Analytic Gradients of Approximate Coupled Cluster Methods with Quadruple Excitations
arXiv:2005.10178 · doi:10.1021/acs.jctc.0c00522
Abstract
The analytic gradient theory for both iterative and non-iterative coupled-cluster approximations that include connected quadruple excitations is presented. These methods include, in particular, CCSDT(Q), which is an analog of the well-known CCSD(T) method which starts from the full CCSDT method rather than CCSD. The resulting methods are implemented in the CFOUR program suite, and pilot applications are presented for the equilibrium geometries and harmonic vibrational frequencies of the simplest Criegee intermediate, CHOO, as well as to the isomerization pathway between dimethylcarbene and propene. While all methods are seen to approximate the full CCSDTQ results well for "well-behaved" systems, the more difficult case of the Criegee intermediate shows that CCSDT(Q), as well as certain iterative approximations, display problematic behavior.
arXiv comment37 pages, 1 figure, 2 tables
References in corpus (6)
- Towards standard methods for benchmark quality ab initio thermochemistry --- W1 and W2 theory
- W4 theory for computational thermochemistry: in pursuit of confident sub-kJ/mol predictions
- W3 theory: robust computational thermochemistry in the kJ/mol accuracy range
- High-Performance Tensor Contraction without Transposition
- Atomization energies of the carbon clusters Cn (n=2--10) revisited by means of W4 theory as well as density functional, Gn, and CBS methods
- Accelerating the Convergence of Higher-Order Coupled Cluster Methods II: Coupled Cluster Equations and Dynamic Damping