# Confused about perturbation theory with path integrals

1. Jun 21, 2015

### Hazz

1. The problem statement, all variables and given/known data
Hi, I am just trying to wrap my head around using path integrals and there are a few things that are confusing me. Specifically, I have seen examples in which you can use it to calculate the ground state shift in energy levels of a harmonic oscillator but I don't see how you can find the shift for any other energy level. I am aware you don't need to use a path integral for this but I am hoping to get a better understanding of them by doing this.

2. Relevant equations
I know that you can calculate the energy shift in the ground state using the partition function by using

$E_0=-lim_{\beta\rightarrow\infty}\frac{1}{\beta} log K_E[J] |_{J=0}$

I know if I am using the Hamiltonian
$\mathcal{H}=\frac{\hat{p}^2}{2m}+\frac{1}{2}m\omega^2\hat{q}^2+\gamma\hat{q}^4$

then to first order you can just write $K_E[J]$ as

$K_E[J]=K_E^0[J]-\gamma \int d\tau (\frac{\delta}{\delta J(\tau)})^4K^0_E[J]|_{J=0}$

where

$K_E^0[J]=\frac{C}{2}\int d\tau d\tau'J(\tau)G(\tau,\tau')J(\tau')$

and G is the green's function.

3. The attempt at a solution

So you can evaluate this all and work it out for the ground state which I don't have a problem with, but what confuses me is how this works for anything that isn't the ground state. Clearly the $\beta\rightarrow\infty$ limit is always going to put you there.

I understand that you probably need to use

$\langle 0 |\hat{a}e^{-\beta H} \hat{a}^\dagger |0\rangle$

In some capacity but I don't really see where to proceed from there. Do you have to worry about commuting $\hat{a}^\dagger$ through the exponential?

I don't really know where to look for guidance with this sort of thing so any help would be appreciate. Thank you very much!

2. Jun 27, 2015