0
$\begingroup$

Consider a $U(1)$ symmetry of a complex scalar field realized as $$\phi\to\phi^\prime=e^{iQ\theta}\phi.$$ where $Q$ is the generator of the symmetry. The conserved Noether's charge (in $D$-spatial dimensions) corresponding to this symmetry is given by $$Q_N=\int j^0(\textbf{x},t) d^D\textbf{x}=iQ\int d^D\textbf{x}[(\partial_0\phi)\phi^*-(\partial_0\phi^*)\phi].$$ So it turns out that $Q_N\propto Q$ but not equal to $Q$ itself.

But in A. Zee's book QFT in a Nutshell, page 198, the Noether charge corresponding to a symmetry $$U(\theta)=e^{iQ\theta}$$ is written as $$Q_N=\int j^0(\textbf{x},t)d^D\textbf{x}=Q$$ But as I've shown $Q_N\propto Q$. So is it a mistake there? Or am I missing something?

Improve this question
$\endgroup$

1 Answer 1

Reset to default
7
$\begingroup$

Your confusion resides in the notation. To make it clear I'll use a hat over the operators. The Noether symmetry is $$ \hat\phi'\equiv\mathrm e^{i\theta\hat Q}\hat\phi\mathrm e^{-i\theta\hat Q}\equiv \mathrm e^{i\theta Q}\hat\phi $$ where $Q\in\mathbb R$ is a scalar. The generator of this symmetry is $\hat U(\theta)=\mathrm e^{i\theta\hat Q}$. The infinitesimal change in $\phi$ is $$ \theta\delta\hat\phi=[\hat Q,\hat\phi]=iQ\phi $$

The conserved charge is $$ \hat Q=\int \hat j^0(\textbf{x},t) d^D\textbf{x}=iQ\int d^D\textbf{x}[(\partial_0\hat\phi)\hat\phi^*-(\partial_0\hat\phi^*)\hat\phi]. $$

Sometimes some authors drop the factor of $Q$ in the definition of $\hat Q$, so that the infinitesimal change in $\phi$ reads $$ \theta\delta\hat \phi=i\phi $$

Improve this answer
$\endgroup$
6
  • $\begingroup$ @AFT Isn't the U(1) transformation on a classical field is implemented by $\phi^\prime=e^{iQ\theta}\phi$ where $\phi$ is not an operator? I don't understand why you added hats in this relation. My understanding was that $\hat{\phi}^\prime=e^{i\hat Q\theta}\hat{\phi}e^{-i\hat Q\theta}$ relation holds in quantum theory where $\hat{\phi}$ is an operator. Is that wrong? $\endgroup$
    – SRS
    Commented Jan 17, 2017 at 13:08
  • 1
    $\begingroup$ I though that your question was about quantum fields, not classical fields (because of the tag quantum-field-theory and because you mentioned the unitary operator $U(\theta)=\mathrm e^{i\theta Q}$). But now I am confused: in the OP is $\phi$ a function or an operator? $\endgroup$
    – AccidentalFourierTransform
    Commented Jan 17, 2017 at 13:11
  • $\begingroup$ You're right. My question was about quantum fields. But the transformation I wrote in the question, now I think, is valid only in classical field theory. But in your answer, you wrote $\hat{\phi}^\prime=e^{iQ\theta}\hat{\phi}$ even for quantum fields with $Q$ being a number. In short, you wrote two ways of transforming the quantum field $\hat{\phi}$ in your first equation. I think the latter implementation holds for classical $\phi$ only. $\endgroup$
    – SRS
    Commented Jan 17, 2017 at 13:24
  • 1
    $\begingroup$ I didnt really write two different ways of transforming $\hat\phi$. Recall that any symmetry transformation can always be written as $\hat\phi'=U(\alpha)\hat\phi U^\dagger(\alpha)$ for some unitary operator $U$ and some parameters $\alpha$. For example, the translation symmetry reads $\hat\phi'(x)= e^{iPa}\hat\phi(x)e^{-iPa}=\hat\phi(x-a)$. In this sense, $\hat\phi'=e^{i\hat Q\theta}\hat\phi e^{-i\hat Q\theta}$ is the general expression for the transformation; and $\hat\phi'=\mathrm e^{iQ\theta}\phi$ is the specific rule for the U(1) symmetry. They are not two different transformations. $\endgroup$
    – AccidentalFourierTransform
    Commented Jan 17, 2017 at 14:54
  • 1
    $\begingroup$ @SRS you can't derive a transformation law. The transformation $\phi'=e^{i\hat Q\theta}\hat\phi e^{-i\hat Q\theta}$ is just the general expression of a unitary transformation. All symmetries can be written as $\phi'=e^{i\hat Q\theta}\hat\phi e^{-i\hat Q\theta}$ for suitable operators $\hat Q$ and paramenters $\theta$. The particular case of a $U(1)$ symmetry is realised as $\phi'=\mathrm e^{iQ\theta}\phi$, and this defines the symmetry transformation. $\endgroup$
    – AccidentalFourierTransform
    Commented Mar 23, 2017 at 21:24

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.