Thursday, July 29, 2021

Interactions between large molecules pose a puzzle for reference quantum mechanical methods

Yasmine S. Al-Hamdani, Péter R. Nagy, Andrea Zen, Dennis Barton, Mihály Kállay, Jan Gerit Brandenburg and  Alexandre Tkatchenko (2021)
Highlighted by Jan Jensen

Figure 1 from the paper (c) The authors. Reproduced under the CC-BY licence

CCSD(T) and DMC are two gold-standard methods that should give the same results, and usually do. However, this study finds three systems for which the disagreement is unexpected large, up to 7.6 kcal/mol. It's not clear why and and it's not clear which method is correct. Since we use these methods to develop and benchmark other methods this is a real problem. 

Now, there could be many reasons for the discrepancy and the authors have considered all of them and discounted most of them. The remaining reasons, such as higher order terms in the CC expansion, are practically impossible to check at presents. It also hard to believe that they would make such a large contributions to the interaction energy of two closed shell systems. 

But there must be some reason for the discrepancy and when it is found we will most likely have learned something new about these methods.

This work is licensed under a Creative Commons Attribution 4.0 International License.