Release notes Amsterdam Modeling Suite 2018

The SCM team is happy to announce our new major release, the Amsterdam Modeling Suite 2018 (AMS2018).
Our molecular DFT code ADF has some exciting new functionality, especially for those studying organic electronics.
ReaxFF has many new tools for acceleration, force field fitting, and trajectory analysis. Our COSMO-RS module now includes quicker sigma profile estimation, quick property predictions, and a solvent mixture optimization tool.

Most importantly, with this release we introduce the new, powerful driver AMS. With AMS you can quickly set up advanced Molecular Dynamics (MD) and Potential Energy Surface (PES) tasks for all the atomistic codes SCM develops, as well as external engines. It’s also really easy to switch between methods and set up workflows. During this major refactoring project to separate engines and drivers, we also simplified the input blocks and made them more uniform for our different compute engines. A selected list of new features and improvements across the entire AMS suite is listed below.

AMS driver

AMS is a powerful driver to quickly set up MD and PES tasks with any atomistic code providing energies, forces, and optionally Hessians. Within AMS2018, it is currently tightly coupled to DFTB, BAND, and UFF. It supports MOPAC and tighter integrations for ADF is underway. ReaxFF can be used both as a tightly coupled engine and as a binary called through AMS. Currently the non-AMS ReaxFF engine contains advanced methods such as acceleration methods, Monte Carlo, molecule gun, etc. which will eventually be moved to the AMS driver.

Coupling to external codes is in progress, and we strongly encourage people to develop and publish their own interfaces. The AMS set up also empowers collaborators to swiftly implement and test new algorithms for MD, Monte Carlo, and transitions state searches.

2D PES scan

With AMS you easily switch computational engine, enabling e.g. workflows or high-throughput screening to gradually increase accuracy by switching from force fields to tight-binding to density functional theory codes.

fast nanoparticle spectra with poltddft

Amsterdam Density Functional (ADF)

The new features and improvements in our molecular DFT code are mostly focused on excited states, bringing more powerful tools to study nanoparticles and organic electronics. See also ADF release notes.

Periodic DFT: BAND & Quantum ESPRESSO

In BAND, effective masses can now be calculated for any k-point. Through AMS, many useful properties are now available in BAND (see also BAND release notes):

  • Elastic tensor + moduli (bulk, shear, Young’s); geometry optimization under pressure
  • Linear Transit and PES scans: multiple degrees of freedom, any periodicity
  • Molecular Dynamics

Quantum ESPRESSO has been updated to version 6.3, including preliminary support for CPMD. The binaries include phonons and electronic-phonon coupling (EPW), to be run from the command line.

ReaxFF

Many new useful tools have been implemented in ReaxFF:

COSMO-RS

The capabilities, range and speed for property predictions in our COSMO-RS module have been vastly improved (see also release notes).

Through SMILES strings or xyz files, instantaneous property predictions are now available for molecules of any size. The most accurate predictions are still expected for the original ADF DFT + COSMO-RS parameterization, but a very quick pre-screening can be achieved by these quick property predictors.

Graphical User Interface

The GUI will support many of the newly implemented features, including support for AMS. Other noteworthy improvements (see also GUI release notes):

  • Grand Canonical Monte Carlo (GCMC) for ReaxFF (e.g. battery discharge)
  • Faster bond guesses
  • Better graph code
  • Change units in input panes by right-clicking
  • Support for intermediate structures in ASE-NEB
  • Export CIF & BGF
  • Quick property predictions via COSMO-RS module
  • LDOS (STM images)
  • Crystal Orbital Overlap Populations (COOP)
  • Visualize NMR spin-spin coupling
  • More information for fragment analysis (SFOs)
  • Open multiple files
  • Visualize nD-PES scans