Comment

The title compound, (I), is a hydrochloride salt of adenine, which is one of the two common purine bases found in ribose and deoxyribose nucleic acids.

The unit cell was determined in 1974 (Iwasaki, 1974); however, it was not possible unequivocally to establish the correct space group, either Pna2₁ or Pnam (non-standard setting of Pnma), as refinement in each gave similar R values (0.043 and 0.045, respectively). The structure was also determined at room temperature by Kistenmacher & Shigematsu (1974), and refined in the centrosymmetric space group Pnma, giving an R value of 0.035. In this space group, mirror symmetry is imposed on the adenine dication, with some atoms having large r.m.s. displacements normal to the mirror plane. However, it was argued that purines commonly show some bending about the C2-C3 bond axis (Sletten & Jessen, 1969), which is inconsistent with the analysis in the centrosymmetric space group. Hence, it was suggested that the true space group could be Pn2₁a (non-standard setting of Pna2₁).We have redetermined the crystal structure at 150 K, to gain more precise data for our molecular modelling studies. The structure was refined in both Pnma and Pna2₁, giving R values of 0.0241 and 0.0229, respectively, despite the statistical averages for the normalized structure factors (E values) being more consistent with a centrosymmetric than a non-centrosymmetric distribution. However, when refined in the non-centrosymmetric space group, all the ring H atoms deviate by between 13-15° from the mean ring plane to which they are attached. These are large deviations when compared with other adeninium crystal structures, which include adeninium sulfate (Langer & Huml, 1978), adeninium dinitrate (Hardgrove et al., 1983) and adeninium diperchlorate monohydrate (Bendjeddou et al., 2003). In addition, analysing the non-centrosymmetric structure with PLATON (Spek, 2003) to search for missing or higher symmetry gave the centrosymmetric structure at 100% confidence level. Hence, using the superior low-temperature data, we can conclude that the most likely space group of (I) is Pnma.

In this low-temperature determination, the precision of the unit-cell dimensions was improved by an order of magnitude, and the unit-cell volume decreased by ca 14 Å³, consistent with the determination at low temperature. In general, the metric parameters are not significantly different, within standard deviations, from those found at room temperature. The adenine molecule is protonated at N1 and N3, with the C-N bond lengths in the rings in the range 1.308 (2)-1.375 (2) Å, and the C2-C3, C3-C4 and C4-N5 bond lengths being 1.379 (2), 1.409 (2) and 1.310 (2) Å, respectively. In the crystal structure, the cations are linked through N-HN hydrogen bonds to form extended chains in the a-axis direction. These chains are, in turn, linked by N-HCl hydrogen bonds to form sheets (Fig. 2) lying parallel to the (040) family of lattice planes. Four of the H atoms on the adenine cation are involved in N-HCl hydrogen bonds (see Table 1) and, in addition, atoms H4 and H6 are involved in weaker bifurcated N-HCl hydrogen bonds, with NCl distances of 3.2936 (15) and 3.5155 (16) Å, respectively. There are two independent Cl^- ions within the hydrogen-bonded sheets: Cl1, which is involved in one conventional and three weaker bifurcated N-HCl hydrogen bonds, and Cl2, which is involved in three conventional N-HCl hydrogen bonds. In the N-HN and N-HCl hydrogen-bonded sheets, all acceptors and donors are used.

Figure 1 View of (I), showing the atom-labelling scheme. Displacement ellipsoids are drawn at the 50% probability level.

Figure 2 View of the hydrogen-bonded sheet motif present in (I), with the hydrogen bonds shown as dotted lines; DA distances greater than 3.3 Å have been omitted for clarity.

Experimental

As part of an experimental polymorph screen on adenine, (I) was obtained by evaporation of a solution of equimolecular amounts of thymine/adenine, and cytosine/adenine in dilute hydrochloric acid, giving colourless block-shaped crystals.

Crystal data

C5H7N52+·2Cl- Mr = 208.06 Orthorhombic, Pnma a = 13.4405 (11) Å b = 6.4774 (5) Å c = 9.3684 (7) Å V = 815.61 (11) Å3 Z = 4 Dx = 1.694 Mg m-3 Mo K radiation Cell parameters from 5209 reflections = 2.7-28.1° = 0.74 mm-1 T = 150 (2) K Block, colourless 0.74 × 0.26 × 0.24 mm

Data collection

Bruker SMART APEX diffractometer Narrow-frame scans Absorption correction: multi-scan (SADABS; Sheldrick, 1996) Tmin = 0.609, Tmax = 0.842 6736 measured reflections 1076 independent reflections 1064 reflections with I > 2(I) Rint = 0.016 max = 28.3° h = -17 17 k = -8 8 l = -12 12

Refinement

Refinement on F2 R[F2 > 2(F2)] = 0.024 wR(F2) = 0.063 S = 0.99 1076 reflections 94 parameters All H-atom parameters refined w = 1/[2(Fo2) + (0.0337P)2 + 0.5379P] where P = (Fo2 + 2Fc2)/3 (/)max < 0.001 max = 0.37 e Å-3 min = -0.25 e Å-3

Table 1 Hydrogen-bonding geometry (Å, °) D-HA D-H HA DA D-HA N1-H2Cl1 0.94 (3) 2.11 (3) 3.0274 (14) 167 (2) N3-H3Cl2i 0.89 (3) 2.25 (3) 3.0693 (14) 153 (2) N4-H4Cl1ii 0.90 (2) 2.53 (2) 3.2936 (15) 143.8 (19) N4-H4Cl2ii 0.90 (2) 2.56 (2) 3.1695 (14) 126.1 (18) N5-H6N2iii 0.85 (2) 2.28 (2) 2.899 (2) 129.6 (19) N5-H6Cl1 0.85 (2) 2.82 (2) 3.5155 (16) 140.3 (18) N5-H7Cl2i 0.88 (3) 2.22 (3) 3.0985 (16) 175 (2) Symmetry codes: (i) x,y,1+z; (ii) ; (iii) .

H atoms were refined independently using an isotropic model.

Data collection: SMART (Bruker, 2000); cell refinement: SAINT (Bruker, 2000); data reduction: SAINT; program(s) used to solve structure: SHELXS97 (Sheldrick, 1990); program(s) used to refine structure: SHELXL97 (Sheldrick, 1997); molecular graphics: SHELXTL (Bruker, 2000) and MERCURY (Bruno et al., 2002); software used to prepare material for publication: SHELXL97.