Neutron scattering studies of some NaFeAs and BaFe2As2 derivatives and Ce1-xYbxCoIn5

Abstract

Within this thesis I present several neutron scattering works on unconventional superconductors, including derivatives of iron pnictides BaFe$_2$As$_2$, NaFeAs and the heavy fermion compound CeCoIn$_5$. Similar to the cuprates, superconductivity in these systems appear in proximity to magnetically ordered phases and strong magnetic fluctuations persist in paramagnetic superconductors. Neutron scattering is an ideal probe to study spin-spin correlations and hence the interplay between magnetism and superconductivity.

In the iron pnictide parent compounds BaFe$_2$As$_2$ and NaFeAs I demonstrate that while applying uniaxial stress may affect the structural and magnetic transition temperatures, the effect is too small to account for the large temperature range over which resistivity anisotropy has been observed, confirming the resistivity anisotropy is intrinsic rather than being linked to a electronically ordered phase (stripe magnetic order) with broken rotational symmetry. Doping Ni into BaFe$_2$As$_2$ to the underdoped regime, I found uniaxial stress can be used to enhance the ordered moment. This is a novel manifestation of the competition between superconductivity and magnetic order since superconductivity is suppressed with uniaxial stress.Increasing Ni concentration towards optimal doping where long-range magnetic order is suppressed, I found the anisotropy of spin fluctuations as a function of energy transfer can be described as a power law decay. This result suggest spin fluctuations rather than orbital ordering drive the Ising-nematic correlations and agrees with proposals of the system being close to magnetic and nematic quantum critical points.

Using polarized neutron scattering I studied the evolution of spin anisotropy in NaFe$_{1-x}$Co$_x$As including the the parent ($x=0$), under-doped ($x=0.015$) and over-doped ($x=0.05$) compounds. Spin anisotropy in NaFeAs can be described by two spin anisotropy terms similar to BaFe$_2$As$_2$ with $c$-axis being the easy-axis, in addition the likely presence of a longitudinal mode highlight the itinerant aspect of magnetism in iron pnictides. In the under-doped compounds polarization reveals the double neutron spin resonance modes seen in this compound has different characteristics, with the first mode being anisotropic and the second mode being isotropic. In the over-doped compound the single resonance is isotropic. In optimal hole-doped (Ba,K)Fe$_2$As$_2$ without static magnetic order, spin anisotropy at $E =3$ meV persists to $\sim 100$ K, while the resonance has a $c$-axis polarized component as well as an isotropic component. While the energy scales of spin anisotropy in these systems are low, they are nevertheless comparable to the superconducting gap and may be an important piece of the puzzle.

CeCoIn$_5$ is believed to be an unconventional superconductivitor with $d_{x^2-y^2}$-wave pairing symmetry similar to the cuprates. Therefore it is anticipated that the neutron spin resonance in this system should also display a prominent downwards dispersion similar to the cuprates. On the contrary I found the resonance mode displays a prominent upwards dispersion that is robust against Yb doping, which induces significant changes to the Fermi surface. This result challenges the predominant view that the resonance can be understood from an itinerant perspective, but instead suggest that a robust nearest-neighbor coupling between Ce$^{3+}$ ions is responsible for the upward dispersion within magnon-like scenario.