AmericanChemicalSociety.com

Summary. In the field of homogenous catalysis, few reports have appeared offering general, chemically mild methods for the introduction of nitrogen- or oxygen-containing functional groups into simple hydrocarbon substrates. The combination of small-molecule activation with C—H bond functionalization represents a significant advance in converting inexpensive chemical feed stocks (e.g. hydrocarbons) to valuable functional molecules with minimal or complete absence of waste generation. We are developing multi-electron redox transformations by mid-to-late, first row transition metal complexes targeting methods for the functionalization of C–H bonds. We have synthesized a new class of electrophilic complexes featuring transiently-formed, or metastable metal-ligand multiple bonds capable of mediating C–H functionalization. The hard donor, weak-field pyrrole-based ligand scaffolds were selected for this study primarily for their ability to confer a range of steric and electronic environments to the bound transition metal ion. The coordination chemistry was investigated for the redox-active dipyrromethane and tris(pyrrolyl)ethane supported transition metal complexes to establish the electronic structure for the pyrrole-based ligand donors. Dipyrromethene (or semi-porphyrin) supported ferrous complexes were found to effect a range of intra- and intermolecular C–H bond functionalization reactions - allowing for the construction of new C–N, C–C and C–O bonds from unactivated C–H bonds, culminating in catalytic amination reactions. While electron-releasing ligands (phosphines, carbenes, and β-diketiminates) have been elegantly employed to stabilize metal-ligand multiple bond formation on mid-late transition metal platforms, we posit that the weak-field ligand environment is critical to the  overall efficacy of the dipyrromethene-supported catalyst system. The following is a brief summary of the research performed during the PRF funding period during the past 24 months.

Coordination chemistry of pyrrole-based ligand scaffolds (Specific Aim #1).

Our strategy to effect sequential bond activation and functionalization requires a ligand system or systems that (1) confer stability to metal ions over a range of possible oxidation states (e.g. M^II ↔ M^IV), (2) maintain a minimal coordination environment at the metal center and create a sterically enforced pocket to protect metal-ligand multiply-bonded linkages, and (3) create an electrophilic metal environment (limit ligand electron-donicity). Pyrrole-derived ligands (Scheme 1): monoanionic dipyrromethene, dianionic dipyrromethane and trianionic tris(pyrrolyl)ethanes are good candidates to satisfy all of the desired design criteria. The pyrrole substituents exhibit attenuated π-basicity of the nitrogen donors from conjugation with the pyrrole π-system, satisfying criterion 3 and making these platforms more amenable for coordination to late transition metal ions.

Figure 1 shows representative transition metal complexes of the three primary ligand platforms used in this study. In this fashion, divalent complexes of Mn→Zn featuring the monoanionic dipyrromethene, dianionic dipyrromethane, and trianionic tri(pyrrolyl)ethane ligands have been prepared. In general, the pyrrole groups are found to coordinate in an η¹-fashion, accommodating a range of coordination geometries at the metal ion as confirmed by X-ray diffraction studies. Given the weak-field nature of the pyrrolide donors, the resulting paramagnetic complexes were characterized by standard spectoscopic means (¹H NMR, UV-Vis, SQUID magnetometry, EPR where appropriate, Mössbauer for Fe-containing complexes, cyclic voltammetry, mass spectrometry, and combustion analysis).

Figure 1. Solid state molecular structures for (left) (^Mesdpme)FeCl(thf), (center) (^Mesdpma)Co(py)₂, and (right) [^MestpeNi(py)]^– (thermal ellipsoids set at 50%).

The dianionic (^Mesdpma) ligand afforded high-spin, pseudo-tetrahedral complexes for Mn^II, Fe^II, and Co^II, but afforded a low-spin, square-planar Ni^II with pyridine as the ancillary ligands. Despite the ligand's dianionic nature, the pyrrolide donors proved to be weak σ-donors in addition to being poor π-donors. Surveying the (^Mesdpma)M(py)₂ complexes by differential pulse voltammetry revealed two one-electron oxidation events (Figure 2a), indicating the oxidations are independent of the bound metal, its spin state or geometry (i.e., the voltammagrams neatly overlay for the Co^II, Ni^II and Zn^II complexes). Thus we concluded the origins of the redox for the (^Mesdpma) complexes is entirely ligand based (corroborated by DFT analysis), indicating there is very little covalency between the (^Mesdpma) platform and the bound metal ion. The dipyrromethene (^Mesdpme) is a very intense chromophore, akin to their porphyrin analogues. Binding a transition metal ion causes a very modest red-shift in the ligand π→π* electronic transition (Figure 2b).  Cyclic voltammetry on (^Mesdpme)FeCl(THF) reveals a fully reversible Fe^III/II redox wave centered at -400 mV (Figure 2b).

Figure 2. (a) Differential pulse voltammagrams of solutions in THF of (^Mesdpma)M(py)₂; (b) UV-vis spectra of (^Mesdpme)M for M = H, Li(THF), FeCl(THF); inset displaying CV for (^Mesdpme)FeCl(THF).

Reactivity Studies. The dipyrrolide redox activity in (^Mesdpma)M complexes has precluded our attempts to observe oxidative group transfer reactions. However, the more redox-stable dipyrromethene complexes have proven to be well-suited to canvassing reactivity. We have shown the (^Mesdpme)FeCl(THF) complex  to be active for a range of intramolecular group transfer reactions. For example, the complex readily decomposes organic azides (electron-rich and electron-poor)  to afford products consistent with formal nitrene delivery to a ligand C–H bond (Scheme 2). The stoichiometric reactions are easily followed by ¹H NMR, as the desymmetrization of the ligand due to the C–H bond amination is readily apparent even with the paramagnetically shifted proton resonances. Reaction of (^Mesdpme)FeCl(THF) with dioxygen does afford a small amount of ligand hydroxylation, but the reaction was substantially improved (nearly quantitative) when the urea-H₂O₂ was used as an O-atom source. Reaction of (^Mesdpme)Fe(OTf)(THF) with several equivalents of N₂CHCO₂Et showed that 1-4 carbene additions (:CHCO₂Et) to the ligand framework occurred by ESI-MS, with the majority of the products showing 3 carbene insertions.

Intermolecular C–H bond amination. To generate a dipyrromethene Fe-complex capable of intermolecular C–H activation and functionalization, we substituted the (^Mesdpme) mesityl flanking units with tert-butyl substituents. Reacting (^tbudpme)FeCl(THF) with one equivalent of 1-azidoadamantane at -35°C in toluene (BDE: 89.7 kcal/mol), cyclohexane or linear hexane (BDE: 95.5 kcal/mol), or styrene consumes the azide and cleanly affords the respective secondary amine or aziridine (from styrene) products (Scheme 3, products analyzed by ¹H NMR, GC-MS). Preliminary reactivity screens suggest the intermolecular C–H bond functionalization could be carried out under the non-optimized catalytic conditions using alkyl azides (^tBuN₃, AdN₃) with 2.5 mol% (^tbudpme)FeCl(THF) in neat hydrocarbons at 50-60°C.