Assessing the accuracy of integral equation theories for nano-sized hydrophobic solutes in water.

Abstract

Integral equation theories provide an efficient route for computing the solvation free energy (SFE) of molecular systems in water. The accuracy of those theories is usually tested against small molecules via comparison of SFE with reference data. However, tests against larger molecules in the nanometer regime are scarce in literature despite recent applications to such systems. Here, we thus study the accuracy and validity of a commonly used integral equation theory, namely, a three-dimensional reference interaction site model (3D-RISM), by considering the following problems: (1) solvation of a small to large Lennard-Jones particle, (2) binding of planar hydrophobic systems with varying size and hydrophobicity, and (3) self-assembly of amphiphilic molecules into a nanocapsule. The energy representation method is also utilized for comparison. The results show that the 3D-RISM method works successfully for small molecules, while the accuracy degrades systematically with system size and hydrophobicity. The size-dependent error in SFE does not cancel adequately between two solute configurations, resulting in a substantial error in the free energy difference. It is also shown that the free energy profiles for hydrophobic association exhibit a fictitious high-energy barrier, suggesting that care must be taken for studying such systems. The numerical difficulties observed above are discussed based on the relation between hypernetted-chain approximation, classical density functional theory with quadratic expansion, and the size-dependent error arising from the cavity region of the system.