The Synthesis of Organic Molecules of Intrinsic Microporosity Designed to Frustrate Efficient Molecular Packing

Abstract Efficient reactions between fluorine‐functionalised biphenyl and terphenyl derivatives with catechol‐functionalised terminal groups provide a route to large, discrete organic molecules of intrinsic microporosity (OMIMs) that provide porous solids solely by their inefficient packing. By altering the size and substituent bulk of the terminal groups, a number of soluble compounds with apparent BET surface areas in excess of 600 m2 g−1 are produced. The efficiency of OMIM structural units for generating microporosity is in the order: propellane>triptycene>hexaphenylbenzene>spirobifluorene>naphthyl=phenyl. The introduction of bulky hydrocarbon substituents significantly enhances microporosity by further reducing packing efficiency. These results are consistent with findings from previously reported packing simulation studies. The introduction of methyl groups at the bridgehead position of triptycene units reduces intrinsic microporosity. This is presumably due to their internal position within the OMIM structure so that they occupy space, but unlike peripheral substituents they do not contribute to the generation of free volume by inefficient packing.


Synthesis
The biphenyl fluorinated core precursor 1 was prepared in moderate yield (34 %) by the hexaethylphosphorous triamide mediated coupling of pentafluorobenzonitrile by adapting ap rocedure described in ap atent. [25] The terphenyl core precursor 2 was isolated as an unexpected by-product in low yield (6 %) but in sufficient quantity to facilitatei ts use as ac ore for adduct synthesis. Each adduct was prepared by mixingt he core precursor 1 or 2 with an excesso fa na rm unit chosen from 3-9 ( Figure 1) in anhydrous DMF at 65 8Cf or 48 h in the presence of potassium carbonate. Each adduct ( Figure 2) was purified either by column chromatography or by repeated trituration to remove more soluble impurities. Owing to the lack of symmetry of many of the substituted precursors, severalo ft he reported OMIMs/adducts were isolated as mixtures of inseparable regioisomers ( Table 1). The molecular mass, homogeneity and discrete nature of each adduct was confirmed by matrix-assisted laser desorption ionisation mass spectrometry (MALDI-MS) and gel permeation chromatography (GPC). For each adduct, the polydispersityi ndex obtained from GPC was less than 1.1.
The potential of 1 and 2 as OMIM cores wasf irst tested by their reactionw ith excess 3a to ascertain their reactivity towards aromatic nucleophilic substitution. After purification, crystalso ft he resultant adducts (10 and 14)w ere achieved by slow diffusion of methanol into chloroform solutions. X-ray crystallography coupled with 19 FNMR spectroscopy ( Figure 3) revealed these adducts to be the tetra-substituted adduct (10), with an ear orthogonal relationship between the two long  (14), with one residual fluorine atom on the central aryl ring. This substitution pattern was found consistently with all catechol adducts of 1 and 2 in this study.B ridgedp roducts, in which as ingle arm bridgest wo cores (Supporting Information, Figure S1) were often found as trace impurities;h owever,t hese were readily removed by columnc hromatography.D etailed synthetic procedures and spectroscopicd ata for all novel precursors and adducts are given in the Supporting Information.

Discussion
The simplest adduct (10)w as found to be effectively non-microporous, with an apparent BET surfacea rea (SA BET )o f 7m 2 g À1 as measured by nitrogen sorption at 77 Ka nd it was highly insoluble, suggesting that the constituent moleculesa re able to packt ogetheri na ne fficient manner, despite its awkward displaced cruciform structure. As predictedb ym olecular modelling, [21b] the introduction of bulky tert-butyl groups to the periphery of the adducts (11,12)i mproveda pparent microporosity (SA BET = 41 and 67 m 2 g À1 ,r espectively). Both adducts werei solated as mixtureso fr egioisomers and were found to be highly soluble in common solvents (for example dichloromethane and THF), which combined with the increasedp orosity,s uggestsadisruptiono ft he cohesivei nteractions between the constituent molecules. To furthers tudy this effect, 2, 3-dihydroxy-5,5,8,8-tetramethyl-6,6,7,7-tetrahydronapthalene( 3d)w as prepared and combined with 1 to give 13, which demonstrates comparable properties to 12.S imple adducts of core 2 were also prepared using three catechol precursors 3a-c.M uch like its biphenyl analogue, terphenyl-based adduct 14 was poorly soluble in commons olvents and possesses an egligible apparent SA BET of 13 m 2 g À1 .H owever,t he addition of two tert-butyl groups per arm gave ahighly soluble material( 16)w ith ag reater apparent SA BET of 102 m 2 g À1 over its biphenyl analogue (12,6 7m 2 g À1 ), suggesting that the higher functionality of the terphenyl core generates am ore porousmaterial.
Notably, triptycene peripheralu nits substituted with methyl groups at their bridgehead positions (5b,d )g ave OMIMs with lower apparent values of SA BET relative to their non-methyl containing counterparts. For example, OMIM-11 has an appar-ent SA BET of 351 m 2 g À1 ,a sc ompared to 423 m 2 g À1 for OMIM-10. The values of SA BET for the tert-butyl substituted analogues (OMIM-12a nd OMIM-13) also differ by as imilar amount (75 m 2 g À1 ). It appearst hat the space adjacent to the bridgehead in triptycene terminated OMIMs is directly contributing to the porosity of the material, hence,f illing this space with am ethyl group reduces the amount of intrinsic microporosity that can be generated during the amorphous packing of the molecules. Indirect evidence for this feature of amorphous packing comesf rom the single-crystal XRD analysiso fO MIM-1( Figure 5), which shows solvent-filledc hannels defined partially by the triptycene bridgehead positions. Similarl ocal ordering mayo ccur within the amorphous packing of triptycenebased OMIMs. Further XRD analysiso fO MIM-1, OMIM-5, and OMIM-8i sp resented in the Supporting Information, Figure S2-S4.
An investigation to determine the suitability of various rigid structuralb uildingu nits as peripheral arms for OMIM construction was performed by combining catechold erivatives of benzenopentacene (6) [20b] spirobifluorene (7), [28] propellane (8), [29]   Chem. Eur.J.2016Eur.J. , 22,2466Eur.J. -2472 www.chemeurj.org and hexaphenylbenzene (9) [30] with cores 1 and 2.B enzenopentacene containing OMIM-6 and OMIM-16, were found to possess apparent SA BET in the range 10-15 %l ower than their more symmetrical isomeric triptycene-based counterparts OMIM-5a nd OMIM-14. It is probable that this is due to greater intermolecular interactions between the long struts of the benzenopentacene arms as comparedt ot hose of triptycene. From an OMIM comparison, it can be concluded that the efficiency of unsubstituted structural units for generating microporosity is in the following order:p ropellane > triptycene > hexaphenylbenzene > spirobifluorene > naphthyl = phenyl. This is consistent with findings from previously reported packing simulation studies. [21] Conclusions Molecular adducts, designed to possess well-definedc oncavities, were synthesised using the reaction between fluorinated biphenyl and terphenyl cores and peripheral arms chosen from ad iverse range of rigid structural units. The intrinsic microporosityg enerated by the inefficient packing of these adducts was evaluated using nitrogen sorptiona t7 7K.T he use of small, planar,n on-substituted arms gave insoluble materials with negligible surfacea reas (< 30 m 2 g À1 ). However,b yu sing arms composed of bulky rigid structural units, OMIMs were prepared with apparent SA BET in the range 333-612 m 2 g À1 .T he efficiency of unsubstituted structuralu nits for generating microporosity is in the following order:p ropellane > triptycene > hexaphenylbenzene > spirobifluorene > naphthyl = phenyl. Substitution of these arms with bulky groups further enhanced microporosity (up to 726 m 2 g À1 ), which is presumably due to reducingi ntermolecular cohesive interactions. In contrast, the introduction of methyl groups at the bridgehead position of triptycene units reduced intrinsic microporosity.I nt his case, the internal positiono ft he methyl groups within the OMIM structure means that they occupy space but, unlike peripheral substituents, cannot contribute to the generation of free volumeb yfrustrating packing.
Full experimental details and spectroscopic data for all of the new compounds are given in the Supporting Information. CCDC 955894, 973327, 1406070, 1406071, 1406072, 1406073, and 1406074 contain the supplementary crystallographic data for this paper.T hese data are provided free of charge by The Cambridge Crystallographic Data Centre.