Neutron Reflection Analysis Programs

Adrian R. Rennie
Materials Physics, Uppsala University

Introduction

It is difficult to provide efficient data analysis software that is well-suited to a wide range of experiments. In neutron reflection studies the most general way to specify a model is in terms of layers of specified scattering length density. However, these layers may not be well matched to molecular or structural parameters in the systems that are to be investigated. For example, it may be better to think of a surface excess in an adsorption experiment rather than particular densities in layers. Computer programs can easily make these calculations and in many cases least squares minimisation is more effective with a good choice of parameters. For this reason a range of programs have been written specifically adapted to a certain types of experiments that can supplement the more general fitting programs such as drydoc and wetdoc that have models defined in terms of layers of specified scattering length density and thickness.

Executable versions of these programs are available for use under Microsoft Windows and Macintosh OSX. The latest versions can be downloaded as zip files:

Windows: http://material.fysik.uu.se/Group_members/adrian/refprog_2013.zip

MacOSX: http://material.fysik.uu.se/Group_members/adrian/RefOSX_2013.zip

Some old versions are still available.

Windows: http://material.fysik.uu.se/Group_members/adrian/refprog_Oct2012.zip

MacOSX: http://material.fysik.uu.se/Group_members/adrian/RefOSX_Oct2012.zip

Windows: http://material.fysik.uu.se/Group_members/adrian/refprog2012.zip

Mac OSX: http://material.fysik.uu.se/Group_members/adrian/RefOSX2012.zip

These programs use the Fitfun data fitting package developed at the ILL by Dr Ron Ghosh and the Clickfit GUI can be used to run the programs.

Programs

The programs can be divided in to two broad categories: (a) programs that treat single data sets in turn and (b) those that make combined fits to multiple data sets simultaneously. The programs in category (a) can also be used to make fits to sequences of data files automatically with output of fitted parameters to a comma separated variable file suitable for further processing with spreadsheet programs. These programs are litsted in the two tables below.

Fitting single data sets and sequences of similar data sets

To run these programs with a graphical user interface (GUI) the procedure Clickfit should be used. This GUI specifically allows series of data sets to be selected for processing in an automatic sequence.

Name	Purpose	Single/Multiple Data Sets	Version	Date
bike	Fit a surfactant bilayer model defined by scattering length and volumes of 'heads', 'tails' and solvent. This is designed for studies at solid/liquid interfaces and includes the possibility of an extra oxide or gel layer. There is also the possibility to include a Q dependent slope in the background.	Single	3.1	24 March 2013
byban	This program models repeating bilayers that give rise to Bragg peaks with a term to model fluctuations.	Single	1.5	24 March 2013
catamaran	A program to model a monolayer formed from two components with defined scattering length (developed from the single component program mono.	Single	1.3	24 March 2013
drydoc	Fit model defined by up to 4 layers of specified scattering length density and thickness. Interfacial roughness can also be included at each boundary.	Single	5.2	23 March 2013
jakt	Fit a two layer model defined by scattering lengths and volumes of heads, tails and solvent and area per molecule in the layer. The total thickness and number of solvent molecules are further parameters. A variant yacht fits the tail region thickness.	Single	1.3	24 March 2013
lprof	Fit model defined by a decaying profile of scattering length density. Options include half-Gaussian, exponential and scaling law functions.	Single	3.2	23 March 2013
mono	Fit a single layer model defined by a scattering length and area per molecule of the material in the layer. The thickness is a further parameter. This model is useful for simple determination of surface excess.	Single	1.7	24 March 2013
scooter	Fits a surfactant bilayer model and is analogous to bike except that there is mixing of a fraction of tails in to the head region rather than mixing of heads in to the tail region	Single	1.4	24 March 2013
spots	Fit model defined by up to 8 layers of specified scattering length density and thickness. This program is particularly intended to calculate the average reflectivity for interfaces that may have regions of different structure that are large compared to the coherence length. Interfacial roughness can also be included at each boundary.	Single	1.5	24 March 2013
trail	This is a modification of 'tram' (see below) that allows the surface structure for up to three different regions to be modelled under the condition that the areas are large (so that reflectivity rather than scattering length density are averaged. It uses a repeating bilayer model on top of a different oxide/bilayer structure to reflectivity data. The 4 layers on the surface are defined by scattering length density and thickness. The repeating bilayer is described by an area per head group and volumes of 'heads'. and 'tails'.	Single	1.4	24 March 2013
tram	Fit a repeating bilayer model on top of a different oxide/bilayer structure to reflectivity data. The 4 layers on the surface are defined by scattering length density and thickness. The repeating bilayer is described by an area per head group and volumes of 'heads'. and 'tails'.	Single	2.0	24 March 2013
velo	Fit a sinusoidal variation in density on top of a different oxide/bilayer structure to reflectivity data. The 4 layers on the surface are defined by scattering length density and thickness. The oscillatory density is described by an average scattering length density, an amplitued of oscillation, a number of oscillations and a phase term.	Single	1.6	24 March 2013
yacht	Fit a two layer model defined by scattering lengths and volumes of heads, tails and solvent molecules and area per molecule in the layer. The tail region thickness and number of solvent molecules are further parameters. This is a simple variant of jakt.	Single	1.3	24 March 2013

Fitting multiple data sets simultaneously

To run these programs with a graphical user interface (GUI) the procedure Clickfit_m should be used.

Name	Purpose	Single/Multiple Data Sets	Version	Date
cprof	Fit model defined by a decaying profile of scattering length density to data sets measured with different scattering contrasts. Options include half-Gaussian, exponential and scaling law functions. This is the multiple data set analogue to lprof	Multiple	2.1	24 March 2013
flyg	Fit a two layer model to a solid/solvent or air/solvent interface with solvent penetration in to each layer (to different amounts).	Multiple	3.0	24 March 2013
lipid	Fit a two layer model defined by scattering length density and thickness of 'tails', 'heads' and solvent. This is designed for studies of surfactants and lipids at air/liquid interfaces and includes the possibility of mixing solvent with the 'heads' and air with the 'tails'.	Multiple	1.7	24 March 2013
mbike	Fit a surfactant bilayer model defined by scattering length and volumes of 'heads', 'tails' and solvent. This is designed for studies at solid/liquid interfaces and includes the possibility of an extra oxide or gel layer.	Multiple	3.3	24 March 2013
mjakt	Models monolayers of molecules defined as heads and tails at an air/liquid interface. This is a multiple data set version of jakt which fits a total layer thickness.	Multiple	1.3	24 March 2013
myacht	Models monolayers of molecules defined as heads and tails at an air/liquid interface. This is a multiple data set version of yacht that fits the thickness of the tail layer.	Multiple	1.3	24 March 2013
rail	Models a partially submerged layer of molecules at an air/liquid interface	Multiple	1.3	24 March 2013
stige	Repeating bilayer structure (as in trikk) but allows gradient in both surfactant density (area per molecule) and in solvent spacing.	Multiple	1.4	24 March 2013
tog	Fit an oxide and gel layer model to a solid/solvent interface.	Multiple	2.6	24 March 2013
trikk	Repeating bilayer structure (as in tram)	Multiple	1.4	24 March 2013
wetdoc	Fit a model defined by up to 4 layers of specified scattering length density to multiple data sets. The model is essentially identical to that used in the single data set program 'drydoc'.	Multiple	3.0	24 March 2013
wetlay	Fit multiple bilayer of specified scattering length density at solid/liquid interface	Multiple	1.9	24 March 2013

The programs that fit multiple data sets can be used to provide a simultaneous fit to data sets either for different samples or for parts of the reflectivity measured from the same sample in different files.

Updates

Updates made on 8 August 2009 allow different Q ranges for each data set to be selected. In order to use this feature with clickfit_m, it is neccessary to download an updated clickfit_m.bat.
The programs have all been rebuilt in April 2011 to use the ffn74 and ffm77 versions of the fifun libraries. These have corrected output in the .lis files and removed the 'timeout' that caused programs to stop automatically when left without input. To use this last feature, it is necessary to have updated versions of clickfit.bat and clickfit_m.bat.
From February 2011 versions of these programs have been built for Macintosh OSX. The programs can be downloaded as a zip file RefOSX2012.zip
Programs were rebuilt in August and September 2011 to use the new features in Fitfun that include plotting error bars and the option of automatic ouput of the plot file in an automated sequence of fits.
Programs to fit multiple data sets were rebuilt on 15 August 2012 to correct mistakes in writing output of model reflectivity curves.
The new .mft format from ILL (cosmos output) was implemented in programs rebuilt on October 2012.
TPrograms were updated to use clickfit2013 and the fitfun2013 distribution in March 2013. This allows larger data arrays and the possibility to interrupt the fit iterations and display the current best fit.

Please address any comments or report any errors to Professor Adrian R. Rennie at:

Materials Physics, Uppsala University, Sweden.

Adrian.Rennie@fysik.uu.se