|Home | About | Journals | Submit | Contact Us | Français|
The molecules of the title compound, C20H23N3O6, are almost completely planar, apart from the H atoms bonded to tetrahedral C atoms. A combination of five hydrogen bonds, one of the N—HO type and two each of the C—HO and C—Hπ(arene) types, links the molecules into complex sheets.
Benzodiazepines are a very commonly prescribed class of psychoactive drugs with varying sedative, hypnotic, anxiolytic (anti-anxiety), anticonvulsant, muscle-relaxant and amnesic properties. As part of a programme for the synthesis of new derivatives of this type, we have developed the reaction between 2-(2-amino-4-nitroanilino)ethanol and the α,β-unsaturated ketone precursor 3-dimethylamino-4-methoxypropiophenone and, from this procedure, we have now isolated the title compound, (I), as a reaction intermediate, along with other products. We report here the molecular and supramolecular structure of (I) and we compare its structure with that of analogue (II) (Low et al., 2003 ), as well as with those of unsubstituted 2-amino-4-nitroaniline, (III) (Kolev et al., 2007 ), and the N-acetyl derivatives (IV)–(VI) (Clark et al., 2000 ) (see scheme).
The non-H atoms in the molecules of compound (I) are almost coplanar, as indicated by the relevant torsion angles (Table 1 ). The r.m.s deviation of the non-H atoms from the mean plane through all of the non-H atoms is only 0.067 Å, and the biggest individual deviations are 0.156 (3) Å for atom C13 and 0.138 (2) Å for atom C16, on opposite sides of the mean plane. For the peripheral substituents, methoxy atom C17 deviates from the plane of the C11–C16 ring by only 0.002 (3) Å, while nitro atoms O51 and O52 deviate from the plane of the C31–C36 ring by −0.007 (2) and 0.029 (2) Å, respectively. By contrast, the torsion angle defining the orientation of the 2-hydroxyethyl substituent in compound (II), analogous to the C32—N21—C22—C23 torsion angle in (I), is 83.1 (3)°, indicating an overall conformation far from planarity.
The bond distances involving the trisubstituted aryl ring show a number of unusual values. Firstly, in the nitro group the C35—N51 bond is short for its type (standard value = 1.468 Å; Allen et al., 1987 ), while the N51—O51 and N51—O52 bonds are both somewhat long (standard value = 1.217 Å; Allen et al., 1987 ). Secondly, the two exocyclic bonds, C31—N31 and C32—N21, differ somewhat in length. These values are consistent with the quinonoid behaviour often observed in 4-nitroaniline derivatives (Ferguson et al., 2001 ; Glidewell et al., 2002 , 2004 , 2005 ; Howie et al., 2004 ), but the C—C distances within the C31–C36 ring of (I) do not support this, as the shortest C—C distance is for the C34—C35 bond, while the longest is for the C31—C32 bond. Although the original report on (II) (Low et al., 2003 ) did not discuss the intramolecular geometry, the corresponding bond distances show a pattern very similar to that in (I). For compounds (IV)–(VI), the authors reported that all bond distances were close to standard values (Clark et al., 2000 ). In fact, the same behaviour for the exocyclic bonds is observed in (IV)–(VI) as is found in (I) and (II), but with the interesting qualifier that the C—C distances in the rings clearly support the quininoid form in (V) and (VI), but not in (IV). In unsubstituted compound (III), the quinonoid form is dominant in the electronic ground state, but with classical aromatic delocalization dominant in the first electronic excited state (Kolev et al., 2007 ).
Methoxy atom C17 in (I) is coplanar with the adjacent aryl ring, within experimental uncertainty. Accordingly, the two exocyclic angles at atom C14 differ by ca 9.0°, as is usual (Seip & Seip, 1973 ; Ferguson et al., 1996 ), while the C—O—C angle in the methoxy group is significantly larger than the tetrahedral value.
The supramolecular aggregation in compound (I) depends upon five hydrogen bonds of the N—HO, C—HO and C—Hπ(arene) types (Table 2 ), and these combine to generate sheets of considerable complexity. However, the formation of the sheets can readily be analysed in terms of two one-dimensional substructures. The first is built from N—HO and C—HO hydrogen bonds, while the second is formed solely by the two C—Hπ(arene) hydrogen bonds. In the first substructure the N—HO hydrogen bond links molecules related by translation into a C(8) (Bernstein et al., 1995 ) chain running parallel to the  direction (Fig. 2 ). By contrast, a chain of rings is present in the structure of compound (IV); here molecules related by an n-glide plane in the space group P21/n are linked into a C(4)C(7)[(7)] chain of rings, although the aggregation was not described as such by the original authors (Clark et al., 2000 ), and this may be contrasted with the simple translational chain in compound (I). Similarly, the two independent molecules in the structure of compound (V) form an exactly analogous motif, but this time built from molecules related by translation. The chain in compound (I) is augmented by the two C—HO hydrogen bonds to form a chain containing fused (12) and (14) rings (Fig. 2 ). Just as the individual molecules of (I) are effectively planar, so too is the chain formed from molecules related by translation; the broad chain runs parallel to  and its plane lies approximately parallel to (10).
In the second substructure, atom C2 in the molecule at (x, y, z) acts as hydrogen-bond donor to the C31–C36 ring in the molecule at (1 − x, 1 − y, −z), so linking a pair of molecules related by inversion. Similarly, atom C23 at (x, y, z) acts as hydrogen-bond donor to the C11–C16 ring in the molecule at (1 − x, y, − z), thereby linking a pair of molecules related by the twofold rotation axis along (, y, ). Propagation of these two hydrogen bonds thus generates a chain of edge-fused rings running parallel to the  direction, in which the component molecules are related by the inversion centres at (, , ), where n represents an integer, and by the twofold rotation axes along (, y, + ), where again n represents an integer (Fig. 3 ). The combination of the ribbon along  and the chain of rings along  generates a complex sheet lying parallel to (100). Two such sheets, related to one another by the C-centring operation, pass through each unit cell, but there are no direction-specific interactions between adjacent sheets. In particular, aromatic π–π stacking interactions are absent.
The constitution of compound (II) differs from that of (I) in two significant ways. Firstly, the ester function is absent, so that a free hydroxyl group is present, and secondly, the compound crystallizes as a stoichiometric 1:1 solvate with dimethylformamide (Low et al., 2003 ). The two N—H bonds are utilized in linking the major component to the solvent molecule, but they play no further role in the hydrogen-bonding scheme. The supramolecular aggregation is thus dominated by an O—HO hydrogen bond, which is necessarily absent from the structure of (I), and this, together, with three C—HO hydrogen bonds, generates a ribbon containing no fewer than five types of ring, two of them centrosymmetric. Again, aromatic π–π stacking interactions are absent, as are significant C—Hπ(arene) hydrogen bonds, so that the hydrogen-bonded structure of (II) is one-dimensional, as opposed to the two-dimensional structure of (I). The molecules of (VI) (Clark et al., 2000 ) are linked into the C(4) chain characteristic of simple amides, while in (III), a combination of four N—HO hydrogen bonds and one N—HN hydrogen bond links the molecules into a complex three-dimensional framework structure (Kolev et al., 2007 ).
A solution containing equimolar quantities (3.55 mmol of each reagent) of 2-(2-amino-4-nitroanilino)ethanol and 3-dimethylamino-4-methoxypropiophenone in acetic acid (5 ml) was heated under reflux for 25 min. The mixture was then cooled to ambient temperature and the solvent was removed. The resulting solid products were separated by column chromatography on silica gel, using dichloromethane–ethyl acetate (3:2 v/v) as eluent. The title compound, (I), was obtained from the last fraction (yield 20%, m.p. 420–421 K). Analysis found: C 59.6, H 6.0, N 10.2%; C20H23N3O6 requires: C 59.8, H 5.8, N 10.5%. Crystals of (I) suitable for single-crystal X-ray diffraction were grown by slow evaporation of a solution in ethanol.
All H atoms were located in difference maps. H atoms bonded to C atoms were then treated as riding atoms in geometrically idealized positions, with C—H distances of 0.95 (aromatic), 0.98 (CH3) or 0.99 Å (CH2), and with U iso(H) = kU eq(C), where k = 1.5 for the methyl groups, which were permitted to rotate but not to tilt, and 1.2 for all other H atoms bonded to C atoms. For the H atoms bonded to N atoms, the coordinates were freely refined, with U iso(H) = 1.2U eq(N), giving N—H distances of 0.80 (2) and 0.85 (3) Å.
Data collection: COLLECT (Nonius, 1999 ); cell refinement: DIRAX/LSQ (Duisenberg et al., 2000 ); data reduction: EVALCCD (Duisenberg et al., 2003 ); program(s) used to solve structure: SHELXS97 (Sheldrick, 2008 ); program(s) used to refine structure: SHELXL97 (Sheldrick, 2008 ); molecular graphics: PLATON (Spek, 2009 ); software used to prepare material for publication: SHELXL97 and PLATON.
Crystal structure: contains datablocks global, I. DOI: 10.1107/S0108270109032235/sf3112sup1.cif
The authors thank the Servicios Técnicos de Investigación of the Universidad de Jaén and the staff for the data collection. BI and YC thank COLCIENCIAS and the Universidad del Valle for financial support. JC thanks the Consejería de Innovación, Ciencia y Empresa (Junta de Andalucía, Spain), the Universidad de Jaén (project No. UJA_07_16_33) and the Ministerio de Ciencia e Innovación (project No. SAF2008-04685-C02-02) for financial support.
Supplementary data for this paper are available from the IUCr electronic archives (Reference: SF3112). Services for accessing these data are described at the back of the journal.