see selected recent publications [2] [3] [4] [7] and [11]
As part of our interest in phosphorus ylides, we have discovered that their protonation with organic acids resulted in the formation of hydrocarbon-soluble phosphonium salts. For example, reaction of triphenylphosphonium methylide with 2,4,6-trimethylphenol provided a crystalline solid that proved to be the first example of a phosphonium aryloxide to be isolated and characterised. The X-ray structure (Figure 13) revealed extensive C-H…O hydrogen bonding between phosphonium donors and aryloxide acceptor groups. In general, these phosphonium aryloxides may be viewed as model intermediates for the hydrolysis of moisture-sensitive ylides, and also have potential application as reagents in organic synthesis, although we have not explored this latter possibility as yet. Of more interest to date has been the development of this strategy in order to study and understand the process of aggregation in the solid state via weak intermolecular interactions. We have characterised a wide range of similar phosphonium aryloxides, all of which exhibit the same structural motif in the solid state and we have demonstrated the control of their supramolecular structure by varying the steric requirements of the cations and anions. This has in turn enabled us to rationalise the complex hydrogen-bonding patterns in inorganic phosphonium salts which contain similar structural motifs in polymeric arrays. We also have characterised a number of these compounds by neutron diffraction in order to establish fully the structural parameters of hydrogen bonds. This work has emphasised the relatively strong nature of these interactions; in one case the C-H…O hydrogen bonding was the shortest (by some 5%) yet to be observed by neutron diffraction. More recently, we have extended this work to include the partial deprotonation of multifunctional organic acids. In these cases the supramolecular structure can be propagated by strong O-H…O interactions as well as the C-H…O interactions observed in simple examples leading to some beautiful and unusual solid-state structures (Figure 14). Another aspect of this work has been the synthesis and structural characterisation of the first examples of phosphonium amides and phosphides. In addition to being novel in themselves, have also allowed the definitive solid-state characterisation of intermolecular C-H…N and, for the first time C-H…P hydrogen bonds. Having laid down the ground rules for the supramolecular aggregation our intent is to develop this programme further: (i) as a fundamental study into weak intermolecular interactions; (ii) as a means of understanding more complex structural motifs in related solids; and (iii) as a route to the controlled synthesis of materials possessing potentially useful properties.
In the section above, hydrogen bonding in solids was propagated by the interaction of C-H, and O-H acidic cations with anionic centres. However, we are also interested in hydrogen bonding in which two or more neutral molecules cocrystallise via X-H…Y interactions, where X-H is an acidic donor group and Y is a basic acceptor group. Our studies in this area have so far concentrated on two types of hydrogen bond acceptors: first, phosphane oxides and iminophosphoranes; and second, cyclic ureas. Although these appear unrelated, calculations show that they have a very similar dipolar charge distribution in their P=O, P=NR or C=O bonds making them all highly efficient hydrogen bond acceptors with a wide range of OH-, NH- and CH-acidic donors. Detailed below are two areas in which we have obtained results recently.
We have studied hydrogen bonded adducts of the neutral C2H12B10 carboranes. Cocrystallisation of ortho, meta, and para-carboranes with HMPA resulted in three 1:1 adducts (Figure 15). In each case, the different orientation of the C-H groups results in each adduct adopting a different supramolecular arrangement. Furthermore, because the carborane icosohedra are ‘locked’ into position via the hydrogen bonding these structures provided the first definitive structures of unsubstituted carboranes. In the absence of hydrogen bonding or substitution, the high symmetry of the icosohedra, coupled with the very similar X-ray scattering power of carbon and boron results in unresolvable disorder of these atoms which has in the past prevented analysis of individual bond lengths within the parent carboranes. As an interesting aside, we have found that reaction of the (Me2N)3P=NH, the imino analogue of HMPA, with carboranes has resulted in the isolation and characterisation of a range of very different products. In addition to interacting with the acidic C-H groups as HMPA does the NH may attack a BH group. In the presence of atmospheric moisture this results in deboronation of the closo cage to give a nido carborane and various other decompostion products. In collaboration with Dr Mark Fox, we have been able to structurally characterise the initial adduct (Figure 16) and a number of intermediates on the way to decomposition.
A second area in which we have recently been active is in the characterisation of hydrogen bonded adducts of cyclic ureas. Several attractions of well-defined adducts of this type are perceived, including their solubility in hydrocarbon media without disruption of desirable solution properties of the organic acid, and the potential to combine two or more useful additives in one crystalline, stoichiometric adduct. We have synthesised a range of N-substituted cyclic ureas and characterised a number of adducts invloving orgnanic acids (e.g., Figure 17 and Figure 18). Work is ongoing in this area to synthesise and characterise new and potentially useful adducts.
| next | back |