Crystal structures of five (2-chloroquinolin-3-yl)methyl ethers: supramolecular assembly in one and two dimensions mediated by hydrogen bonding and π–π stacking

Five closely related (2-chloroquinlin-3-yl)methyl ethers all exhibit different patterns of direction-specific intermolecular interactions, leading to the formation of different types of chain in four of them and sheets in the fifth.


Structural commentary
As noted above, the molecular constitutions of compounds (I)-(III) are very similar: those of compounds (I) and (II) differ only in the presence of a 6-methyl substituent in (II) which is absent from (I), while those of compounds (II) and (III) differ only in the presence of a bromo substituent in (II) which is absent from (III). Despite these close similarities, compounds (I)-(III) all crystallize in different space groups, P2 1 /n and Pbca, respectively, for (I) and (II), both with Z 0 = 1, and P2 1 2 1 2 1 with Z 0 = 4 for (III). A search for possible addi- The molecular structure of compound (I) showing the atom-labelling scheme. Displacement ellipsoids are drawn at the 30% probability level.

Figure 2
The molecular structure of compound (II) showing the atom-labelling scheme. Displacement ellipsoids are drawn at the 30% probability level.

Figure 3
The structure of a type 1 molecule of compound (III), showing the atomlabelling scheme. Displacement ellipsoids are drawn at the 30% probability level. tional crystallographic symmetry in compound (III) found none: comparison of the atomic coordinates for the Cl atoms within the selected asymmetric unit shows that while the x-coordinates of atoms Cl12 and Cl32 differ by ca 0.5 and their z-coordinates are almost identical, the y-coordinates of these two atoms differ by ca 0.13; similarly the x-coordinates of atoms Cl22 and Cl42 again differ by ca 0.5 but now the y-coordinates are almost identical, while the z-coordinates differ by ca 0.18. Hence it is not possible to identify even pseudosymmetry here. For compound (III), it will be convenient to refer to the molecules containing atoms N11-N14 as molecules of types 1-4, respectively. Compounds (IV) and (V) both crystallize with Z 0 = 1, in space groups P2 1 and P2 1 /c, respectively.
In compounds (I)-(III), the non-H atoms are almost coplanar, as shown by the relevant torsional and dihedral angles (Table 1). It is interesting to note that the orientation of the The structure of a type 2 molecule of compound (III), showing the atomlabelling scheme. Displacement ellipsoids are drawn at the 30% probability level.

Figure 5
The structure of a type 3 molecule of compound (III), showing the atomlabelling scheme. Displacement ellipsoids are drawn at the 30% probability level.

Figure 6
The structure of a type 4 molecule of compound (III), showing the atomlabelling scheme. Displacement ellipsoids are drawn at the 30% probability level.

Figure 7
The molecular structure of compound (IV) showing the atom-labelling scheme. Displacement ellipsoids are drawn at the 30% probability level.

Figure 8
The molecular structure of compound (V) showing the atom-labelling scheme. Displacement ellipsoids are drawn at the 30% probability level. ester function in compound (I) differs from that in compounds (II) and (III) ( Table 1 and Figs. 1-6): this difference may arise, at least in part, from the participation of the carbonyl O atom of the ester unit in short C-HÁ Á ÁO interactions in all of the molecules of compounds (II) and (III) but not in compound (I) ( Table 2). The non-H atoms in compound (IV) are also nearly coplanar, with a dihedral angle between the mean planes of the quinoline and naphthalene units of 7.39 (12) . By contrast, while the quinoline and pyridine units in compound (V) are nearly parallel (Fig. 8), with a dihedral angle between their mean planes of only 3.10 (9) , they are by no means coplanar, as indicated by the values of the torsional angles C2-C3-C37-O31, 92.08 (18), C3-C37-O31-C33, 165.21 (13) and C37-O31-C33-C32, À90.17 (17) . This again may perhaps be ascribed in part to the strong hydrogen bonds present in the crystal structure of (V) ( Table 2).
None of the molecules of compounds (I)-(V) exhibits any internal symmetry and hence all are conformationally chiral. For compounds (I), (II) and (V), the centrosymmetric space groups accommodate equal numbers of the two conforma-tional enantiomers, but only one such enantiomer is present in each crystal of compound (IV): the absolute configuration of the enantiomer present in the crystal selected for data collection was established by means of the Flack x parameter (Flack, 1983), although this has no chemical significance. For compound (III), the value of the Flack x parameter gives evidence of partial inversion twinning.

Supramolecular interactions
The supramolecular assembly in compounds (I)-(V) is determined by a variety of direction-specific intermolecular interactions, including bothstacking interactions and hydrogen bonds of C-HÁ Á ÁN, C-HÁ Á ÁO and C-HÁ Á Á types, as well as O-HÁ Á ÁN hydrogen bonds in compound (V) only. In compound (III), there are two fairly short intermolecular C-HÁ Á ÁN contacts involving C-H bonds from methyl groups bonded to the quinoline nucleus: not only are such bonds of low acidity, but these methyl groups are likely to be undergoing very rapid rotation about the adjacent C-C bonds  Table 1 Selected torsional and dihedral angles ( ) for compounds (I)-(III).
There are no hydrogen bonds of any kind in the crystal structure of compound (I), but molecules are linked into chains bystacking interactions. The fused aryl ring of the molecule at (x, y, z) and the brominated ring of the molecule at (Àx + 1, Ày + 1, Àz + 1) make a dihedral angle of 1.04 ; the ring centroid separation is 3.6168 (10) Å , and the shortest perpendicular distance from the centroid of one ring to the plane of the other is 3.4132 (6) Å , with a ring-centroid offset of ca 1.20 Å . For the heterocyclic ring at (x, y, z) and the brominated aryl ring at (Àx, 1 À y, 1 À z), the corresponding values are 1.52 (9) , 3.7454 (11) Å , 3.4357 (8) Å and ca 1.49 Å . The combination of these two stacking interactions links the molecules of (I) into a chain running parallel to the [100] direction ( Fig. 9). Two chains of this type pass through each unit cell but there are no direction-specific interactions between adjacent chains.
The only short C-HÁ Á ÁO contact in the structure of compound (II) has a C-HÁ Á ÁO angle of only 136 (Table 2), and so it is unlikely to be of major structural significance (Wood et al., 2009). However, there is a weakstacking interaction between molecules related by a 2 1 screw axis. The pyridyl ring at (x, y, z) and the brominated aryl ring at (Àx + 1 2 , y + 1 2 , z) make a dihedral angle of 3.87 (10) : the shortest perpendicular distance from the centroid of one ring to the plane of the other is 3.3816 (9) Å , but the ring-centroid separation is 3.882 (12), resulting in a ring-centroid offset of ca 1.78 Å . Thus there is only a very modest overlap of these rings and a consequently weak stacking interaction: if this interaction is, in fact, regarded as significant, it links the molecules into a -stacked chain running parallel to [010].
Within the selected asymmetric unit for compound (III), three of the four independent molecules, those of types 2, 3 and 4 (cf. Figs. 3-6), are linked by twostacking interactions, but the type 1 molecule does not participate in any such interaction. One of these stacking interactions involves the pyridyl ring of the type 2 molecule and the fused aryl ring of the type 3 molecule, while the other involves the pyridyl ring of the type 3 molecule and the fused aryl ring of the type 4 molecule. The dihedral angles between the ring planes within these two interactions are 3.11 (18) and 0.96 (7) , respectively, the ring-centroid separations are 3.553 (2) Å and 3.544 (2) Å , and the shortest perpendicular distances from the centroid of one ring in each interaction to the plane of the other ring are 3.4014 (15) and 3.3820 (15) Å , corresponding to ring-centroid offsets of ca 1.03 and ca 1.06 Å , respectively. The only short C-HÁ Á ÁN contact within the crystal structure of compound Part of the crystal structure of compound (I) showing the formation of a -stacked chain along [100]. For the sake of clarity, H atoms have been omitted. Atoms marked with an asterisk (*), a hash (#), a dollar sign ($) or an ampersand (&) are at the symmetry positions (Àx, Ày + 1, Àz + 1), (Àx + 1, Ày + 1, Àz + 1), (x À 1, y, z) and (x + 1, y, z), respectively.

Figure 10
Part of the crystal structure of compound (III) showing the formation of two independent chains running parallel to the [010] direction and formed separately by the molecules of types 1 and 3. For the sake of clarity, H atoms not involved in the motifs shown have been omitted. Atoms marked with an asterisk (*), a hash (#) or a dollar sign ($) are at the symmetry positions (Àx, y À 1 2 , Àz + 1 2 ), (x, y À 1, z) and (Àx + 1, y À 1 2 , Àz + + 1 2 ), respectively.
(III) has an HÁ Á ÁN distance which is not significantly less than the sum of the van der Waals radii, but there are four independent C-HÁ Á ÁO hydrogen bonds present in the structure although all are probably weak as they have quite small C-HÁ Á ÁO angles (Table 2). However, the pattern of these contacts is of interest as it precludes the possibility of any additional crystallographic symmetry in this structure where Z 0 = 4. One of the C-HÁ Á ÁO interactions involves only mol-ecules of type 1 which are related by the 2 1 screw axis along (0, y, 1 4 ), forming a C(6) (Bernstein et al., 1995) running parallel to the [010] direction (Fig. 10); an entirely similar chain is formed by type 3 molecules related to one another by the 2 1 screw axis along ( 1 2 , y, 1 4 ). However, the molecules of types 2 and 4 which are related by the 2 1 screw axis along ( 1 2 , y, 1 4 ) together form a C 2 2 (12) chain parallel to [010] (Fig. 11), which runs antiparallel to the chains formed by the molecules of types 1 and 3. Hence the patterns of supramolecular assembly in compounds (I)-(III), as well as their crystallization characteristics, show significant differences.
There are no hydrogen bonds of the C-HÁ Á ÁN or C-HÁ Á ÁO types in the crystal structure of compound (IV) and, despite the large number of independent aromatic rings, there are nostacking interactions. The only direction-specific intermolecular interaction is a weak C-HÁ Á Á(arene) contact involving molecules related by translation.
The supramolecular assembly in compound (V) is, however, rather more elaborate, resulting in part from the presence of additional hydrogen-bond donors and acceptors in the unfused pyridine unit. An intramolecular O-HÁ Á ÁO hydrogen bond (Table 2) gives rise to an S(7) (Bernstein et al., 1995) motif, and an intermolecular O-HÁ Á ÁN hydrogen bond links molecules related by the n-glide plane at y = 3 4 , forming a C(7) chain running parallel to the [101] direction (Fig. 12). In addition, inversion-related pairs of molecules are linked bystacking interactions involving the unfused pyridine ring of one molecule and the quinoline unit of the other (Fig. 13). Thus the unfused pyridine ring of the molecule at (x, y, z) and Part of the crystal structure of compound (III) showing the formation of a chain running parallel to the [010] direction and containing alternating molecules of types 2 and 4. For the sake of clarity, H atoms not involved in the motifs shown have been omitted. Atoms marked with an asterisk (*) or a hash (#) are at the symmetry positions (Àx + 1, y + 1 2 , Àz + 1 2 ) and (x, y + 1, z), respectively.

Figure 12
A stereoview of part of the crystal structure of compound (V) showing the formation of a C(7) chain formed by O-HÁ Á ÁN hydrogen bonds and running parallel to [101]. For the sake of clarity, H atoms bonded to C atoms have been omitted.

Database survey
The structures of a number of fairly simple 2-chloroquinolione derivatives related to compounds (I)-(V) have been reported in recent years. A structural study of a closely related group of six simply substituted 2-chloroquinolines (Hathwar et al., 2010) focused on supramolecular aggregation via C-HÁ Á ÁCl hydrogen bonds and attractive ClÁ Á ÁCl interactions. However, it must be pointed out firstly that it is now well established (Brammer et al., 2001;Thallapally & Nangia, 2001) that Cl atoms bonded to C atoms are extremely poor acceptors of hydrogen bonds, even from strong donors such as O-H or N-H; and secondly, that for none of the compounds in this group were the shortest intermolecular ClÁ Á ÁCl distances less than the sum of the van der Waals radii (Bondi, 1964;Nyburg & Faerman, 1985;Rowland & Taylor, 1996): indeed, the concept of the van der Waals radius was nowhere mentioned by the original authors. Two of the six compounds in this group contained 3-hydroxymethyl substituents and, in each of these, the molecules are linked into C(6) chains by means of O-HÁ Á ÁN hydrogen bonds.

Synthesis and crystallization
For the synthesis of compounds (I)-(V), a mixture of 0.4 mmol of the appropriate quinoline derivative, 2-chloro-3-(chloromethyl)quinoline for compounds (I), (IV) and (V) or 2chloro-3-(chloromethyl)-5-methylquinoline for compounds (II) and (III) and 0.4 mmol of the appropriate hydroxy compound, methyl 5-bromo-2-hydroxybenzoate for (I) and (II), methyl 2-hydroxybenzoate for (III), 1-hydroxynaphthalene for (IV), or 3-hydroxy-4,5-bis(hydroxymethyl)-2methylpyridinium chloride for (V), were dissolved in N,Ndimethylformamide (3-5 ml) together with potassium carbonate (2 mmol) and these mixtures were stirred at ambient temperature for 6-9 h, with monitoring by TLC. When each reaction was complete, ice-cold water (5 ml) was added and the resulting solid products were collected by filtration, washed with water and dried in air. Crystals suitable for singlecrystal X-ray diffraction were obtained by slow evaporation, at ambient temperature and in the presence of air, of solutions in dichloromethane, with yields in the range 86-97%.

Refinement
Crystal data, data collection and structure refinement details are summarized in Table 3. All H atoms were located in difference Fourier maps. C-bound H atoms were then treated as riding atoms in geometrically idealized positions: C-H distances 0.95-0.99 Å with U iso (H) = 1.5U eq (C) for the methyl groups, which were permitted to rotate but not to tilt, and 1.2U eq (C) for other C-bound H atoms.

Special details
Geometry. All e.s.d.'s (except the e.s.d. in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell e.s.d.'s are taken into account individually in the estimation of e.s.d.'s in distances, angles and torsion angles; correlations between e.s.d.'s in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell e.s.d.'s is used for estimating e.s.d.'s involving l.s. planes.