2,6-Diacetylpyridine–resorcinol (1/1)

The title co-crystal, C9H9NO2·C6H6O2, is composed of one 2,6-diacetylpyridine molecule and one resorcinol molecule as the asymmetric unit. In the 2,6-diacetylpyridine molecule, the two carbonyl groups are antiperiplanar to the pyridine N atom. In the crystal, the 2,6-diacetylpyridine and resorcinol molecules are connected by two O—H⋯O hydrogen bonds, forming planar chains of alternating components running along [120].

The title co-crystal, C 9 H 9 NO 2 ÁC 6 H 6 O 2 , is composed of one 2,6diacetylpyridine molecule and one resorcinol molecule as the asymmetric unit. In the 2,6-diacetylpyridine molecule, the two carbonyl groups are antiperiplanar to the pyridine N atom. In the crystal, the 2,6-diacetylpyridine and resorcinol molecules are connected by two O-HÁ Á ÁO hydrogen bonds, forming planar chains of alternating components running along [120].   Table 1 Hydrogen-bond geometry (Å , ). We thank Dr Guido Wagner for quantum-mechanical calculations of the relative stability of the 2,6-diacetylpyridine conformations and Professor Dr E. Egert for helpful discussions.

Quoc-Cuong Ton and Michael Bolte Comment
The co-crystallization process between two components which possess either donor or acceptor hydrogen bond properties in order to obtain the AAA-DDD (A=Acceptor, D= Donor) hydrogen bond pattern containing two strong O-H···O hydrogen bonds and one weak C-H···O hydrogen bond (see Fig. 1, III) was the initial motivation of this research. Therefore, 2,6-diacetylpyridine, I, (CSD REFCODE: BARKAH) and resorcinol, II, (CSD REFCODE: RESORA03) have been chosen for this purpose. Compounds I and II can exist in three possible conformations (Anwar et al., 2007).
Considering all possible hydrogen bonds between the two components, forming the complex as mentioned above is the most unfavourable constellation. The calculations of the three molecular conformations of 2,6-diacetylpyridine using quantum-mechanical calculations (Gaussian 03) predict that conformer Ia is the most stable, followed by Ib, and then Ic.
The determination of the relative stability of resorcinol using quantum-mechanical density functional theory (DFT) said that conformer IIa is the most stable, followed by IIb, and then IIc. Two out of three conformers of resorcinol have been observed in neutron powder experiments (Bacon & Lisher, 1980). Beyond that all three conformations have been found in diverse multi-component-complexes (Boldog et al. 2004;MacGillivray et al. 2000;Friščić & MacGillivray, 2009;Matsumoto et al., 2006) where resorcinol showed these conformations. Another possibility of building a finite hydrogen bond network between the two components is highlighted as an example (V in Fig. 1), where different conformers are involved. The energy for the conversion of the relative stable conformers Ia and IIb to the least energetically favoured conformational states Ic and IIc is estimated to be approximately 60 kJ/mol. The co-crystal of the title compound ( Fig. 2) in the constellation of Ia and IIb adopts a chain motif (IV in Fig. 1) (Fig. 3). The desired complex (III in Fig. 1) was not formed.

Experimental
The starting compounds 2,6-diacetylpyridine and resorcinol were purchased from Aldrich and Alfa Aesar which were used for co-crystallization experiments without purification. The starting compounds were dissolved in a 1:1 molecular ratio in ether and setlaid aside at room temperature. After several weeks adequate single crystals were obtained.

Refinement
All H atoms were refined using a riding model with fixed individual displacements parameters [U iso (H) = 1.2 U eq (C) or U iso (H) = 1.5 U eq (C methyl , O)] with C aromatic -H = 0.95 Å, C methyl = 0.98 Å, and O-H = 0.84 Å. The methyl and hydroxyl groups were allowed to rotate but not to tip.    Refinement. Refinement of F 2 against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F 2 , conventional R-factors R are based on F, with F set to zero for negative F 2 . The threshold expression of F 2 > σ(F 2 ) is used only for calculating R-factors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F 2 are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger.