[lammps-users] accuracy of REAX

Dear all:

I recently installed reax package and tried to benchmark it with the Fig. 3 in J. Phys. Chem. A 2001, 105, 9396-9409.

The C-H total energy curve matches quite well with the paper result. But error exists with the C-C and H-H total energy curve when the interatomic distance is below 2.0A.

My question is have the developer checked the accuracy of the reax package by benchmark it with the reaxFF paper results? Will there be a future updated version? When will it come out?

Would I attach my comparison figure with this email?

Thank you so much!


There is no "one" ReaxFF paper and the parameters
change frequently, but I'll let Aidan comment.



Here a response from the chief architect of ReaxFF. Thanks Adri.


p.s. I will add a corresponding comment to the LAMMPS documentation

------ Forwarded Message