Laurel L. Schafer , University of British Columbia
Selective activation of sp3 hybridized C-H bond adjacent to an amine, and the subsequent functionalization of this bond, is an attractive approach for the synthesis of a-chiral amines. When mediated in a catalytic fashion, this strategy results in the 100% atom economic preparation of structurally diverse amines from primary amine and alkene starting materials in one C-C bond forming reaction while avoiding all amine protection/deprotection sequences. Preliminary results in our laboratory have attained such a transformation (the cyclization of an aminoalkene to give a substituted cyclohexylamine) in up to 80% yield. Preliminary investigations suggest that multi-metallic species are responsible for the interesting reactivity observed and we have postulated a bridging imido as a metal-protected amine to permit C-H activation adjacent to the nitrogen.
Since submission of the proposal we have disclosed a Zr pyridonate complex for this emerging catalytic technology that has been coined hydroaminoalkylation (preliminary work funded from other sources). Notably the Ti congener of the same pyridonate complex is not catalytically viable for the same transformations. Indeed, the complexes from our program that are capable of mediating this C-C bond forming transformation were initially developed for catalytic hydroamination. Since the submission and funding of this proposal Prof. Sven Doye has shown that other Ti hydroamination catalysts, including the commercially available Ti(NMe2)4 and Ti(Bn)4 complexes can be used for this transformation in an intramolecular and intermolecular fashion. Most recently he has reported kinetic investigations using Ti(NMe2)4 as the catalyst system and the turnover limiting step has been proposed to be the intramolecular C-H activation step required before product elimination. We have established Zr(NMe2)4 can mediate this transformation with select substrates, however, it is not broadly useful for this transformation. We have also verified that our Zr pyridonate complexes are only useful for the intramolecular version of this reaction and importantly, they are selectively reactive with primary amines substrates. Thus, initial work on this project, which has been undertaken by two doctoral students, has focused on what features control intramolecular hydroaminoalkylation reactivity, including metal, ligand structure, concentrations, solvents and reaction temperatures. These investigations have been pursued using a 2 pronged approach. One student has focused on mechanistic investigations, which includes the stoichiometric preparation of reactive intermediates and reaction kinetics. This work is presently being prepared for publication. In our second approach we are methodically varying the ligand environment about the reactive metal center, to test our current working hypothesis that sterically accessible metal complexes must be used to promote the formation of reactive dimers. Furthermore, the electronic impact of metal complex upon reactivity is an area of on-going investigation.
We have identified 1-amino-5-hexene derivatives as our preferred intramolecular screening reaction, as the reaction outcome can be either 6-membered N-containing heterocycles resulting from hydroamination or 5-membered amine substituted carbocycles resulting from the targeted hydroaminoalkylation. Thus, with this preferred screening reaction we are able to identify metal complexes that show chemoselectivity for hydroaminoalkylation over hydroamination reactivity. Various group 4 bis(N,O) chelated complexes have been screened using this protocol and we have found that for bis(amidate) ligated systems Ti complexes are preferred over their Zr counterparts, which show no sign of hydroaminoalkylation. This observation for bis(amidates) is in contrast to the previously reported bis(pyridonate) system. The most suitable complex was that of Ti bis(N-(2,6-dimethylphenyl)pivalamidate) complex with a starting material consumption of 55.4% with no hydroamination products observed by 1H NMR spectroscopy. It seems that less sterically bulky amidate ligands are preferred for hydroaminoalkylation, in contrast to hydroamination catalysis. However, these group 4 bis(amidate) complexes are not as reactive as commercially available and previously reported Ti(NMe2)4.2
Thus, less bulky titanium bis(pyridonate) complexes were synthesized from commercially available pyridones and screened with three different primary aminoalkene substrates making 5- and 6-membered rings. Electron-withdrawing groups such as Cl and CF3 groups did not favour hydroaminoalkylation, and it seems that substituents at the 6-position of pyridones are unfavourable for hydroaminoalkylation. The most promising pyridone ligands are those with substituents at the 3-position (a-to carbonyl). Such ligands show promise for 5-membered ring formation via hydroaminoalkylation, however, their reactivity for 6-membered cyclization is not yet satisfactory.
Mono(η5-cyclopentadienyl)-mono(amidate)-bis(dimethylamido) Ti precatalysts for intramolecular hydroaminoalkylation have been prepared, however these complexes show no hydroamination reactivity and only limited hydroaminoalkylation reactivity. Furthermore, mono(η5-cyclopentadienyl)-mono(pyridonate)-bis(dimethylamido) Ti precatalysts were screened, but their reactivity and selectivity were not as promising as bis(pyridonate) Ti complexes.
To date, we have found that in the presence of less sterically demanding ligands, Ti is indeed the metal of preference for this transformation. We have also established that recently disclosed ureate catalyst systems are completely selective for hydroamination and there is no observed hydroaminoalkylation. Complementary N,N chelated Ti complexes, aminopyridinate complexes, have been prepared in collaboration with Prof Rhett Khempe of the U. of Bayreuth Germany and these new precatalyst systems also show chemoselectivity for hydroamination over hydroaminoalkylation.
Kinetic investigations show a first order dependence upon substrate and non-linear dependence upon catalyst concentration. Using KIE of deuterated substrates a turnover limiting C-H activation step is presently postulated. Not surprisingly an Eyring analysis reveals a very large and negative entropy of activation, consistent with an highly order transition state. At present we favour a catalytic mechanism that procedes via formation of bridged species and on-going efforts target the discrete preparation of organometallic reactive intermediates.
Future directions include the development of diastereoselective hydroaminoalkylation catalysts. Using this catalytic manifold, outstanding diastereoselectivities in the ring-closing step can be achieved and it may be possible to access the trans-diastereomer which is very difficult to obtain using traditional reductive amination approaches. With such stereoselectivity in hand, applications in the preparation of select amine substituted carbocycles can be pursued and indeed such targets have attracted the interest of potential industrial partners for this research. Finally, this project will target the development of asymmetric catalysts to give enantiopure a-chiral amines. Thus, this PRF grant provided critical seed funding to establish a new catalytic technology and pursue fundamental investigations of this alternative reaction manifold.