Indenyl rhodium N-Heterocyclic carbene complexes for catalytic C-H borylation

Abstract

Metal-catalysed C-H activation offers the ability to access key synthetic targets in more straightforward reactions than previously used methods. However, undirected activation pathways face issues of selectivity and low rates of reaction that make substituting simple hydrocarbons difficult. Indenyl (Ind) and fluorenyl ligands offer increased reactivity compared to cyclopentadienyl groups, which have been used previously in C-H borylation, and combining these donors with electron-donating NHC ligands was investigated for the borylation of arenes and alkanes. Additionally, the effects of tethered systems were explored to see whether the catalytic ability is enhanced. [Rh(Ind)(SIPr)(C2H4)], [Rh(Ind)(SIPr)(COE)] and [Rh(Ind)(SIPr)(CO)] (SIPr = 1,3- bis(2,6-diisopropylphenyl)-4,5-dihydroimidazol-2-ylidene, COE = cis-cyclooctene) were synthesised and characterised by multinuclear NMR spectroscopy and X-ray diffraction. Only the ethylene and cyclooctene complexes were found to be reactive under photolytic conditions and towards silanes. Photolysis led to the loss of coordinated alkenes and the formation of a cyclometallated species due to C-H activation of the NHC substituents. With reducing silanes or hydrogen, a rhodium dihydride complex was observed, that is hypothesised to form via the reaction of the cyclometallated species, while less reducing silanes led to the formation of the oxidative addition product. Both [Rh(Ind)(SIPr)(C2H4)] and [Rh(Ind)(SIPr)(COE)] were found to be catalytically competent for the borylation of benzene, while the carbonyl complex was found to be unreactive under these conditions. Borylation of a selection of arenes showed that the selectivity was comparable to previously reported rhodium catalysts, which is dominated by steric effects, however, the reactivity was lower compared to previously reported catalysts such as [RhCp*(C6Me6)]. Borylation of decane and octane showed that the cyclooctene complex was capable of borylating alkanes, albeit in low yields. Stoichiometric experiments monitored by NMR spectroscopy provided evidence that the catalysis proceeds via rhodium boryl hydride species, with the previously identified cyclometallated species also likely to play a role. The synthesis of fluorenyl-tethered saturated-NHC ligands required the development of homobimetallic synergic bases in order to bring about a ring-opening deprotonation of a spirocyclic intermediate. The structure of [Li2(μ2 ‐Ph){μ2 ‐N(SiMe3)2}] was crystallographically characterised as a coordination polymer, and reaction with the spirocyclic compound led to the formation of dialkali metal complexes of a fluorenidetehered NHC ligand that incorporated a bridging amide group. The use of these bimetallic complexes as ligand transfer reagents gave rhodium carbonyl and ethene complexes in low yields. Initial testing of these complexes in the borylation of benzene found that the carbonyl species was inactive while the ethene complex was less active than the related monodentate species. Overall, this research has demonstrated that NHC ligands can be used to develop Rhcomplexes capable of C-H activation, the oxidative addition of silanes and the catalytic borylation of hydrocarbons. This supports the idea that a [Rh(Ind)(NHC)] fragment (16 electron for η 5 -indenyl, or 14 electron with η 3 -indenyl) can mimic the reactivity of the previously successful [Rh(Cp)(L)] and [Rh(Cp*)] fragments. Although the compounds synthesised in this thesis were not better catalysts than literature examples, they hold much promise because the incorporation of a tuneable NHC ligand on the metal centre can lead to future improvements, especially considering the potential importance of cyclometallated species in C-H activation reactivity