# How to evaluate this sum of coupling coefficients?

I would like to evaluate the following summation of Clebsch-Gordan and Wigner 6-j symbols in closed form:

$$\sum_{l,m} C_{l_2,m_2,l_1,m_1}^{l,m} C_{\lambda_2,\mu_2,\lambda_1,\mu_1}^{l,m} \left\{ \begin{array}{ccc} l & l_2 & l_1 \\ n/2 & n/2 & n/2 \end{array}\right\} \left\{ \begin{array}{ccc} l & \lambda_2 & \lambda_1 \\ n/2 & n/2 & n/2 \end{array}\right\}$$

with $n \in \left[0,\infty\right)$, $l,l_1,l_2,\lambda_1,\lambda_2 \in \left[0,n\right]$, $m \in \left[-l,l\right]$, $m_1 \in \left[-l_1,l_1\right]$, $m_2 \in \left[-l_2,l_2\right]$, $\mu_1 \in \left[-\lambda_1,\lambda_1\right]$ and $\mu_2 \in \left[-\lambda_2,\lambda_2\right]$. All indices are integers and n must be also even.

I have been using Varshalovich's Book, but can't find any identities that have been useful to simplify this. I am hoping that the result is something like $\delta_{l_2,\lambda_2}\delta_{m_2,\mu_2}\delta_{l_1,\lambda_1}\delta_{m_1,\mu_1}$, but I'm not sure that that will be the case. Any ideas of how to evaluate this?

• Is $n$ any integer? – Vibert Jun 25 '13 at 23:08
• Well Mathematica has ClebschGordan and SixJSymbol functions but I can't get it to simplify your expression. Even evaluating simple cases is taking me a long time. Maybe somebody with more Mathematica and/or combinatorics knowledge than me can find a trick. – Michael Brown Jun 26 '13 at 0:57
• @Vibert: $n \geq 0$ is an even integer, $0 \leq l \leq n$ is an integer, all other indices take integer values and their limits follow from the definition of the CG coefficients and 6j symbols. Sorry about not stating that before. – okj Jun 26 '13 at 1:47
• Specifically: $$n \in \left[0,\infty\right)$$ $$l,l_1,l_2,\lambda_1,\lambda_2 \in \left[0,n\right]$$ $$m \in \left[-l,l\right]$$ $$m_1 \in \left[-l_1,l_1\right]$$ $$m_2 \in \left[-l_2,l_2\right]$$ $$\mu_1 \in \left[-\lambda_1,\lambda_1\right]$$ $$\mu_2 \in \left[-\lambda_2,\lambda_2\right]$$ ($n$ is an even integer, all other indices are integers) – okj Jul 2 '13 at 12:47