For a quantum system with time-reversal symmetry, other than the absence of a magnetic field, can we infer anything else about the system?

  • 3
    $\begingroup$ The symmetry itself tells you that eigenfunctions are real--- there are constraints on the random matrix theory, what exactly are you interested in? "anything else" is too vague. $\endgroup$
    – Ron Maimon
    Mar 30, 2012 at 7:18
  • $\begingroup$ Perhaps a better way to phrase the question: other than the presence of a magnetic field, what other conditions would break time-reversal symmetry? $\endgroup$
    – leongz
    Apr 2, 2012 at 23:06
  • $\begingroup$ That's a nice one! There is the idea that you can have a spontaneous PT symmetry breaking in High Tc superconductors, due to Zee and collaborators ca 1989. Other than that, I don't know. That would be a better question, I'd +1 it in a heartbeat--- perhaps you can ask that separately. $\endgroup$
    – Ron Maimon
    Apr 2, 2012 at 23:52

1 Answer 1


We can infer some general topological properties of the system for special cases. This has been used in the recent "trending topic" of condensed matter physics, topological insulators. For simplicity, I will restrict myself in the following to $2\text{D}$ systems although one can generalize everything to $3\text{D}$.

The time-reversal operator is an antiunitary operator that admits the following representation:

$\hat{\Theta} = \exp \left(i\pi\hat{S}_y/\hbar \right)K$

where $K$ means complex conjugation and $\hat{S}_y$ is the spin operator along the $\hat{y}$ axis. Consider a fermionic Hamiltonian for spin $s=1/2$ electrons. Then

$\hat{\Theta} =-\hat{1}$ [*]

In this case, Kramers theorem applies:

Let $\hat{\mathcal{H}}$ be a $T$-invariant (fermionic) Hamiltonian. Then, all the eigenstates of the Hamiltonian are twofold degenerate.

The proof of this statement is simple once you have understood [*]. As a consequence, $T$-invariant fermionic systems must have topologically protected twofold degenerate states. The $T$-invariant Hamiltonian satisfies

$ \hat{\Theta}\hat{\mathcal{H}} (\mathbf{k}) =\hat{\mathcal{H}} (-\mathbf{k}) \hat{\Theta}$

and can be classified by a new topological index, called the $\mathbb{Z}_2$ index. The $\mathbb{Z}_2$ index, $\nu$, is an integer given by the number of edge states modulo $2$ and distinguishes the $\nu = 0$ or insulating phase from the $\nu = 1$, the topological insulator. Thus, the equivalence classes of $T$-invariant Hamiltonians for insulators can be classified by its $n = 0$ Thouless-Kohmoto-Nightingaleden Nijs invariant [i.e. its $C=0$, first Chern index] and the additional index $\nu$. This gives a $\mathbb{Z} \times \mathbb{Z}_2$ symmetry for the $2\text{D}$ band structures.

After all, what can we infer from the Hamiltonian? For example, that we have time reversal invariant electronic states with a bulk electronic band gap that supports the transport of charge and spin in gapless edge states. This is exactly what we get in the so-called quantum spin Hall phase.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.