Start with a sodium 2-bromobenzenethiolate and 4-chloro-3-nitrobenzonitrile, that will react with heat to a sulfur-bridged bicyclic compound. Then you reduce the nitro-moiety with SnCl2 (or Fe, but nothing that reduces the benzonitrile) and react the aromatic amine and the Br in a third step with K2CO3 in DMF (sometimes a Cu(0)-catalyst is used here). Et voila, you're done...
this has been described in the literature: Marivingt-Mounir, Cecile; Mettey, Yvette; Vierfond, Jean-Michel, Journal of Heterocyclic Chemistry (1998), 35(4), 843-845. We don't have the journal, so I can't send it, but perhaps your library can organise the article for you. The yields seem fairly low, however, and there are patents which report higher yields by different routes. See this one here: Guo, Wei et al
From Faming Zhuanli Shenqing Gongkai Shuomingshu, 1583732, 23 Feb 2005.
You can also buy it from Molecula http://molekula.com/chemicals/product/64214784
We have bought things from them before and they were reliable.