SHARC Release 4.0.2

This is release version 4.0.2 of our ab initio nonadiabatic molecular dynamics code SHARC (Surface Hopping including ARbitrary Couplings). It is developed in the González group at the University of Vienna and in the Truhlar group at the University of Minnesota and its purpose is the investigation of excited-state dynamics.

SHARC4.0.2 is a bugfix release. Check the individual files for details and the release notes of release 4.0 for new features.

SHARC Release 4.0.1

This is release version 4.0.1 of our ab initio nonadiabatic molecular dynamics code SHARC (Surface Hopping including ARbitrary Couplings). It is developed in the González group at the University of Vienna and in the Truhlar group at the University of Minnesota and its purpose is the investigation of excited-state dynamics.

SHARC4.0.1 is a strict bugfix release. Check the individual files for details and the release notes of release 4.0 for new features.

SHARC Release 4.0

This is release version 4.0 of our ab initio nonadiabatic molecular dynamics code SHARC (Surface Hopping including ARbitrary Couplings). It is developed in the González group at the University of Vienna and in the Truhlar group at the University of Minnesota, and its purpose is the investigation of excited-state dynamics.

New features in SHARC Version 4.0

The SHARC Version 4.0 constitutes a significant milestone in the development of the package.
The main goal was to redesign the complete framework of the communication between the SHARC dynamics driver and the electronic structure data providers. To this end, a complete refactoring of the interfaces was carried out.
The new interfaces are developed in an object-oriented way, using inheritance to simplify development.
The communication protocol was generalized and made more rigorous.
For the user, the main advantages are

• better performance for fast dynamics using model potentials
• more systematic setup routines
• the possibility to combine and nest interfaces to achieve a broad variety of workflows.

These “interfaces that can call other interfaces” are called hybrid interfaces in SHARC.

The hybrid interfaces enable methods like quantum mechanics/molecular mechanics (QM/MM) to describe solvated molecules, excitonic configuration interaction (ECI) to describe multichromophoric systems, adaptive sampling and automatic data storage to collect electronic structure data (for machine learning or other purposes), automatic numerical differentiation (to get gradients, nonadiabatic couplings, or other derivatives), or umbrella sampling for various sampling tasks.

Full Changelog: https://github.com/sharc-md/sharc4/commits/v4.0