PARAMETERISING WEAK INTERACTIONS. S.J. Maginn, Molecular Simulations, 240/250 The Quorum, Barnwell Rd., Cambridge CB5 8RE, UK G.R. Desiraju, Dept. of Chemistry, University of Hyderabad, Hyderabad 500134, India

Forcefield parameters to account for a range of so-called "weak" interactions, such as C-H....O, have been successfully derived and tested.

Many crystal structures depend for their integrity on one or more of a range of so-called "weak" interactions. Examples of these include C-H....O, C-H....N, N....Cl etc.

No currently widely available molecular mechanics forcefield contains parameters to account for these interactions. Consequently, this is a weakness of the emerging forcefield-based crystal structure prediction methodologies (e.g. those of Gavezzotti (1), and Karfunkel and Gdanitz (2)) - using such forcefields, structures containing "weak" interactions can never be accurately predicted.

A range of crystal structures containing C-H....O and C-H....N interactions has been used, along with the tools in the Cerius2 molecular modelling suite (such as molecular mechanics minimisation, and the Karfunkel and Gdanitz structure prediction method) (3), to produce a customised forcefield which accounts for these, and which it is hoped will lead to more accurate structure predictions in future.

(1) A. Gavezzotti; J. Am. Chem. Soc. 113 (1991), p4622

(2) H.R. Karfunkel, R.J. Gdanitz; J. Comp. Chem. 13 (1992), p1171

(3) Cerius2, from Molecular Simulations Inc., 9685 Scranton Rd., San Diego, CA 92121-3752, USA, or 240/250 The Quorum, Barnwell Rd., Cambridge CB5 8RE, UK