Abstract
Computer simulation increasingly complements experimental efforts to describe nanoscale structure formation. Molecular mechanics simulations and related computational methods fundamentally rely on the accuracy of classical atomistic force fields for the evaluation of inter- and intramolecular energies. One indispensable component of such force fields, in particular for large organic molecules, is the accuracy of molecule-specific dihedral potentials which are the key determinants of molecular flexibility. We show in this work that non-local correlations of dihedral potentials play a decisive role in the description of the total molecular energy—an effect which is neglected in most state-of-the-art dihedral force fields. We furthermore present an efficient machine learning approach to compute intramolecular conformational energies. We demonstrate with the example of α-NPD, a molecule frequently used in organic electronics, that this approach outperforms traditional force fields by decreasing the mean absolute deviations by one order of magnitude to values smaller than 0.37 kcal/mol (16.0 meV) per dihedral angle.
Highlights
Molecular dynamics (MD) methods are widely used for the simulation of inorganic and organic materials at the atomistic level, for example in the field of computational biology and drug design and in the field of organic electronics for organic light emitting diodes, organic solar cells and other technologically relevant applications
The standard approach to model the energy as a sum of terms depending only on one dihedral angle leads to significant deviations between the energy as determined by electronic structure methods and the force field energy
These deviations result in part from branched molecular structures and delocalized electronic states which result in correlations of quantum mechanical origin between dihedral angles that are difficult to capture in state-of-the-art force fields
Summary
To investigate the accuracy of different dihedral force fields, we parameterize molecule-specific force fields for a set of molecules (see Fig. 2) using semi-empirical PM7 calculations as implemented in MOPAC 201614,15. For branched aromatic molecules, such as TPA/DPNA and more complex derivatives such as α-NPD the force field approach fails to predict the internal molecular energy, indicating that the total energy of the molecule cannot be written as additive terms of the dihedral energies. This can be attributed in part to the fact that for branched molecules with a particular dihedral conformation the electronic energy is a complex function of the internal degrees of freedom. We investigate an artificial neural network which is trained using random dihedral configurations of a molecule to predict the molecular energy of all possible configurations
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
