Reverse Monte Carlo

Reverse Monte Carlo (RMC) is a computational modelling technique used primarily in the fields of condensed‑matter physics, materials science, and chemistry to generate three‑dimensional atomic configurations that are consistent with experimental measurements, such as neutron or X‑ray diffraction, pair‑distribution functions, and other structural data. Unlike conventional Monte Carlo simulations, which explore the thermodynamic ensemble of a system based on a prescribed interatomic potential, RMC starts from experimental observations and iteratively adjusts a model structure to minimize the discrepancy between calculated and measured data.

Methodology

1. Initial configuration – An initial atomic or molecular arrangement is created, often randomly or based on known crystallographic information.
2. Forward calculation – The model’s predicted observable (e.g., structure factor, radial distribution function) is computed from the configuration using standard scattering or statistical‑mechanics formulas.
3. Error evaluation – A merit function, typically a chi‑squared (χ²) statistic, quantifies the difference between the calculated observable and the experimental data.
4. Monte Carlo move – A trial move (e.g., displacement of an atom, swapping of atomic species) is proposed.
5. Acceptance criterion – The trial move is accepted if it reduces the merit function; otherwise it may be accepted with a probability that depends on the increase in error, analogous to the Metropolis algorithm.
6. Iteration – Steps 2–5 are repeated until convergence, defined by the merit function reaching a predetermined tolerance or by stagnation of further improvements.

The algorithm does not require an explicit interatomic potential; constraints such as coordination numbers, bond‑angle limits, or density can be imposed to guide the solution toward physically reasonable structures.

Applications

• Amorphous and glassy materials – RMC has been employed to reconstruct the atomic networks of silica glass, metallic glasses, and amorphous semiconductors.
• Liquids – Structural models of liquid water, metallic liquids, and complex fluids have been derived using RMC fits to neutron and X‑ray scattering data.
• Nanostructured systems – The technique aids in interpreting scattering from nanoparticles, porous materials, and disordered alloys.
• Magnetic and electronic structure – Extensions of RMC incorporate magnetic scattering data to model spin configurations, and variants combine RMC with ab‑initio calculations for electronic structure refinement.

History
The RMC method was introduced in the early 1990s by McGreevy and Pusztai as a means of directly interpreting total‑scattering data without relying on predefined structural models. Since its inception, the approach has been refined with algorithms such as Hybrid Reverse Monte Carlo (HRMC) that integrate energy terms from empirical potentials, and with parallel‑computing implementations to handle large systems.

Limitations

• Non‑uniqueness – Because many distinct atomic configurations can reproduce the same experimental data, RMC solutions are not unique; additional constraints or complementary measurements are often required to obtain chemically realistic models.
• Dependence on data quality – The accuracy of the resulting structure is directly linked to the precision and completeness of the experimental input.
• Computational cost – For systems containing thousands of atoms, the iterative forward calculations can become computationally demanding, though modern parallel algorithms mitigate this issue.

Related techniques

• Molecular dynamics (MD) – Simulates atomic trajectories based on interatomic potentials, providing dynamical information absent from static RMC models.
• Metropolis Monte Carlo – Generates equilibrium ensembles using Boltzmann weighting; RMC differs by targeting experimental observables rather than thermodynamic probabilities.
• Maximum entropy methods – Alternative approaches for inferring structures from limited data, often used in conjunction with RMC to impose additional statistical constraints.

References
(Representative literature)

• McGreevy, R. L., & Pusztai, L. (1991). Reverse Monte Carlo simulation: A new technique for the determination of disordered structures. Molecular Simulation, 6(4), 359–367.
• Gereben, O., & Pusztai, L. (2007). Reverse Monte Carlo modelling of disordered materials: From the structure factor to the pair distribution function. Journal of Physics: Condensed Matter, 19(33), 335212.

External links

• Software packages implementing RMC, such as RMCProfile and RMC++.

This entry presents an overview of the Reverse Monte Carlo method as established in the scientific literature.