A Revision of the Azulene Structure

Abstract

In a previous communication from this laboratory¹, a crystal structure for azulene based on the non-centrosymmetric space group Pa was reported. It was assumed that the (0k0) halving was accidental and it was easy to place the two molecules with respect to the glide plane in such a way that an almost exact halving of these spectra was obtained. This structure led to reasonably good agreements in all the principal zones, the discrepancies between observed and calculated structure factors being about 12 per cent for (h0l), 14 per cent for (hk0) and 20 per cent for (0kl). A convincing electron density map for the projection on (010) was obtained, and it was confidently expected that a further analysis of the full three-dimensional crystal data would lead to a complete refinement of the structure, with significant values for the bond-lengths and electron distribution. Our preliminary structure was further supported by an independent investigation made by Takeuchi and Pepinsky², who obtained almost identical results. Neither of us saw any evidence of disorder in our X-ray photographs, although Günthard³ has shown from a comparison of the entropies of azulene and naphthalene that there may be some disorder in the azulene crystal.