Approaching the FCI Limit
with Local Correlation: MP2 to CCSDTQ

Portrait of Andy Jiang, speaker, in front of a wall displaying the CCQC logo
Date & Time:
-
Location:
iSTEM Building 2, Room 1218

Local pair natural orbital coupled-cluster methods have seen much success recently, allowing for the application of “chemically accurate” approaches such as CCSD(T) to significantly larger systems. Notably, the DLPNO-CCSD(T) approach of Professor Frank Neese and coworkers has been utilized to run computations on whole proteins such as crambin (636 atoms) and insulin (789 atoms). At the CCSD(T) level of theory, local approximations have errors which are controllable to within 0.25 kcal/mol through tight parameters. Inspired by the work of Neese et al., we developed analogous models for higher orders of coupled-cluster theory through full quadruples (DLPNO-CCSDTQ). These methods allow for an accurate estimate of quadruples contributions such as (Q) and Q-(Q) on the order of 0.01 kcal/mol, as well as the T-(T) difference within 0.05 kcal/mol.

Type of Event:
Research Areas:
Andy Jiang
Department:
Graduate Student, Department of Chemistry
University of Georgia