TUTORIAL 8 : Exercise 1

Ideal gas in a planar pore

In this exercise we will simulate ideal gas partitioning in a slit confinement, in either “non-wetting” or “wetting” surface conditions.

Navigate to directory tutorial_8/slit-ideal …

Check the FIELD file:

slit-ideal > more FIELD

The initial FIELD file for this tutorial should look as shown below.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
Ideal gas in a slit
CUTOFF 1.0
UNITS K
NCONFIGS 1
ATOMS 1
LJ core 1.0  0.0
MOLTYPES 1
IG
MAXATOM 256
FINISH
VDW 1
LJ core  LJ core  hs   0.0  0.0
external 1
LJ core  hs       0.0  1.0  1.0
CLOSE

The important points to note are:

the cutoff = 1.0 (very small)

the energy unit is K (reduced units)

the particle-particle hard sphere interaction is zero (ideal gas!)

the external particle-wall hard square interaction is in the form of a ‘repulsive shoulder’:

width = 1.0 Angstrom (from a wall) and strength = 1.0 kT.

Check the CONTROL file:

slit-ideal > more CONTROL

The initial CONTROL file for this tutorial is

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
NVT simulation of ideal gas in the slit geometry
use ortho
finish
seeds 12 34 56 78               # Seed RNG seeds explicitly to the default

slit walls soft 2
#slit mfa 0

noewald all
temperature       1.0           # Corresponds to T*=1.0
pressure          1.0           # katm (*10^9)

steps           1000000 #0         # Number of moves to perform in simulation
equilibration       0           # Equilibration period: statistics are gathered after this period
print            100000         # Print statistics every 'print' moves
stack              1000         # Size of blocks for block averaging to obtain statistics

sample zdensity    120  10
sample coordinates 100000

archiveformat    dlpoly2 #+dcd
revconformat     dlpoly2        # REVCON file is in DL_POLY CONFIG format

acceptatmmoveupdate   100000000

maxatmdist  10.0
move atom 1  100               # Move atoms 100% of the time
LJ core

check   1000000
start

Run the simulation and, when done, plot the density profile along z (ZDENSITY.000):

[tutorial_8]$ gnuplot
gnuplot> plot [x=-6:6] [y=0.:3.0] 'ZDENSY.000' u 1:2 w l t "Density(z)"
_images/tutorial8-slit-ig0.png

Questions to ask yourself:

  • Is it “wetting” or “non-wetting” case?
  • Why are there two levels in density?
  • Can you calculate the exact density levels?
  • How does the external potential look like?
  • How to reconstruct the external potential (approximately) from the density data?
  • What will happen if you increase the cell dimentions in X and Y in CONFIG?

If we increase the ‘steps’ number in CONTROL by a factor of 10 (uncomment the extra zero), we get a much better precision, as below.

_images/IG-256p-rep1kT1A.png

Additional exercise:

Change the sign for the external interaction strength (in FIELD), and rerun the simulation.

  • How do you think the density will look this time?

Hint: \(\rho(z) = \rho(0)\exp(-\beta U(z))\)

Next exercise:

TUTORIAL 8 : Exercise 2 - Water adsoprtion in a slit