1
$\begingroup$

I am dealing with a Chiral-symmetric Hamiltonian such that

$$ ๐‘†๐ป๐‘†^{โˆ’1}=โˆ’๐ป. $$

Two of its eigenstates have zero eigenvalue and fulfill $๐‘†โˆฃ๐œ“_{\alpha}โŸฉ=๐‘’^{๐‘–๐œ™_{\alpha}}โˆฃ๐œ“_{\alpha}โŸฉ$, while the rest have finite eigenvalues and are pairwise symmetric. When I diagonalize my Hamiltonian numerically however, the zero energy states get mixed and the resulting eigenstates are therefore not symmetric.

When I have a symmetry that commutes with my Hamiltonian, $[๐ป,๐‘ˆ]=0$, and two degenerate eigenstates, to obtain the symmetric ones I can just add a small perturbation to the Hamiltonian proportional to $๐‘ˆ$ which breaks the degeneracy and does not modify the eigenstates. Is there a similar trick one can do to obtain symmetric degenerate eigenstates for symmetries that anticommute with the Hamiltonian?

$\endgroup$

1 Answer 1

0
$\begingroup$

Can you not apply $S$ to your numerical eigenstates? If so $(1+S)/2 $ and $(1-S)/2$ project out the $S\to \pm 1$ eigenstates.

$\endgroup$

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