I came across with the following questions in reading Altland and Simons page 186


Is $\alpha=l$ the same orbital quantum number related to angular momentum $\hat{L}\to l(l+1)$?

They talk about that fermions couple to a magnetic field by their orbital momentum. Why did not they include that orbital momentum coupling? Are they talking about $$\sim \vec{L}\cdot\vec{B}$$ ?

They write down the hamiltonian $$H_0=\sum_{\alpha, \sigma}a^\dagger_{\alpha\sigma}\epsilon_\alpha a_{\alpha\sigma}$$

What is the expression for $\epsilon_\alpha$ ?



$H_0$ is the "non magnetic" Hamiltonian operator, which means the part which is not coupled to the magnetic field. In the purpose of the exercise, you can think the quantum number $\alpha$ as the set of momentum-spin $k,\sigma$. For instance if you have an electron system in a cubic lattice with hopping between nearest neighbours sites, you have $\varepsilon_{k\sigma} = -2t\sum_{\mu} \cos{(\vec{k} \cdot \hat{e}_{\mu})}$ (where $\vec{k}$ is the momentum and $\hat{e}_{\mu}$, $\mu=x,y,z$ is the versor of an axis). Anyway the explicit form is not necessary to solve the exercise.

Solution to your doubt

Your doubt is justified: if you want to give a full quantum treatment of the electron gas coupled to a magnetic field, you not only have to take into account the coupling with the spin $-\mu_B\vec{S}\cdot \vec{B}$, but also the coupling to the orbital angular momentum $-\mu_B\vec{L}\cdot \vec{B}$, which indeed affects $H_0$.

However one can prove that the magnetic susceptibility can be written as $\chi = \chi_L + \chi_S$, where $\chi_S$ is what you get if you consider the spin coupling only, while $\chi_L$ is what you get if you consider the orbital coupling only. A reference is Landau Vol. 5 (which is really precise, but a little tough to understand). In conclusion, the spirit of the exercise is to compute only one of these two contributions, in particular $\chi_S$.


It turns out that $\chi_S$ is positive (paramagnetic term), while $\chi_L$ is negative (diamagnetic term), and that $\chi_L = -\chi_S/3$ if you assume no underlying lattice structure. In conclusion the free electron gas is paramagnetic, and the effect of the term you are neglecting is to reduce the paramagnetic response by a factor $2/3$.


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.