I am writing a project on electronic structure with a molecular approach and a surface approach (plane waves). I have knowledge of the molecular part, but I am having difficulties describing the project for plane waves. I intend to use Gaussian for molecular structure and Quantum Espresso for plane waves.
I have the methodology for the molecular part and would like to know how to describe the same items for plane waves. Could someone please help me?
Below, I am placing the methodology for the molecular part:
Construction of systems and structural studies Reactivity index calculations (for the molecular part, Condensed Fukui Indices to Atoms are used) Reactivity index calculations (plane waves ???)
Opto-electronic properties (for the molecular area TD-DFT)
In particular, the aim is to evaluate data associated with the energy and spatial distribution of frontier orbitals, local and global density of states, reactivity indices, optical properties of the materials composing the chemical species, in order to establish simple rules for the preparation of materials with optimized properties.
Adsorption study
It is intended to evaluate adsorption processes of chemical species and reactions with the systems of interest through two different approaches: i) calculations of molecular electronic structure and ii) calculations of surface electronic structure.
Calculations of electronic structure Optimization of geometry of adsorbed systems will be performed in a DFT (and/or Hartree-Fock) approach with Grimme corrections to better describe interactions between unbound systems.
And how does the calculation of electronic structure for the surface work?
Since no one's responded, I'll give a partial answer even though there are others better qualified for a full answer. Usually for plane-wave surface calculations you would look at the projected density of states, and possibly the orbital/site projected wavefunction character, to gain information about electronic structure. For a classic solid state physics picture, you would calculate a band structure. It should be possible to use TD-DFT to get optical properties. As far as I know, Fukui indices don't really have a clear analogue in a plane-wave code. You may want to take your molecule and run it in both Gaussian and Quantum ESPRESSO so you can compare, and then move to the surface with QE.
Like the previous responder I can only give a partial answer or rather only a comment. Although many attempts have been made in the past, there are no genuine methods to treat surfaces. In a molecular package you treat atoms on an individual basis and you enter their coordinates. In a bulk package, a bandstructure program, you can do so only within the bulk unit cell. The rest of the geometry is controlled by the lattice, representing a 3-dimensional infinite periodicity. For the surface this periodicity needs to be broken in the direction perpendicular to the surface. Plane waves are just mathematics and the powertool there is the Fourier transform. In its integral representation it can handle any problem, also a cluster of atoms and mathematically this is even an easy process. It is a matter of numerical efficiency that this s not done in practice. The problem is the contribution of the nuclear cores to the potential. Since this contribution is singular at the core site, it requires an excessive amount of plane waves to require enough
accuracy. A technical way out is to replace the potentials by pseudopotentials, free from singularities, but still yielding good enough results. In the family of bandstructure methods, CLOPW is a good alternative, bu it has never been developed to a complete package, If you manage to use/develope the plane wave representation into a proper instrument to handle true surfaces that would be a good thing, but as far as I know that has not been done.
- Badges
- Science topic
- Similar topics
- Engineering
- Petroleum Engineering
- Oil