Bis(benzyltrimethylammonium) bis[(4SR,12SR,18RS,26RS)-4,18,26-trihydroxy-12-oxido-13,17-dioxaheptacyclo[14.10.0.03,14.04,12.06,11.018,26.019,24]hexacosa-1,3(14),6,8,10,15,19,21,23-nonaene-5,25-dione] sesquihydrate: dimeric structure formation via [O—H—O]− negative charge-assisted hydrogen bonds (–CAHB) with benzyltrimethylammonium counter-ions

The title compound forms a supramolecular dimeric entity via [O—H—O]− negative charge-assisted hydrogen bonds (–CAHB) following a reaction with benzyltrimethylammonium fluoride salt.

The reaction between bis-ninhydrin resorcinol and benzyltrimethylammonium fluoride in ethanol has produced the title compound, 2C 10 H 16 N + Á2C 24 H 13 O 8 À Á-1.5H 2 O, which contains a unique centrosymmetric supramolecular dimeric entity, where two deprotonated ligands are held together via two strong and short [OÁ Á ÁO = 2.4395 (13) Å ] [O-H-O] À bonds of the type negative chargeassisted hydrogen bonds (-CAHB). The central aromatic rings of the ligands create parallel-displacedstacking at an interplanar distance of 3.381 (1) Å , which helps stabilize the dimer. In the crystal, two symmetry-related solvent water molecules with a site occupancy of 0.75 are attached to the carbonyl groups of the dimer by weaker O-HÁ Á ÁO hydrogen bonds, forming chains along [101].

Chemical context
The vasarene family consists of self-assembled, vase-shaped compounds and their analogues, which are prepared by a onepot reaction between cyclic vicinal polycarbonyl compounds and polyhydroxy aromatics (Na et al., 2005;Almog et al., 2009). The supramolecular behaviors of these structures have been an ongoing study in our group, particularly their intriguing feature of selective affinity towards ion-pairs of type M + F À , M being a large monovalent cation (Almog et al., 2012). A recent study has shown that the multiple oxygen-containing functional groups of these ligands (hemiketals, carbonyls and hydroxyls) play a key role in this supramolecular binding mechanism by forming dimeric entities via strong [O-H-O] À hydrogen-bonding (Bengiat et al., 2016).

Structural commentary
The dimer was formed following the reaction of bis ninhydrin resorcinol (1) with benzyltrimethylammonium fluoride, in which the fluoride acted as a base removing a proton from the hemiketal hydroxyl group (Scheme). Several factors help in stabilizing this dimeric entity. The first is thestacking of ISSN 2056-9890 the middle aromatic rings that are parallel-displaced but could almost be considered as a 'sandwich' conformation due to the minor angle of displacement (15 ). The interplanar distance between the two rings is also quite short [3.381 (1) Å ] supporting the strength of this interaction (Janiak, 2000).
The two [O-H-O] À negative charge-assisted hydrogen bonds (CAHB), although deviating from linearity [164 (2) ], are still considerably strong and short -with an OÁ Á ÁO distance of 2.4395 (13) Å , corresponding to low-barrier hydrogen bonds (LBHB) (Cleland et al., 1998). Additional hydrogen bonding (Table 1) between the remaining hydroxyl groups O7-H7O, O3-H3O and the etheric hemiketal oxygen atoms O1 and O5, respectively, assist in stabilizing the dimer (Fig. 1). Fig. 2 shows that the steric benzyl groups of the cations remain beside the ligands and parallel to each other, with two water molecules hydrogen bonded to the carbonyl groups on the ligands (O1W-H2W1Á Á ÁO8). Two cell units also display parallel-displacedstacking between the aromatic rings of the 'side-walls' of the ligands with an interplanar distance of 3.349 (1) Å (Fig. 3).

Refinement details
Crystal data, data collection and structure refinement details are summarized in Table 2. The site occupancy of the water was set at 0.75 during the refinement process, as when defining a value of 1 the R-factor increased considerably by 0.7%. Hydroxyl H atoms of the ligand molecules and H atoms of the water molecule were located in a different Fourier map and all H-atom parameters were refined except for those of the water molecule for which only the U-parameters were refined. Other H atoms were placed in calculated positions with C-H = 0.93 (aromatic) and 0.96 A (methyl), and refined in a riding-model approximation with U iso (H) = 1.2U eq (C) for aromatic and aliphatic H atoms and 1.5U eq (C) for the methyl H atoms.

Figure 3
The parallel-displacedstacking between two aromatic rings on the 'side-walls' of the ligands of two different cell units showing the interplanar distance between the rings. The cations, solvent molecules and aromatic hydrogen atoms have been removed for clarity.  Computer programs: SMART and SAINT (Bruker, 2002), SHELXS97, SHELXL97 and SHELXTL (Sheldrick, 2008), ORTEPIII (Burnett & Johnson, 1996) and Mercury Data collection: SMART (Bruker, 2002); cell refinement: SAINT (Bruker, 2002); data reduction: SAINT (Bruker, 2002); program(s) used to solve structure: SHELXS97 (Sheldrick, 2008); program(s) used to refine structure: SHELXL97 (Sheldrick, 2008); molecular graphics: ORTEPIII (Burnett & Johnson, 1996) and Mercury (Macrae et al., 2008); software used to prepare material for publication: SHELXTL (Sheldrick, 2008).  where P = (F o 2 + 2F c 2 )/3 (Δ/σ) max < 0.001 Δρ max = 0.26 e Å −3 Δρ min = −0.42 e Å −3 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. 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 R-factors(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.