0
$\begingroup$

I found the following expression in a paper:

$$ \frac{d}{dt}\arg\langle\phi_+|\dot{\phi_-}\rangle $$

where the $\arg$ term is the argument of the complex number given by inner product between two eigenstates $\phi_+$ and $\phi_-$, and the dot indicates a time-derivative.

Does the product rule apply in the usual sense in this case? i.e., does the following hold? If not, why?

$$ \frac{d}{dt}\arg\langle\phi_+|\dot{\phi_-}\rangle = \arg\langle\dot\phi_+|\dot{\phi_-}\rangle + \arg\langle\phi_+|\ddot{\phi_-}\rangle $$ I could not find much about $\arg$ terms in quantum mechanics and bra-ket literature, and so would appreciate any help you might have to offer.

$\endgroup$
5
  • $\begingroup$ @AaronStevens thanks for the comment, but would you mind clarifying what you meant by “they” (the bra-ket without arg?) and “normal vectors” (orthogonal)? $\endgroup$ Commented Mar 14, 2019 at 6:26
  • $\begingroup$ @AaronStevens thank you for clarifying. Yes, that is how I interpreted it, and that means that the time derivative is a kind of rate of change of angle. However, I am trying to make more sense of this interesting notation by wondering whether the product rule can be applied the way I wondered about in my question. $\endgroup$ Commented Mar 14, 2019 at 7:38
  • $\begingroup$ Correction: taking the magnitude of the dot product is one way to interpret it. The way I think is natural to the paper is to take the fact that the dot product is a complex number, making the arg the argument of the resulting complex number. $\endgroup$ Commented Mar 14, 2019 at 7:41
  • 1
    $\begingroup$ You should definitely make the edit in the question then. $\endgroup$ Commented Mar 14, 2019 at 11:32
  • $\begingroup$ I just went ahead and edited it :) $\endgroup$ Commented Mar 14, 2019 at 21:37

1 Answer 1

5
$\begingroup$

No it definitely doesn't work that way. Take $z\equiv \langle \phi_+ |\dot{\phi}_-\rangle$. It is true you can use the product rule on derivatives of $z$. But $Arg$ is a nontrivial function and you need to use a chain rule.

$$z=|z|e^{iArg (z)}$$ $$\log z = \log |z|+iArg(z)$$ where $Arg$ is only defined up to an integer times $2\pi$ but it won't matter since we're taking a derivative, $$\frac{d}{dt}Arg(z)=\frac{d}{dt}\text{Im}\log z=\text{Im}\frac{\dot{z}}{z}$$ So the product rule appears in $\dot{z}$ but otherwise it looks nothing like it.

$\endgroup$

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.