I have the free action in position space

$$S_0[J,\phi] = \int \! d^4x[\frac{1}{2} \phi (\partial_\mu \partial^\mu - m^2 + i\epsilon)\phi + \frac{\hbar}{i}J\phi].$$

Knowing that the Fourier transform of $ \phi(x) $ gives it's representation in momentum space, and likewise for $J(x)$ how do I find the free action in momentum space?

I've tried plugging the Fourier transforms into the free action, but can't seem to simplify the expression. I think I have an algebra issue.

This is how far I've got:

$$ S_0[J,\phi] = \int \! d^4x [\frac{1}{2}(\int \! \frac{d^4k}{(2\pi)^2} \tilde{\phi}(k) e^{-ik_\mu x^\mu} )(\partial_\mu \partial^\mu - m^2 + i\epsilon) (\int \! \frac{d^4k}{(2\pi)^2} \tilde{\phi}(k) e^{-ik_\mu x^\mu}) + \frac{\hbar}{i} (\int \! \frac{d^4k}{(2\pi)^2} \tilde{J}(k)e^{-ik_\mu x^\mu}) (\int \! \frac{d^4k}{(2\pi)^2} \tilde{\phi}(k) e^{-ik_\mu x^\mu})] $$

I've tried to simplify this further using a convolution and various other attempts but nothing has worked so far. Help would be very much appreciated.

  • $\begingroup$ Please show everyone how far you got, and at what point you can't simplify further. $\endgroup$
    – kaylimekay
    Feb 3, 2021 at 5:46
  • $\begingroup$ Thanks, I've just updated it $\endgroup$
    – Ravi
    Feb 4, 2021 at 22:02

1 Answer 1


You're integrating over $k$ twice, which is a bad idea. If you introduce a second dummy variable $q$, then the first term becomes something like

$$\int d^4x \int d^4k\int d^4q \left[ \tilde \phi(k)\tilde \phi(q)(-k^2-m^2+i\epsilon) e^{-ix\cdot (k+q)}\right]$$

Now the only thing which depends on $x$ is that exponential. Noting that $\int d^4x \ e^{-ix\cdot a}= (2\pi)^4 \delta(a)$, you should be able to take it from here.


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.