Crystal structure solution from neutron powder diffraction data by a new Monte Carlo approach incorporating restrained relaxation of the molecular geometry

We report the development of a new aspect of the Monte Carlo method for crystal structure solution from powder diffraction data and demonstrate the advantages of this new approach by applying it to solve the low-temperature crystal structure (phase II) of perdeuterated pyrene from neutron powder diffraction data. In previous applications of our Monte Carlo technique, the structural fragment was constrained to have a standard geometry throughout the calculation, whereas in the new approach, restrained relaxation of the structural fragment from standard geometry is considered for each structure sampled during the calculation, The new approach gives rise to a substantially improved discrimination (in terms of the weighted profile R factor R-wp) between the correct structure solution and incorrect structures sampled during the Monte Carlo calculation [for the calculation with standard geometry of the structural fragment, R-wp ranged from circa 27.5-30% for 'wrong' structures to 21.2% for the best structure solution (i.e. discrimination similar to 6.3%); for the calculation with relaxation of the geometry of the structural fragment, R-wp ranged from circa 24-26% for 'wrong' structures to 11.6% for the best structure solution (i.e. discrimination similar to 12.4%)]. The work reported here represents the first application of the Monte Carlo method for structure solution using neutron powder diffraction data; the new methodology (applicable to both neutron and X-ray powder diffraction data) represents a substantial enhancement of the scope and potential of the method, particularly for cases in which the discrimination in R-wp between the correct structure solution and incorrect structure solutions is expected to be inherently small.