Quantum chemists seek to develop reliable computational methods which deliverchemical accuracy (4 kJ/mol). The predominant limitation in many of the state of the art methods is that a huge 1-particle basis set is required to achieve this accuracy. Recently , new explicitly correlated methods have been developed which include the inter-electron distance explicitly in the basis functions, and are capable of delivering chemical accuracy using the relatively small standard basis sets available. However, the high ly promising explicitly-correlated coupled-cluster method does not currently match this potential due to the approximations used in the evaluation of the integrals introduced in explicitly-correlated methods. We propose to address this limitation through t he use of auxiliary basis function techniques such that the potential of this method is realised. We propose to implement these developments within the computational package DIRCCR12-OS which is able to perform explicitly-correlated coupled-cluster calcula tions in both open and closed shell systems, and makes use of highlyefficient computational techniques such as parallelisation. The resulting computational tool will be capable of performing near basis set limit coupled-cluster calculations at a similar co st to regular coupled cluster calculations, yielding unprecedented accuracy for medium sized open and closed shell systems. We propose to use this new tool to investigate pi stacking van der Waals complexes whose interaction energies are of the order of 10 kJ/mol, and are important in many areas of chemistry such as the structure of proteins. Weak interactions such as these have hitherto been inaccessible to reliable prediction due to the basis set truncation error.
Call for proposal
See other projects for this call