(S)-(−)-1-Phenylethanaminium 4-(4,4-difluoro-1,3,5,7-tetramethyl-3a,4a-diaza-4-borata-s-indacen-8-yl)benzoate

The title compound, C8H12N+·C20H18BF2N2O2 −, crystallizes with a significant amount of void space [4.0 (5)%] in the unit cell. The structure displays N—H⋯O hydrogen bonding between the components. The plane formed by the benzoic acid moiety of the BODIPY-CO2 − is twisted by 80.71 (6)° relative to the plane formed by the ring C and N atoms of the tetramethyldipyrrin portion of the molecule.

The title compound, C 8 H 12 N + ÁC 20 H 18 BF 2 N 2 O 2 À , crystallizes with a significant amount of void space [4.0 (5)%] in the unit cell. The structure displays N-HÁ Á ÁO hydrogen bonding between the components. The plane formed by the benzoic acid moiety of the BODIPY-CO 2 À is twisted by 80.71 (6) relative to the plane formed by the ring C and N atoms of the tetramethyldipyrrin portion of the molecule.

Related literature
For the use of crystalline materials that contain emissive transition metal complexes for sensing small molecules, see: McGee & Mann (2007); Smith & Mann (2009). The boron dipyrrin family of dyes could be an alternative to these often costly transition metal complexes and can also be easily modified at the meso position, see: Erten-Ela et al. (2008);Ulrich et al. (2008). We have found that to sense small molecules effectively, empty channels must be present in the crystal structure to allow the analyte molecules to penetrate the crystalline lattice, see:   Table 1 Hydrogen-bond geometry (Å , ). We hypothesized, based on previous work of Lancaster et al., 2006, Imai et al., 2007, 2008, and Brock et al., 1991, that the co-crystallization of an optically pure chiral amine with the sensing chromophore could facilitate inefficient packing as the molecular chirality requires a non-centrosymmetric space group. To this same end, increasing the number of specific intermolecular interactions (i.e. hydrogen bonding, (Tominaga et al., 2011)) could increase the probability of inefficient packing while supporting and strengthening a lattice containing significant amounts of void space. Our initial attempt to apply this hypothesis experimentally was successful and is reported here. X-ray quality crystals of the amine-BODIPY adduct showed both specific intermolecular interactions as well as channels of void space.
The BODIPY-CO 2 H crystallizes with the chiral amine (S)-(-)-α-methylbenzylamine in the solid-state by forming an ammonium moiety and a deprotonated carboxylic acid pair linked by hydrogen bonds (Figure 1). Information regarding the hydrogen bonding within this structure can be found in Table 1. It is also possible that this hydrogen is partially occupied on both N3 and O2, but this model did not significantly improve the refinement statistics.
PLATON/VOID (Spek, 2009) was used to determine the amount and location of the solvent accessible void space within the structure. Voids of 52 (1) Å 3 were found in the unit cell which corresponds to 4.0 (5) % of the total unit cell volume.
Molecules such as methanol (37 (1) Å 3 ) and water (22 (1) Å 3 ) have molecular volumes consistent with their possible incorporation in the void cavities (Buss et al., 1998). Figure 2 shows the packing of the structure and the location of the void space as represented by the red spheres. These isolated void channels run parallel to the b axis with spokes of void space perpendicular from the main channel. These spokes form around the hydrogen bonding that occurs throughout the structure.
supplementary materials sup-2 The hydrogen bonding pattern can be described as two identical parallel tape interactions that are joined by an additional interaction to form a two-dimensional lattice of hydrogen bonding. The first nearly linear tape is described by graph set notation as C 2 2 (6), the second zigzag tape as C 1 2 (4), and the ring interaction that results from the hydrogen bond that spans the two tapes as R 3 4 (10).

Refinement
Attempts to place one hydrogen atom on the carboxylic acid moiety and two on the amine moiety gave poorer refinement statistics. The better model is that shown in Figure 1 as an ion pair.
All aromatic H atoms were placed in ideal positions and refined as riding, with C-H = 0.95 Å and U iso (H) = 1.2U eq (C).
Methyl H atoms were placed in ideal positions and refined as riding, with C-H = 0.98 Å and U iso (H) = 1.5U eq (C). Those methyl H atoms on the BODIPY-CO 2 molecule were modeled over two postions 60 ° from one another with half occupancy.
Hydrogen atoms located on the ammonium moiety were placed in ideal positions and refined as riding, with N-H = 0.87 Å and U iso (H) = 1.5U eq (N).
The chirality of the ammonium bearing molecule is known to be S at C27 as the S isomer of the chiral amine was introduced as a co-crystallant. The Flack x parameter (Flack, 1983) based on refinement with 2071 Friedel pairs was 0.4 (10), indicating no conclusions can be drawn regarding the absolute structure and Friedel pairs were merged before final refinement of the structure.
supplementary materials sup-3 Figures   Fig. 1. Labeled diagram of the asymmetric unit. Ellipsoids are drawn at 50% probability and methyl hydrogen disorder has been removed for clarity. Fig. 2. View of the packing down the a axis. Areas of void space are drawn as red spheres. Thermal ellipsoids are drawn at 50% probability and all hydrogen atoms have been removed for clarity.

Special details
Experimental. Cell errors are from iterative updates since the crystal is believed to have been moving during the data collection. Refinement. Refinement of F 2 against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F 2 , conventional R-factors R are based on F, with F set to zero for negative F 2 . The threshold expression of F 2 > σ(F 2 ) is used only for calculating Rfactors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F 2 are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger.
Errors in the CIF check pertaining to the Flack parameter should be ignored. Not only is this structure a light atom structure, but the chirality of the amine is known explicitly from the synthesis of the material to be S at C27.
This same argument can be used to ignore the other errors regarding the Friedel data. Merging the data with the MERG 4 command did not significantly decrease the number of errors received in the cif report, and in fact made the number of significant errors increase.
Consequently, the structure was not refined with the MERG 4 command.