1. Could you explain the significance of your article to the non-specialist?

This work presents a new approach to highly accurate density functionals (that describe all kinds of matter at a microscopic, quantum mechanical level) with a focus on general applicability and usefulness for day-to-day computations of molecular electronic structure. The performance of these functionals is significantly improved compared to common methods and we are close to CCSD(T) accuracy (the 'gold standard' of quantum chemistry) at two orders of magnitude less computational cost. The proposed method now holds several "world records" for best performance of density functionals for important chemical problems (heat of formations, van der Waals complexes, protein folding). On the other hand this article reveals the importance of dispersion (van der Waals) effects, not only for non-covalently bound complexes but also for general intra-molecular cases. These effects become increasingly important when larger molecular systems (e.g. in nanosciences) are investigated but have so far mostly have been overlooked. 

2. What has motivated you to conduct this work?

The accuracy of the original double-hybrid density functionals was already very good (in fact better than with other functionals) but careful analysis of the failures (outliers) revealed an incomplete account of dispersion that could be corrected. 

3. Where do you see this work developing in the future?

We propose our new method as a successor to the famous B3LYP density functional that has influenced chemistry a lot. The new B2PLYP-D method is more accurate and robust and allows a broader range of applications now including e.g. also large biomolecular species. The very promising results give hope, that it will be applied successfully in computational chemistry research for more diverse chemical problems and also other properties (e.g. in spectroscopy). 

4. Are there any particular challenges facing future research in this area?

At the moment the main focus has been on main group chemistry. The next, urgent question will be whether the approach is also similarly successful for transition metal chemistry. Only preliminary tests in that direction have been undertaken so far.