Studies of condensed matter excitations in bicollinear magnetic materials and disordered perovskites using inelastic neutron scattering techniques

Abstract

Neutron scattering is an important research technique that has furnished a wealth of information about condensed matter systems ever since Ernest O.

Wollan and Clifford G. Shull first demonstrated its ability to probe the structure of polycrystalline samples, such as NaCl, on the Clinton Pile—a graphite moderated nuclear reactor built for research purposes at Oak Ridge, Tennessee. Since those early successful demonstrations, the experimental technique has been continuously improved such that it is now regularly used as a research tool by scientists in a plethora of different fields to glean information about the structure or dynamics of their samples. An example of these advances in the use of neutron scattering came when Bertram N. Brockhouse invented the triple-axis spectrometer, which was an instrument that facilitated the study of lattice dynamical behaviour within condensed matter systems. Without this innovation it would be much more challenging to study phonons—the quasiparticles representing the collective vibrations of atoms—within the perovskites and iron chalcogenides that were researched for this thesis.

This thesis details how neutron scattering has been used as a tool to study the structure and lattice dynamical behaviour of the relaxor ferroelectric, PbMg1/3Nb2/3O3, and the iron-based chalcogenide, Fe1+yTe. The study of PbMg1/3Nb2/3O3 was to investigate the structure of the higher energy phonon modes, complementing previous research that successfully characteristed the lower energy acoustic and optical phonons of this crystal, and to use information obtained from neutron spectroscopy data to compare the lattice dynamical behaviour of this crystal with ordered perovskites, such as SrTiO3. An interest in studying the higher energy phonons has been aroused because the disorder within PbMg1/3Nb2/3O3, such as the presence of short-range polar nano-regions between the ferroelectric Curie temperature and the so-called Burns temperature (the temperature at which the short-range polar order emerges), leads to the energy broadening and dampening of the transverse optical phonons and triggers their precipitous collapse into the acoustic phonon branch. Previous neutron scattering studies suggested that the band of phonon scattering located at energy scales where the higher energy longitudinal and transverse optical phonons were predicted to be present were due to the occupancy disorder on the B-site of the perovskite between the Mg and Nb atoms. However, in this thesis investigations of the longitudinal and transverse scattering channels revealed lattice dynamical behaviour similar to those expected in SrTiO3, which has no occupancy disorder on the B-site. The hypothesis of disorder within the higher energy modes is further negated by studying the structure factors at various reciprocal lattice vectors that reveals a close agreement with a model for the lattice dynamics that involves the motion of the oxygen and B-site ions. The second study within this thesis investigates the softening of the transverse acoustic phonons within the Fe1+yTe system using time-of-flight neutron spectroscopy. This iron-based system has a complex structural, magnetic and electronic order that depends on the concentration of interstitial Fe atoms within the crystal structure and is often compared with other iron-based systems that share similar physical properties, such as FeSe, underdoped Ba(Fe0.94Co0.03)2As2 and optimally doped Ba(Fe0.94Co0.06)2As2. Studies of these systems revealed softening of the acoustic phonons close to the Brillouin zone centre (q∼ 0) that is indicative of the role that nematic order of the electronic charge has on the lattice dynamics within their crystal structures. However, investigations into the softening of the transverse acoustic phonon within low-y Fe1+yTe reveal that the softening within this system is mostly contained to the Brillouin zone edge and that the entire phonon branch is being driven to lower energies by scattering channels that open up as the tetragonal–monoclinic phase transition temperature is approached. The origin of the scattering channels is unknown, but the electronic nematic order could still be playing a role, albeit one that differs in effect to that observed in the other iron-based systems.

The final study is a theoretical investigation into the origin of the bicollinear double stripe antiferromagnetic order that is observed in Fe1+yTe with a low interstitial iron concentration. The hypothesis is that this complicated magnetic ground state comes about due to the interplay between the spin and orbital degrees of freedom within the crystal due to the Jahn-Teller effect and the small energy scale associated with the splitting of the dXz and dY z orbitals in the magnetic (S=1) Fe atoms. This builds on previous theoretical studies which proposed models of the magnetic interactions within the FeTe single crystals that negated the possibility of an interplay between the spin and orbital degrees of freedom and hoped to explain the double stripe magnetic order by a model that included either the spin or orbital physics but not both. A Rayleigh-Schrödinger perturbative model has been developed that includes the second and fourth order virtual hopping of the electrons on neighbouring magnetic Fe atoms surrounded by tetrahedra of non-magnetic Te atoms. A domain space for each of the Fe atoms in the S=1 spin state has 3 degrees of freedom for the spin projection and 2 degrees of freedom for the orbital projection meaning that the resultant Hamiltonian that couples two neighbouring magnetic Fe atoms has a 36×36 matrix structure. This effective Hamiltonian leads to interactions that include spin-only terms, orbital pseudo-spin only terms and terms that describe the interaction between both the spin and orbital physics. The spin operators include single-ion anisotropy, an Ising term that correlates spin projections, an Ising-like term that correlates “mid-planeness” and a spin exchange term. The orbital pseudo-spin operators include a single-site “orbital transverse field” term and a two-site “orbital transverse field” term. The interplay between the spin and orbital degrees of freedom is produced by the multiplication of the spin-only and orbital pseudo-spin only terms in the effective Hamiltonian.