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.

28 CHAPTER 2. THE CANONICAL ENSEMBLE<br />

specify one of them, the other two are determined. We have found two recipes<br />

for these functions. If the energy U is the independent variable, the entropy<br />

S(U) is constructed by counting st<strong>at</strong>es with ɛs = U. The temper<strong>at</strong>ure T(U)<br />

follows from <br />

∂S<br />

∂U V,N = T −1 . On the other hand, if the temper<strong>at</strong>ure T is<br />

the independent variable, we calcul<strong>at</strong>e the Boltzmann factor and the partition<br />

function Z(T ). The energy U(T) then follows from the deriv<strong>at</strong>ive of log(Z),<br />

while the entropy follows from a calcul<strong>at</strong>ion of the probabilities P(s). Note th<strong>at</strong><br />

in calcul<strong>at</strong>ing the partition function we often use the multiplicity function, since<br />

Z(T ) = <br />

E<br />

E − k g(E)e B T (2.24)<br />

which is a simple Laplace transform<strong>at</strong>ion. This transform<strong>at</strong>ion can be inverted<br />

and used to obtain the multiplicity function g(E) from the partition function,<br />

but th<strong>at</strong> requires an integr<strong>at</strong>ion in the complex temper<strong>at</strong>ure plane. Interesting,<br />

since complex temper<strong>at</strong>ure corresponds to time. We obtain factors like e ıEt .<br />

Field theory?? But th<strong>at</strong> is for another course.<br />

2.3 Work and pressure.<br />

A system which remains in a given equilibrium st<strong>at</strong>e is r<strong>at</strong>her boring. We would<br />

like to change the st<strong>at</strong>e of a system. This can be done by varying some of the<br />

external parameters. For example, if a system is in contact with a reservoir,<br />

we can change the temper<strong>at</strong>ure of the reservoir. The system will adjust itself<br />

until it is again in thermal equilibrium with the reservoir. This is an example<br />

of a process in which the system is brought from an equilibrium st<strong>at</strong>e with<br />

temper<strong>at</strong>ure Ti to an equilibrium st<strong>at</strong>e with temper<strong>at</strong>ure Tf . Wh<strong>at</strong> is happening<br />

in between, though? If we suddenly switch the temper<strong>at</strong>ure of the reservoir, it<br />

will take some time before thermal equilibrium is restored again. The st<strong>at</strong>es in<br />

between are non-equilibrium st<strong>at</strong>es and the entropy of the system will increase.<br />

We can also change the temper<strong>at</strong>ure very slowly, in such a way th<strong>at</strong> <strong>at</strong> each time<br />

the system is in thermal equilibrium with the reservoir. A process for which a<br />

system is always in thermal equilibrium in all of the intermedi<strong>at</strong>e st<strong>at</strong>es is called<br />

reversible. If we reverse the change in temper<strong>at</strong>ure of the reservoir, the system<br />

will go back to its initial st<strong>at</strong>e. If we suddenly switch the temper<strong>at</strong>ure back and<br />

forth, the system will end up in a different st<strong>at</strong>e with higher entropy. For the<br />

l<strong>at</strong>ter to happen we have to allow other variables to change too, if only T,S, and<br />

U can change the st<strong>at</strong>es are completely determined by T. Reversible processes<br />

are nice, since they can easily be described in m<strong>at</strong>hem<strong>at</strong>ical terms. It is much<br />

harder to deal with irreversible processes.<br />

An important extensive parameter of a system is its volume. We will in this<br />

chapter still assume th<strong>at</strong> the number of particles does not change. A change<br />

in volume will in general affect the energy eigenvalues of the quantum st<strong>at</strong>es in<br />

this system and we write ɛs(V ). If we introduce changes in the volume of the<br />

system, we need a force to produce these changes. We assume th<strong>at</strong> this force is<br />

homogeneous, and hence the only thing we need is an external pressure p. The

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

Saved successfully!

Ooh no, something went wrong!