Gromos53a6 has good result for lipids, taking into account the Berger et al parameters, also:
O. Berger, O. Edholm, F. Jahnig, Molecular dynamics simulations of a fluid bilayer of dipalmitoylphosphatidylcholine at full hydration, constant pressure, and constant
temperature, Biophys. J. 72 (1997) 2002–2013.
For coarse grained, we have obtained good results with Martini force field.
But, I do not know which kind of compounds are yours, or as Alexey said: in which kind of property are you interested in.
In this case, you need a field describing the thermodynamic properties, like AMBER-ii, TraPPE, OPLS.
However, if the phase separation is in aqueous solution, none of them does not help. Perhaps then it is better to use AMBER-i. This is previous unpublished version of AMBER-ii. In it is a more realistic interaction of alkanes with water.
William Jorgensen, the author of TIP3P and OPLS, always shows in his talks a plot, where he compares benzene dimer interaction energies of the state-of-the-art quantum chemical methods. Then he points to the fact that in the course of time, the QM energies converge to the OPLS interaction energy:-)
if you have the option of using commercial software try COMPASS FF. This ab initio force field was parameterized and validated using condensed-phase properties in addition to various ab initio and empirical data for molecules in isolation.
and it works reassonable well with hydrocarbons....