04.08.2013 Views

Statistical Mechanics - Physics at Oregon State University

Statistical Mechanics - Physics at Oregon State University

Statistical Mechanics - Physics at Oregon State University

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

2.3. WORK AND PRESSURE. 29<br />

pressure p is the n<strong>at</strong>ural intensive parameter which has to be combined with the<br />

extensive volume. The work W performed on the system to reduce the volume<br />

of the system from V to V − ∆V (with ∆V small) is p∆V .<br />

An example of a reversible process in which the volume of the system is<br />

reduced is a process where the system starts in a given st<strong>at</strong>e ˆs and remains in<br />

th<strong>at</strong> st<strong>at</strong>e during the whole process. It is clearly reversible, by definition. If we<br />

reverse the change in volume we will still be in the same quantum st<strong>at</strong>e, and<br />

hence return to the same st<strong>at</strong>e of the system <strong>at</strong> the original value of the volume.<br />

The energy of the system is ɛˆs(V ) and the change in this energy is equal to the<br />

work done on the system<br />

<br />

∂ɛˆs<br />

p∆V = ∆ɛˆs(V ) = −∆V<br />

(2.25)<br />

∂V N<br />

Now consider an ensemble of systems, all with initial volume V and initial<br />

temper<strong>at</strong>ure Ti. The number of subsystems in a st<strong>at</strong>e s is proportional to the<br />

Boltzmann factor <strong>at</strong> Ti. We change the volume in a similar reversible process<br />

in which each subsystem remains in the same quantum st<strong>at</strong>e. The change in<br />

energy is given by<br />

∆U = <br />

∆ɛsP (s) = p∆V <br />

P (s) = p∆V (2.26)<br />

s<br />

because the probabilities P(s) do not change. Since ∆V is a reduction in<br />

volume we find p = − <br />

∂U<br />

∂V . We are working with two independent variables,<br />

however, and hence we have to specify which variable is constant. Since P(s)<br />

does not change, the entropy does not change according to the formula we<br />

derived before. The temper<strong>at</strong>ure, on the other hand, will have to change! Hence<br />

<br />

∂U<br />

p = −<br />

(2.27)<br />

∂V<br />

If we take other extensive variables into account we have to keep those the<br />

same too while taking this deriv<strong>at</strong>ive. The last formula is identical to the formula<br />

for pressure derived in thermodynamics, showing th<strong>at</strong> the thermodynamical and<br />

the st<strong>at</strong>istical mechanical description of volume and pressure are the same. In<br />

thermodynamics we assumed th<strong>at</strong> mechanical work performed on the outside<br />

world was given by ¯ dW = −pdV . In general we have dU = T dS − pdV , but <strong>at</strong><br />

constant entropy this reduces to dU = −pdV and the formula for the pressure<br />

follows immedi<strong>at</strong>ely. In our present st<strong>at</strong>istical mechanical description we used<br />

the same formula, −pdV , for the work done by a system in a specific quantum<br />

st<strong>at</strong>e ˆs. But now we assume th<strong>at</strong> the probabilities for the quantum st<strong>at</strong>es do<br />

not change, again leading to dU = −pdV . In thermodynamics the entropy is<br />

always used as a m<strong>at</strong>hem<strong>at</strong>ical tool, and in order to arrive <strong>at</strong> the formula for the<br />

pressure we simply set dS = 0. In st<strong>at</strong>istical mechanics the entropy is a quantity<br />

connected to the quantum st<strong>at</strong>es of a system. The fact th<strong>at</strong> the entropy did not<br />

change in deriv<strong>at</strong>ion for the formula for the pressure is a simple consequence of<br />

the requirement th<strong>at</strong> the quantum st<strong>at</strong>es keep their original probabilities.<br />

s<br />

S,N

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!