Unified role of Green's function poles and zeros in topological insulators

Andrea Blason,Michele Fabrizio
DOI: https://doi.org/10.1103/PhysRevB.108.125115
2023-09-06
Abstract:Green's function zeros, which can emerge only if correlation is strong, have been for long overlooked and believed to be devoid of any physical meaning, unlike Green's function poles. Here, we prove that Green's function zeros instead contribute on the same footing as poles to determine the topological character of an insulator. The key to the proof, worked out explicitly in 2D but easily extendable in 3D, is to express the topological invariant in terms of a quasiparticle thermal Green's function matrix $G_*(i\epsilon,\mathbf{k})= 1/\big(i\epsilon-H_*(\epsilon,\mathbf{k})\big)$, with hermitian $H_*(\epsilon,\mathbf{k})$, by filtering out the positive definite quasiparticle residue. In that way, the topological invariant is easily found to reduce to the TKNN formula for quasiparticles described by the non-interacting Hamiltonian $H_*(0,\mathbf{k})$. Since the poles of the quasiparticle Green's function $G_*(\epsilon,\mathbf{k})$ on the real frequency axis correspond to poles and zeros of the physical-particle Green's function $G(\epsilon,\mathbf{k})$, both of them equally determine the topological character of an insulator.
Mesoscale and Nanoscale Physics,Strongly Correlated Electrons
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: **The physical meaning of the zeros of the Green's function in strongly - correlated topological insulators and their contribution to topological invariants**. Specifically, the paper points out that the zeros of the Green's function have long been considered to have no physical meaning, while in fact these zeros play an equally important role in determining the topological properties of insulators as the poles of the Green's function. The paper proves this by introducing the quasiparticle thermal Green's function matrix \(G^*(i\epsilon, k)=\frac{1}{i\epsilon - H^*(\epsilon, k)}\), where \(H^*(\epsilon, k)\) is a Hermitian matrix, and filtering out the positive - definite quasiparticle residues. ### Main contributions: 1. **Physical meaning of the zeros of the Green's function**: - The paper proves that the zeros of the Green's function, like the poles, contribute to topological invariants. - This finding challenges the traditional view that only the poles of the Green's function have physical meaning. 2. **Expression of topological invariants**: - The paper derives an expression that represents topological invariants as the winding number \(W(G^*)\) of the quasiparticle Green's function \(G^*(i\epsilon, k)\). - This winding number can be simplified to the TKNN formula, which is applicable to quasiparticles described by the non - interacting Hamiltonian \(H^*(0, k)\). 3. **Topological transitions in strongly - correlated systems**: - The paper explores how the system undergoes topological transitions when the zeros of the Green's function cross the chemical potential in strongly - correlated systems. - This phenomenon provides the physical mechanism of non - adiabatic processes and explains why topological invariants can change without closing the charge energy gap. ### Mathematical derivations: The paper derives the above conclusions in detail through the following steps: 1. **Representation of the Green's function**: \[ G(i\epsilon, k)=\frac{1}{i\epsilon - H_0(k)-\Sigma(i\epsilon, k)} \] where \(H_0(k)\) is the non - interacting Hamiltonian and \(\Sigma(i\epsilon, k)\) is the self - energy matrix. 2. **Quasiparticle Green's function**: \[ G(i\epsilon, k)=\sqrt{Z(\epsilon, k)}\frac{1}{i\epsilon - H^*(\epsilon, k)}\sqrt{Z(\epsilon, k)} \] where \(Z(\epsilon, k)\) is the positive - definite quasiparticle residue matrix and \(H^*(\epsilon, k)\) is the Hermitian quasiparticle Hamiltonian. 3. **Calculation of the winding number**: \[ W(G)=\frac{1}{24\pi^2}\int d\epsilon \, dk \, \epsilon_{\mu\nu\rho}\text{Tr}\left(G(i\epsilon, k)\partial_\mu G^{-1}(i\epsilon, k)G(i\epsilon, k)\partial_\nu G^{-1}(i\epsilon, k)G(i\epsilon, k)\partial_\rho G^{-1}(i\epsilon, k)\right) \] 4. **Simplification to the TKNN formula**: \[ W(G)=\frac{i}{2\pi}\int dk \, \epsilon_{ij}\sum_\alpha\theta(-\epsilon_\alpha(0, k))\langle\partial_i\alpha(0, k)|\partial_j\alpha(0, k)\rangle