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

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

26 CHAPTER 2. THE CANONICAL ENSEMBLE<br />

At this point we seem to have a contradiction. The basic assumption of st<strong>at</strong>istical<br />

mechanics tells us th<strong>at</strong> all accessible quantum st<strong>at</strong>es are equally probably.<br />

The Boltzmann distribution on the other hand gives the st<strong>at</strong>es unequal<br />

probabilities! This will always occur when we specify some intensive parameter<br />

like T. In such a case one always needs a reservoir, and the basic assumption<br />

applies to the st<strong>at</strong>es of the reservoir plus system. If we specify the st<strong>at</strong>e of the<br />

system, this corresponds to a configur<strong>at</strong>ion, i.e. a set of st<strong>at</strong>es, of the combin<strong>at</strong>ion<br />

reservoir plus system. The basic assumption only applies to st<strong>at</strong>es, not<br />

to configur<strong>at</strong>ions. For a configur<strong>at</strong>ion one always has to include a multiplicity<br />

function counting the number of st<strong>at</strong>es available for th<strong>at</strong> configur<strong>at</strong>ion.<br />

The Boltzmann factor is not correct in the limit ɛs → ∞, since th<strong>at</strong> implies<br />

taking a large fraction of the energy of the reservoir and putting it in the system.<br />

The upper limit of ɛs for which the Boltzmann factor is valid can be derived<br />

from the consider<strong>at</strong>ions in the beginning of this section. This is not a practical<br />

problem, though, since we can make the reservoir very large. The probability<br />

of finding st<strong>at</strong>es with an energy which is too large for our formalism to be valid<br />

then becomes exponentially close to zero!<br />

2.2 Energy and entropy and temper<strong>at</strong>ure.<br />

The expect<strong>at</strong>ion value of the energy, or average energy, of the system is given<br />

by<br />

U = <br />

ɛsP (s) (2.11)<br />

s<br />

where we have dropped the subscript S for system in the probability function.<br />

This can be rel<strong>at</strong>ed to the partition function by<br />

<br />

∂ log(Z)<br />

=<br />

∂T<br />

1<br />

<br />

∂Z<br />

=<br />

Z ∂T<br />

1 ɛs<br />

− ɛs<br />

k e B T<br />

Z kBT 2 (2.12)<br />

or<br />

V,N<br />

V,N<br />

U(T, V, N) = kBT 2<br />

<br />

∂ log Z)<br />

∂T<br />

V,N<br />

The entropy can be obtained from U(T) through<br />

<br />

∂S<br />

∂T<br />

<br />

∂S<br />

=<br />

∂U<br />

<br />

∂U<br />

∂T<br />

which leads to<br />

V,N<br />

V,N<br />

T<br />

S(T, V, N) = Sref (V, N) +<br />

Tref<br />

V,N<br />

1<br />

T<br />

s<br />

= 1<br />

T<br />

(T, V, N) (2.13)<br />

<br />

∂U<br />

∂T V,N<br />

(2.14)<br />

<br />

∂U<br />

dT (2.15)<br />

∂T V,N<br />

since we assume th<strong>at</strong> all other variables remain constant. It is useful to have an<br />

explicit form for the entropy, however. Based on

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

Saved successfully!

Ooh no, something went wrong!