MSD for diffusion at infinite dilution

[Amir]

Good day,

I have been trying to simulate self-diffusion of TIP4P2005 water in MIL125-Ti MOF at the limit of zero loading (i.e. infinite dilution).
I have 20 water molecules in my system but I have set up my simulation in such a way that adsorbate molecules do not interact with each other, thus resembling a case when we have only 1 molecule in the system. My problem with the results is two folds:

1.Considering I have only one type of molecule (i.e. water) in the system, both self-diffusion of water and self-diffusion of total fluid must be the same. However, the MSDs that I get for these two are not the same (as shown in Figure 1 of the attached excel file) which cannot be correct.

2. Because water molecules do not interact with each other, I expect my MSDs (and hence diffusivity coefficients) to be the same when they are averaged over all 20 molecules in the system and when they are calculated per molecule. But in my results they are different (as shown in Figure 2 of the attached excel file).

I was wondering, if I you could kindly give me a hint as where I might be making a mistake. or could this be a bug in the code?
Below I have provided a copy of my input file. The MSD results are also attached.

Thank you for your help in advance.
Best regards,
Amir   

=========================
Simulation Input file:

SimulationType                   MolecularDynamics
NumberOfCycles                   200000000
NumberOfInitializationCycles     10000
NumberOfEquilibrationCycles      50000
PrintEvery                       1000
PrintPropertiesEvery             1000
RestartFile                      no

# Restart and crash-recovery
ContinueAfterCrash               no        # Write a binary file (binary restart.dat).
WriteBinaryRestartFileEvery      1000      # The output frequency of the crash-recovery file.

Ensemble                         NVT

Forcefield                    Water_MIL125_Ti
CutOffVDW                     12.0
ChargeMethod                  Ewald  #CoulombShifted or Ewald
CutOffChargeCharge            18.0
EwaldPrecision                1e-6

UseChargesFromCIFFile         yes
OmitAdsorbateAdsorbateCoulombInteractions yes
OmitAdsorbateAdsorbateVDWInteractions yes

Framework 0
FrameworkName  MIL125-Ti
UnitCells 2 2 2
ExternalTemperature  298.15
TimeStep             0.001   

ComputeMSD yes
SampleMSDEvery 1000
PrintMSDEvery 1000
ComputeIndividualMSD yes

Movies yes
WriteMoviesEvery    1000

Component 0 MoleculeName             TIP4P2005
            StartingBead              0
            MoleculeDefinition       TIP4P2005
            IdealGasRosenbluthWeight 1.00536
            TranslationProbability   0.5
            RotationProbability      0.5
            ReinsertionProbability   1.0
            CreateNumberOfMolecules  20

=========================

[Amir]

Apparently, the website doesn't allow me to upload my attachments! I can provide them via direct contact to anyone who is happy to have a look. Thanks you :)

[David Dubbeldam]

What do you mean with "Total fluid"? Self-diffusion of water inside a framework is not the same as self-diffusion of pure water.
You can not compute infinite dilution diffusion with just a single water molecule. The limit of low density is using lots of particles but using a larger volume.