and Iridium(III) Complexes Supported by a Diphenolate Imidazolyl

Organometallics 2010, 29, 89–100 DOI: 10.1021/om900803r

89

Iridium(I) and Iridium(III) Complexes Supported by a Diphenolate Imidazolyl-Carbene Ligand David R. Weinberg, Nilay Hazari,† Jay A. Labinger,* and John E. Bercaw* Arnold and Mabel Beckman Laboratories of Chemical Synthesis, California Institute of Technology, Pasadena, California 91125. † Current address: Chemistry Department, Yale University, P.O. Box 208107, New Haven, CT 06520. Received September 14, 2009

Deprotonation of 1,3-di(2-hydroxy-5-tert-butylphenyl)imidazolium chloride (1a) followed by reaction with chloro-1,5-cyclooctadiene Ir(I) dimer affords the anionic Ir(I) complex [K][{OCO}Ir(cod)] (2: OCO = 1,3-di(2-hydroxy-5-tert-butylphenyl)imidazolyl; cod = 1,5-cyclooctadiene), the first Ir complex stabilized by a diphenolate imidazolyl-carbene ligand. In the solid state 2 exhibits squareplanar geometry, with only one of the phenoxides bound to the metal center. Oxidation of 2 with 2 equiv of [FeCp2][PF6] generates the Ir(III) complex [{OCO}Ir(cod)(MeCN)][PF6] (3). Reaction of 3 with H2 results in the liberation of cyclooctane and a species capable of catalyzing the hydrogenation of cyclohexene to cyclohexane. Displacement of cyclooctadiene from 3 can be achieved by heating in acetonitrile to form [{OCO}Ir(MeCN)3][PF6] (4) or by reaction with either PMe3 or PCy3 to generate [{OCO}Ir(PMe3)3][PF6] (5) or [{OCO}Ir(PCy3)2(MeCN)][PF6] (6), respectively. 6 reacts with CO in acetonitrile to give an equilibrium mixture of 6 and [{OCO}Ir(PCy3)2(CO)][PF6] (7) and with chloride to generate [{OCO}Ir(PCy3)2Cl] (8). The solid-state structure of 8 shows that the diphenolate imidazolylcarbene ligand is distorted from planarity; DFT calculations suggest this is due to an antibonding interaction between the phenolates and the metal center in the highest occupied molecular orbital (HOMO) of the complex. 8 undergoes two successive reversible one-electron oxidations in CH2Cl2 at -0.22 and at 0.58 V (vs ferrocene/ferrocenium); EPR spectra, mass spectroscopy, and DFT calculations suggest that the product of the first oxidation is [{OCO}Ir(PCy3)2Cl]þ (8þ), with the unpaired electron occupying a molecular orbital that is delocalized over both the metal center and the diphenolate imidazolyl-carbene ligand.

Introduction Both Ir(I) and Ir(III) complexes containing N-heterocyclic carbene (NHC) ligands have proven capable of catalyzing a variety of transformations, including hydrogenations,1-5 *Corresponding author. E-mail: [email protected]. (1) Chang, Y.-H.; Fu, C.-F.; Liu, Y.-H.; Peng, S.-M.; Chen, J.-T.; Liu, S.-T. Dalton Trans. 2009, 861–867. (2) Zhu, Y.; Burgess, K. Adv. Synth. Catal. 2008, 350, 979–983. (3) Baskakov, D.; Herrmann, W. A.; Herdtweck, E.; Hoffmann, S. D. Organometallics 2007, 26, 626–632. (4) Vazquez-Serrano, L. D.; Owens, B. T.; Buriak, J. M. Inorg. Chim. Acta 2006, 359, 2786–2797. (5) Nanchen, S.; Pfaltz, A. Helv. Chim. Acta 2006, 1559–1573. (6) Chen, T.; Liu, X.-G.; Shi, M. Tetrahedron 2007, 63, 4874–4880. (7) Chianese, A. R.; Mo, A.; Datta, D. Organometallics 2009, 28, 465– 472. (8) Chianese, A. R.; Crabtree, R. H. Organometallics 2005, 24, 4432– 4436. (9) Bauer, E. B.; Andavan, G. T. S.; Hollis, T. K.; Rubio, R. J.; Cho, J.; Kuchenbeiser, G. R.; Helgert, T. R.; Letko, C. S.; Tham, F. S. Org. Lett. 2008, 10, 1175–1178. (10) Corber an, R.; Lillo, V.; Mata, J. A.; Fernandez, E.; Peris, E. Organometallics 2007, 26, 4350–4353. (11) Frey, G. D.; Rentzsch, C. F.; von Preysing, D.; Scherg, T.; M€ uhlhofer, M.; Herdtweck, E.; Herrmann, W. A. J. Organomet. Chem. 2006, 691, 5725–5738. (12) Grasa, G. A.; Moore, Z.; Martin, K. L.; Stevens, E. D.; Nolan, S. P.; Paquet, V.; Lebel, H. J. Organomet. Chem. 2002, 658, 126–131. r 2009 American Chemical Society

hydrosilylations,6-8 hydroaminations,9 borations,10-12 alkylations,1,13 cycloadditions,14 cyclopropanations,15 hydrogen isotope exchanges,16 and polymerizations.17,18 The strong σ-donation of NHC ligands often imparts stability, whereas their large trans effects tend to increase the rates at which other ligands in the complex exchange.19 Increased rates of ligand exchange may be important for catalysis because Ir(III) complexes are often relatively substitutionally inert.20 NHC ligands can be readily synthesized with a wide variety of different substituents on the nitrogen atoms, (13) Gnanamgari, D.; Sauer, E. L. O.; Schley, N. D.; Butler, C.; Incarvito, C. D.; Crabtree, R. H. Organometallics 2009, 28, 321–325. (14) Peng, H. M.; Webster, R. D.; Li, X. Organometallics 2008, 27, 4484–4493. (15) Winkelmann, O.; N€ather, C.; L€ uning, U. J. Organomet. Chem. 2008, 693, 923–932. (16) Brown, J. A.; Irvine, S.; Kennedy, A. R.; Kerr, W. J. Chem. Commun. 2008, 1115–1117. (17) Jin, G.-X.; Xiao, X.-Q. J. Organomet. Chem. 2008, 693, 3363– 3368. (18) Zhang, Y.; Wang, D.; Wurst, K.; Buchmeiser, M. R. J. Organomet. Chem. 2005, 690, 5728–5735. (19) Lee, M.-T.; Hu, C.-H. Organometallics 2004, 23, 976–983. (20) Cotton, F. A.; Wilkinson, G.; Murillo, C. A.; Bochmann, M. Advanced Inorganic Chemistry, 6th ed.; Wiley & Sons: New York, 1999; pp 1039-1063. Published on Web 12/07/2009

pubs.acs.org/Organometallics

90

Organometallics, Vol. 29, No. 1, 2010

allowing modulation of reactivity. Chelating carbene ligands have been utilized in some Ir catalysts to further modify reactivity and increase stability.3,5,7-10,12 We have chosen to examine diphenolate imidazolylcarbene ligands (1), which have recently been shown to support a variety of transition metal complexes and catalysts,21-28 although none involve group 9 transition metals. We envisioned that the combination of the “soft” carbene and the “hard” phenolate donors could stabilize Ir in more than one oxidation state;29 furthermore, the redox chemistry could be particularly interesting, as phenolate ligands may be oxidized to generate stable ligand-centered radicals.30,31

Results and Discussion Synthesis and Characterization of an Ir(I) Complex. The NHC precursor 1,3-bis(2-hydroxy-5-tert-butylphenyl)imidazolium chloride (1a) was prepared by a procedure similar to those used for analogous species.32 It was then deprotonated with 3 equiv of potassium hexamethyldisilazide and treated with [IrCl(cod)]2 to cleanly generate [K][{OCO}Ir(cod)] (2: OCO = 1,3-di(2-hydroxy-5-tert-butylphenyl)imidazolyl; cod = 1,5-cyclooctadiene; eq 1). 2 was crystallized from benzene as the 18-crown-6 etherate, and the structure determined crystallographically (Figure 1). The geometry about the Ir(I) center is square planar: the NHC ligand is κ2(C,O)coordinated with one “dangling” phenoxide O interacting

(21) Bellemin-Laponnaz, S.; Welter, R.; Brelot, L.; Dagorne, S. J. Organomet. Chem. 2009, 694, 604–606. (22) Zhang, D.; Liu, N. Organometallics 2009, 28, 499–505. (23) Manna, C. M.; Shavit, M.; Tshuva, E. Y. J. Organomet. Chem. 2008, 693, 3947–3950. (24) Zhang, D. Eur. J. Inorg. Chem. 2007, 4839–4845. (25) Zhang, D.; Aihara, H.; Watanabe, T.; Matsuo, T.; Kawaguchi, H. J. Organomet. Chem. 2007, 692, 234–242. (26) Xu, X.; Yao, Y.; Zhang, Y.; Shen, Q. Inorg. Chem. 2007, 46, 3743–3751. (27) Yagyu, T.; Oya, S.; Maeda, M.; Jitsukawa, K. Chem. Lett. 2006, 35, 154–155. (28) Aihara, H.; Matsuo, T.; Kawaguchi, H. Chem. Commun. 2003, 2204–2205. (29) Liddle, S. T.; Edworthy, I. S.; Arnold, P. L. Chem. Soc. Rev. 2007, 36, 1732–1744. (30) Kurahashi, T.; Kikuchi, A.; Tosha, T.; Shiro, Y.; Kitagawa, T.; Fujii, H. Inorg. Chem. 2008, 47, 1674–1686. (31) Mukherjee, A.; Lloret, F.; Mukherjee, R. Inorg. Chem. 2008, 47, 4471–4480. (32) Bellemin-Laponnaz, S.; Welter, R.; Brelot, L.; Dagorne, S. J. Organomet. Chem. 2009, 694, 604–606. (33) Rodman, G. S.; Mann, K. R. Inorg. Chem. 1988, 27, 3338– 3346. (34) The 13C NMR assignments of the imidazolyl -CH2 peak and the cod -CH peak were made from a 1H-13C HMQC spectrum of 2. (35) Karabiyik, H.; Kilinc-arslan, R.; Ayg€ un, M.; C-etinkaya, B.; B€ uy€ ukg€ ung€ or, O. Z. Naturforsch., B: Chem. Sci. 2005, 60, 837–842.

Weinberg et al.

with the crown ether-encapsulated Kþ ion, and there is normal (η2)2-cod coordination.

The 1H NMR spectrum for 2 in THF-d8 (without 18-crown-6 ether) is shown in Figure 2; the spectra in CD3CN and DMSO-d6, as well as those in THF-d8 and in C6D6 in the presence of 18-crown-6 ether, are all very similar. There is clearly fluxional behavior: from the solid-state structure we would expect two sets of phenolate aromatic protons and tert-butyl groups, at least (depending on the rate of rotation about the dangling phenolate N-C bond) two environments each for the cod -CH and imidazolyl -CH2 groups, and at least four environments for the cod -CH2 protons;33 however, there is only one set of peaks for the phenolates and the cod -CH protons and two signals each for both the cod and imidazolyl -CH2 groups. The 13C NMR spectrum exhibits only one set of peaks for the -CH2 groups in the imidazolyl ring, the aromatic carbons, and the tert-butyl groups.34 1 H NMR spectra were obtained at various temperatures between -70 and 60 C. All the peaks broaden significantly below -40 C; between -40 and -70 C, the two imidazolyl -CH2 peaks coalesce, and the cyclooctadiene -CH peak splits. The latter observation suggests that exchange between the coordinated and dangling phenolates (as depicted in eq 1) is freezing out; in that interpretation, the two peaks observed at room temperature in the 1H (but not the 13C) NMR spectrum for the imidazolyl -CH2 groups would be ascribed to slow rotation about the nitrogen-phenolate bonds.35 That would also imply further nonequivalence of the cod protons, but all those signals are relatively broad; any additional splitting may be thus masked. There is no obvious reason why the two imidazolyl -CH2 peaks should converge at low temperature; possibly this is an accidental degeneracy associated with differential temperature dependence of the chemical shifts. Synthesis and Chemistry of Ir(III) Complexes. Oxidation of 2 is effected by ferrocenium in acetonitrile; the resulting 1H NMR spectrum, showing one set of phenolate signals, two for the imidazolyl -CH2 groups, and six peaks for coordinated cod (the latter are shifted significantly downfield from their positions in 2), is consistent with the octahedral Ir(III) complex[{mer-OCO}Ir(cod)(MeCN)][PF6] (3, Scheme 1). A molecular ion corresponding to 3 was observed in the mass spectrum. Simply heating 3 at 90 C in acetonitrile results in the displacement of cod (observed by 1H NMR spectroscopy when the reaction is performed in CD3CN) to generate [{mer-OCO}Ir(MeCN)3][PF6] (4, Scheme 1), confirmed by mass spectrometry. When 4 is generated in CH3CN, isolated,

Article


91

Figure 1. Structural drawing of 2 with 18-crown-6 ether coordinated to potassium. Displacement ellipsoids are drawn at the 50% probability level. A molecule of benzene in the unit cell is not shown. (a) Perspective taken perpendicular to the square plane of the complex. (b) Perspective taken facing down the C15-Ir bond; the 18-crown-6 ether was removed for clarity. Selected bond lengths (A˚) and angles (deg): Ir-C15 2.041(3); Ir-O1 2.0423(19); Ir-C(24) 2.093(3); Ir-C25 2.145(3); Ir-C28 2.159(3); Ir-C29 2.174(3); N1-C15 1.369(4); N2-C15 1.348(4); O2-K 2.576(2); C15-Ir-O1 87.56(10); C15-Ir-C24 93.31(12); O1-Ir-C24 153.01(11); C15-Ir-C25 100.71(10); C15-Ir-C25 39.06(12); C15-Ir-O1 87.56(10); C15-Ir-C24 93.31(12); O1-Ir-C24 153.01(11); C15-Ir-C25 100.71(12); O1-Ir-C25 165.77(10); C24-Ir-C25 39.06(12); C15-Ir-C28 164.27(12); O1-Ir-C28 88.10(10); C24-Ir-C28 97.40(12); C25-Ir-C28 80.84(11); C15-Ir-C29 156.85(12); O1-Ir-C29 86.99(10); C24-Ir-C29 81.68(12); C25-Ir-C29 89.62(12); C28-Ir-C29 37.61(12).

Figure 2. 1H NMR spectrum of 2 in THF-d8 at 298 K. Note some benzene was isolated with 2, and it is present in solution. Also, the “-CH3” peak is truncated, as it has a much greater intensity than the other peaks.

and then redissolved in CD3CN, two peaks for Ir-coordinated CH3CN appear in the 1H NMR spectrum, at 2.53 and 2.68 ppm with an approximate ratio of 2:1. 4 was kept in CD3CN at room temperature for two weeks; during this time the smaller peak, probably corresponding to CH3CN coordinated trans to the carbene donor, disappeared completely from the 1H NMR spectrum. The larger peak also

decreased in size, but only slightly, while the free CH3CN peak grew. The peaks at 2.53 and 2.68 ppm are not present when 4 is generated in CD3CN. The displacement of cod by acetonitrile indicates relatively weak binding to Ir(III), not unexpected for a high-valent olefin complex; the facile exchange of the acetonitrile trans to the carbene is more noteworthy for Ir(III).

92


Weinberg et al. Scheme 1

Treatment of 3 with 1 atm of H2 in THF at room temperature results in the formation of cyclooctane; the 1H NMR spectrum showed only trace amounts of cyclooctene, no Ir-hydride signals, and broad peaks attributable to the diphenolate carbene ligand. The Ir-containing product can catalyze the hydrogenation of cyclohexene to cyclohexane. Removing the solvent and other volatiles under reduced pressure and heating the resulting solid in CD3CN at 90 C affords 4, suggesting that the diphenolate imidazolylcarbene ligand remains bound to Ir, but the identity of the catalyst is otherwise unclear. Although no Ir hydride signals were observed, they may be present in small amounts; most Ir-catalyzed hydrogenations involve either Ir(I) or hydridoiridium(III) as catalyst precursors.36-43 Reaction of 3 with an excess (greater than 3 equiv) of PMe3 or PCy3 proceeds slowly at room temperature, but goes to completion after 12 h at 90 C, via displacement of cod to give [{OCO}Ir(PMe3)3][PF6] (5) and [{OCO}Ir(PCy3)2(MeCN)][PF6] (6), respectively (Scheme 2). The structure of 5 was assigned on the basis of NMR data, including the observation of doublet and triplet 31P signals in 2:1 intensity ratio, along with HRMS. In contrast, the 31P NMR spectrum of 6 shows only one singlet for coordinated PCy3, and only one set of phenolate peaks is observed in the 1H NMR spectrum, indicating the mer, trans configuration shown. Steric factors likely cause the PCy3 ligands to bind trans to each other and prevent a third PCy3 ligand from binding to the metal center. Complex 6 reacts reversibly with CO in acetonitrile; an equilibrium mixture of 6 and [{OCO}Ir(PCy3)2(CO)][PF6] (7) is established over 3 days at 90 C under 1.3 atm CO, with the ratio of 6 to 7 approximately 2:1 (eq 2). 7 is converted back to 6 when the solution is degassed and reheated to 90 C. 7 exhibits an infrared CO-stretching signal at 2064 cm-1, which shifts to 2016 cm-1 (2018 cm-1 calculated) when 13CO is used; 7-13CO also displays the expected triplet in the 13C NMR spectrum and doublet in the 31P NMR spectrum. The relatively high νCO, while within the range of (36) Vazquez-Serrano, L. D.; Owens, B. T.; Buriak, J. M. Inorg. Chim. Acta 2006, 359, 2786–2797. (37) Nagaraja, C. M.; Nethaji, M.; Jagirdar, B. R. Organometallics 2007, 26, 6307–6311. (38) Roseblade, S. J.; Pfaltz, A. Acc. Chem. Res. 2007, 40, 1402–1411. (39) Crabtree, R. H. The Organometallic Chemistry of the Transition Metals, 3rd ed.; Wiley & Sons: New York, 2001; pp 80-114, 222-258. (40) Lee, D.-H.; Patel, B. P.; Clot, E.; Eisenstein, O.; Crabtree, R. H. Chem. Commun. 1999, 297–298. (41) Ogo, S.; Kabe, R.; Hayashi, H.; Harada, R.; Fukuzumi, S. Dalton Trans. 2006, 4657–4663. (42) Sablong, R.; Osborn, J. A. Tetrahedron: Asymmetry 1996, 7, 3059–3062. (43) Chan, Y. N. C.; Meyer, D.; Osborn, J. A. J. Chem. Soc., Chem. Commun. 1990, 869–871.

Scheme 2

other cationic diphosphine Ir(III) carbonyl complexes,44-46 indicates relatively weak π-donation from Ir to the carbonyl π*-orbital, consistent with the fact that 7 and 6 equilibrate.

Determination of a precise equilibrium constant is precluded by the formation of yellow crystals above the surface of the solution during the reaction; surprisingly, these result from cocrystallization of 6 and 7 as a 4:1 mixture (Figure 3). The two complexes have indistinguishable structural parameters aside from the identity of the sixth ligand. While this type of substitutional disorder appears to be rare, similar (44) O’Connor, J. M.; Pu, L.; Rheingold, A. L. J. Am. Chem. Soc. 1990, 112, 6232–6247. (45) Clark, H. C.; Reimer, K. J. Inorg. Chem. 1975, 14, 2133–2140. (46) Thorn, D. L.; Fultz, W. C. J. Phys. Chem. 1989, 93, 1234–1243. (47) Rotar, A.; Varga, R. A.; Silvestru, C. Acta Crystallogr. 2008, E64, m45.

Article


93

Figure 3. Three perspectives of the structure of 6/7, with displacement ellipsoids at the 50% probability level; [PF6]- is not shown. In the unit cell, there is a 78% occupancy of positions N3A, C60A, and C61A of 6 and a 22% occupancy of positions C3B and O3B of 7. The Cipso-O1-O2-Cipso torsion angle is 2.9. Selected bond lengths (A˚) and angles (deg): Ir1-C1 1.9694(19); Ir1-O2 2.0331(14); Ir1-O1 2.0390(13); Ir1-N3A 2.055(8); Ir1-C3B 2.14(4); C3B-O3B 0.95(4); Ir1-P2 2.4348(5); Ir1-P1 2.4437(5); C1-Ir1-O2 91.57(8); C1-Ir1-O1 92.21(7); O2-Ir1-O1 176.10(5); C1-Ir1-N3A 178.8(2); O2-Ir1-N3A 87.8(2); O1-Ir1-N3A 88.5(2); C1-Ir1-C3B 174.1(10); O2-Ir1-C3B 94.4(10); O1-Ir1-C3B 81.9(10); N3A-Ir1-C3B 6.7(12); C1-Ir1-P2 92.79(6); O2-Ir1-P2 91.00(5); O1-Ir1-P2 89.77(4); N3A-Ir1-P2 86.2(2); C3B-Ir1-P2 87.5(11); C1-Ir1-P1 94.97(6); O2-Ir1-P1 89.17(5); O1-Ir1-P1 89.56(4); N3A-Ir1-P1 86.0(2); C3B-Ir1-P1 84.8(11); P2-Ir1-P1 172.229(16); Ir1-C3B-O3B 178(4); Ir1-N3A-C60A 175.1(7); N3A-C60A-C61A 177.1(7); Cipso-O1-O2-Cipso 2.71(22).

examples do exist.47-51 The diphenolate carbene ligand in 6/7 is relatively planar; the Cipso-O1-O2-Cipso torsion angle, a useful measure of the degree of distortion, is only 2.9. Displacement of acetonitrile from 6 by chloride generates air-stable [{OCO}Ir(PCy3)2Cl] (8, eq 3), confirmed by singlecrystal X-ray diffraction (Figure 4). Again, the solid-state structure is not straightforward: the unit cell contains two conformers of 8 (80 and 800 ), along with two molecules of CH2Cl2. In both conformers the metal, chloride, two oxygens, and the carbene carbon are nearly coplanar, but the rings of the diphenolate imidazolyl ligand are significantly distorted out of this plane, from C2v-symmetry toward C2-symmetry, to an extent that differs for the two molecules. The Cipso-O1-O2-Cipso torsion angle is 15.72(26) for (48) Norman, N. C.; Orpen, A. G.; Quayle, M. J.; Robins, E. G. Acta Crystallogr. 2000, C56, 50–52. (49) Laungani, A. C.; Keller, M.; Breit, B. Acta Crystallogr. 2008, E64, m24–m25. (50) Marimuthu, T.; Bala, M. D.; Friedrich, H. B. Acta Crystallogr. 2008, E64, 6772. (51) de Silva, N.; Nichiporuk, R. V.; Dahl, L. F. Dalton Trans. 2006, 2291–2300.

80 and 35.19(26) for 800 , a considerably greater distortion for the latter.

The chloro-Ir(III) complex 8 can be further oxidized; the cyclic voltammogram in CH2Cl2 (Figure 5; similar results are obtained in THF and dimethylformamide) displays two reversible oxidations (ΔEp =98 mV for both), at -0.22 and 0.58 V. Coulometric oxidation at 0.25 V resulted in the passage of 0.92 faraday per mole, indicating that the wave at -0.22 V corresponds to a one-electron oxidation of 8; further coulometric oxidation at 0.96 V resulted in the passage of 0.99 faraday per mole, indicating that the wave at 0.58 V also corresponds to a one-electron oxidation.

94


Weinberg et al.

Figure 4. Structural drawing of the two conformers in the unit cell of 8 with displacement ellipsoids drawn at the 50% probability level; two molecules of CH2Cl2, not shown, are also in the unit cell. (a) Conformer 80 ; the perspective on the left is directed approximately bisecting the P2A-Ir1-O1A angle; that on the right is directed down the C15A-Ir1 bond. The Cipso-O1-O2-Cipso torsion angle is 15.72(26). Selected bond lengths (A˚) and angles (deg): Ir(1)-C(15A) 1.944(2); Ir(1)-O(2A) 2.0257(17); Ir(1)-O(1A) 2.0343(16); Ir(1)-P(2A) 2.4194(7); Ir(1)-P(1A) 2.4365(7); Ir(1)-Cl(1) 2.4584(5); C(15A); Ir(1)-O(2A) 91.66(8); C(15A)-Ir(1)-O(1A) 93.06(8); O(2A)-Ir(1)-O(1A) 175.24(6); C(15A)-Ir(1)-P(2A) 94.16(7); O(2A)-Ir(1)-P(2A) 90.58(5); O(1A)-Ir(1)-P(2A) 88.45(5); C(15A)-Ir(1)-P(1A) 93.67(7); O(2A)-Ir(1)-P(1A) 92.81(5); O(1A)-Ir(1)-P(1A) 87.51(5); P(2A)-Ir(1)-P(1A) 171.367(19); C(15A)-Ir(1)-Cl(1) 179.44(8); O(2A)-Ir(1)-Cl(1) 87.99(4); O(1A)-Ir(1)-Cl(1) 87.29(4); P(2A)-Ir(1)-Cl(1) 85.40(2); P(1A)-Ir(1)-Cl(1) 86.79(2). (b) Conformer 800 with approximately the same perspectives. The Cipso-O1-O2-Cipso torsion angle is 35.19(26). Selected bond lengths (A˚) and angles (deg): Ir(2)-C(15B) 1.944(2); Ir(2)-O(1B) 2.0307(17); Ir(2)-O(2B) 2.0380(18); Ir(2)-P(1B) 2.4102(8); Ir(2)-P(2B) 2.4260(8); Ir(2)-Cl(2) 2.4529(5); C(15B)-Ir(2)-O(1B) 92.75(9); C(15B)-Ir(2)-O(2B) 90.14(9); O(1B)-Ir(2)-O(2B) 177.00(6); C(15B)-Ir(2)-P(1B) 93.81(8); O(1B)-Ir(2)-P(1B) 92.32(5); O(2B)-Ir(2)-P(1B) 86.70(5); C(15B)-Ir(2)-P(2B) 94.60(8); O(1B)-Ir(2)-P(2B) 88.27(5); O(2B)-Ir(2)-P(2B) 92.29(6); P(1B)-Ir(2)-P(2B) 171.53(2); C(15B)-Ir(2)-Cl(2) 177.33(8); O(1B)-Ir(2)-Cl(2) 85.42(4); O(2B)-Ir(2)-Cl(2) 91.66(4); P(1B)-Ir(2)-Cl(2) 84.32(2); P(2B)-Ir(2)-Cl(2) 87.30(2).

EPR spectra were recorded for the solutions generated by the above controlled potential electrolysis of 8. The EPR spectrum obtained after the first oxidation of 8 (Figure 6) is indicative of an S=1/2 species with g||=2.19 and g^ ≈ 1.94; even at temperatures as low as 7 K, no hyperfine splitting was resolved. Mass spectroscopy of this sample showed the expected molecular ion for 8. The same EPR spectrum was obtained from oxidation of 8 with less than 1 equiv of [FeCp2][PF6]. Characterization of products generated in the second oxidation was complicated by product instability; a mixture of species was suggested by both EPR and NMR spectroscopies, while a molecular ion for 8 was still detected in the mass spectrum.

The EPR signal in Figure 6 most probably corresponds to [{OCO}Ir(PCy3)2(Cl)]þ (8þ), generated by chemical or (reversible) electrochemical one-electron oxidation of 8. The electronic configuration of radical-cation 8þ is not obvious: oxidation of Ir(III) complexes to Ir(IV) and oxidation (52) Benisvy, L.; Bill, E.; Blake, A. J.; Collison, D.; Davies, E. S.; Garner, C. D.; McArdle, G.; McInnes, E. J. L.; McMaster, J.; Ross, S. H. K.; Wilson, C. Dalton Trans. 2006, 258–267. (53) Sokolowski, A.; M€ uller, J.; Weiherm€ uller, T.; Schnepf, R.; Hildebrandt, P.; Hildenbrand, K.; Bothe, E.; Wieghardt, K. J. Am. Chem. Soc. 1997, 119, 8889–8900. (54) Diversi, P.; de Biani, F. F.; Igrosso, G.; Laschi, F.; Lucherini, A.; Pinzino, C.; Zanello, P. J. Organomet. Chem. 1999, 584, 73–86.

Article


95

Figure 5. Cyclic voltammogram of a 3 mM CH2Cl2 solution of 8 at a scan rate of 100 mV/s with 0.3 M [NBu4][BF4] as the supporting electrolyte.

Figure 6. EPR spectrum at 7 K of the solution resulting from controlled potential oxidation of 8 at 0.25 V (0.92 faraday per mole).

of phenolates to phenoxyl radicals are both well precedented,52-63 and while both typically occur at potentials greater than that for the oxidation of 8 (-0.22 V), the latter value does fall within the range observed for both types of oxidation. The EPR spectrum might be expected to exhibit hyperfine coupling to Ir (191Ir, 37%; 193Ir, 63%: both I=3/2) and P for Ir(IV), or to (55) Panda, M.; Das, C.; Lee, G.-H.; Peng, S.-M.; Goswami, S. Dalton Trans. 2004, 2655–2661. (56) Acharyya, R.; Basuli, F.; Peng, S.-M.; Lee, G.-H.; Wang, R.-Z.; Mak, T. C. W.; Bhattacharya, S. J. Organomet. Chem. 2005, 690, 3908– 3917. (57) Li, X.; Chen, Z.; Zhao, Q.; Shen, L.; Li, F.; Yi, T.; Cao, Y.; Huang, C. Inorg. Chem. 2007, 46, 5518–5527. (58) Wu, F.-I.; Su, H.-J.; Shu, C.-F.; Luo, L.; Diau, W.-G.; Cheng, C.-H.; Duan, J.-P.; Lee, G.-H. J. Mater. Chem. 2005, 15, 1035–1042. (59) Hoganson, C. W.; Babcock, G. T. Biochemistry 1992, 31, 11874– 11880. (60) Halfen, J. A.; Jazdzewski, B. A.; Mahapatra, S.; Berreau, L. M.; Wilkinson, E. C.; Que, L., Jr.; Tolman, W. B. J. Am. Chem. Soc. 1997, 119, 8217–8227. (61) Chaudhuri, P.; Hess, M.; M€ uller, J.; Hildenbrand, K.; Bill, E.; Weyherm€ uller, T.; Wieghardt, K. J. Am. Chem. Soc. 1999, 121, 9599– 9610. (62) Hay-Motherwell, R. S.; Wilkinson, G.; Hussain-Bates, B.; Hursthouse, M. B. J. Chem. Soc., Dalton Trans. 1992, 3477–3482. (63) Larsen, S. K.; Pierpont, C. G. J. Am. Chem. Soc. 1988, 110, 1827– 1832. (64) Drago, R. S. Physical Methods in Chemistry; W. B. Saunders: Philadelphia, PA, 1977; pp 467-509.

H and N for a phenoxyl radical;46,64 there is precedent, though, for Ir(IV) complexes for which hyperfine splitting cannot be resolved even at low temperatures, and highly delocalized singly occupied orbitals can also fail to show coupling.56 The g-values differ significantly from the free electron value of 2.0023 and from values for previously characterized phenoxyl radicals,46,53,58-66 suggesting either a metal-centered radical or a delocalized orbital with metal character. The reversibility of the second one-electron oxidation of 8 suggests that [{OCO}Ir(PCy3)2Cl]2þ (82þ) is the initial product; either an Ir(V) species or an Ir(IV) species with a ligandcentered radical could be consistent with oxidation potentials observed previously for Ir(IV) complexes and phenolates.52,53,59,60,62,63,67 Unfortunately, the appearance of multiple signals in both the EPR and NMR spectra indicates that 82þ is unstable, precluding characterization or a detailed electronic description of 82þ. (65) Hulsebosch, R. J.; van den Brink, J. S.; Nieuwenhuis, S. A. M.; Gast, P.; Raap, J.; Lugtenburg, J.; Hoff, A. J. J. Am. Chem. Soc. 1997, 119, 8685–8694. (66) Kim, S. H.; Aznar, C.; Brynda, M.; Silks, L. A. “P.”; Michalczyk, R.; Unkefer, C. J.; Woodruff, W. H.; Britt, R. D. J. Am. Chem. Soc. 2004, 126, 2328–2338. (67) Castillo-Blum, S. E.; Richens, D. T.; Sykes, A. G. Inorg. Chem. 1989, 28, 954–960.

96


Weinberg et al.

Figure 9. (a) Depiction of the HOMO for the fully optimized gas-phase structure of 9. (b) Depiction of the SOMO for the fully optimized gas-phase structure of 9þ. Figure 7. Illustration (a) and fully optimized gas-phase structural drawings (b) of 9.

Figure 8. Illustration (a) and fully optimized gas-phase structural drawings (b) of 9þ.

Calculational Studies and OCO Geometries. DFT calculations were performed on [{OCO0 }Ir(PH3)2Cl] (9) and [{OCO0 }Ir(PH3)2Cl]þ (9þ), where OCO0=1,3-di(2-hydroxyphenyl)imidazolyl, as simpler (and less sterically crowded) models for 8 and 8þ. The optimized structures of 9 and 9þ are shown in Figures 7 and 8; the parameters of 9 agree well with those determined experimentally for 8 (see the Supporting Information for a side-by-side comparison of the bond lengths and angles in 80 , 800 , 9, and 9þ), with the exception of the deviation from planarity of the OCO ligands (see below). The HOMO of 9 and the SOMO of 9þ are fairly similar (Figure 9); both have considerable metal character but are substantially delocalized onto the OCO0 ligand, with significant π-antibonding interactions between iridium and the phenolate oxygens as well as within the ligand π-orbitals. The degree of metal character and the highly delocalized nature of the calculated SOMO appear consistent with the experimental g-value as well as the lack of hyperfine splitting in the low-temperature EPR spectrum of 8þ.

An intriguing structural feature of these complexes is the dependence of the geometry of the OCO ligand on electronic and steric properties, as has been observed for related ligands with other metals.68-71 As noted earlier, the ligand is virtually planar (C2v symmetry) for the mixed carbonylacetonitrile complexes in the structure of 6/7, but is significantly distorted in chloro complex 8. Analogous differences can be readily seen in the side-on views of the calculated structures for 9 and 9þ (Figures 7 and 8): the ligand is distorted from planarity to C2 symmetry in the former but has C2v symmetry in the latter. This is reflected in the calculated Cipso-O1-O2-Cipso torsion angles, which are 0.3 in 9þ, similar to that found experimentally for 6/7, and 62.62 in 9, significantly greater than the two (significantly different between themselves!) experimentally observed values in 80 (15.72(26)) and 800 (35.19(26)). In order to assess the degree to which the difference between the calculated and experimental values might be due to the substantially greater steric bulk of the real molecule compared to the model, we calculated a geometry for complex 8, with the actual ligands, that was optimized using QMMM. The molecular orbitals and in particular the HOMO of 8 (shown in the Supporting Information) are virtually identical to those of the model compound 9. However, the calculated structure of 8 (Figure 10) has a much smaller Cipso-O1-O2-Cipso torsion angle of 9.55, closer to the experimental value for 80 , suggesting that the steric bulk of the PCy3 ligands in 8 does play a significant role. One might expect the nonplanar C2-geometry to be electronically favored in 8 and 9 because of the π-antibonding interaction between the phenolate ligands and the metal center in the HOMO: increasing the Cipso-O1-O2-Cipso torsion angle would decrease the overlap and reduce this unfavorable contribution. This argument is consistent with the C2v conformation found by calculation for 9þ, where the orbital containing the unfavorable interaction is only singly occupied, so that its energetic cost may be outweighed by increased bonding interactions in lower energy orbitals in a planar geometry. While we have not carried out any calculations on 6/7, the fact that they are cationic (and contain an (68) Tonks, I. A.; Henling, L. M.; Day, M. W.; Bercaw, J. E. Inorg. Chem. 2009, 48, 5096–5105. (69) Agapie, T.; Henling, L. M.; DiPasquale, A. G.; Rheingold, A. L.; Bercaw, J. E. Organometallics 2008, 27, 6245–6256. (70) Agapie, T.; Day, M. W.; Bercaw, J. E. Organometallics 2008, 27, 6123–6142. (71) Agapie, T.; Bercaw, J. E. Organometallics 2007, 26, 2957–2959.

Article


97

Figure 10. Optimized gas-phase structure of 8 determined by QMMM calculations. The Cipso-O1-O2-Cipso torsion angle is 9.55. Hydrogen atoms have been removed for clarity.

Figure 11. Relative energies for optimized structures of 9 when the Cipso-O1-O2-Cipso torsion angle is set at various angles between 0 and 90.

Figure 12. Relative energies for optimized structures of 9þ when the Cipso-O1-O2-Cipso torsion angle is set at various angles between 0 and 35.

electron-withdrawing CO ligand in 7) might similarly attenuate the antibonding contribution and favor planarity. The drastic difference in the calculated Cipso-O1-O2-Cipso torsion angles for 8 and 9 is likely due to the steric bulk of the PCy3 ligands, which favors the sterically less demanding planar structure relative to the electronically preferred C2-symmetry. To assess the overall effect of the Cipso-O1-O2-Cipso torsion angle on the energy of the molecule, we performed linear transit calculations on 9 and 9þ with the Cipso-O1-O2-Cipso torsion angle held constant at various angles between 0 and 90 for 9 and 0 and 35 for 9þ. The dependence of the relative energy on torsion angle (Figures 11 and 12) confirms the expected preferences;planarity for 9þ, distortion for 9;but also indicates that the potential well is fairly shallow, with only a small energetic cost for changing the degree of distortion. In this light, it seems reasonable to account for the dramatic differences in torsion angle between model 9 and real molecule 8, as well as those between the two conformers 80 and 800 , in terms of steric crowding and crystal packing, respectively.

techniques or in a glovebox under a nitrogen atmosphere as described previously.72 (Under standard glovebox conditions purging was not performed between uses of petroleum ether, diethyl ether, benzene, toluene, and tetrahydrofuran; thus when any of these solvents were used, traces of all these solvents were in the atmosphere and could be found intermixed in the solvent bottles.) All NMR solvents were purchased from Cambridge Isotopes Laboratories, Inc. The solvents for air- and moisturesensitive reactions were dried by passage through a column of activated alumina followed by storage under dinitrogen. All other materials were used as received. 2-Amino-4-tertbutylphenol, oxalyl chloride, triethylamine, BH3-THF (1 M in THF), triethyl orthoformate, potassium hexamethyldisilazide (potassium bis(trimethylsilyl)amide), and [FeCp2][PF6]73 were purchased from Aldrich. Chloro-1,5-cyclooctadiene Ir(I) dimer was purchased from Strem Chemicals, Inc. 1H, 13C, 31P, and 19F NMR spectra were recorded on Varian Mercury 300 spectrometers at room temperature, unless indicated otherwise. Chemical shifts are reported with respect to residual internal protio solvent for 1H and 13C{1H} NMR spectra. H3PO4 and CFCl3 were used as external standards to reference 31P and 19F NMR spectra, respectively, at 0 ppm. High-resolution mass spectra (HRMS) were obtained at the California Institute of Technology Mass Spectral Facility. Elemental analyses were obtained by Midwest Microlab, LLC, located in Indianapolis, IN, and Columbia Analytical Services (formerly Desert Analytics) located in Tucson, AZ. All electrochemical measurements were carried out on a Bioanalytics BAS100B/W. These measurements were performed under a nitrogen or argon atmosphere in 0.3 M tetrabutylammonium tetrafluoroborate solutions. Ferrocene was used as an

Conclusions The diphenolate imidazolyl-carbene ligand 1 has been shown to be capable of stabilizing both Ir(I) and Ir(III) complexes, including a complex capable of catalyzing olefin hydrogenations and complexes with interesting structural and electronic properties. The fact that 1 can support Ir(I) complexes, as well as Ir(III) complexes capable of undergoing two reversible one-electron oxidations, is promising for the incorporation of this ligand into redox-active catalysts.

Experimental Section General Considerations. All air- and/or moisture-sensitive compounds were manipulated using standard Schlenk

(72) Burger, B. J.; Bercaw, J. E. In New Developments in the Synthesis, Manipulation, and Characterization of Organometallic Compounds; Wayda, A., Darensbourg, M. Y., Eds.; American Chemical Society: Washington, DC, 1987; Vol. 357. (73) Ferrocenium(III) hexafluorophosphate was obtained from the supplier with a small amount of [BF4]-; hence, all [PF6]- salts synthesized are contaminated with a small amount of the [BF4]- salt.

98


internal standard, and all potentials are referenced to the ferrocene/ferrocenium couple. For cyclic voltammetry experiments, the working electrode was a glassy carbon disk (2 mm diameter); the counter electrode was a coiled platinum wire; and a silver wire was used as a pseudoreference electrode. In the bulk electrolysis experiments, the glassy carbon disk was replaced with a platinum wire basket for the working electrode. The spectrometer used for all EPR experiments was an X-band Bruker EMX with a standard TE102 cavity. Approximately 250 μL samples were transferred to an EPR tube, frozen with liquid nitrogen, and then placed in a modified TE102 cavity fit with a liquid helium-powered cryostat. This allowed EPR spectra to be obtained at temperatures between 7 and 77 K. Single-crystal X-ray diffraction samples were prepared by decanting or pipetting off any solvent and then coating the crystals with Paratone N oil. The crystals were then transferred to a microscope slide. Samples were selected and mounted on a glass fiber with Paratone N oil. Data collection was carried out on a Bruker Smart 1000 CCD diffractometer. The structures were solved by direct methods. Density functional calculations were carried out using Gaussian 03 revision D.01.74 Calculations on the model systems (with minimal steric bulk) were performed using the nonlocal exchange correction by Becke75,76 and nonlocal correlation corrections by Perdew,77 as implemented using the bvp86 keyword in Gaussian. The following basis sets were used: LANL2DZ78-80 for Ir atoms, Stuttgart-Dresden81,82 for phosphorus atoms, and 6-31G** basis set for all other atoms. Pseudopotentials were utilized for Ir and P atoms, using the LANL2DZ ECP for Ir and the Stuttgart-Dresden potential for P, as these gave the best agreement between the calculated and experimental structure. For calculations on full experimental systems, QMMM calculations were performed using ONIOM(bvp86:UFF).83 In these calculations the tertiary-butyl groups on the tridentate diphenolate imidazolyl-carbene ligands and the cyclohexyl groups on the phosphine ligands were calculated at the UFF level and the rest of the molecule was calculated using DFT. Initial geometries were obtained using the coordinates from X-ray structures, and all optimized structures were verified using frequency calculations to check that (74) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision C.02; Gaussian, Inc.: Wallingford, CT, 2004. (75) Becke, A. D. Phys. Rev. A: At., Mol., Opt. Phys. 1988, 38, 3098– 3100. (76) Becke, A. D. J. Chem. Phys. 1988, 88, 1053–1062. (77) Perdew, J. P. Phys. Rev. B 1986, 33, 8800–8802. (78) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270–283. (79) Wadt, W. R.; Hay, P. J. J. Chem. Phys. 1985, 82, 284–298. (80) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299–310. (81) Andrae, D.; H€aussermann, U.; Dolg, M.; Stoll, H.; Preuss, H. Theor. Chim. Acta 1990, 77, 123–141. (82) Bergner, A.; Dolg, M.; K€ uchle, W.; Stoll, H.; Preuss, H. Mol. Phys. 1993, 80, 1431–1441. (83) Svensson, M.; Humbel, S.; Froese, R. D. J.; Matsubara, T.; Sieber, S.; Morokuma, K. J. J. Phys. Chem. 1996, 100, 19357–19363. (84) Flukiger, P.; Portmann, H. P. L.; Weber, J. MOLEKEL; Swiss Centre for Scientfic Computing: Manno, Switzerland, 2000; Vol. 1.

Weinberg et al. they did not contain any imaginary frequencies. Isosurface plots were made using the Molekel program.84 Synthesis of 1,3-Di(2-hydroxy-5-tert-butylphenyl)imidazolium Chloride (1a). For preparation of the precursor, N,N0 -di(2-hydroxy-5-tert-butylphenyl)oxalamide, 4.0 mL (5.9 g, 0.046 mol) of oxalyl chloride was combined with 35 mL of CH2Cl2 under argon in a Schlenk flask. Then 100 mL of CH2Cl2 was used to dissolve 15.0 g (0.091 mol) of 2-amino-4-tert-butylphenol in a second Schlenk flask, and 12.6 mL (9.1 g, 0.090 mol) of triethylamine was added. Both solutions were then cooled to 273 K. The solution of 2-amino-4-tert-butylphenol and triethylamine was slowly transferred by cannula into the solution of oxalyl chloride, producing a white gas. Additional CH2Cl2 was used to wash in any remaining 2-amino-4-tert-butylphenol and triethylamine. The resulting solution was stirred for 2 h, and then 150 mL of H2O was added. CH2Cl2 was used to extract the organic products, the solvent was removed under vacuum, and the resulting solid was washed with both ethyl acetate and petroleum ether and dried under vacuum, giving a white solid (9.77 g, 0.0255 mol, 56.2% yield). 1H NMR (300 MHz, N,N-dimethylformamide (DMF)-d7): δ 10.61 (s, 2H), 10.04 (s, 2H), 8.45 (d, 4J = 2.34 Hz, 2H), 7.11 (dd, 3J = 8.55 Hz, 4 J = 2.33 Hz, 2H), 6.99 (d, 3J = 8.55 Hz, 2H), 1.31 (s, 18H). 13 C{1H} NMR (75 MHz, DMF-d7): δ 157.7, 145.4, 142.5, 125.1, 122.4, 117.0, 114.8, 31.4. HRMS (FABþ): m/z calcd for C22H28O4N2 þ H, 385.2127; found, 385.2108 (M þ H). Anal. Calcd for C22H28O4N2: C, 68.73; H, 7.34; N, 7.29. Found: C, 68.53; H, 7.25; N, 7.33. A 1.00 g (2.60 mmol) sample of N,N0 -di-(2-hydroxy-5-tertbutylphenyl)oxalimide was weighed into an oven-dried Schlenk flask and dissolved in 26.5 mL of BH3-THF (1 M in THF, 26.5 mmol). The solution turned bright orange, accompanied by vigorous bubbling; it was heated to 70 C and stirred for 19 h, then cooled to room temperature, yielding a transparent brown solution. While still stirring, methanol was added until all bubbling ceased; then 5.1 mL of 12 M HCl was added, and the solvent removed under reduced pressure. The resulting solid was redissolved in methanol, and solvent was again removed under reduced pressure; this process was repeated two more times, resulting in a white solid (the dihydrochloride salt of the diimine, which was not characterized). In a Schlenk flask under argon, the solid was suspended in 10.5 mL of triethylorthoformate and heated to 100 C with vigorous stirring. The solid still did not dissolve, even when another 10 mL of triethylorthoformate was added. The suspension was then heated for 2 h at 120 C, during which a tan color appeared, then filtered while still warm, and the resulting solid was washed with diethyl ether. Further solid deposited from the filtrate upon cooling. This solid was collected, combined with the first fraction, and dried under reduced pressure to yield 1a as a white solid (777 mg, 1.93 mmol, 74.2%). 1H NMR (300 MHz, dimethylsulfoxide (DMSO)-d6): δ 10.74 (br s, 2H), 9.49 (s, 1H), 7.38 (d, 4J = 2.09 Hz, 2H), 7.27 (dd, 3J = 8.69 Hz, 4J = 2.09 Hz, 2H), 7.05 (d, 3J=8.69 Hz), 4.57 (s, 4H), 1.27 (s, 18H). 13 C{1H} NMR (75 MHz, DMSO-d6): δ 156.7, 148.0, 142.3, 125.7, 123.0, 120.3, 116.5, 49.8, 34.0, 31.2. HRMS (FABþ): m/z calcd for C23H31O2N2, 367.2386; found, 367.2390 (Mþ). Anal. Calcd for C23H31O2N2Cl: C, 68.55; H, 7.75; N, 6.95. Found: C, 68.30; H, 7.78; N, 6.85. Synthesis of [K][{OCO}Ir(cod)] (2). In a glovebox, 1.5 g (4.0 mmol) of 1a and 2.4 g (12 mmol) of potassium hexamethyldisilazide were dissolved in a minimal amount of THF in an Erlenmeyer flask, and the solution was stirred for 30 min. In a separate flask, chloro-1,5-cyclooctadiene Ir(I) dimer (1.33 g, 1.98 mmol) was dissolved in a minimal amount of THF and slowly added to the solution of 1a and potassium hexamethyldisilazide. The resulting bright orange solution was stirred for 30 min, and the solvent was removed under reduced pressure. The resulting orange solid was stirred in methanol for 12 h under argon, solvent was again removed under reduced pressure, the

Article resulting solid was dissolved in THF and stirred for 15 min, the solvent was removed under reduced pressure. The resulting solid was dissolved in benzene and filtered, followed by lyophilization of the filtrate to give a bright orange powder (2.2 g, 3.1 mmol, 79%). 1H NMR (300 MHz, THF-d8): δ 6.81 (m, 4H), 6.43 (d, 3 J=8.62 Hz, 2H), 3.90 (app. t, 3J=8.81 Hz, 2H), 3.66 (app. t, 3J= 8.81 Hz, 2H), 3.51 (br s, 4H), 1.84 (br s, 4H), 1.38 (m, 4H), 1.26 (s, 18H). 1H NMR (300 MHz, CD3OD): δ 7.19 (d, 4J=2.58 Hz, 2H), 6.98 (dd, 3J=8.37 Hz, 4J=2.58 Hz, 2H), 6.66 (d, 3J=8.37 Hz, 2H), 4.53 (br s, 2H), 4.27 (br s, 2H), 2.40 (br s, 2H), 3.74 (br s, 2H), 2.92 (br s, 2H), 1.92 (br s, 2H), 1.55 (br s, 2H), 1.33 (br m, 20H). 13C{1H} NMR (75 MHz, THF-d8): δ 196.8, 161.1, 134.3, 132.6, 123.9, 119.6, 51.6, 34.3, 32.4, 32.2. HRMS (FABþ): m/z calcd for C31H40N2O2Ir, 665.2719; found, 665.2745 (Mþ), 557.1913 (Mþ - cod). Crystallization of 2 with 18-Crown-6 Ether. In a glovebox, saturated solutions of 2 (40 mg, 0.099 mmol) and 18-crown-6 ether (26.3 mg, 0.0995 mmol) in benzene were prepared. Another 0.5 mL of benzene was added to each solution, and the solution of 18-crown-6 ether was slowly added to the solution of 2 in a 20 mL vial. The vial was sealed and left to stand; orange crystals formed over the course of approximately one month. Synthesis of [{OCO}Ir(cod)(MeCN)][PF6] (3). In a glovebox, a saturated acetonitrile solution of 1.18 g (3.55 mmol) of [FeCp2][PF6] was slowly added to a saturated acetonitrile solution of 1.00 g (1.42 mmol) of 2. The resulting solution was stirred for 30 min and filtered. The solvent was removed from the filtrate under reduced pressure, and the resulting solid was washed with petroleum ether and diethyl ether and dried under reduced pressure to give 550 mg (0.65 mmol, 46% yield) of yellow solid. 1H NMR (300 MHz, CD3CN): δ 7.02 (dd, 3J=8.41 Hz, 4J=2.29 Hz, 2H), 6.95 (d, 3J=2.29 Hz, 2H), 6.87 (d, 3J=8.41 Hz, 2H), 6.35 (m, 2H), 5.12 (m, 2H), 4.51 (m, 2H), 4.29 (m, 2H), 2.59 (br m, 2H), 2.46 (br m, 2H), 2.35 (br m, 2H), 2.17 (br m, 2H), 1.30 (s, 18H). 13C{1H} NMR (75 MHz, THF-d8): δ 152.3, 141.4, 128.3, 127.4, 123.8, 120.3, 115.4, 94.0, 49.3, 35.1, 34.9, 31.8, 26.9. 31 P{1H} NMR (CD3CN, 121 MHz): δ -141 (septet, 1J = 700 Hz). 19F{1H} NMR (CD3CN, 282 MHz): δ -72.2 (d), -151.2 (s, [BF4]-).73 HRMS (FABþ): m/z calcd for C33H43N3O2Ir: 706.2984; found, 706.2987 (Mþ), 665.2721 (Mþ - CH3CN), 557.1781 (Mþ - (CH3CN þ cyclooctadiene)). In Situ Generation of [{OCO}Ir(MeCN)3][PF6] (4). In a glovebox, 100 mg (0.118 mmol) of 3 was dissolved in about 25 mL of acetonitrile and transferred into a 50 mL Schlenk bomb. The bomb was sealed and heated at 90 C for about 12 h. It was then cooled to room temperature, and the solvent was removed under reduced pressure. This solid was extracted with diethyl ether, and again the solvent was removed under reduced pressure. A 45 mg (0.054 mmol, 46%) amount of brown solid was isolated. The peaks in the 1H NMR at 2.53 and 2.68 ppm disappear over time in CD3CN, while the CH3CN peak at 1.97 ppm increases by approximately the same amount. 1H NMR (300 MHz, CD3CN): δ 6.94 (m, 4H), 6.67 (d, 3J=8.37 Hz, 2H), 4.50 (s, 4H), 2.68 (s,

and Iridium(III) Complexes Supported by a Diphenolate Imidazolyl

Recommend Documents