The Ir(III) fragment {Ir(PCy(3))(2)(H)(2)}(+) has been used to probe the role of the metal centre in the catalytic dehydrocoupling of H(3)B⋅NMe(2)H (A) to ultimately give dimeric aminoborane [H(2)BNMe(2)](2) (D). Addition of A to [Ir(PCy(3))(2)(H)(2)(H(2))(2)][BAr(F)(4)] (1; Ar(F) = (C(6)H(3)(CF(3))(2)), gives the amine-borane complex [Ir(PCy(3))(2)(H)(2)(H(3)B⋅NMe(2)H)][BAr(F)(4)] (2 a), which slowly dehydrogenates to afford the aminoborane complex [Ir(PCy(3))(2)(H)(2)(H(2)B-NMe(2))][BAr(F)(4)] (3). DFT calculations have been used to probe the mechanism of dehydrogenation and show a pathway featuring sequential BH activation/H(2) loss/NH activation. Addition of D to 1 results in retrodimerisation of D to afford 3. DFT calculations indicate that this involves metal trapping of the monomer-dimer equilibrium, 2 H(2)BNMe(2) ⇌ [H(2)BNMe(2)](2). Ruthenium and rhodium analogues also promote this reaction. Addition of MeCN to 3 affords [Ir(PCy(3))(2)(H)(2)(NCMe)(2)][BAr(F)(4)] (6) liberating H(2)B-NMe(2) (B), which then dimerises to give D. This is shown to be a second-order process. It also allows on- and off-metal coupling processes to be probed. Addition of MeCN to 3 followed by A gives D with no amine-borane intermediates observed. Addition of A to 3 results in the formation of significant amounts of oligomeric H(3)B⋅NMe(2)BH(2)⋅NMe(2)H (C), which ultimately was converted to D. These results indicate that the metal is involved in both the dehydrogenation of A, to give B, and the oligomerisation reaction to afford C. A mechanism is suggested for this latter process. The reactivity of oligomer C with the Ir complexes is also reported. Addition of excess C to 1 promotes its transformation into D, with 3 observed as the final organometallic product, suggesting a B-N bond cleavage mechanism. Complex 6 does not react with C, but in combination with B oligomer C is consumed to eventually give D, suggesting an additional role for free aminoborane in the formation of D from C.

Efficient catalysts for the dehydrocoupling or dehydropolymerisation of H(3)B·NMe(x)H((3-x)) (x = 1, 2) have been developed by variation of the P-Rh-P angle in {Rh(Ph(2)P(CH(2))(n)PPh(2))}(+) fragments (n = 2-5).

Rhodium(III) dihydrido complexes [Rh(L 2 )(H) 2 (acetone)][BAr F4 ] (Ar F =C 6 H 3 (CF 3 ) 2 ) containing the potentially hemilabile ligands L 2 = 9,9-dimethyl-4,5-bis(diphenylphosphino)xanthene (Xantphos) and [Ph 2 P(CH 2 ) 2 ] 2 O (POP 0 ) have been prepared from their corresponding norbornadiene rhodium(I) precursors. In solution these complexes are fluxional by proposed acetone dissociation, which can be trapped out by addition of MeCN to form [Rh(L 2 )(H) 2 (NCMe)][BAr F 4 ], which have been crystallographically characterized. Addition of alkene (methyl acrylate) to these complexes results in reduction to a rhodium(I) species and when followed by addition of the aldehyde HCOCH 2 CH 2 SMe affords the new acyl hydrido complexes [Rh(L 2 )(COCH 2 CH 2 SMe)H][BAr F 4 ] in good yield. The solid-state and solution structures show a tight binding of the POP 0 and Xantphos ligands, having a trans-arrangement of the phosphines with the central ether linkage bound. This is similar to the previously reported complex [Rh(DPEphos)-(COCH 2 CH 2 SMe)H][BAr F 4 ] (DPEphos=[Ph 2 P(C 6 H 4 )] 2 O). Unlike the DPEphos complex, the Xantphos and POP 0 ligated complexes are not effective catalysts for the hydroacylation reaction between methyl acrylate and HCOCH 2 CH 2 SMe. This is traced to their inability to dissociate the central ether link in a hemilabile manner to reveal a vacant site necessary for alkene coordination. Consistent with this lack of availability of the vacant site, these complexes also are stable toward reductive decarbonylation. Complexes [Rh(Ph 2 P(CH 2 ) n PPh 2 )(acetone) 2 ][BAr F 4 ] (n = 2-5) have also been studied as catalysts for the hydroacylation reaction between methyl acrylate and HCOCH 2 CH 2 SMe at 22°C. As found previously, for n=2 this affords the product of alkene hydroacylation, but as the chain length is progressively increased to n=5, the reaction also progressively changes to favor the product of aldehyde hydroacylation. This is suggested to occur by a decrease in the accessibility of the metal site on increasing the bite angle of the chelate ligand, so that alkene coordination to a putative Rh(III)-acyl hydrido intermediate is progressively disfavored and aldehyde coordination (followed by hydride transfer) is progressively favored. These, and previous, results show that the overall conversion in the hydroacylation reaction can be controlled by the hemilabile nature of the chelating phosphine in the catalyst (e.g., DPEphos versus Xantphos), and the course of the reaction can also be tuned by changing the bite angle of the phosphine, cf. Ph 2 P(CH 2 ) 2 PPh 2 and Ph 2 P(CH 2 ) 5 PPh 2 .

[Rh(nbd)(PCyp(3))(2)][BAr(F) (4)] (1) [nbd = norbornadiene, Ar(F) = C(6)H(3)(CF(3))(2), PCyp(3) = tris(cyclopentylphosphine)] spontaneously undergoes dehydrogenation of each PCyp(3) ligand in CH(2)Cl(2) solution to form an equilibrium mixture of cis-[Rh{PCyp(2)(eta(2)-C(5)H(7))}(2)][BAr(F) (4)] (2 a) and trans-[Rh{PCyp(2)(eta(2)-C(5)H(7))}(2)][BAr(F) (4)] (2 b), which have hybrid phosphine-alkene ligands. In this reaction nbd acts as a sequential acceptor of hydrogen to eventually give norbornane. Complex 2 b is distorted in the solid-state away from square planar. DFT calculations have been used to rationalise this distortion. Addition of H(2) to 2 a/b hydrogenates the phosphine-alkene ligand and forms the bisdihydrogen/dihydride complex [Rh(PCyp(3))(2)(H)(2)(eta(2)-H(2))(2)][BAr(F) (4)] (5) which has been identified spectroscopically. Addition of the hydrogen acceptor tert-butylethene (tbe) to 5 eventually regenerates 2 a/b, passing through an intermediate which has undergone dehydrogenation of only one PCyp(3) ligand, which can be trapped by addition of MeCN to form trans-[Rh{PCyp(2)(eta(2)-C(5)H(7))}(PCyp(3))(NCMe)][BAr(F) (4)] (6). Dehydrogenation of a PCyp(3) ligand also occurs on addition of Na[BAr(F) (4)] to [RhCl(nbd)(PCyp(3))] in presence of arene (benzene, fluorobenzene) to give [Rh(eta(6)-C(6)H(5)X){PCyp(2)(eta(2)-C(5)H(7))}][BAr(F) (4)] (7: X = F, 8: X = H). The related complex [Rh(nbd){PCyp(2)(eta(2)-C(5)H(7))}][BAr(F) (4)] 9 is also reported. Rapid ( approximately 5 minutes) acceptorless dehydrogenation occurs on treatment of [RhCl(dppe)(PCyp(3))] with Na[BAr(F) (4)] to give [Rh(dppe){PCyp(2)(eta(2)-C(5)H(7))}][BAr(F) (4)] (10), which reacts with H(2) to afford the dihydride/dihydrogen complex [Rh(dppe)(PCyp(3))(H)(2)(eta(2)-H(2))][BAr(F) (4)] (11). Competition experiments using the new mixed alkyl phosphine ligand PCy(2)(Cyp) show that [RhCl(nbd){PCy(2)(Cyp)}] undergoes dehydrogenation exclusively at the cyclopentyl group to give [Rh(eta(6)-C(6)H(5)X){PCy(2)(eta(2)-C(5)H(7))}][BAr(F) (4)] (17: X = F, 18: X = H). The underlying reasons behind this preference have been probed using DFT calculations. All the complexes have been characterised by multinuclear NMR spectroscopy, and for 2 a/b, 4, 6, 7, 8, 9 and 17 also by single crystal X-ray diffraction.