Indexed on: 06 Jul '12Published on: 06 Jul '12Published in: Journal of Computational Chemistry
Calculation of large complex systems remains to be a great challenge, where there is always a trade-off between accuracy and efficiency. Recently, we proposed the extended our own n-layered integrated molecular orbital (ONIOM) method (XO) (Guo, Wu, Xu, Chem. Phys. Lett. 2010, 498, 203) which surmounts some inherited limitations of the popular ONIOM method by introducing the inclusion-exclusion principle used in the fragmentation methods. The present work sets up general guidelines for the construction of a good XO scheme. In particular, force-error test is proposed to quantitatively validate the usefulness of an XO scheme, taking accuracy, efficiency and scalability all into account. Representative studies on zeolites, polypeptides and cyclodextrins have been carried out to demonstrate how to strive for high accuracy without sacrificing efficiency. As a natural extension, XO is applied to calculate the total energy, fully optimized geometry and vibrational spectra of the whole system, where ONIOM becomes inapplicable.