Dear experts,
What is the correct approach to model a cylindrical nano-pore or a slit-pore in molecular simulations (MC, MD).
I have seen numerous papers applied the rigid pore (frozen pore) in their work. However, it will create an unrealistic temperature gradient between the pore (zero Kelvin since it is rigid) and the fluid (let's say 300K). I afraid it will influence the subsequent results such as adsorption or heat flux calculation.
If one does not make the pore to be rigid, the pore will collapse when the thermostat is applied to the pore. In a few case, for example graphite, people keeps 1 graphene layer to be frozen and allow the other 2 layers to be non-rigid. Nonetheless, this semi-rigid approach cannot be applied for all cases.
Fixing atoms of the nanopore (or applying harmonic restraints) does not set up a temperature gradient. Your assumption that a rigid structure is at zero Kelvin is not correct. Constrained coordinates cannot absorb energy and, hence, cannot affect the temperature. The temperature is related to the kinetic energy in each degree of freedom — fixed positions are not degrees of freedom. Does the rigid water model usually used in the MD simulations mean the water is at 0 K? No, the degrees of freedom associated with the internal motions of the water can neither absorb nor release energy.
On the other hand, as you suggest, atoms with constrained positions cannot transport thermal energy and, hence, act like perfect thermal insulators (not as a 0 K bath). To allow transport of thermal energy, you should harmonically restrain the atoms rather than perfectly constraining them. If the thermal transport is very important, you can set up a simulation to measure it.
Jeff
Do you need a realistic nanopore or just a steric confinement?.
In the first case if the molecules that forms the pore are correctly parametrized in MD force field you should not have any "melting". You can constraint the external solid layer, apply a thermostat (if you need) to the layer in the middle and leave free the layer in contact with the fluid. I applied this approach in the determination of slip length for water on a wall coated with alkyl chains and the result was robust under a relatively large variation of the arbitrary choice of the protocol (e.g. how many layers are themostatted).
In the second case, if you are not interested in the control wall boundary condition but you want just something that confine the system, harmonical contrained atoms could work.
Mauro
Fixing atoms of the nanopore (or applying harmonic restraints) does not set up a temperature gradient. Your assumption that a rigid structure is at zero Kelvin is not correct. Constrained coordinates cannot absorb energy and, hence, cannot affect the temperature. The temperature is related to the kinetic energy in each degree of freedom — fixed positions are not degrees of freedom. Does the rigid water model usually used in the MD simulations mean the water is at 0 K? No, the degrees of freedom associated with the internal motions of the water can neither absorb nor release energy.
On the other hand, as you suggest, atoms with constrained positions cannot transport thermal energy and, hence, act like perfect thermal insulators (not as a 0 K bath). To allow transport of thermal energy, you should harmonically restrain the atoms rather than perfectly constraining them. If the thermal transport is very important, you can set up a simulation to measure it.
Jeff
- Badges
- Science topic