Reverse Monte Carlo Studies of Inorganic Functional Materials Based on Neutron Scattering

Abstract

In this thesis, the crystal structure and atomic dynamics of four inorganic functional materials were studied based on powder neutron total scattering data. The studies focused on the structural features that depend on the changes of temperature, especially when the materials go across various phase transitions. The reverse Monte Carlo (RMC) modelling method was used to generate supercell configurations based on the neutron data. This modelling method is different from many other crystallographic analysis that result in average positions of atoms. The supercell configuration generated from RMC contain atoms at instantaneous positions. This makes it possible to study the arrangement and the fluctuations of the atoms in the structures as the temperature changes. In this study, one negative thermal expansion (NTE) material and three multiferroics are studied. The NTE material ScF3 is interesting because although it has a very simple cubic structure, up to the time the author wrote this thesis, it still lacks a experiment-based theory which can draw a complete picture of the mechanism of its NTE phenomenon. CuO is interesting as a newly-discovered high-temperature multiferroic material. It will bring significant benefits if its mechanism of the magneto-electric phase can be understood. BiFeO3 has been a hot multiferroic material to material scientists for many years. Even though having been studied for so many years, many of the aspects of this material are still mysterious to scientists. This includes lots of ’ghost’ phases at certain temperatures which may or may not be introduced by impurities, and also about the couplings between the magnetism and the intrinsic dielectric polarization. Studying the powder sample with the RMC method is helpful to find the intrinsic changes of the atomic positions and their fluctuations when going through various temperatures. YMnO3 is also a multiferroic material that has studied a lot. Its attraction lies in the intrinsic strong coupling between the magnetism and the dielectric polarization. The RMC method is performed on this material in this study with the hope to reveal the evolution of the atomic fluctuations in a large range of temperature with the multiferroic phase lies in it. The first material studied in this work is ScF3 which is a negative thermal expansion (NTE) material. It is an attractive material from the aspects of both the internal mechanism of the NTE and its potential applications. It has a simple cubic structure but has NTE in a very large range of temperature. The work in this thesis reveals that the tension effect plays an important role in the formation of the NTE in this material. When the temperature increases, the transverse motions of F atoms become large and they stretch the Sc–F–Sc chain and then shorten the Sc–Sc lengths resulting in the decrease of the cell parameters. The transverse motions of the F atoms were found to be mostly harmonic and the anharmonic components are not visible, contrary to some recently published ideas. The second material studied is CuO. CuO is well-known to be an antiferromagnetic material at low temperatures. It was discovered to be multiferroic about ten years ago. Its ferroelectricity exists in a small range of temperature about 100 K below the room temperature. The ferroelectric phase is also antiferromagnetic but with a different magnetic ordering from the pure antiferromagnetic phase. Detailed studies on the atomic structure and the fluctuations of the atoms were performed by RMC. Results show that the Cu and O atoms are restrained from having much fluctuations along the average positions. This is likely caused by the specific atomic arrangement. Anomalies were found near the multiferroic phase from the distribution of the pseudodipole moments of OCu4 and CuO4 units (the ’pseudo-dipole moment’ here is used to denote the displacement from an cation atom to the center of the anion neighbours without taking account of the charges). It indicates the coupling between the magnetic ordering and the structure-induced polarisation in this material. The anomalies can also be observed around the temperature where the ferroelectricity appears and the arrangement of the magnetic ordering changes, indicating that the magnetic structure has an effect on the structure in this material. The proportions of the three types of distortions including the rotation of the polyhedral units, the bending and bond stretching in the units were obtained based on the RMC-refined configurations and the results were discussed. The third material studied here is BiFeO3. RMC-refined configurations were obtained from neutron data in a large temperature range from 15 K to 800 K. BiFeO3 as a famous and an attractive multiferroic material to study mostly because its antiferromagnetic phase transition occurs above room temperature. This is very rare among other multiferroics. Few anomaly was observable in the range of temperature studied even near the antiferromagnetic phase transition temperature. The conflicts on the anomalous phases and properties at some temperature reported in some literatures seem not to be intrinsic but induced either from interfaces or various boundaries. Fairly consistent evolution of thermal motions of the atoms in this material were observed. From the results, it seems that the atomic fluctuations in this material are very large. Results also show that when the temperature increases, there seem to be little tendency for the atoms to arrange to form a more symmetric structure in spite that the highest temperature studied is only a few hundreds below the phase with cubic structure. The last material studied is the multiferroic material YMnO3. This material is famous for its large magneto-electric coupling. From RMC-refined supercell configurations, YO8 and MnO5 polyhedra were analysed as pseudo-dipoles. Anomalies were seen from the average values as well as the fluctuations of YO8 and MnO5 pseudo-dipole moments when the temperature goes across the antiferromagnetic phase transition indicating the features of the magnetoelectric coupling. The so-called isosymmetric phase transition in the reported studies does not appear to affect the pseudo-dipole moments of MnO5 and only partly affect the moments of YO8.