$C_V=\frac{1}{kT^2}[\overline{E^2} -\overline E ^2]$ where $\overline{E}^2=-\frac{1}{Z}\frac{\partial Z}{\partial \beta}$ and $\overline{E^2}=-\frac{1}{Z}\frac{\partial^2 Z}{\partial^2 \beta}$.

$Z=\sum _ie^{-\beta\cdot E_i}$ of finite sum is the partition funciton and $\beta=\frac{1}{kT}$.

The question is what is $\lim_{T\rightarrow0^+}C_V$.

According to my calculation, as $T\rightarrow 0^+$, $[\overline{E^2} -\overline E ^2]$ approach some finite number $K$. Thus $C_V$ approach $\infty$.

I didn't believe in the begining, so I tried to plot the numerical solution, I also put the simplified function into Wolfram.

However, all the result I got was $\lim_{T\rightarrow 0^+}C_V=\infty$.

And it did not make too much sense,

I found two reference

  1. http://stp.clarku.edu/notes/chap6.pdf 6.17 seem to support my argument.

  2. http://www-personal.umich.edu/~lsander/ESP/chap4.pdf page 5 otherwise.

and my professor's answer sheet says it's suppose to be $0$.

What's the correct answer?

  • $\begingroup$ Eq. 6.17 no longer holds when temperature decreases. Gas would condense, right? $\endgroup$ – K_inverse Nov 13 '18 at 7:50
  • $\begingroup$ @K_inverse Yeah, but obviously there is a thermal stastical treatment, I couldn't figure out A. how to show $C_V=0$ with thermal statistics derivation. B. how to proof the above calculation for $C_V$ fails. $\endgroup$ – user9976437 Nov 13 '18 at 8:08

At very low temperature, the states with the highest probability are those at very low energy, close to the ground state. In the limit $T \rightarrow 0^+$ it is only the ground state which contributes to the canonical sum over the states. If the ground state is non-degenerate, the distribution probability must have zero-variance.

Notice that such an argument fails for a classical perfect gas just because the classical description of the system assigns a macroscopic degeneracy even to the lowest energy state (zero velocity for all particles but with a huge positional degeneracy).


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.