Abstract
We demonstrate an exact local transformation which maps a purely Fermionic many-body system to a system of spinful bosons and spinless fermions, demonstrating a possible path to a non-Fermi-liquid state. We apply this to the half-filled Hubbard model and show how the transformation maps the ordinary spin half Fermionic degrees of freedom exactly and without introducing Hilbert space constraints to a chargelike quasicharge fermion and a spinlike quasispin Boson while preserving all the symmetries of the model. We present approximate solutions with localized charge which emerge naturally from the Hubbard model in this form. Our results strongly suggest that charge tends to remain localized for large values of the Hubbard U.
Introduction
In the study of condensed matter physics, the behavior of electrons in standard metals is typically described by Fermi-liquid theory, where electrons act as quasiparticles carrying both spin and charge. However, in strongly correlated materials—such as high-temperature superconductors—this paradigm fails. Physicists have long hypothesized "spin-charge separation," a phenomenon where spin and charge degrees of freedom decouple and move as independent entities, known as spinons and holons.
While spin-charge separation is well-established in one-dimensional systems, extending this description to two or three dimensions has proven mathematically daunting. Previous attempts, such as "slave-boson" methods, often expand the Hilbert space beyond physical reality, necessitating complex constraints to filter out unphysical states. This paper by Stellan Östlund and Mats Granath introduces an exact, local transformation that achieves spin-charge separation without such constraints, offering a clearer path to understanding non-Fermi-liquid states.
Methodology: A Constraint-Free Mapping
The core of this work is a nonlinear local operator transformation. The authors map standard Fermionic operators into two distinct types of quasiparticles:
- Quasicharge ($\hat{c}$): Spinless Fermions.
- Quasispin ($q$): Spin-1/2 Bosons.
Unlike traditional slave-particle approaches, this transformation is a one-to-one mapping that stays strictly within the physical Hilbert space. By incorporating a specific phase factor ($(-1)^r$) based on lattice site coordinates, the authors preserve essential symmetries, including SU(2) spin symmetry, without requiring mathematical "handcuffs" or unphysical state filtering. This allows the Hubbard Model—a fundamental model for many-body interactions—to be rewritten in a form that separates the Hamiltonian into a spin-only sector and a charge-coupled sector.
Key Insights and Theoretical Framework
The transformation reveals several profound properties of the system:
- The Quasispin Identity: Quasispin is shown to be the sum of physical spin and "pseudospin" (related to site occupancy). This provides a transparent mechanism for tracking how magnetic order influences charge dynamics.
- Nagaoka Ferromagnetism: The mathematical framework naturally recovers the Nagaoka theorem. In a system with strong repulsion, a single hole can lower its energy by aligning surrounding spins ferromagnetically, facilitating its movement.
- Symmetry Preservation: The mapping is constructed to respect the underlying symmetries of the lattice, ensuring that the resulting bosonic and fermionic degrees of freedom are physically meaningful.
Results: Charge Localization and Stability
By applying this transformation to the Hubbard model, the authors found that the interaction term ($U$) effectively counts the number of "quasicharge" particles. At the "half-filling" limit (one electron per site), the model reduces exactly to the well-known Heisenberg model of magnetism.
A critical finding of the study is the emergence of localized charge states. For large values of the Hubbard $U$ (strong electron repulsion), it is energetically more favorable for a charge to remain "trapped" within a small ferromagnetic cluster inside an antiferromagnetic background, rather than propagating as an extended wave.
The stability of these localized clusters suggests that strongly correlated materials may not behave as standard metals at low temperatures, as the charge carriers are effectively pinned by the magnetic environment.
Conclusion and Future Directions
The exact transformation presented by Östlund and Granath provides a rigorous foundation for studying spin-charge separation in higher dimensions. While the transformation itself is exact, the authors utilized mean-field approximations to solve the resulting equations for 2D lattices, which averages out some magnetic fluctuations. Future work may extend this approach to intermediate interaction strengths ($U$) and investigate the real-time dynamics of these separated quasiparticles. Nonetheless, the demonstration of stable localized charge states marks a significant step toward a complete theory of non-Fermi-liquid behavior.
Summary updated