Non-Abelian bosonization and modular transformation approach to superuniversality

TL;DR

This paper develops a non-Abelian bosonization and modular transformation framework to explain superuniversality in quantum Hall phase transitions, showing that critical exponents are independent of specific transitions within a large N expansion.

Contribution

It introduces a new effective theory with emergent $U(N)$ gauge symmetry for quantum phase transitions and demonstrates superuniversality of critical exponents across various quantum Hall states.

Findings

Critical exponents are independent of the transition type within a large N expansion.

The new theory applies to both integer and fractional quantum Hall transitions.

Superuniversality may persist beyond the large N limit according to duality conjectures.

Abstract

Quantum Hall inter-plateaux transitions are physical exemplars of quantum phase transitions. Near each of these transitions, the measured electrical conductivity scales with the same correlation length and dynamical critical exponents, i.e., the critical points are superuniversal. In apparent contradiction to these experiments, prior theoretical studies of quantum Hall phase transitions within the framework of Abelian Chern-Simons theory coupled to matter found correlation length exponents that depend on the value of the quantum critical Hall conductivity. Here, we use non-Abelian bosonization and modular transformations to theoretically study the phenomenon of superuniversality. Specifically, we introduce a new effective theory that has an emergent $U(N)$ gauge symmetry with any $N > 1$ for a quantum phase transition between an integer quantum Hall state and an insulator. We then use modular transformations to generate from this theory effective descriptions for transitions between a large class of fractional quantum Hall states whose quasiparticle excitations have Abelian statistics. We find the correlation length and dynamical critical exponents are independent of the particular transition within a controlled 't Hooft large $N$ expansion, i.e., superuniversal! We argue that this superuniversality could survive away from this controlled large $N$ limit using recent duality conjectures.