I am following the following lecture notes: http://web.mit.edu/6.763/www/FT03/Lectures/Lecture9.pdf

In the last slide, we see how a gauge change for the EM field impact the phase of the wavefunction. I remind:

$$ \psi(x,t)=\sqrt{n(x,t)}e^{i \theta(x,t)}$$ $$ \mathbf{J}=qn(x,t) \left( \frac{\hbar}{m} \mathbf{\nabla}(\theta(x,t))-\frac{q}{m} A(x,t) \right)$$

If we change the E.M potential like the following, the physical description is the same:

$$A'= A+\mathbf{\nabla} \chi $$ $$\phi'= \phi - \frac{\partial \chi}{\partial t} $$

Thus in the Gauge "prime", the physical quantities are the same: $n(x,t)=n'(x,t)$ and $\mathbf{J}=\mathbf{J'}$.

From those equalities, we find:

$$qn(x,t)\left( \frac{\hbar}{m} \mathbf{\nabla}(\theta'(x,t))-\frac{q}{m} A'(x,t) \right)=qn(x,t)\left( \frac{\hbar}{m} \mathbf{\nabla}(\theta(x,t))-\frac{q}{m} A(x,t) \right)$$

It implies:

$$ \mathbf{\nabla}(\theta'-\theta-\frac{q}{\hbar} \chi)=0$$

A particular solution for this is: $\theta'=\theta+\frac{q}{\hbar} \chi$, but we could expect other. Why is it this particular solution that is only considered in the slides ?


Consider first a time-independent gauge transformation. Then the vanishing of the gradient implies only that $$ \theta'-\theta+\frac{q}{\hbar}\chi= {\rm constant}. $$ But we also know that $\theta-\theta'=0$ if $\chi=0$. Therefore the constant is zero. For a time dependent gauge transformation, one needs to include the gauge covariance of the Josephson acceleration equation to make sure that the "constant" is also time independent

  • $\begingroup$ Hello. Thank you for your answer. I totally agree with the first part (time independant). It mostly answers my question. However I am not sure to understand your explanation for the "time dependance". Could you elaborate more ? Why doesn't the first answer works for time dependance ? I'm not sure to follow. Thanks ! $\endgroup$
    – StarBucK
    Jul 9 '20 at 20:20

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.