First, I'm using VASP. It's known that surface defects can degrade desirable properties of some quantum dots. There's a molecule which is suspected to regenerate the dots by interacting with the defects. The QDs I'm interested in are cubical and around 20 nm, so I'm worried it'll be too many electrons for the cluster I'm on to handle. Should I:
1. Decrease dimensions to 1nm x 1nm x 1nm, place in a box with ~15 angstroms on all sides, and study the adsorptions from there?
2. Create a slab and study the adsorptions that way?
3. Do something else, possibly molecular dynamics.
I'm worried that the slab will fail to capture confinement effects, which may or may not be important for what I'm modeling. Similarly, I'm worried that going from 20 nm to 1 nm may be too far of a deviation from reality. What do more experienced modelers think?
If your computational resources are limited, reduce the dimensions is the best way to speed up your work.
I would do all the three that you have listed. I worked on a perovskite system, where we took all the three routes. We did slab calculation to study the ligand binding. Took a ~2.4 nm (3x unit cell) cube in vacuum (real is about ten times bigger). Also did the classical MD using Bader charges and explicit solvent model.
20 nm in many systems is large enough to be treated as bulk. To verify this, you can do the following test: Create a minimal slab (without any molecule), terminated along a certain crystalline direction, let's say c. Then systematically calculate the thickness dependence of the free energy of the slab. Repeat this procedure for the other two crystalline directions. From these calculations, estimate the optimal thickness at which the Free energy approaches the bulk value. If for all three directions, it is less than 20 nm. Then you can basically, use a slab to model your quantum dots and their molecular adsorption properties.
- Badges
- Science topic