If I have the Lagrangian density:

$$\mathcal{L}=\left(\partial_{\mu} \phi^{*}\right)\left(\partial^{\mu} \phi\right)-m^{2} \phi^{*} \phi$$

How can I show it is invariant under the following gauge transforms (do I just sub them in)?

$$\phi^{\prime}=e^{i \theta} \phi$$

$$\phi^{*^{\prime}}=e^{-i \theta} \phi^{*}$$

And then from this find the conserved Noether current?

  • $\begingroup$ After that you find exactly the same Lagrangian... $\endgroup$
    – Turgon
    Jun 12, 2019 at 6:45
  • 1
    $\begingroup$ A covariant derivative with a gauge field has to be introduced. $\endgroup$
    – p6majo
    Jun 12, 2019 at 7:06
  • $\begingroup$ Hi desirooo: Are you talking about global gauge transformations? $\endgroup$
    – Qmechanic
    Jun 12, 2019 at 7:18

1 Answer 1


In Field Theory, a Lagrangian is said to be invariant if you only introduce a quadridivergence of a field which only depends on space-time I.E.:

Be $\mathcal{L}=\mathcal{L}[\phi(x),\partial_\mu\phi(x),x]$ a Lagrangian of scalar field $\phi$ where $x=(t,\vec{x})$ is a space time coordinate. Then for an infinitesimal transformation $\phi\rightarrow\phi+\alpha\delta\phi$, if the Lagrangian becomes $\mathcal{L}\rightarrow\mathcal{L}+\alpha\delta\mathcal{L}=\mathcal{L}+\alpha\partial_\mu J^\mu(x)$ where $J^\mu$ is any function of space-time, then the Lagrangian is said to be invariant.

With Noether's theorem, the conserved current is given by $j^\mu\equiv J^\mu(x)-\frac{\partial\mathcal{L}}{\partial(\partial_\mu\phi)}\delta\phi$.

In your example you have $\mathcal{L}=\partial_\mu\phi^*\partial^\mu\phi - m^2\phi^*\phi$ and your infinitesimal translation is $\phi\rightarrow\phi+i\theta\phi$ and $\phi^*\rightarrow\phi^*-i\theta\phi^*$. By plugging into the Lagrangian we get

$$\mathcal{L}\rightarrow\partial_\mu(\phi^*-i\theta\phi^*)\partial^\mu(\phi+i\theta\phi)-m^2(\phi^*-i\theta\phi^*)(\phi+i\theta\phi)$$ $$=\partial_\mu\phi^*\partial^\mu\phi-i\theta\partial_\mu\phi^*\partial^\mu\phi+i\theta\partial_\mu\phi^*\partial^\mu\phi+\theta^2\partial_\mu\phi^*\partial^\mu\phi-m^2(\phi^*\phi-i\theta\phi^*\phi+i\theta\phi^*\phi -\theta^2\phi^*\phi)$$ We then neglect terms of order $O(\theta^2)$ taking $\theta$ as an infinitesimal parameter. Thus we have:


Because all other terms cancels each other! Thus the Lagrangian is invariant under this transformation and the function $J^\mu=$constant which we can choose to be $0$. The conserved current is (one for each field):

$$j^\mu_1=-\frac{\partial\mathcal{L}}{\partial(\partial_\mu\phi^*)}\delta\phi^*$$ $$=-(\partial^\mu\phi)(-i\phi^*)=i(\partial^\mu\phi)\phi^*$$

Similarly: $j_2^\mu=-i(\partial^\mu\phi^*)\phi$


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.