Crystallographic Evidence for a Symmetrical Structure for the Purpurate Ion

Abstract

THE accepted structure of the purpurate ion, C₈H₄N₅O₆⁻, consists of two trihydroxopyrimidine rings (probably in the triketo form) linked at the 5-positions by the group = N−. Two possibilities at once arise. Either (a) the bridge angle is about 120°, in which case the ring planes must be steeply inclined to one another owing to the steric effect of the carbonyl oxygen atoms. The double bond must then be localized on one side of the bridge, and the rings will not be equivalent (only one carries a negative charge). Although there is chemical evidence¹ that the rings are not equivalent, this is based on the supposed non-equivalence of the rings in alloxantin, and a recent determination of crystal structure² has shown that the latter is in fact centrosymmetric. Alternatively, (b) the bridge may be linear and the rings coplanar. This would permit delocalization of electrons over an eleven-atom conjugated system, but would require that the non-bonding electrons of the central nitrogen atom occupied a pure p-orbital, a situation for which there is no precedent.