1
$\begingroup$

In Callen's Thermodynamics textbook, he writes that $$\left(\frac{\partial u}{\partial s}\right)_v = \left(\frac{\partial U}{\partial S}\right)_{V,N}$$

where $u = U/N$, $s = s/N$, and $v = V/N$ and, moreover, $U$ is first order homogeneous so that $$ U (S,V,N) = U(SN/N,VN/N,N) = NU(S/N,V/N,1) \equiv Nu(s,v)$$ where the last equality represents another way to define $u$. I'm not able to see how to argue for the given equality of derivatives though. I get stalled out: $$\left(\frac{\partial u}{\partial s}\right)_v \equiv \left(\frac{\partial (U/N)}{\partial (S/N)}\right)_{V/N} \stackrel{?}{=} \left(\frac{\partial U}{\partial S}\right)_{V,N}$$

I want to be clear that I understand why (namely because the quantity so obtained on the RHS is intensive) this equality must hold, but I can't see how to show it mathematically as I get lost in the morass of partial derivatives and chain rules.

$\endgroup$
4
  • $\begingroup$ It might be easier to go in the other direction. $\endgroup$ Commented May 9, 2023 at 19:36
  • $\begingroup$ @Chemomechanics I will try that right now; one moment! $\endgroup$
    – EE18
    Commented May 9, 2023 at 19:38
  • $\begingroup$ @Chemomechanics I see it now, thank you! Note to self -- try both ways in the future :) $\endgroup$
    – EE18
    Commented May 9, 2023 at 20:03
  • $\begingroup$ @EE18 Calculus works both ways: $$ \left(\frac{\partial u}{\partial s}\right)_v \equiv \left(\frac{\partial (U/N)}{\partial (S/N)}\right)_{V/N} = \left(\frac{\partial U}{\partial S}\right)_{V,N} $$ With $N=\text{const}$ we go from the middle result to the last. $\endgroup$
    – Themis
    Commented May 10, 2023 at 18:51

3 Answers 3

1
$\begingroup$

The key point here is the homogeneity of extensive thermodynamic properties. For $U(S,V,N)$ this means that if we increase $S$, $V$, $N$, simultaneously by the same factor, then $U$ increases by the same factor as well: $$ U(\lambda S, \lambda V, \lambda N) = \lambda U(S,V,N) $$ It follows that the partial derivatives of $U$ are intensive. This can be shown by calculus for any homogeneous function with degree 1 but in this case it is clear since these derivatives are known intensive properties: $$ \left(\frac{\partial U}{\partial S}\right)_{VN} = T;\quad \left(\frac{\partial U}{\partial V}\right)_{VN} = -P;\quad \left(\frac{\partial U}{\partial S}\right)_{VN} = \mu. $$ As intensive properties they do not depend on the absolute $U$, $V$ and $N$, but have the same value in all systems with the same ratios $S:V:N$. Therefore, they have the same values in a system with $s=S/N$, $v=V/N$, $n=N/N=1$: $$ \left(\frac{\partial U}{\partial S}\right)_{VN} = \left(\frac{\partial u}{\partial s}\right)_{v} = T. $$

$\endgroup$
0
$\begingroup$

Observe that $$ \left(\frac{\partial U}{\partial S}\right)_{V,N} \equiv \left(\frac{\partial (Nu(S/N,V/N))}{\partial S}\right)_{V,N} = N\left(\frac{\partial (u(S/N,V/N))}{\partial S}\right)_{V,N} $$ $$\stackrel{(1)}{=} N \left[\left(\frac{\partial (u(S/N,V/N))}{\partial (S/N)}\right)_{V/N}\left(\frac{\partial (S/N)}{\partial S}\right)_{V,N} + \left(\frac{\partial (u(S/N,V/N))}{\partial (V/N)}\right)_{S/N}\left(\frac{\partial (V/N)}{\partial S}\right)_{V,N} \right]$$ $$ N \left[\left(\frac{\partial (u(S/N,V/N))}{\partial (S/N)}\right)_{V/N}\frac{1}{N} + 0 \right] = \left(\frac{\partial (u(S/N,V/N))}{\partial (S/N)}\right)_{V/N}$$ $$\equiv \left(\frac{\partial u}{\partial s}\right)_v $$ where in (1) we have used the chain rule for partial derivatives and in the last step we have simply used the definitions of $s$ and $v$.

$\endgroup$
0
$\begingroup$

I would consider the very simple and immediate proof based on the definition of the partial derivative and the homogeneity of $U$: $$ \left(\frac{\partial U}{\partial S}\right)_{V,N} = \lim_{\Delta S \rightarrow 0} \frac{U(S+\Delta S,V,N)-U(S,V,N)}{\Delta S}=\lim_{\Delta s \rightarrow 0} \frac{U(N s+ N \Delta s,N v,N)-U(N s,N v,N)}{N\Delta s}= \lim_{\Delta s \rightarrow 0} \frac{u(s+\Delta s,v)-u(s,v)}{\Delta s}=\left(\frac{\partial u}{\partial u}\right)_{v} $$

$\endgroup$

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.