1
$\begingroup$

Given the equation of state $$p+a\left(\frac{N}{V}\right)=\frac{Nk_BT}{V-bN} \tag 1$$ Taking into account of the fact that a realistic model requires $p \geq 0, V \geq Nb, N>0$ classify the solutions of the equation of state as a function of temperature.

My attempt is that

$$\left[p+a\left(\frac{N}{V}\right)\right]\left[\frac{V-bN}{V} \right]=\frac{Nk_BT}{V} $$ Let $n=\frac{N}{V}$,

$$\left[p+an\right]\left[1-bn \right]=nk_BT$$

$$p+an-pbn-abn^2=nk_BT$$

Let $\tilde{n}=\sqrt{ab}n$, $\tilde{n}^2=abn^2$,

$$\tilde{n}^2+\left(p\sqrt{\frac{b}{a}}-\sqrt{\frac{a}{b}}\right)\tilde{n}-p+\frac{k_BT}{\sqrt{ab}}=0$$

Let $\tilde{p}=\sqrt{\frac{b}{a}}p$,

$$\tilde{n}^2+\left(\tilde{p}-\sqrt{\frac{a}{b}}\right)\tilde{n}-\sqrt{\frac{a}{b}}p+\frac{k_BT}{\sqrt{ab}}=0$$

Let $\tilde{t}=\frac{k_BT}{\sqrt{ab}}$,

$$\tilde{n}^2+\left(\tilde{p}-\sqrt{\frac{a}{b}}\right)\tilde{n}-\sqrt{\frac{a}{b}}p+\tilde{t}=0$$

Now the discriminant is $b^2-4ac$ so

$$\left(\tilde{p}-\sqrt{\frac{a}{b}} \right)^2+4\sqrt{\frac{a}{b}}p-4\tilde{t}$$

$$\tilde{p}^2-2\sqrt{\frac{a}{b}}\tilde{p}+\frac{a}{b} +4\sqrt{\frac{a}{b}}p-4\tilde{t}$$

$$\left(\tilde{p}+\sqrt{\frac{a}{b}} \right)^2=4\tilde{t}$$

so ignoring the negative time solution we have

$$\tilde{p}=-\sqrt{\frac{a}{b}} +2\sqrt{\tilde{t}}$$

but I don't see what Im meant to be doing next.

After I that I need to show that there is a region of the $T-V^{-1}$ plane where this equation of state is not thermodynamically stable and determine the boundary of this region.

I know that for stability we need that

$$\left(\frac{\partial p}{\partial V} \right)_{T,N} \leq 0$$

but I dont see how this is usable in the current context.

$\endgroup$

1 Answer 1

3
$\begingroup$

First of all you can directly write $p$ as a function of $V$ without solving a quadratic equation:

$$ p = \frac{N k_B T}{V- N b } - a \frac{N}{V} $$

Then you take a look at the interesting points

$$\left(\frac{\partial p}{\partial V}\right)_{T,N} = 0 $$

The solutions to this equation tell you where the bulk modulus (or its inverse, the compressibility) might change the sign. That's how you can use this criterion to determine the boundaries of unstable regions.

EDIT: If I am not mistaken, the equation above has solutions

$$ \vartheta_\pm = b \frac{1}{1\pm \sqrt{\frac{k_B T}{a}}} = b \frac{1}{1\pm \sqrt{t}}$$,

with $\vartheta = V/N$ and the reduced temperature $t=T k_B /a $. In terms of this reduced variable and $v = \vartheta / b$ we can write

$$ \left(\frac{\partial p}{\partial \vartheta}\right)_{T,N} = \frac{a}{b^2} \left(\frac{1}{v^2}- \frac{t}{(v-1)^2} \right) \overset{!}{\leq} 0$$

So you know the critical points where the sign might change and you know the equation for checking which sign is present. The only thing you have left to do is check which sign is present in which region along the $\vartheta$ axis. This is only math left to do.. Since you know the sign along the $\vartheta$ axis you will automatically know the sign along the $\vartheta^{-1}$ axis.

$\endgroup$
8
  • $\begingroup$ Could you expand on your last sentence please? (I am only familiar with compressibility). Ive solved the second equation but got $V=\frac{2abN \pm \sqrt{4k_BTa}bN}{2(a-k_BT)}$ $\endgroup$
    – Trajan
    Commented May 12, 2015 at 9:56
  • $\begingroup$ How can we do this in the $T-V^{-1}$ plane? $\endgroup$
    – Trajan
    Commented May 12, 2015 at 10:00
  • $\begingroup$ How did you know that the sign change is the criterion of this? I can find no mention of it my notes $\endgroup$
    – Trajan
    Commented May 12, 2015 at 10:02
  • $\begingroup$ Imagine you have gas whose pressure increases when you extend its volume. You could build a perpetual motion machine with this ;) That's why you expect this derivative to be negative. I will let you try a little bit and extend my answer a little bit later if you don't mind. A hint: it may also be useful to think about the temperature up to which your solution is really a solution. This will give you a critical temperature. Usually its easier to work with reduced quantities. $\endgroup$ Commented May 12, 2015 at 10:13
  • $\begingroup$ So $\left(\frac{\partial p}{\partial \vartheta}\right)_{T,N} {\leq} 0 \iff \left(\frac{1}{v^2}- \frac{t}{(v-1)^2} \right) \leq 0 \iff (-\frac{1}{v}+\frac{1}{v^2}) \leq t$ So if this is true stable if not unstable. I cannot see how this answers both parts though $\endgroup$
    – Trajan
    Commented May 13, 2015 at 13:22

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.