REVERSE MONTE CARLO MODELLING

The reverse Monte Carlo (RMC) method for modelling structural disorder began in Oxford in 1988 as a method for creating three-dimensional models of liquid structures from neutron total scattering data.  The method has been developed considerably since then and its scope has been broadened to include, amongst others, crystalline and amorphous structures and magnetic materials.  Robert McGreevy (ISIS), one of the originators of the method, has produced a comprehensive review on the subject [1].  In essence the method is very strightforward; a three-dimensional configuration of atoms is constructed and randomly-chosen atoms are moved one at a time a random amount using periodic boundary conditions.  Each time an atom is moved correlation functions are calculated which are compared to experimental data using a simple c2 function.  A move is accepted depending on the change to c2.  The process is repeated until c2 has converged and the configuration can be said to represent the structure of the material giving rise to the experimental scattering functions.  This configuration may then be used to provide additional information about the nature of the structure of the material at that temperature, degree of doping, pressure etc.

I have been involved in a number of RMC developments, including early work which used a combination of neutron and x-ray total scattering structure factors [2] and the initial work to incorporate magnetic scattering [3].  I have been investigating how chemical constraints, such as those which would describe local chemical units may be used in RMC methods [4] and how to include the effects of experimental resolution in RMC codes [5].  The latter has produced a number of RMC programs designed primarily for the analysis of data from the GEM instrument at ISIS, including MCGRtof [6] (an inverse method for producing G(r) from i(Q) taking into account instrumental resolution) and RMCprofile [7] (which also fits the Bragg profile within the RMC refinement). 

I have used RMC modelling to investigate a large number of materials, including several polymorphs of silica, superionic conductors such as AgI, AgBr and Ag2Te and a variety of other disordered crystalline systems, such as SF6, and C4F8.  References to these and other compounds may be found in my general publication lists.

References.  

 

1. Reverse Monte Carlo modelling R. L. McGreevy J. Phys.: Condensed Matter 13 (2001) R877-913 
2. Structural modelling of glasses using reverse Monte Carlo simulation D. A. Keen and R. L. McGreevy Nature 344 (1990) 423-5
3. Determination of disordered magnetic structures by RMC modelling of neutron diffraction data  D. A. Keen and R. L. McGreevy J. Phys.: Condensed Matter 3 (1991) 7383-94
4. Refining disordered structural models using reverse Monte Carlo methods: Application to vitreous silica D. A. Keen Phase Transitions 61 (1997) 109-24
5. Application of the reverse Monte Carlo method to crystalline materials M. G. Tucker, M. T. Dove and D. A. Keen J. Appl. Cryst. 34 (2001) 630-8
6. MCGRtof: Monte Carlo G(r) with resolution corrections for time-of-flight neutron diffractometers M. G. Tucker, M. T. Dove and D. A. Keen J. Appl. Cryst. 34 (2001) 780-2
7.  RMCProfile: Reverse Monte Carlo for polycrystalline materials M G Tucker, D A Keen, M T Dove, A L Goodwin and Q Hui J. Phys.: Condensed Matter 19 335218 (2007)

 

 

Home