Step 4b - Calculate g(r) and S(Q)
The experimental data we’ll be fitting to is neutron scattering data, so we will need to calculate neutron-weighted structure factors.
Layer ⇨ Create ⇨ Correlations ⇨ RDF and Neutron S(Q)
The new layer contains the following modules:
| Module | Purpose |
|---|---|
GR |
Calculates partial g(r) between every pair of atom types, and sums them into the total G(r) |
SQ |
Calculates partial S(Q) between every pair of atom types by Fourier-transforming a source set of g(r), and sums them into the total F(Q) |
NeutronSQ |
Takes the S(Q) calculated by an
SQ module to generate the neutron-weighted partial and total structure factors |
Note that the ordering of these three modules in the layer is important - calculating the neutron-weighted structure factors requires the unweighted structure factors from the
SQ module, which in turn requires the radial distribution functions from the
GR module. Modules are run “top to bottom” in the list, fulfilling the requirement that the
GR module runs before the
SQ module etc.
We will now need to set a few parameters in the
NeutronSQ module, in particular informing it of the isotopic composition of our system and loading in reference data.
A
Set up Isotopes
The
NeutronSQ module will use isotopic natural abundances to calculate the neutron weights for all species unless we tell it otherwise. We’ll first define the correct isotopologue for our argon species, and then tell
NeutronSQ to use it. The experimental measurement was made using Ar36 since its coherent scattering cross-section (24.9 fm) is considerably higher than that of the naturally-occurring mix (1.91 fm).
Go to the
Ar species tab
Click on the
Isotopologues section
Click to create a new isotopologue definition assigning the default (natural) isotope to each atom type present in the species
Change the entry for the Ar atom type from
Natural (bc = 1.909 fm)to36 (bc = 24.9 fm)
For sanity’s sake, you may also want to double-click on the name of the isotopologue in order to change it to something more meaningful (‘Ar36’ perhaps)
Now we’ll go to our calculation layer and set the isotopologue for our
NeutronSQ module:
Go to the
G(r) / Neutron S(Q) layer tab
Select the
The Isotopologue keyword currently shows that all species will “Default to Natural” isotopologues
Click the button for the Isotopologue keyword to open its full options
Press the button to populate the list with the default isotopic selection for each species
Change the isotopologue for the argon species from
NaturaltoAr36(assuming that you changed the name earlier!)
The ‘Natural’ isotopologue for each species is defined internally by Dissolve, and is always available. It does not appear in the list of defined isotopologues on the species tab.
Import Reference Data
The
NeutronSQ module itself looks after any related experimental reference data that we might wish to compare our calculated data to, and which we’ll now set up.
Go to the
Select the
For the Reference keyword open the file
yarnell.sq, and set the format of the data toxy
The data, along with its Fourier transform, can now be seen in the module’s page. You may notice that the data have been normalised to the average squared value of the atomic scattering and oscillate around 1.0 - we will need to tell Dissolve to convert the data back to absolute units and make it oscillate around zero.
Select the
Set the ReferenceNormalisedTo style to
SquareOfAverage
This tells Dissolve that the data have been normalised, and allows Dissolve to remove that normalisation in order to get the data in the correct units
Open the options for the file import at the extreme right of the ReferenceData keyword
Find the Manipulations group and set the RemoveAverage value to
9.0
This will instruct Dissolve to work out the average value of the data from a Q of 9.0 Å-1 onwards, and subtract it from the data