A microscopic basis is provided for the spin-fermion model used to describe the physics of the underdoped cuprates. In this way, a spin-fermion coupling is shown to take care of the local no double occupancy constraint, which is ignored in the weakly coupled regime. This ingredient is proved to be essential to capture the physics of strong correlations, however. We elaborate further on our recent proposal for the strong-coupling version of the spin-fermion model that may prove to be the correct starting point to deal with the strong correlation physics displayed by the underdoped cuprates.
We study a system of doped antiferromagnet in three dimensions at finite temperatures using the t-J model, a canonical model of strongly correlated electrons. We employ the slave-fermion representation of electrons, in which an electron is described as a composite of a charged spinless holon and a chargeless spinon. We introduce two kinds of U(1) gauge fields on links as auxiliary fields, one describing resonating valence bonds of antiferromagnetic nearest-neighbor spin pairs and the other for nearest-neighbor hopping amplitudes of holons and spinons in the ferromagnetic channel. To perform a numerical study of the system, we integrate out the fermionic holon field by using the hopping expansion in powers of the hopping amplitude, which is legitimate for the region in and near the insulating phase. The resultant effective model is described in terms of bosonic spinons, two U(1) gauge fields, and a collective field for hole pairs. We study this model by means of Monte Carlo simulations, calculating the specific heat, spin correlation functions, and instanton densities. We obtain a phase diagram in the hole concentration-temperature plane, which is in good agreement with that observed recently for clean and homogeneous underdoped samples.
Different experiments on strongly correlated materials have shown phenomena which are not consistent with our conventional understandings. We still do not have a general framework to explain these properties. Developing such a general framework is much beyond the scope of this thesis, but here we try to address some of challenges in simpler models that are more tractable. In correlated metals it appears as strong correlations have different effect on different parts of fermi surface. Perhaps most striking example of this is normal state of optimally doped cuprates; the quasiparticle peaks on the nominal fermi surface do not appear uniformly. We try to track such phenomena in heavy fermion systems, which are correlated fermi liquids. In these systems, a lattice of localized electrons in f or d orbitals is coupled to the conduction electrons through an antiferromagnetic coupling. Singlets are formed between localized and conduction electrons. This singlet naturally have non-zero internal angular momentum. This nontrivial structure leads to anisotropic effect of strong correlations. Internal structure of Kondo singlet can also lead to quantum Hall effect in Kondo insulator, and formation of isolated points on the fermi surface with fractionalized quasiparticles. In the second part we study a phase transition in Heisenberg model between two insulating phases, Neel ordered and certain spin liquid state, popular in theories of the cuprates. The existence of such a transition has a number of interesting implications for spin liquid based approaches to the underdoped cuprates and clarifies existing ideas for incorporating antiferromagnetic long range order into such a spin liquid based approach. This transition might also be enlightening, despite fundamental differences, for the heavy fermion critical points where a second order transition between the heavy fermion phase and a metallic phase with magnetic antiferromagnetic order is observed.