I'm having trouble following the proof of the adiabatic theorem (apparently due to Messiah) on Wikipedia.

At one stage we have:

$U(t_1,t_0)=1+{1\over i}\int_{t_0}^{t_1}H(t)dt+{1\over i^2}\int_{t_0}^{t_1}\int_{t_0}^{t'}dt'dt''H(t')H(t'')+\ldots$

which I'll write as $1+H_1+H_2+\ldots$.

The argument then goes

$\zeta=\left<0|(1+iH_1)(1-iH_1)|0\right>+$ other terms.

So $\zeta = \left<0|H_1^2|0\right>+$ other terms.

But if we're computing to second order in $H$, shouldn't we keep terms to second order all the way through the computation? In which case we really need:

$\zeta=\left<0|(1+iH_1+H_2)(1-iH_1+H_2)|0\right>+$ other terms.

So $\zeta = \left<0|H_1^2+2H_2|0\right>$+other terms?

Why is apparently OK to drop $H_2$?


1 Answer 1


Yes, you are right that you should keep all terms to a given order. However if you look at the full expression that is being calculated:

$\zeta= \left<0|(1+iH_1+H_2)(1-iH_1+H_2)|0\right>-\left<0|(1+iH_1+H_2)|0\right> \left<0|(1-iH_1+H_2)|0\right>$

You will find that the $H_2$ terms cancel out identically.

Surprisingly, this answer is controversial, so let's go through this more slowly. From Wikipedia:

$\zeta = \langle 0|\hat{U}^\dagger(t_1,t_0)\hat{U}(t_1,t_0)|0\rangle - \langle 0|\hat{U}^\dagger(t_1,t_0)|0\rangle\langle 0|\hat{U}(t_1,t_0)|0\rangle$

Dan expands $U$ as $U \approx 1 + iH_1 + H_2$.

Then Dan correctly expands the first term for $\zeta$ to second order in H as:

$\zeta = \langle 0| H_1^2 + 2H_2|0\rangle +...$

and asks why it is OK to drop $H_2$? The answer is that the second part of $\zeta$ is, to the same order in H,

$...-\langle 0|H_1|0\rangle \langle 0|H_1|0\rangle - 2\langle 0|H_2|0\rangle$

Therefore, the $\langle 0|H_2|0\rangle $ terms cancel exactly from $\zeta$.

  • $\begingroup$ Do they cancel? $\left<0|H_2^2|0\right>$ isn't the same as $\left<0|H_2|0\right>\left<0|H_2|0\right>$. $\endgroup$
    – Dan Piponi
    May 22, 2012 at 20:40
  • $\begingroup$ Dan is being nice. They don't cancel. Can you delete the answer? $\endgroup$
    – Ron Maimon
    May 22, 2012 at 20:55
  • $\begingroup$ The $<0|H_2|0>$ terms cancel. The $H_2^2$ terms are of higher order than $H_1^2$ terms. From the question it appeared the questioner understood $H_2$ was of the same perturbative order as $H_1^2$. $\endgroup$
    – user1631
    May 22, 2012 at 21:02
  • $\begingroup$ $<0|H_2^2|0>$ would be 4th order in H. The question specifically referenced terms up to second order in H. What is the confusion? $\endgroup$
    – user1631
    May 22, 2012 at 21:05
  • $\begingroup$ @user1631: The confusion was due to two things 1. the Wikipedia person not making clear that H_1 is an infinitesimal perturbation. 2. the fact that you write "the H_2 terms cancel exactly" in your answer, when you mean H_1 terms. $\endgroup$
    – Ron Maimon
    May 22, 2012 at 21:19

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.