see selected recent publications [1] [5] [6] [8] [9] and [12]
Phosphorus ylides (R3P=CR’2) were transformed from a chemical curiosity into one of the most significant classes of organic reagents in the 1950s when Georg Wittig discovered the reaction that now bears his name. Since that time a high level of interest has been maintained in their structure and reactivity. In 1995 we reported the first synthesis and X-ray structure of a group 1-ylide complex, [Li(NCH2Ph)2× CH2PPh3]2 (Figure 1). Subsequently, we have prepared a number of such complexes of group 1 amides and alkoxides all of which exhibit coordination of an ylidic carbon atom to the metal centre. These complexes are unique in using carbon as a donor atom in s-block metal chemistry and give rise to some interesting coordination chemistry. For example, coordination of an ethylide to lithium leads to a rare example of a chiral carbon centre attached directly to an alkali metal centre. DFT calculations indicate that this coordination is surprisingly strong (i.e., comparable to that of a phosphane oxide) and and solution measurements demonstrate that it is maintained in solution. We are currently developing aspects of this coordination chemistry further by introducing other heteroatoms (including other ylidic carbon centres) into the neutral ylidic ligands.
In addition to exhibiting novel coordination chemistry, such complexes may have a significant role to play in organic chemistry. It is well known that the presence of lithium salts can alter the stereochemical outcome of Wittig reactions. This is attributed to partial coordination of the ylide to lithium. Our complexes are well defined and hydrocarbon soluble and so we have been investigating their reactions with carbonyl compounds. Our initial results are encouraging; stoichiometric complexes between ylides and lithium amides/aryloxides are able to perform Wittig reactions and the E:Z ratios of the alkene products differ significantly from those obtained under similar conditions with free ylides. Currently, our efforts are directed towards developing these new reagents and gaining insight into the mechanism of Wittig reactions involving lithium, about which little beyond empirical observation is known.
We have very recently extended this coordination chemistry to include examples of Group 2 complexes of phosphorus ylides. We have characterised a series of complexes by X-ray crystallography including monomeric bis-ylide complexes of calcium-, strontium- and barium amides (Figure 2). These complexes are the first structurally characterised examples of s-bonded organometallic compounds of the heavy alkaline earth metals. The use of such complexes, for example, as organic reagents or as potential anionic polymerisation initiators remains to be explored.
In tandem with our work on phosphorus ylides, we have also investigated the coordination chemistry of iminophosphoranes (R3P=NH), which are isoelectronic with phosphorus ylides and expected to show some similarities in their interactions with s-block metals. Indeed, we have shown that iminophosphoranes complexes analogous to those described above for ylides can be prepared for lithium and sodium aryloxides, magnesium Grignard reagents (Figure 3), and calcium, strontium and barium amides. Prior to this work no s-block metal complexes of neutral iminophosphoranes had been reported. Unlike phosphorus ylides, however, iminophosphoranes are prone to deprotonation. For example, lithium amides readily form N-metallated species making complexes of the neutral ligands inaccessible. Accordingly, we have also investigated the chemistry of these anions, which have proved to be highly versatile ligands, in particular for lithium (Figure 4) and magnesium (Figure 3). As above, our work is closely related to organic synthesis. For example, in this context, N-lithioiminophosphoranes are employed as synthons for RN2-. Our synthetic and structural work has shown, contrary to previous results, that the active species in such reactions is in fact a LiBr adduct of the N-lithioiminophosphorane both in solution and in the solid state (Figure 4). Results such as these have significant implications in the development of reagents since there is a proven relationship between solution structure and the reactivity/selectivity of a given reagent. We have also utilised similar N-lithioiminophosphoranes as transmetallation reagents for the generation of new N-metallated compounds such as N-cuproiminophosphoranes (Figure 5).
Recently, the synthesis of multimetallated organics has led to some spectacular results in s-block metal chemistry. Unusual coordination situations and supramolecular architectures provide many academically interesting results but the general concept of encapsulating an essentially ionic and ‘inorganic’ core in a hydrocarbon shell can also lead to insight into more commonly occurring situations offering potential for developing new reagents. Our recent work in this area has concentrated on the tetralithiation of calixarenes, a large class of phenolic macrocyclic molecules. In strictly moisture-free conditions, reaction of tbutlycalix[4]arene with an excess of nbutyllithium in the presence of HMPA (hexamethylphosphoramide) leads to a ‘dimeric’ lithium complex which contains a Li8O8 core (Figure 6). However, if moisture is introduced into the reaction (either deliberately or serendipitously!) a monomeric species results which contains a Li5O5 core (Figure 7). The fifth lithium and oxygen atoms in the Li5O5 core come from the capture of one molecule of LiOH per calixarene. The Li+ cation resides in the oxygen-lined lower cavity defined by the calixarene and the OH- anion is located in the lithium-lined upper cavity defined by four HMPA molecules. These two cores, both based on a square pyramidal structural motif are unique in organolithium chemistry. The capture of a small, inorganic molecule in two complementary cavities of a large, hydrocarbon-soluble species has obvious implications for the design of phase transfer reagents. Furthermore, it is interesting to speculate in the light of our results (and those of others) on the exact composition and structure of many supposedly pure organolithium species. We intend to continue this work on a number of fronts by studying: (i) the multimetallation of related cyclic and acyclic molecules; (ii) the capture of other small molecules by the system described above; and (iii) the feasability of using small inorganic species as templates around which s-block-metallated species can aggregate.
| next | back |