Abstract: When deriving a formula for the Hall conductivity of interacting electrons in our original paper [Neupert et al., Phys. Rev. B 86, 165133 (2012)], we have relied on an unjustified implicit assumption that a certain gauge choice could be made. Only under this condition would the formula follow. If this condition fails, the formula we derived does not lead to the exactly quantized value of the Hall conductance in fractional Chern insulators.
Introduction
In the study of condensed matter physics, the Hall conductivity serves as a fundamental transport coefficient, particularly in systems exhibiting topological order. In fractional Chern insulators (FCIs), this conductivity is expected to be precisely quantized. In a previous work (Neupert et al., 2012), we proposed an "elementary formula" intended to simplify the calculation of Hall conductivity for these interacting electron systems. However, subsequent critiques by Simon, Harper, Read, and Haldane identified a lack of gauge invariance in the proposed expression. This paper serves as a formal correction and clarification of the theoretical limits of that formula.
The Problem of Gauge Invariance
In quantum mechanics, a "gauge" represents a choice of coordinate system for quantum states. For a physical observable such as Hall conductivity to be considered a fundamental identity, it must be gauge invariant—meaning its value remains unchanged regardless of the mathematical perspective or coordinate system chosen.
If a formula for a physical quantity changes under a gauge transformation, it indicates that a hidden assumption has been made regarding the state of the system. The critique of our original work pointed out that our derivation relied on such a transformation, which implies that the formula was not a universal identity but rather a result dependent on a specific, unproven mathematical frame.
Theoretical Re-evaluation: The Position Operator
The core of the error lies in the treatment of the position operator ($X$) within the interacting system. In the original derivation, the operator was decomposed into two distinct components:
- $T$: An operator that shifts the momentum of the system.
- $A$: An operator acting as a "gauge field" that does not shift momentum.
Our original derivation assumed that it was always possible to select a gauge where the $T$ operator is strictly off-diagonal in momentum space. By making this assumption, the contribution of $T$ to the Hall conductivity was effectively ignored, leading to the simplified formula. In this re-evaluation, we acknowledge that this "gauge-fixing" condition was an implicit assumption. We have not provided a proof that such a choice is possible for the ground states of interacting systems, and in many cases, it likely is not.
Numerical Validity vs. Mathematical Exactness
While the formula (4.15 in the original paper) is not mathematically exact due to this gauge dependence, it remains a significant numerical tool. In all systems studied numerically thus far, the formula yields results that are remarkably close to the quantized values.
This suggests that while the $T$ operator cannot be formally ignored in a gauge-invariant proof, its contribution in practical numerical simulations of FCIs appears to be minimal. Thus, the formula serves as a high-quality numerical approximation rather than a fundamental exact identity.
Conclusion
We retract the claim that the elementary formula for Hall conductivity in interacting systems is mathematically exact. The derivation relied on an unjustified gauge-fixing assumption regarding the position operator. Consequently, the formula does not provide a universal proof of exact quantization in fractional Chern insulators. Future theoretical work is required to determine if specific classes of Hamiltonians satisfy the necessary gauge conditions or to develop a truly gauge-invariant version of the elementary expression. Despite these theoretical limitations, the formula remains a useful and accurate tool for numerical investigations into the Hall effect in complex materials.
Summary updated