Consider a single particle on a cubic lattice with $N$ sites each with a site length of $a$. The particle has no kinetic energy but only feels a spatially varying continuous potential $v(\boldsymbol{r})$.

One can directly write down the partition function as sum over possible states (particle at different sites):

$$Z=\sum_i^N\mathrm{e}^{-\beta v(\boldsymbol{r_i})}$$

But what happens if we want to evaluate the partition function from the continuous representation, i.e.

$$Z=\frac{1}{h^3}\left(\int \mathrm{d}\boldsymbol{p}\right)\left(\int \mathrm{d}\boldsymbol{r} \mathrm{e}^{-\beta v(\boldsymbol{r})}\right)=\frac{1}{h^3}\left(\int \mathrm{d}\boldsymbol{p}\right)a^3\sum_i^N \mathrm{e}^{-\beta v(\boldsymbol{r_i})}$$

Does that mean that $\frac{a^3}{h^3}\int\mathrm{d}\boldsymbol{p} = 1$?

Edit: As suggested by AccidentalFourierTransform we could integrate over the first Brilloin zone, this would exactly be in agreement with the above statement:

$\int\limits_{-\infty}^\infty \mathrm{d} \boldsymbol{p} = \hbar^3\int\limits_{-\infty}^\infty \mathrm{d} \boldsymbol{k} = \hbar^3\int\limits_{-\pi/a}^{\pi/a}\mathrm{d}\boldsymbol{k} = \frac{h^3}{a^3}$

Could someone comment on that or give an explanation? Is this a valid approach? Can e.g. atoms sitting on the lattice sites be considered as waves? And why do we not need to integrate over higher momenta as the Brilloin zone?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.