Premium
Computing free energy hypersurfaces for anisotropic intermolecular associations
Author(s) -
Strümpfer Johan,
Naidoo Kevin J.
Publication year - 2009
Publication title -
journal of computational chemistry
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.907
H-Index - 188
eISSN - 1096-987X
pISSN - 0192-8651
DOI - 10.1002/jcc.21317
Subject(s) - intermolecular force , chemistry , anisotropy , reaction coordinate , statistical physics , measure (data warehouse) , energy (signal processing) , thermodynamics , dimer , chemical physics , computational chemistry , physics , molecule , quantum mechanics , computer science , data mining , organic chemistry
We previously used an adaptive reaction coordinate force biasing method for calculating the free energy of conformation (Naidoo and Brady, J Am Chem Soc 1999, 121, 2244) and chemical reactions (Rajamani et al., J Comput Chem 2003, 24, 1775) amongst others. Here, we describe a generalized version able to produce free energies in multiple dimensions, descriptively named the free energies from adaptive reaction coordinate forces method. To illustrate it, we describe how we calculate a multidimensional intermolecular orientational free energy, which can be used to investigate complex systems such as protein conformation and liquids. This multidimensional intermolecular free energy W ( r , θ 1 , θ 2 , ϕ ) provides a measure of orientationally dependent interactions that are appropriate for applications in systems that inherently have molecular anisotropic features. It is a highly informative free energy volume, which can be used to parameterize key terms such as the Gay‐Berne intermolecular potential in coarse grain simulations. To demonstrate the value of the information gained from the W ( r, θ 1 , θ 2 , ϕ ) hypersurfaces we calculated them for TIP3P, TIP4P, and TIP5P dimer water models in vacuum. A comparison with a commonly used one‐dimensional distance free energy profile is made to illustrate the significant increase in configurational information. The W ( r ) plots show little difference between the three models while the W ( r, θ 1 , θ 2 , ϕ ) hypersurfaces reveal the underlying energetic reasons why these potentials reproduce tetrahedrality in the condensed phase so differently from each. © 2009 Wiley Periodicals, Inc. J Comput Chem 2010