It has been suggested that when dodecane is confined between mica sheets separated by <5-10 molecular dimensions at ambient conditions, it undergoes a phase transition to an ordered solid-like structure1. However, even a brief search of the literature shows that there is much debate surrounding what occurs when dodecane is confined between sheets of mica. This is particularly true in the case of experiment, in which resolution at the scales required to observe such phenomena is extremely difficult to attain and, as such, the structure of the confined molecules is generally inferred from surface force apparatus experiments (SFA). Unfortunately, even with this technique the results of different laboratories differ, with some observing an order of magnitude increase in viscosity2, suggesting a phase transition, and others not3. Conversely, molecular simulation is ideally suited to the study of phenomena at the nanoscale and, as such, several techniques have been applied to the study of nano-confined systems1,4,5. With respect to dodecane, most workers have observed the formation of a layered herringbone structure, consistent with a many-order-of-magnitude increase in viscosity, but have done so using fairly simplistic models (E.g. mica represented by an fcc lattice of Lennard-Jones spheres), or have performed the simulations in a manner which may be criticized as biased towards the formation of such structures. Recently some workers, such as Jabbarzadeh et al.4, have considered factors which may cause such structures either not to form, or to break down once formed, e.g. an irregular/amorphous surface structure in the former case and rapid shear of the surfaces in the latter. However, they still use an artificially enforced system shape as well as a simplistic mica model (see above). Thus, while their suggested causes may explain the variation in observed behavior, the question remains as to whether or not such behavior exists in a real mica system or whether they are merely artifacts of the choice of model and simulation technique applied.
Given the level of interest in this phenomenon, and the ongoing disagreement, we feel that it is appropriate to revisit the molecular simulation of these systems. In particular, rather than a simple fcc lattice representation of the mica surfaces, we use the fully atomist model of Heinz et al.6, leading to the most realistic simulations of this system yet. Using this model, we have performed simulations similar to those of Jabbarzadeh et al.4 and Cui et al.1 in that we consider dodecane confined between infinitely periodic mica sheets. Since the periodic nature of this system, together with the fixed number of molecules used in the simulations, may be considered to bias the formation of an ordered structure we have also used the simulation technique of Gao et al.5 for nano-confined systems. In this case our system consists of parallel mica sheets surrounded by bulk dodecane which is free to move in and out of the gap between the mica sheets, thus removing the bias towards an ordered structure.
Based on previously published work, we discuss our results in terms of surface structure, surface energy and the strength of mica-dodecane interaction and the roles they play in the formation of herringbone ordered layering. We then explain why, and to what extent, these features are reproduced in our atomistically detailed simulations before considering other, neglected, factors, which may play an important role in these systems.
1S. T. Cui, P. T. Cummings and H. D. Cochran, J. Chem. Phys. 114, 7189 (2001)
2J. Klein and E. Kumacheva, J. Chem. Phys. 108, 6996 (1998); 108, 7010 (1998)
3A. L. Demirel and S. Granick, J. Chem. Phys. 115, 1498 (2001)
4A. Jabbarzadeh, P. Harrowell and R. I. Tanner, J. Chem. Phys. 125, 034703 (2006)
5J. Gao, W. D. Luedtke, and U. Landman, J. Chem. Phys. 106, 4309 (1997)
6H. Heinz, H. Koerner, K. L. Anderson, R. A. Vaia and B. L. Farmer, Chem. Mater. 17, 5658 (2005)