally, we are studying the effect of including other organic molecules in the liquid phase. To begin, parameters for the interaction of an ether oxygen atom with the substrate were determined to facilitate the simulation of alkane-ether mixtures. Experimental work shows ideal mixing between ethers and their alkane analogs at the solid-liquid interface. Our Monte Carlo simulation work this summer shows the same behavior. In equimolar mixtures of ether and alkane analogs (for example, pentane and diethyl ether or nonane and dibutyl ether), the mole fractions of the two components in the first molecular layer at the substrate interface are the same within the uncertainty of the simulations. In contrast, when the size of the molecules differs (for example, nonane and diethyl ether), the mole fraction of the larger molecule is 16% (±2%) greater than the total mole fraction of that component. Analysis of these simulations is ongoing.

The second year of this project will focus on completing the analysis to the simulations discussed above. In addition to the analysis, several additional series of simulations are required to further explore these mixtures. The next step is to run simulations of molecules with other molecular architectures, such as rings, as well as additional classes of compounds, such as alcohols.

Financial support from the PRF for this program was used to support two undergraduate students during the summer of 2011. Michael Haslam simulated the mixtures of decane isomers and has begun work look at additional molecular architectures. Selena Watkins parameterized the substrate-molecule interactions of the ether oxygen atom and simulated the ether-alkane mixtures. Both students presented posters of their work at the 2011 MERCURY Conference on Undergraduate Computational Chemistry Research.