The spin-rotationally invariant SU(2) approach to the Hubbard model is extended to accommodate the charge degrees of freedom.Both U(1) and SU(2) gauge transformation are used to factorize the charge and spin contribution to the original electron operator in terms of the emergent gauge fields.It is shown that these fields play a similar role as phonons in the BCS theory: they provide the "glue" for fermion pairing.By tracing out gauge bosons the form of paired states is established and the role of antiferromagnetic correlations is explicated.
Superconductivity in the strongly correlated electronic systems described by the Hubbard model for slightly less than half-filled band is shown to arise from the condensation of charge + e holons which are bosons. The holon effective mass is found to be very close to that of the band electrons and they have a strong nearest neighbour repulsion induced by the spin fluctuations. They interact with an U(1) part of the collective SU(2) gauge field (named S-gauge field) of the spin system. This results in spontaneous symmetry breaking of the holon field and the appearance of the Anderson-Higgs phenomenon. The interesting possibility of the SU(2) gauge field-spinon coupled system retaining some gapless excitation spectrum inspite of the spontaneous symmetry breaking in the holon field is pointed out.
In this thesis I present a microscopic model for the high-To cuprates, and analyze the phase diagram predicted by this model.The model is based on the t-J Hamiltonian, which describes the physics of a single copper-oxide plane.An important feature of the t-J model is that close to half filling the motion of the electrons is controlled by the empty sites in the system.This leads to spin-charge separation, which I take into account using the slave-boson formalism.In this approach the electrons are split into spinons and holons, which carry, respectively, the spin and the charge degrees of freedom of the electrons.Superconductivity will occur when the spinons form Cooper pairs and the bosons are Bose condensed.I first analyze the model in the mean-field approximation.Novel flux phases are found at low doping, indicating that the system tries to simulate a Nel ordered state to minimize its magnetic energy.At intermediate doping the mean-field analysis predicts a superconducting gap with a d-wave symmetry.7 A drawback of the mean-field analysis is that it grossly overestimates the superconducting transition temperature, especially at low doping.One can improve on the mean-field results by including a fluctuating gauge field in the analysis.This gauge field mediates the interactions between the spinons and the holons, which are strongly coupled.I show in Chapter 3 that the fluctuating gauge field suppresses the tendency of the holons to Bose condense.8 In order to obtain this result I developed a new technique based on a summation over Feynman paths which is a very convenient approach in the strong-coupling limit t > J.Chapter 4 focuses on the pair-breaking effects of the gauge field, which can be analyzed by studying the contribution of the gauge field to the total free energy.I show that there is a significant reduction of free energy due to gauge field fluctuations, which 7 M.
This thesis deals with different aspects of high temperature superconductivity in hole-doped cuprates. We assume the t - J model to describe the CuO2 planes of cuprates and we use a spin-charge gauge approach with spin-charge separation to describe holes in terms of a spinless fermion carrying the charge (holon) and a neutral boson carrying spin 1/2 (spinon), coupled by a slave-particle gauge field. In this framework we consider the effects of the presence of a finite density of incoherent holon pairs in the normal state as a precursor of superconductivity. We show that it is possible to take such pairs into account through a strongly direction dependent spectral weight for holons, which suppresses quasi-particles starting from the anti-nodal directions, and a wave function renormalization for the slave particle gauge field. In this way we prove that, when temperature is reduced, the formation of holon pairs causes both the deviation from linearity of the resistivity and the deviation from the constant value of the Knight shift. Moreover we point out the need for a negative next-nearest-neighbor hopping term ((t - t' - J model) to get a good continuum limit and to evaluate the Knight shift. These results are obtained through a Green's function, which follows naturally from the formalism and analytically interpolates between a Fermi liquid-like behavior and a d-wave superconductor when the coherence length of the holon pair order parameter is increased. The system preserves a finite Fermi surface until the superconducting transition, where it reduces to four nodes. Finally when, at lower temperatures, both the holon pairs and the RVB spinon pairs condense we enter the superconducting phase. In this phase we study the magnetic excitations showing how to extend to spin waves the hourglass-shape dispersion, found for spinons near the antiferromagnetic vector, if a suitable local mechanism of attraction between spinon and anti-spinon is assumed. The resulting spin wave Green's function, whose imaginary part is directly comparable with experiments, generates the hourglass with a finite gap between the two branches. Since the U(1) slave-particle gauge field gains mass in the superconducting phase via Anderson-Higgs mechanism, we propose the necessary mechanism of attraction comes from the unbroken Z2 subgroup of the U(1) gauge group.
High Tc, cuprates have a number of features that are anomalous from the point of view of the conventional theory of metals, i.e. the Fermi liquid theory. These include the high Tc itself, the linear resistivity in optimally doped cuprates, the "spin gap" phenomena, and antiferromagnetic fluctuations. Some of the basic features of the high Tc, phase diagram can be readily understood from slave boson mean field theories of the t-J model that yield a variety of phases like the uRVB phase (strange metal), d-wave RVB phase (spin gap), and the superconducting phase. This thesis is concerned with improving this picture by considering important fluctuations around the mean fields. This leads to certain gauge theories in which neutral spin 1/2 fermions ("spinons") and charged spinless bosons ("holons") are coupled to gauge fields. If the gauge field is not "confining", the spin and charge degrees of freedom are separated to some extent, and an unconventional picture for transport and magnetic properties is expected. Following an introduction and overview, the first half of the main body of this thesis analyzes a simple model of a degenerate two-dimensional Bose liquid interacting with a fluctuating gauge field, with the goal of studying the charge degree of freedom in the cuprates. It is shown that the fluctuating gauge field efficiently destroys superfluidity even in the Bose degenerate regime. The nature of the resulting normal state is discussed in terms of the geometric properties of the imaginary-time paths of the bosons. Charge response functions are studied numerically (by path integral Monte Carlo methods); it is found that the transport scattering rate behaves as ... , consistent with the experiments on the cuprates in the normal state, and that the density correlations of our model resemble the charge correlations of the t-J model. The second half considers the magnetic properties of cuprates from the point of view of a gauge theory of spinons, with emphasis on the underdoped regime. Despite the spin gap, there is a substantial antiferromagnetic correlation in the underdoped cuprates, as evidenced by the q-space scan of Neutron scattering cross section and different temperature dependences of the Copper and Oxygen site relaxation rates, features which are not captured well by mean field theories. As a concrete illustration of the gauge-fluctuation restoration of the antiferromangetic correlation and the feasibility of the 1/N perturbation theory, a U(1) gauge theory of id Heisenberg spin chain is worked out exactly, and then perturbatively. The difference between the behavior of uniform and staggerd spin correlation functions is discussed in terms of conserved and nonconserved currents. The 1/N-perturbative treatment of the gauge theory in 2+1D (which can be motivated from the mean field Flux phase of the Heisnberg model) leads to a dynamical mass generation corresponding to an antiferromagnetic ordering, but it is argued that in a similar gauge theory with an additional coupling to a Bose (holon) field, symmetry breaking does not occur, but antiferromagnetic correlations are improved, which is the situation in the underdoped cuprates.
We study the slave-boson $t\ensuremath{-}J$ model of cuprates with high superconducting transition temperatures, and derive its low-energy effective field theory for the charge-spin separated state in a self-consistent manner. The phase degrees of freedom of the mean field for hoppings of holons and spinons can be regarded as a U(1) gauge field,
Ai. The charge-spin separation occurs below a certain temperature,
TCSS, as a deconfinement phenomenon of the dynamics of
Ai. Below a certain temperature
TSG(<TCSS), the spin-gap phase develops as the Higgs phase of the gauge-field dynamics, and
Ai acquires a mass
mA. The effective field theory near
TSG takes the form of a Ginzburg-Landau theory of a complex scalar field $\ensuremath{\lambda}$ coupled with
Ai, where $\ensuremath{\lambda}$ represents d-wave pairings of spinons. Three dimensionality of the system is crucial to realize a phase transition at
TSG. By using this field theory, we calculate the dc resistivity $\ensuremath{\rho}.$ At
T>TSG, $\ensuremath{\rho}$ is proportional to T. At
T<TSG, it deviates downward from the T-linear behavior as $\ensuremath{\rho}\ensuremath{\propto}T{1\ensuremath{-}{c(T}_{\mathrm{SG}}\ensuremath{-}{T)}^{d}}.$ When the system is near (but not) two dimensional, due to the compactness of the phase of the field $\ensuremath{\lambda},$ the exponent d deviates from its mean-field value
1/2 and becomes a nonuniversal quantity which depends on temperature and doping. This significantly improves the comparison with the experimental data.