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.

72 CHAPTER 4. STATISTICS OF INDEPENDENT PARTICLES.<br />

< no >=<br />

1<br />

e ɛo−µ<br />

k B T − 1<br />

(4.15)<br />

Again, this only depends on the values of the temper<strong>at</strong>ure and chemical potential<br />

(as given by the external reservoir) and the energy of the orbital. No other<br />

properties of the orbital play a role. The distribution function for Bosons is<br />

therefore<br />

fBE(ɛ; T, µ) =<br />

1<br />

e ɛ−µ<br />

k B T − 1<br />

(4.16)<br />

This function is called the Bose-Einstein distribution function. The only difference<br />

with the Fermi-Dirac distribution function is the minus sign in the<br />

denomin<strong>at</strong>or. This is a small difference, but with large consequences. The<br />

Bose-Einstein distribution function has the following properties:<br />

lim<br />

ɛ→∞ fF D(ɛ; T, µ) = 0 (4.17)<br />

lim fF D(ɛ; T, µ) = ∞ (4.18)<br />

ɛ↓µ<br />

and it is the last infinity th<strong>at</strong> is the cause of all differences. The big difference<br />

between the two different distribution functions is the maximal value they can<br />

<strong>at</strong>tain. For fermions the distribution function never exceeds the value of one,<br />

while for bosons there is no upper limit.<br />

Limit of small occup<strong>at</strong>ion numbers.<br />

It is possible to choose T and µ in such a manner th<strong>at</strong> the value of the distribution<br />

function is always much less than one for all orbital energies. In order to<br />

use this argument it is essential th<strong>at</strong> the orbital energy has a lower limit (which<br />

is zero in our case of particles in a box). This requirement is quite n<strong>at</strong>ural.<br />

All realistic Hamiltonians are bounded from below, and the energy spectrum<br />

always has a minimum. The only exception is the Dirac Hamiltonian in first<br />

quantiz<strong>at</strong>ion, but in th<strong>at</strong> case the existence of neg<strong>at</strong>ive energy st<strong>at</strong>es is easily<br />

resolved in second quantiz<strong>at</strong>ion, where the neg<strong>at</strong>ive energy st<strong>at</strong>es correspond to<br />

positrons with normal positive energies.<br />

The requirement for small occup<strong>at</strong>ion numbers is th<strong>at</strong><br />

µ(T ) ≪ ɛmin − kBT (4.19)<br />

Note th<strong>at</strong> this does not mean th<strong>at</strong> in the limit T → ∞ this inequality is not<br />

obeyed! The chemical potential also has a temper<strong>at</strong>ure dependence which needs<br />

to be taken into account. When the inequality above is valid, we have<br />

e ɛ−µ<br />

k B T ≫ 1 (4.20)

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

Saved successfully!

Ooh no, something went wrong!