We finally got our paper on hybrid functionals and the local density approximation (lda) accepted. This work presents some basic conditions hybrid XC functionals should satisfy and a discussion on the balance between correlation, HF, and local exchanges. We suggested an adiabatic functional, cam-lda0, which can be used to speed up calculations a little bit. Copies available at the JPCA website, here
This paper started after thinking for a while about the pbe0 xc functional. Our impatience and desperation with cluster computations led us to consider removing the gradient contributions. We then further did this for cam-b3lyp, and noted that the gaps did not change significantly.
I’m proud this is the birth of a new field, “functional dedeveloping”
Update 08/2020: Functional dedeveloping has led us to domain separated DFT. Will post more about it later…