a) with N-heterocyclic carbene complexes

Stable carbenes based on the heterocyclic structure 1, an N-heterocyclic carbene (NHC), have had an enormous impact on modern organometallic chemistry; they bind as strong, basic, s-donor ligands to many metals. We have developed new ligands that combine strongly basic N-heterocyclic carbenes and anionic functional groups such as alkoxides and amides (eg 2) (Chem. Commun. 2001, Angew. Chem. 2003, Chem. Rev. 2009). This allows us for the first time to explore the influence of the carbene group on harder f-block metals, and to exploit the combination of the reactive metal, and C-donor heterocycle in small molecule reactivity studies, and catalysis.

This has allowed the isolation of the first f-element-gallium bond and the first 3d-4f (Nd-Fe 3) bond (JACS 2007, Chem Commun. 2009), along with regioselective carbene silylation and a range of small molecule activation chemistry. The comparison of the unusual tetravalent s-bound cerium carbene complexes 4 with the uranium analogue provides information on the subtle differences in covalency between the lanthanide and actinide series that improves our fundamental understanding of the processes in nuclear waste separation (Chem. Commun. 2007,Chem. Eur. J. 2010). Most recently, we have shown (JACS 2010) that a new type of reaction can occur for redox-innocent organolanthanide NHCs. Here (4) the NHC is used as a reactive centre where a species EX first adds across a metal-NHC bond, leading subsequently to the elimination of the functionalised lanthanide-bound substrate (RE).  The procedure can be used inter alia for silylation, borylation, and phosphination of the metal's σ-bound group and most recently we have demonstrated C-Si and C-C bond forming reactions (JACS 2011). This chemistry also offers routes to new metal-metal bonds, and to functionalise small molecules such as N2, CO and CO2. Finally, a collaboration with Prof. Melanie Sanford (Michigan, US) is exploiting the oxidative stability of the ligands for catalytic C-H oxidation at the other end of the periodic table; 5 is a tuneable PdIV complex that catalyses for hydrocarbon sp2-C-H bond halogenation (JACS 2009).

The uranyl dication, [UO₂]²⁺, is the most prevalent form of uranium and is a soluble and problematic environmental contaminant. Over half of all uranium compounds (including oxide solids) contain the rigorously linear, trans- uranyl dication [UO₂]²⁺ with strongly covalent trans-UO2 bonding. They shows little propensity to participate in the myriad of oxo-group and redox reactions that are characteristic of transition metal analogues such as [MoO₂]²⁺. In collaboration with Dr Jason Love's group, we have shown that the hinged, ditopic Pacman macrocycle, L, can only bind a single uranyl ion, since there is only space for one linear O=U=O unit. This has allowed us to address the reactivity of the oxo groups in a constrained, asymmetric environment, and target oxo group reactivity instead of the normal ligand exchange processes that occur in the equatorial binding plane.

We have since demonstrated a new type of chemical reaction of the uranyl ion, in which it is rendered sufficiently oxidising to break N-Si and C-Si bonds, 6 (Nature 2008, JACS 2007), affording the first covalently oxo-functionalised uranyl, and used this uranyl activation to now show that C-H bonds can be broken by lithiation of the uranyl complex, 7 (Nature Chem. 2010), the first thermal C-H bond activation by a uranyl complex.  This reactivity mimics that of transition metal oxo catalysts for C-H activation chemistry, hinting at new vistas in uranium-catalysed hydrocarbon bond activation chemistry. We are now exploring the chemistry of these oxo-functionalised complexes to make models of relevance to aqueous waste remediation, such as hydroxyls 8 (Angew. Chem. 2011), and to build up bimetallic species with strong metal commmunication through oxo-bridges and interesting electronic/magnetic properties ( first lanthanide-uranyl dimers 9, Angew. Chem. 2011).

These U complexes improve our fundamental understanding of the bonding in these dioxo cations, and eventually might help us better understand the processes involved when soluble uranyl salts are reductively precipitated out of contaminated water, or help us to make better models of the neptunium and plutonium [AnO2]n+ analogues, which display a range of oxo-atom basicity and redox chemistry due to their fn configuration, but whose radioactivity precludes their manipulation in normal labs.

We have shown that the simple U(III) amide can couple ambient pressure CO gas exclusively to the [OCCO]2- ion, which demonstrates unprecedented second C-C bond formation upon warming 10 (Chem. Sci. 2011). Having shown that related simple aryloxide systems bind and activate not only CO, but also N, and CO2 11 (JACS 2011). We are working towards the first low-temperature/pressure catalytic CO homologation process. We have also been exploring use of the labile carbene in bifunctional catalysis, i.e. in which both metal and NHC activate substrates (12). Excellent control of lactide polymerisation has led to the further design of homochiral complexes for stereoselective polylactide and other biodegradable polymer syntheses 13 (Angew. Chem. 2008). We have made a series of chiral bidentate ligands based on phosphine oxides.The ligands are so large that they display 'ligand self-recognition': a racemic mixture of ligands predominantly forms homochiral RRR and SSS LnL3 homoleptic complexes with the smaller lanthanides. This has not been achieved before at a single lanthanide metal centre; lanthanide ions are so large, this assembly process is difficult to control. The resulting C₃-symmetric complexes are highly active catalysts for the polymerisation of polar, biorenewable monomers, such as lactide, into biodegradable polymers. Moreover, lactide is polymerised with a relatively unusual iso-tactic polymer structure. Thus if the racemic mixture of RRR and SSS LnL3 is used to polymerise racemic lactide (the cheapest form of the corn-derived monomer), the R and S monomers are polymerised separately, but in the same pot. Subsequent annealing forms what is known as a stereocomplex polymer, in which the strands form a polymer with a higher melting point than is possible for optically pure R orS polymer samples 14. Current work is focused on further CO2 chemistry with these catalysts.