For the 1D Ising model with the Hamiltonian

$$H=const.-\mu h' \sum_i S_i^z$$

we can write the canonical partition sum as

$$Z_N = \sum_{ \{ S_i^z \}_N } e^{-\beta \mu h \sum_i S^z_i} = \sum_{ \{ S_i^z \}_N } \prod_i e^{-\beta \mu h S^z_i}$$

for which we then later used

$$Z_N=(Z_1)^N$$with the single particle partition sum $Z_1$

We didn't go further into the proof during the lecture, and it's not immediately obvious to me.

My attempt:

For the classical ideal gas I could use the Multinomial Theorem

$$(x_1+x_2+\dots+x_n)^k=\sum_{k_1+\dots+k_n=k} \frac{k!}{k_1!\dots k_n!} \ x_1^{k_1}x_2^{k_2} \dots x_n^{k_n}$$

but for that the factor $\frac{k!}{k_1!\dots k_n!}$ would have to be $1$.

How do I best prove $Z_N=(Z_1)^N $?


In fact, $Z_N = (Z_1)^N$ is true for all non-interacting systems with identical components, where the Hamiltonian can be written as $H = H_1 + ... + H_N$ where $H_k$ only depends on the state of the $k$-th particle (/spin/...) and all $H_k$ are the same.

This can be seen like this.

$$Z_N = \sum_{states \ s_1...s_N} e^{- \beta H} = \sum_{s_1} \sum_{s_2} ...\sum_{s_N} e^{- \beta (H_1+...+H_N)}= \sum_{s_1} e^{- \beta (H_1)} \sum_{s_2} e^{- \beta (H_2)}...\sum_{s_N} e^{- \beta (H_N)} = ( \sum_{s_1} e^{-\beta H_1})^N = Z_1^N$$

Note: For identical particles, a factor $1/N!$ might additionally be necessary.

| cite | improve this answer | |

First of all, the last manipulation of the partition sum is not correct. You cannot pull out a global factor $e^{-\beta\mu h}$.

To give you a hint: It might be helpful to write $$ \sum_{\{S_i\}} = \prod_i\sum_{S_i=\pm 1} = \sum_{S_1=\pm 1}\sum_{S_2=\pm 1}\cdots\sum_{S_N=\pm 1} $$ and exploit that the Boltzmann factor factorizes.

| cite | improve this answer | |

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.