The free fermion Hamiltonian for the 1D quantum Ising model is $$H = -J\sum_i (c_{i}^{\dagger }c_{i+1} +c_{i+1}^{\dagger }c_{i}+c_{i}^{\dagger }c_{i+1}^{\dagger }+c_{i+1}c_{i}-2gc_{i}^{\dagger }c_{i} +g)$$ where the sum is over $i$, the site index.

My question is - what are the basis states of this system? I read somewhere that the matrix representation of the Hamiltonian of this form has $L \times L$ elements where $L$ are the total sites. I don't see how that is. (As opposed to the spin states where the matrix would be $2^{L}\times 2^{L}$)

I eventually want to find the ground state but do not want to resort to a Fourier transform to the momentum states, so I want the matrix in the fermionic basis states form only.


The $2L$-by-$2L$ matrix is the one that is diagonalized by means of a Bogoliubov transformation. After a suitable such transformation $H$ ends up as $\sum_m E_m b_m^\dagger b_m$. I'm afraid that some Fourier stuff will be necessary. You start by writing $$ H= (c^\dagger_i,c_j) \left(\matrix{ A_{ik} & B_{il}\\ C_{jk} & D_{jl}}\right) \left(\matrix{ c_k \\ c^\dagger_l}\right). $$

| cite | improve this answer | |
  • $\begingroup$ But @mike, eventually I will want to introduce disorder (not for the Ising model, but the Kitaev model) such that k is not a good quantum number. $\endgroup$ – user2578520 Mar 20 '17 at 4:56
  • $\begingroup$ Well, for a finite chain, it always possible to do the $2L$-by$2L$ Bogoliubov diagonalization numerically. I think its actually easier for Fermions because diagonlizaing the Hamitonian can be though of as conjugating an element of the Lie algebra of ${\rm SO}(2N)$ into its maximal torus. For Bosons the linear algebra involves the non-compact {\rm Sp}(2N, {\mathbb R})$, and there are all sorts of different cases. $\endgroup$ – mike stone Mar 20 '17 at 15:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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