An investigation to elucidate the factors dictating the crystal structure of seven ammonium carboxylate molecular salts

Hydrogen-bonded ladders typically encountered in ammonium carboxylate salts did not form in the presence of a pyridine acceptor group.


Chemical context
Crystal engineering, the conception and synthesis of molecular solid-state structures, is fundamentally based upon the discernment and subsequent exploitation of intermolecular interactions. Thus, primarily non-covalent bonding is used to achieve the organization of molecules and ions in the solid state in order to produce materials with desired properties. Examples of such materials include organic field-effect transistors, hole collectors in organic photovoltaic cells (Snaith, 2013), laser materials (Tessler, 1999) as well as organic light-emitting diodes and semiconductors (Odom et al., 2003). The two principle forces exploited in the design of molecular solids are hydrogen bonding and coordination complexation (Desiraju, 1989).

Structural commentary
An amine and a carboxylic acid will combine to form a salt if the difference in pK a 's is approximately 3 or greater (Bhogala et al., 2005;Lemmerer et al., 2015). Thus, from the differences in pK a values depicted in Table S1 in the supporting information, all the compounds considered in this work should be in the form of salts and hence possess charge-assisted hydrogen bonds, which are considered to be a stronger and more robust supramolecular synthon than the same between neutral molecules (Lemmerer et al., 2008a). All structures crystallize with a 1:1 ratio of ammonium cation to benzoate anion, with all molecules on general positions. The asymmetric units and atom-numbering schemes are shown in Fig. 1.

Supramolecular features
Salt (I) consists of one 1-phenylethan-1-aminium cation and one isonicotinate anion. The ammonium group forms three charge-assisted hydrogen bonds, shown in Fig. 2a. The first of these bonds involves the O2 atom of the isonicotinate anion (i) (see Table 1) and is designated a. The second involves the O1 atom of the isonicotinate anion in the asymmetric unit and is designated b. The third involves the pyridine ring nitrogen of a third isonicotinate anion (ii) and is designated c. The b and c hydrogen bonds form a ring structure involving two of each kind of bond, consisting of two molecules of both 1-phenylethan-1-aminium and isonicotinate (See Fig. 2a). The graph set of this pattern is R 4 4 (18). A larger R 8 8 (30) ring is formed using all three hydrogen bonds involving four of both 1-phenylethan-1-aminium and isonicotinate ions. Overall, this forms a 2-D sheet as shown in Fig. 2b. As neither of the two expected type II or type III ladders are formed it seems that the additional hydrogen-bond acceptor in the form of the nitrogen atom of the pyridine ring disrupts their formation.
In salt (II), the asymmetric unit consists of one 1-phenylethan-1-aminium cation and one flurbiprofenate anion. Once again the ammonium group of the 1-phenylethan-1-aminium ion forms three charged-assisted hydrogen bonds ( Table 2). The first of these bonds involves the O2 atom of the anion while the other two involve the O1 atoms of the carboxylate group of two separate symmetry-related flurbiprofenate anions. These three hydrogen bonds form a type II ladder system where each of the O1 atoms behaves as a bifurcated hydrogen-bond acceptor, linking the rings (Fig. 3a). This pattern has translational symmetry through a twofold screw axis along the crystallographic b axis which is inherent in the space group P2 1 /n. As no short contacts such as halogen bonding or -halogen interactions are observed, the fluorine atom does not disrupt the formation of the expected hydrogen-bonding patterns. However a peculiarity exists. As the cation was present as a racemate, traditionally type III ladders are expected to dominate as reported by Lemmerer and co-workers (Lemmerer et al., 2008b).
In salt (III), the asymmetric unit consists of one 1-phenylethan-1-aminium cation and one 2-chloro-4-nitro-benzoate anion. The ammonium ion forms three charge-assisted hydrogen bonds to the carboxylate group and not to the nitro group of the 2-chloro-4-nitro-benzoate anion (Table 3). In fact, no relevant non-covalent interactions involving the nitro group are observed. As for compound (II), a type II ladder is formed by the above-mentioned hydrogen bonds, as shown in Fig. 3b. The anions in adjacent rings (related by translation along the b axis) are connected via C-OÁ Á ÁCl halogen bonds [OÁ Á ÁCl = 3.225 (1) Å ; C-OÁ Á ÁCl = 160.5 (1) ]. However, this interaction does not perturb the ladder supramolecular synthons to a significant enough degree to prevent their formation. Once again, both enantiomers of the 1-phenylethan-1-aminium were present and thus type III ladders were expected to form.
benzylammonium cation and the iodine atom (Fig. 3c). This is possible as, due to its size, iodine is very polarizable and thus the delocalized electrons in the aromatic system can create a permanent dipole in the iodine atom in the solid state. The distance of 3.740 (3) Å is similar to other molecules containing iodine interacting non-covalently with aromatic systems reported in the literature (Nagels et al., 2013). Again, as in salt (III), the halogen bonding does not disrupt the formation of the ladder motif. In salt (V), the asymmetric unit consists of one (S)-1cyclohexylethylammonium cation and one 2-chloro-4-nitrobenzoate anion, both on general positions. A type II ladder is formed as shown in Fig. 3d. No hydrogen bonding to the nitro group takes place (Table 5), which is consistent with the results for salt (III). However, a type I ClÁ Á ÁCl halogen bond is observed with a distance of 3.207 (3) Å that connects adjacent ladders along the a axis. As the cation is present as a single enantiomer, the type II ladder formation is in line with the previous studies.
In salt (VI), the asymmetric unit consists of one 2-(1cyclohexenyl)ethylammonium cation and one 4-bromobenzoate anion, both on general positions. A type III ladder is observed (Table 6), unique among the salts here reported (Fig. 3e). As in salt (V), the crystal structure is stabilized by halogen bonding, in this case between bromine and oxygen O1 with a distance of 3.253 (3) Å . The halogen bond connects adjacent ladders related by the two-fold screw axis.
In salt (VII), the asymmetric unit consists of one (S)-1cyclohexylethylammonium cation and one 4-bromobenzoate anion, both on general positions. A type II ladder is formed (Table 7, Fig. 3f). This is expected as the cation is enanti-omerically pure (Lemmerer et al., 2008b). In contrast to the previous salts that have a halogen atom on the anion, no halogen bonding is observed.
In summary, introducing a pyridine functional group instead of a plain benzene in (I) has altered the hydrogen-bonding pattern usually observed in ammonium carboxylate salts, which generally show the typical type II and III patterns as seen in (II)-(VII)

Synthesis and crystallization
All chemicals were purchased from commercial sources (Sigma Aldrich) and used as received without further purification. Crystals were grown via the slow evaporation of methanol or ethanol solutions under ambient conditions. All solutions contained a 1:1 ratio of amine and acid. Detailed masses and volumes are as follows. For (I): (RS)--methylbenzylamine (0.100 g, 0.825 mmol) and isonicotinic acid (0.102 g, 0.825 mmol) in methanol (5 mL

Refinement details
Crystal data, data collection and structure refinement details are summarized in Table 8. For all compounds, the C-bound H atoms were placed geometrically [C-H bond lengths of 1.00 Å (methine CH), 0.99 Å (ethylene CH 2 ), 0.98 Å (methylene CH 3 ) and 0.95 Å (Ar-H)] and refined as riding with U iso (H) = 1.2U eq (C) or U iso (H) = 1.5U eq (C). The N-bound H atoms were located in difference-Fourier maps and their coordinates and isotropic displacement parameters allowed to refine freely for (I)-(VI). For (VII), the N-bound H atoms were geometrically placed (C-H bond lengths of 0.91 Å ) and refined as riding with U iso (H) = 1.5U eq (C).
For the disorder of the atom C4 in the cyclohexene ring of (VI), two alternate positions were found in a difference-Fourier map, and their occupancies refined to final values of 0.77 (2) and 0.23 (2).

Special details
Experimental. Numerical integration absorption corrections based on indexed crystal faces were applied using the XPREP routine (Bruker, 2016) Geometry. All esds (except the esd in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell esds are taken into account individually in the estimation of esds in distances, angles and torsion angles; correlations between esds in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell esds is used for estimating esds involving l.s. planes.

Special details
Experimental. Numerical integration absorption corrections based on indexed crystal faces were applied using the XPREP routine (Bruker, 2016) Geometry. All esds (except the esd in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell esds are taken into account individually in the estimation of esds in distances, angles and torsion angles; correlations between esds in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell esds is used for estimating esds involving l.s. planes.

Special details
Experimental. Numerical integration absorption corrections based on indexed crystal faces were applied using the XPREP routine (Bruker, 2016) Geometry. All esds (except the esd in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell esds are taken into account individually in the estimation of esds in distances, angles and torsion angles; correlations between esds in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell esds is used for estimating esds involving l.s. planes.

Special details
Experimental. Numerical integration absorption corrections based on indexed crystal faces were applied using the XPREP routine (Bruker, 2016) Geometry. All esds (except the esd in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell esds are taken into account individually in the estimation of esds in distances, angles and torsion angles; correlations between esds in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell esds is used for estimating esds involving l.s. planes.

Special details
Experimental. Numerical integration absorption corrections based on indexed crystal faces were applied using the XPREP routine (Bruker, 2016) Geometry. All esds (except the esd in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell esds are taken into account individually in the estimation of esds in distances, angles and torsion angles; correlations between esds in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell esds is used for estimating esds involving l.s. planes.