Search for: All records

We report that the cationic iridium complex (iPrPCP)IrH+ undergoes addition of alkane C-H bonds, which is manifested by catalytic alkane transfer-dehydrogenation to give alkenes and by hydrogen isotope (H/D) exchange (HIE). Contrary to established selectivity trends found for C-H activation by transition metal complexes, strained cycloalkanes, including cyclopentane, cycloheptane, and cyclooctane, undergo C-H addition much more readily than n-alkanes which in turn are much more reactive than cyclohexane. Aromatic C-H bonds also undergo H/D exchange much less rapidly than those of the strained cycloalkanes, but much more favorably than cyclohexane. The order of reactivity toward dehydrogenation correlates qualitatively with the reaction thermodynamics, but the magnitude is much greater than can be explained by thermodynamics. Accordingly, the cycloalkenes corresponding to the strained cycloalkanes undergo hydrogenation much more readily than cyclohexene, despite the less favorable thermodynamics of such hydrogenations. Computational (DFT) studies allow rationalization of the origin of reactivity and the unusual selectivity. Specifically, the initial C-H addition is strongly assisted by 𝛽-agostic interactions, which are particularly favorable for the strained cycloalkanes. Subsequent to 𝛼-C-H addition, the H atom of the 𝛽-agostic C-H bond is transferred to the hydride ligand of (iPrPCP)IrH+, to give a dihydrogen ligand. The overall processes, C-H addition and 𝛽-H-transfer to hydride, generally show intermediates on the IRC surface but they are extremely shallow, such that the 1,2-dehydrogenations are presumed to be effectively concerted although asynchronous. 

Pincer-ligated iridium complexes have been widely developed, and (pincer)Ir(III) complexes, particularly five-coordinate, are central to their chemistry. Such complexes typically bear two formally anionic ligands in addition to the pincer ligand itself. Yet despite the prevalence of halides as anionic ligands in transition metal chemistry there are relatively few examples in which both of these ancillary anionic ligands are halides or even other monodentate low-field anions. We report a study of the fragment (iPrPCP)IrCl2 (iPrPCP = 3 2,6 C6H3(CH2PiPr2)), and adducts thereof. These species are found to be thermodynamically disfavored relative to the corresponding hydridohalides. For example, DFT calculations and experiment indicate that one Ir-Cl bond of (iPrPCP)IrCl2 complexes will undergo reaction with H2 to give (iPrPCP)IrHCl or an adduct thereof. In the presence of aqueous HCl, (iPrPCP)IrCl2 adds a chloride ion to give an unusual example of an anionic transition metal complex ((iPrPCP)IrCl3–) with a Zundel cation (H5O2+). (iPrPCP)IrCl2 is not stable as a monomer at room temperature but exists in solution as a mixture of clusters which can add various small molecules. DFT calculations indicate that dimerization and trimerization of (iPrPCP)IrCl2 is more favorable than the analogous reactions of (iPrPCP)IrHCl, in accord with cluster formation being observed only for the dichloride complex. 

Pincer-ligated iridium complexes have been widely developed, and (pincer)Ir(III) complexes, particularly five-coordinate, are central to their chemistry. Such complexes typically bear two formally anionic ligands in addition to the pincer ligand itself. Yet despite the prevalence of halides as anionic ligands in transition metal chemistry there are relatively few examples in which both of these ancillary anionic ligands are halides or even other monodentate low-field anions. We report a study of the fragment (iPrPCP)IrCl2 (iPrPCP = 3-2,6-C6H3(CH2PiPr2)), and adducts thereof. These species are found to be thermodynamically disfavored relative to the corresponding hydridohalides. For example, DFT calculations and experiment indicate that one Ir-Cl bond of (iPrPCP)IrCl2 complexes will undergo reaction with H2 to give the (iPrPCP)IrHCl or an adduct thereof. In the presence of aqueous HCl, (iPrPCP)IrCl2 adds a chloride ion to give an unusual example of an anionic transition metal complex ((iPrPCP)IrCl3–) with a Zundel cation (H5O2+). (iPrPCP)IrCl2 is not stable as a monomer at room temperature but exists in solution as a mixture of clusters which can add various small molecules. DFT calculations indicate that dimerization of (iPrPCP)IrCl2 is more favorable than dimerization of (iPrPCP)IrHCl, in accord with its observed tendency to form clusters. 

The para-N-pyridyl-based PCP pincer ligand 3,5-bis(di-tert-butylphosphinomethyl)-2,6-dimethylpyridine (pN-tBuPCP-H) was synthesized and metalated to give the iridium complex (pN tBuPCP)IrHCl (2-H). In marked contrast with its phenyl-based congeners (tBuPCP)IrHCl and derivatives, 2-H is highly air sensitive and reacts with oxidants such as ferrocenium, trityl cation, and benzoquinone. These oxidations ultimately lead to intramolecular activation of a phosphino-t-butyl C(sp3)-H bond and cyclometalation. Considering the greater electronegativity of N than C, 2-H is expected to be less easily oxidized than simple PCP derivatives; DFT calculations of direct one-electron oxidations are in good agreement with this expectation. However, 2-H is calculated to undergo metal-ligand-proton tautomerism (MLPT) to give an N-protonated complex that can be described with resonance forms representing a zwitterionic complex (negative charge on Ir) and a p-N-pyridylidene (remote NHC) Ir(I) complex. One-electron oxidation of this tautomer is calculated to be dramatically more favorable than direct oxidation of 2-H (G° = 31.3 kcal/mol). The resulting Ir(II) oxidation product is easily deprotonated to give metalloradical 2• which is observed by NMR spectroscopy. 2• can be further oxidized to give cationic Ir(III) complex, 2+, which can oxidatively add a phosphino-t butyl C-H bond, and undergo deprotonation to give the observed cyclometalated product. DFT calculations indicate that less sterically hindered complexes would preferentially undergo intermolecular addition of C(sp3)-H bonds, for example, of n alkanes. The resulting iridium alkyl complexes could undergo facile -H elimination to afford olefin, thereby completing a catalytic cycle for alkane dehydrogenation that is driven by one-electron oxidation and deprotonation, enabled by MLPT. 

Our laboratory has reported that (CX3Phebox)Ir(H)(OAc) (X = H, F) catalysts are highly active for the acceptorless dehydrogenation of n-alkanes1, particularly in the presence of Lewis acids. In this work we report the synthesis of isoelectronic (Pybox)Os(H)(OAc) and (Pybox)Ru(H)(OAc), and investigation of these complexes for alkane dehydrogenation. DFT calculations predict (Pybox)Ru(H)(OAc) to catalyze acceptorless alkane dehydrogenation with a barrier lower than that for (CH3Phebox)Ir(H)(OAc), while the barrier calculated for (Pybox)Os(H)(OAc) is even lower. The rate-limiting step chem. for the catalytic cycle is calculated to be a net M-H/C-H σ-bond metathesis reaction, although expulsion of H2 from the reaction mixture was found to be rate-determining under typical conditions for acceptorless n-alkane dehydrogenation catalyzed by (CF3Phebox)Ir(H)(OAc). H/D exchange experiments were used to probe the kinetics of C-H activation yielding the order of activity: (Pybox)Os(H)(OAc) > (Pybox)Ru(H)(OAc) > (CF3Phebox)Ir(H)(OAc). Exptl. investigation of catalysis by (Pybox)Ru(H)(OAc) and (Pybox)Os(H)(OAc) is still in progress but the Ru complex, unfortunately, does not appear to be stable at the high temperatures required for acceptorless alkane dehydrogenation. We have also reported that (CH3Phebox)Ir(C2H4)2 catalyzes selective dehydrogenative coupling of ethylene to butadiene via an iridacyclopentane complex.2 In this work we used the precursor (Pybox)OsH4 to investigate the same catalytic reaction and appears to result in and analogous dehydrogenative coupling of ethylene to form butadiene via an osmacyclopentane.