Comment

The biological activity of substituted pyrazoles has been the subject of much research in medicine aimed at authenticating their potential properties. In fact, these compounds constitute an important class of bioactive drug targets in the pharmaceutical industry that includes blockbuster drugs such as celebrex (Penning et al., 1997) and viagra (Terrett et al., 1996). Pyrazole nuclei have attracted pharmacological interest as anti-anxiety, antipyretic, analgesic, anti-inflamatory, antiparasitic and antimicrobial drugs (Elguero et al., 2002), and some related derivatives have been described as potent PDE4B inhibitors (Card et al., 2005). Pyrazole compounds have also been used as ligands for obtaining transition metal complexes, since their heterocyclic nuclei may coordinate to the metal directly via one or both of the vicinal N atoms (Rojas et al., 2004). In a previous paper, we reported preliminary results of the synthesis of a large library of pyrazoles by reaction of -diketohydrazones with substituted arylhydrazines (Bustos et al., 2009). We present here the structure of a new member of this family, the title compound, (I), and compare it with that of the previously reported 3,5-dimethyl-4-[(E)-(4-nitrophenyl)diazenyl]-1-(2,3,4,5,6-pentafluorophenyl)-1H-pyrazole, (II) (Bustos et al., 2007).

Scheme 1 shows the difference between the structurally isomeric molecules, (I) having the nitro group ortho to the N-C bond, while (II) presents the nitro group in the para position. In particular, the nitro group in (I) is rotationally disordered around the C-N bond (Fig. 1), with occupancy factors of 0.615 (3) and 0.385 (3). Although the N-O distances in (I) and (II) present some differences, this may be an artifact of the disorder. The remaining interatomic bonds and angles are almost indistinguishable and do not depart from commonly found values [Cambridge Structural Database (CSD), Version 5.33; Allen 2002]. There are, however, conformational differences at those sites with a low torsional energy barrier. Both facts can be clearly seen in Fig. 2, which presents a least-squares overlap of (I) and (II) where only the central core (labelled atoms in the figure) has been fitted. The rotational differences in the nitro groups and benzene rings are apparent, and can be quantified by the dihedral angles subtended by the various planes [values are given for (I) and (II), respectively]: Cg1-Cg2 = 72.16 (8) and -63.20 (16)°, Cg1-Cg3 = 7.26 (8) and -12.95 (16)°, and X1-Cg3 = 34.02 (12)/43.96 (15) and 4.19 (8)°, where X1 represents the plane(s) of the nitro group and Cg the planes of the rings depicted by the Cg symbols in Fig. 1].

Since there are no conventional hydrogen-bond donors present in (I) or (II), the supramolecular organization in both structures is due to weaker interactions. Fig. 3 presents a view of (I) perpendicular to (011), where the leading interactions (labelled in bold upper case) are clearly visible, viz. two FF contacts (A and B, first and second entries in Table 1) which concatenate pentafluorophenyl groups along the [100] direction in an extremely planar fashion. This can be assessed by inspection of the rightmost part of Fig. 3, where a side-on view of the pentafluorophenyl linkage is presented and quantified by the rather small zigzag dihedral angle of 12.3 (2)°. At this stage, it is interesting to comment on the characteristics of C-FF-C interactions, which, according to their geometric disposition, have historically been divided into types `I' and `II' (see Scheme 2); both cases analysed herein correspond to the first type. Although only type `II' contacts had originally been ascribed a stabilizing effect, further studies have begun to disclose a stabilizing character for many type `I' cases. For further details on this subject, see Baker et al. (2012, and references therein).

The FF interactions define comb-like chains (heavy lines, Fig. 3, left), interdigitated by their centrosymmetrically related counterparts (light lines, Fig. 3), and this particular set-up enables neighbouring rings 1 and 3 to lie almost parallel to each other in a favourable disposition for a - stacking interaction (C in Fig. 3, first entry in Table 2). The result is a family of double-chain strips running along [100].

Fig. 4, in turn, shows a view of the structure down the chain direction, displaying the way in which these strips interact with each other (one of them has been highlighted for clarity). Again, the responsible interactions are identified with bold upper-case labels; they run preferentially along [011] and are all weak contacts of very different types, viz. FF (D) or OF (E) (Table 1, third and fourth entries), or F (F) or O (G) (Table 2, second to fourth entries). The final result is a weakly but evenly connected three-dimensional structure.

It is interesting to analyse in parallel the closely related structure, (II), which in spite of the extreme analogies in molecular structure to (I) presents a substantially different packing scheme. As described by Bustos et al. (2007), the main interactions responsible for the supramolecular arrangement in (II) are two different - contacts linking rings 2 and 3, defining a family of -bonded planes parallel to (100) (Fig. 5). These planes are in turn connected via FF contacts between pentafluorophenyl groups in a head-to-tail fashion (Fig. 6, left), resembling the way in which the link in (I) is achieved but with two significant differences:

(i) The F-bonded strips they give rise to are not planar, but run in a definite zigzag fashion instead [Fig. 6, right; zigzag dihedral angle = 79.1 (2)°].

(ii) FF contact distances between symmetry-related pentafluorophenyl groups are longer in (II) than in (I) [F2F4' = 2.8384 (16) and 2.931 (3) Å, and F1F5' = 2.8328 (18) and 3.009 (3) Å, for (I) and (II), respectively], suggesting they are second-order interactions in (II).

Although this packing diversity is obviously due to differences in the nitro-group disposition, it is not clear if it is also a consequence of steric effects [e.g. the fact that molecule (II) is longer than (I)] or due to intrinsic interactions of the nitro groups; while there are no obvious interactions in (II), there are a few in (I), viz. a conspicuous contact between the minor component of the disordered NO₂ and a neighbouring F atom (Table 1, fourth entry), and some OCg contacts (Table 2, third or fourth entries).

Even if this fact could be left as an open issue, a definite conclusion from what is described above is that the structures discussed here provide further examples of type `I' C-FF-C interactions that should be given due consideration as effective synthons in supramolecular organization.

Experimental

3-[2-(2-Nitrophenyl)hydrazinylidene]pentane-2,4-dione (0.62 g, 2.5 mmol), perfluorophenylhydrazine (0.50 g), HOAc (5 ml) and ethanol (50 ml) were added to a 100 ml round-bottomed flask. The reaction mixture was stirred magnetically and heated under reflux for 36 h. After cooling to room temperature, the solid which formed was filtered off by suction and dried at 318 K for 24 h (yield 77%). Single crystals were obtained by saturation of tetrahydrofuran (20-25 ml) with the crude product, and the insoluble impurities were removed by filtration. The filtered product was treated with charcoal (0.1 g) and the mixture was heated gently near boiling for 5 min, after which the hot solution was filtered again. Finally, while maintaining the heating, the product was filtered again and treated with an EtOH-H₂O mixture (15 ml, 1:1 v/v). The hot solution was covered with a watch glass and allowed to stand at room temperature. After 4-5 d, orange crystals of (I) were separated by filtration and dried at room temperature (m.p. 453-454 K).

The nitro group showed rotational disorder [occupancy factors = 0.615 (3) and 0.385 (3)]. Its refinement required some metric similarity restraints for equivalent N-O and OO distances, but in spite of this some residual effects remained, mostly in the form of rather large Hirshfeld factors. All H atoms attached to C atoms were originally found in a difference Fourier map, but were subsequently repositioned in their expected positions and thereafter allowed to ride, with C-H = 0.96 Å and U_iso(H) = 1.5U_eq(C) for methyl H atoms, and with C-H = 0.93 Å and U_iso(H) = 1.2U_eq(C) for aromatic H atoms. Methyl groups were allowed to rotate around their C-C bond.

Data collection: SMART (Bruker, 2001); cell refinement: SAINT (Bruker, 2002); data reduction: SAINT; program(s) used to solve structure: SHELXS97 (Sheldrick, 2008); program(s) used to refine structure: SHELXL97 Sheldrick, 2008); molecular graphics: SHELXTL (Sheldrick, 2008); software used to prepare material for publication: SHELXTL and PLATON (Spek, 2009).

---

Supplementary data for this paper are available from the IUCr electronic archives (Reference: EM3056 ). Services for accessing these data are described at the back of the journal.

---