What problems occur when want macroscopic properties from microscopic simulation?
Finite-size effects always occur more or less when simulating microscopically but want to know about macroscopic properties.
The interaction range may be larger than the simulated range.
there can occur surface-effects where the simulation box ends which may affect the result.
Why does one not want to simulate molecule in vacuum?
- Large finite-size and surface-effects.
- No solvation interactions eg. dielectric screening, solvent collisions etc.
- spherical shape –> reduced surface area
- too strong electrostatic interactions, very strong H-bonds
- molecule stuck in one conformation, because can only interact with itself
Explain fixed boundary conditions (FBC) of a droplet in water and the problems with this approach.
Fixed boundary condition means there is a limited volume, in this case water droplet with explicit water molecules, in vacuum.
- large surface-effects in the liquid-vacuum interface
- requires pot. so liquid not evaporates
- can only sim. truly microscopic sys. > 5nm, otherwise too much surface and finite-size effects
Explain fixed boundary conditions with a water droplet in implicit solvent.
Having droplet with explicit solvent in implicit solvent, ie. outside droplet only continuum with one dielectric permittivity eps and viscous friction gamma, describing the properties.
- more realistic bulk behavior
- less surface effects and finite-size
- hard to construct the boundary pot. at the interface
Explain the idea of a periodic boundary conditions (PBC).
We have a box with explicit solvent molecules, and this have periodic images repeating in all directions. So when a molecule comes to the interface it will interact with the periodic image of another one and when a particle goes out on one side it will come back in on the other.
- reduced finite-size effects because infinite system and no surface effects
- but can have artifacts due to induced periodicity, in liquids
Explain nearest image cut-off and problems occurring.
This means that every particle is only interacting with the nearest image, if the nearest image of one particle is not the original particle but its periodic image, the interaction should be with this particle instead.
This may cause strange effects of particle accumulating in the corners.
What does spherical cut-off mean and what is the preferred cut-off length?
No anisotropy effects due to accumulation in corners.
Box length L, cut-off radius Rc:
Rc > L/2: including periodic images of particles since cut-off sphere larger than box, the solute comes twice
Rc
