Sign up ×
Physics Stack Exchange is a question and answer site for active researchers, academics and students of physics. It's 100% free.

Context: Solid state physics. Monoatomic linear chain.

Question: To prove that the total momentum of the chain is zero.

Attempted solution: I consider the sum:

\begin{align*} p = \sum_{n=1}^{N} m \dot{u_n} \end{align*}

where $u_n$ is the displacement from the equilibrium position of the $n$-th atom. The displacement is given by the formula:

\begin{align*} u_n = u_0 \exp\left[-i\left(\omega t \pm k n a\right)\right] \end{align*}

where $k$ is the wavevector, $n$ is the $n$-th atom and $a$ is the distance between atoms. If I substitute the above formula to the first sum, I get a result of the form:

\begin{align*} p \sim \sum_{n=1}^{N} \exp(i k n a) \end{align*}

I wonder, how could I prove that this is always zero ? If I treat it as the sum of a geometric series with $\alpha_1=\exp(ika)$ and $\lambda=\exp(ika)$, I still get a result that isn't necessarily zero.

\begin{align*} S_{1\to N} = \alpha_1 \frac{\lambda^N-1}{\lambda - 1} = e^{ika} \frac{e^{ikNa}-1}{e^{ika}-1} \end{align*}

If I further require that the first and last atoms are fixed, then $\exp(ikNa) = \exp(ika)$ and $S_{1\to N} = \exp(ika)$. Then

\begin{align*} p = -i\omega u_0 m \exp(-i\omega t) \exp(ika) = -i\omega m \underbrace{u_0 \exp\left[-i(\omega t - k a)\right]}_{u_1} = -i \omega m u_{1} = 0 \end{align*}

since we assumed that the 1st atom is fixed. Does this sound correct ?

share|cite|improve this question
'If I treat it as the sum of the geometric series...' can you show this? – Kvothe Apr 10 '14 at 21:03

1 Answer 1

up vote 1 down vote accepted

With $$u(n)=u_0e^{-i(\omega t+k n a)}$$ we have $$p=\sum_{n=1}^Nm\frac{d}{dt}u(n)=i\omega m u_0e^{-i\omega t}\left(\frac{1-e^{-iakN}}{1-e^{iak}}\right).$$ With cyclic boundary conditions we have $$u(N+1)=u(1)\Rightarrow k=\frac{2\pi j}{Na}\text{ for }j\in\mathbb{Z}$$ and inserting $k$ into $p$ gives $$p=i\omega m u_0e^{-i\omega t}\left(\frac{1-e^{-2 i \pi j}}{1-e^{\frac{2 i \pi j}{N}}}\right)=0.$$ Your method uses a slightly different boundary condition, but I think it's still valid.

share|cite|improve this answer
I like yours better. Thanks @DumpsterDoofus! – Zet Apr 10 '14 at 22:26

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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