Communication pubs.acs.org/Organometallics
Structure and Bonding of the Manganese(II) Phosphide Complex (tBuPH2)(η5-Cp)Mn{μ-(t-BuPH)}2Mn(Cp)(t-BuPH2) Francesca A. Stokes, Robert J. Less, Joanna Haywood, Rebecca L. Melen, Richard I. Thompson, Andrew E. H. Wheatley,* and Dominic S. Wright* Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 1EW, U.K.
Adam Johannes Johansson and Lars Kloo Department of Chemistry, Royal Institute of Technology, Teknikringen 36, S-100 44 Stockholm, Sweden S Supporting Information *
ABSTRACT: Rather than achieving bis-deprotonation of the phosphine, reaction of Cp2Mn (Cp = cyclopentadienyl) with t-BuPH2 at room temperature yields monodeprotonation of half of the available phosphine in the product (tBuPH2)(η5-Cp)Mn{μ-(t-BuPH)}2Mn(Cp)(t-BuPH2) (1). This complex comprises a Mn(II) phosphide and is a dimer in the solid state, containing a Mn2P2 diamond core. Consistent with the observation of a relatively short intermetal distance of 2.8717(4) Å in 1, DFT analysis of the full structure points to a singlet ground state stabilized by a direct Mn−Mn single bond. This is in line with the diamagnetic character of 1 and an 18-electron count at Mn.
T
undergo the displacement of Cp in the presence of stronger nucleophiles (e.g., L = [R2N]−, [RN(CR)NR]−, [RCC]−), initially to give monosubstituted dimers of the form [CpMn(μL)]2 (eq 2),5 and more recently, reacting further to yield the trisubstituted manganate [(hpp)3MnLi]2 (hppH = 1,3,4,6,7,8hexahydro-2H-pyrimido[1,2,a]pyrimidine).6 The reaction of manganocene as a Brønsted base toward weak organic acids has also been documented (eq 3), with the extent of deprotonation having shown significant variability according to the acid strength. Hence, reaction of Cp2Mn with a variety of 2-aminopyrimidines has yielded octameric amido/imido cage complexes of the general formula [{CpMnNHR}{MnNR}]4 (R = 4,6-Me2pm, 4,6-MeO2pm, 4-Me-6-MeOpm, pm = pyrimidinyl, eq 4).7a,b In contrast, the weaker acids 2-aminopyridine (2-NH2py, py = pyridyl) and 8-aminoquinoline (8-NH2quin, quin = quinolinyl) undergo single deprotonation under similar conditions, giving hexanuclear [Cp2Mn3(NHpy)4]2 and dimeric [CpMn(μ-8-HNquin)]2, respectively (eqs 5 and 6).7b Finally, Cp2Mn does not deprotonate weakly acidic (BnNHCH2)2 (Bn = benzyl), instead giving only the simple adduct (η1-Cp)(η5Cp)Mn{(BnNHCH2)2} (eq 7).7b In spite of advances in the field of Mn(II) amide synthesis, little is known about their phosphide analogues,8 and here we report on the extension of background studies into Mn(II) amides through the 1:1 reaction of Cp2Mn with t-BuPH2 at room temperature. Rather than incurring bis-deprotonation of the phosphine, this reaction results in monodeprotonation of half of the available phosphine to give a diamagnetic Mn(II)
he organometallic chemistry of manganese(II) in general and of manganocene (Cp2Mn, Cp = cyclopentadienyl) in particular is frequently distinct from that of other 3d transitionmetal metallocenes.1 It is established as being capable of adding weak nucleophiles (e.g., [Cp]−) at the metal center to generate manganate anions (e.g., [Cp3Mn]−, eq 1),2 with more general
manganate species having been found to be applicable to synthetic chemistry3 while also being implicated in solvent decomposition processes.4 Meanwhile, further established chemistry of manganocene has focused on its ability to © 2011 American Chemical Society
Received: September 15, 2011 Published: December 19, 2011 23
dx.doi.org/10.1021/om200381p | Organometallics 2012, 31, 23−26
Organometallics
Communication
phosphide dimer based on a metallacyclic Mn2(phosphide)2 core. A THF solution of Cp2Mn was treated with equimolar tBuPH2 at −78 °C. The mixture was subsequently stirred at room temperature overnight to give a yellow-brown solution. Storage of this at −30 °C afforded a yellow crystalline material, 1.9 Care was required during isolation, with the crystals rapidly decomposing under prolonged vacuum (with the release of bonded t-BuPH2). However, careful isolation using only a brief (ca. 10 s) reduction in pressure during drying of the crystals allowed their isolation and storage under argon without the significant loss of t-BuPH2. In contrast to samples placed under vacuum for a prolonged period during isolation, this material gave satisfactory chemical (C, H, P) analysis. NMR spectroscopic analysis proved impossible in hydrocarbon media, owing to poor solubility properties. Following dissolution in d8-THF, 1H, 13C{1H}, and 31P NMR spectroscopy reveal symptoms consistent with the sample being diamagnetic. Thus, sharp resonances attributable to Cp are noted at δ 4.20 and 76.2 ppm, respectively. The observation of doublets attributable to 3JHP coupling at δ 1.53 and 1.07 ppm in the 1H NMR spectrum points to the presence of two distinct tert-butyl environments, with 13C{1H} NMR spectroscopy reinforcing this view by revealing CMe3 signals at δ 33.7 and 31.9 ppm. Integration of the 1H NMR spectrum suggests that the Cp and two t-Bu residues are present in a 1:1:1 ratio. The charged or neutral nature of the two separate coordinated phosphide/phosphine environments seen is suggested by 1H NMR spectroscopy, the associated hydrogen atoms observed at δ 3.76 (1H, 1JPH 301 Hz) and 3.21 (2H, 1JPH = 313 Hz) ppm suggesting [t-BuPH]− and t-BuPH2, respectively. Both 1H and 31 P NMR spectra show the presence of significant amounts of free t-BuPH2. Although there is evidently some unavoidable hydrolysis of the bridging t-BuPH groups in 1, it is also plausible that the presence of some of the free phosphine results from displacement of the neutral t-BuPH2 ligands by THF. This is consistent with the observed lability of these ligands when the complex is placed under vacuum (see above). The proximity in the 1H NMR spectrum of the t-Bu signal at δ 1.07 ppm to that of free phosphine (δ 1.16 ppm) allows us to assign the former as being attributable to the bonded t-BuPH2 ligand in 1. Moreover, the low-field shift of the tert-butyl signal from the phosphide ligand is consistent with previous studies.10 The presence of free phosphine is reiterated by 31P NMR spectroscopy, with a broad signal at δ −75.7 ppm agreeing well with that noted in a reference t-BuPH2 spectrum. A doublet at δ 241.6 ppm (1JPH = 301 Hz) and a triplet at δ 79.7 ppm (1JPH = 313 Hz) unambiguously establish the presence of monodeprotonated phosphide and coordinated phosphine in 1. While spectroscopic results established the formulation of 1, X-ray crystallography was required to clarify the structure and aggregation state. Analysis revealed it to be a mixed phosphide/ phosphine dimer, which crystallized in the tetragonal crystal system P43212 (Figure 1 and Scheme 1). The structure core of dimeric 1 is based on a Mn2P2 metallacycle containing short Mn−P bonds (Mn1−P1 = 2.2218(5) Å, Mn1−P1A = 2.2500(4) Å) that adopts a butterfly configuration folded along the intermetal axis ((Mn1−Mn1A− P1)−(Mn1−Mn1A−P1A) angle 135.8°). The core secondary phosphide ligands are the result of monodeprotonation (H1 and its symmetry equivalent were located directly by Fourier synthesis). The 1:1 Mn to phosphine ratio employed in the reaction supplies sufficient Brønsted base to monodeprotonate
Figure 1. Structure of 1 at the 40% probability level with only Pbonded H atoms shown. Selected bond lengths (Å): Mn1−Mn1A = 2.8717(4), Mn1−P1 = 2.2218(5), Mn1−P1A = 2.2500(4), Mn1−P2 = 2.1937(4). Selected bond angles (deg): P1−Mn1−P1A = 90.499(17), Mn1−P1−Mn1A 7= 9.906(16).
Scheme 1
all of the phosphine present. However, crystallography reveals that incomplete reaction at room temperature has instead allowed the coordination of each Mn center in 1 by an unreacted phosphine substrate molecule (Mn1−P2 = 2.1937(4) Å; like those of the bridging phosphides, the phosphine H atoms were observable in the Fourier difference map), and this observation is consistent with that of two chemically different t-Bu and phosphorus moieties by NMR spectroscopy (see above).11 The phosphide ligands in the metallacyclic core are relatively syn-arranged such that the sterically undemanding phosphido hydrogen centers project endo to the pocket described by the core metallacycle, while the t-Bu units are exo oriented. Two η5-cyclopentadienyl ligands (Mn−Cp(centroid) = 2.06(12) Å) reside in a syn fashion with respect to the phosphide t-Bu components, allowing the metal-solvating secondary phosphines to occupy sterically uncongested positions adjacent to the phosphide hydrogen atoms H1 and H1B. The contrasting actions of the phosphide and phosphine ligands (bridging vs terminal) are reflected in the relative metal−phosphorus bond lengths; the short bridging Mn−P(phosphide) (mean 2.24 Å) nevertheless being extended relative to terminal Mn−P(phosphine) (2.1937(4) Å). IR spectroscopy (Nujol) undertaken on 1 in the solid state reinforces the crystallographic observation of phosphido and phosphine ligands resulting from the incomplete deprotonation of t-BuPH2 at room temperature. In correspondence with the crystallographically observed structure of dimer 1, three P−H stretching modes were clearly noted.9 These were at 2397, 2328, and 2234 cm−1, and they were replaced by a band at 2290 cm−1 upon air exposure (cf. 2288 cm−1 for the free phosphine starting material). 24
dx.doi.org/10.1021/om200381p | Organometallics 2012, 31, 23−26
Organometallics
Communication
Figure 2. Optimized structure of 1 (left; Mn = purple, P = gold) in the closed-shell singlet state, including the bonding highest occupied molecular orbital (right). Hydrogen atoms are omitted for clarity.
row transition metals, the results discussed below are based on B3LYP* calculations.23 This analysis predicts the ground state to be a so-called broken-symmetry or open-shell singlet state, in which each manganese ion has one unpaired electron, in antiferromagnetic coupling. However, the closed-shell singlet state (having no unpaired electrons) lies only 9 kJ mol−1 higher in energy than the open-shell singlet. Considering the accuracy of the methods applied, these two singlet states should be regarded as energetically degenerate. The triplet state (having one unpaired electron on each manganese in ferromagnetic coupling) lies 15 kJ mol−1 higher than the ground state, while the quintet state (M = 5) is 127 kJ mol−1 above the ground state. In spin states of higher multiplicity, i.e. M = 7, 9, and 11, the complex is unstable with respect to metal−ligand (phosphine and cyclopentadienyl) dissociation. Thus, all highspin states seem to be unimportant for understanding the chemical properties of 1. It must thus be concluded that the DFT analysis of the energetic spin splitting predicts the ground state to be of singlet (M = 1) multiplicity, while it is unclear whether this singlet is of closed- or open-shell nature. An analysis of the corresponding geometries may shed further light on this issue. The geometrically optimized structures of both singlets (open and closed shell) are superficially very similar. However, a close inspection of atom−atom distances reveals significant differences; the closed-shell singlet structure displays Mn−Mn = 2.879 Å, Mn−P(phosphido) = 2.263 and 2.286 Å, Mn−P(phosphine) = 2.250 Å, and Mn−C = 2.16−2.23 Å, while the corresponding distances for the open-shell singlet structure are Mn−Mn = 3.248 Å, Mn−P(phosphido) = 2.352 and 2.374 Å, Mn−P(phosphine) = 2.291 Å, and Mn−C = 2.17−2.28 Å. The observation that the molecular units are more closely bound in the closed-shell structure and that the Mn−Mn distance is markedly shorter (cf. 2.8717(4) Å in 1), combines with the expected valence electron structure of Mn(II), to suggest the presence of a direct Mn−Mn single bond in the structure of 1. As shown in Figure 2, modeling clearly points to the highest occupied molecular orbital constituting a direct d−d-mediated Mn−Mn single bond in the closed-shell structure. In summary, the isolation of 1 allows the direct observation of a Mn(II) phosphide complex. Complex 1 exists as a dimer on the basis of the formulation Cp(t-BuPH)Mn·H2P(t-Bu), with the crystallographic observation of an intermetal distance of 2.8717(4) Å suggesting a direct Mn−Mn interaction. This view is reinforced by DFT studies, which strongly point to a singlet ground state stabilized by a direct Mn−Mn single bond. Support of the metal superstructure is provided by two monodeprotonated phosphide ligands that bridge the metal ions. In follow-up studies we aim to fully probe both the
The crystallographically verified structure of 1 is unique; there being, so far as we are aware, no closely analogous examples of Mn(II) phosphides.12 Hence, while silyl phosphide and aryl phosphide Mn(II) complexes are known, 1 represents one of a very limited number of Mn(II) complexes incorporating aliphatic phosphide ligands.13 However, it is noteworthy that it shares with THF-complexed dimer (Me3Si)2PMn(μ-(Me3Si)2P)2Mn(·THF)P(SiMe3)2 the tendency to readily lose solvating molecules in vacuo (see above).14 A simple electron count suggests that the formation of an intermetal single bond is plausible in 1, and the observed internuclear distance of 2.8717(4) Å is consistent with this view (see DFT studies, below). Of the few salient literature examples of phosphide-bridged dimanganese systems, only the Mn(I) complexes trans-(CO) 4 Mn{μ-P(H)Ph} 2 Mn(CO) 4 , 1 5 (CO)4Mn{μ-P(H)Ph}(μ-Br)Mn(CO)4,16 and trans-(CO)4Mn{μ-P(H)Ph}{μ-P(COMe)Ph}Mn(CO)417 reveal bridging secphosphides. Consistent with the lower metal oxidation state present, the MnI−P bonds noted for these species (range 2.346(6)−2.415(1) Å) are significantly extended relative to the Mn(II)−P(phosphide) bonds in 1 (see above). In a similar vein, tert-phosphides (CO)4Mn(μ-PCy2)2Mn(CO)4 (Cy = cyclohexyl)14 and (CO)4Mn{μ-P(H)Ph}(μ-H)Mn(CO)418 have been reported and the dicarbene PhC(OEt)(CO)3Mn(μ-PPh2)2Mn(CO)3C(OEt)Ph reveals MnI−P = 2.379(2)− 2.402(2) Å and an inter-Mn(I) displacement of 3.371(1) Å.19 Although a detailed investigation of the variable-temperature magnetic behavior of 1 is yet to be undertaken, its apparent diamagnetic behavior at room temperature represents an interesting contrast with previous reports of both straightforward half-sandwich Mn(II) complexes20 and also of bimetallic {M2(NCy)4}{Mn(η5-Cp)}2 (M = Sb, As).21 Whereas magnetometry suggested high-spin paramagnetic behavior for {M2(NCy)4}{Mn(η5-Cp)}2 in the presence of hard N-based cyclohexylamide ligands, the inclusion of a softer P-based ligand set in 1 would be expected to support low-spin Mn(II) and an intermetal interaction. Indeed, the observation of relatively short Mn−C(Cp) distances in 1 (in the range 2.1163(8)− 2.1782(17) Å) is consistent with the metal centers being low spin.5 To investigate this further, we have interrogated the full crystallographic structure by quantum-chemical methods using hybrid density-functional methods.22 The geometry of complex 1 was optimized in spin states of all possible multiplicities, i.e. M = 1, 3, 5, 7, 9, and 11. The energetic spin splitting was calculated at the B3LYP/LACV3P +** and B3LYP*/LACV3P+** levels, but since decreasing the amount of Hartree−Fock exchange from the 20% incorporated in the hybrid density functional B3LYP to the 15% used in B3LYP* has been shown to better describe complexes of first25
dx.doi.org/10.1021/om200381p | Organometallics 2012, 31, 23−26
Organometallics
Communication
H2PCMe3). IR (Nujol): 2397, 2328, 2234 (PH and bound PH2 sym + asym); (after 1 min air exposure) 2290 cm−1 (free PH2). X-ray crystal data: C26H52Mn2P4, Mr = 598.44, tetragonal, space group P43212, a = b = 9.1287(1) Å, c = 35.7743(3) Å, α = β = γ = 90°, V = 2981.19(5) Å3, Z = 4, ρcalcd = 1.333 g cm−3, Mo Kα radiation, λ = 0.710 73 Å, μ = 1.075 mm−1, T = 180(2) K, 13 256 data (2515 unique, Rint = 0.0264, θ < 24.69°) collected. wR2 = {∑[w(Fo2 − Fc2)2]/∑[w(Fo2)2]}1/2 = 0.0546, conventional R = 0.0186 on F values of 2485 reflections with F2 > 2σ(F2), S = 1.242, 162 parameters, residual electron density extrema ±0.263 e Å−3. (a) Sheldrick, G. M. Acta Crystallogr., Sect. A 1990, 46, 467. (b) Sheldrick, G. M. SHELXL-97 Program for Crystal Structure Refinement, University of Göttingen, Göttingen, Germany, 1997. (10) (a) Armstrong, D. R.; Feeder, N.; Hopkins, A. D.; Mays, M. J.; Moncrieff, D.; Wood, J. A.; Woods, A. D.; Wright, D. S. Chem. Commun. 2000, 2483. (b) Bashall, A.; Cole, J. M.; García, F.; Primo, A.; Rothenberger, A.; McPartlin, M.; Wright, D. S. Inorg. Chim. Acta 2003, 354, 41. (11) Arif, A. M.; Jones, R. A.; Schwab, S. T. J. Organomet. Chem. 1986, 307, 219. (12) (a) Chen, H.; Olmstead, M. M.; Pestana, D. C.; Power, P. P. Inorg. Chem. 1991, 30, 1783. (b) von Hänisch, C.; Weigend, F.; Clérac, R. Inorg. Chem. 2008, 47, 1460. (13) Jones, R. A.; Koschmieder, S. U.; Nunn, C. M. Inorg. Chem. 1988, 27, 2524. (14) Goel, S. C.; Chiang, M. Y.; Rauscher, D. J.; Buhro, W. E. J. Am. Chem. Soc. 1993, 115, 160. (15) (a) Flörke, U.; Haupt, H.-J. Acta Crystallogr., Sect. C 1993, 49, 374. (b) Flörke, U.; Haupt, H.-J. Acta Crystallogr., Sect. C 1993, 49, 533. (16) Flörke, U.; Haupt, H.-J. Acta Crystallogr., Sect. C 1997, 53, 876. (17) Manojlovic-Muir, L.; Muir, K. W.; Jennings, M. C.; Mays, M. J.; Solan, G. A.; Woulfe, K. W. J. Organomet. Chem. 1995, 491, 255. (18) Flörke, U.; Schwefer, M.; Haupt, H.-J. Z. Kristallogr. 1994, 209, 999. (19) Haupt, H.-J.; Petters, D.; Flörke, U. J. Organomet. Chem. 1998, 558, 81. (20) Kohler, F. H.; Hebendanz, N.; Thewalt, U.; Kanellakopulos, B.; Klenze, R. Angew. Chem., Int. Ed. Engl. 1984, 23, 721. (21) Bashall, A.; Beswick, M. A..; Ehlenberg, H.; Kidd, S. J.; McPartlin, M.; Palmer, J. S.; Raithby, P. R.; Rawson, J. M.; Wright, D. S. Chem. Commun. 2000, 749. (22) See the Supporting Information. (23) (a) Reiher, M.; Salomon, O.; Hess, B. A. Theor. Chem. Acc. 2001, 107, 48. (b) Salomon, O.; Reiher, M.; Hess, B. A. J. Chem. Phys. 2002, 117, 4729.
magnetic susceptibility and the variable-temperature solution behavior of 1 and, thereafter, to achieve the complete deprotonation of 1 so as to build extended and/or cyclic arrangements.
■
ASSOCIATED CONTENT * Supporting Information Text, figures, a table, and a CIF files giving synthetic details, spectroscopic data, and crystallographic data for 1 as well as computational details. This material is available free of charge via the Internet at http://pubs.acs.org. S
■
AUTHOR INFORMATION Corresponding Author *Fax: Int +44-1223-336362. E-mail:
[email protected] (A.E.H.W.);
[email protected] (D.S.W.).
■ ■
ACKNOWLEDGMENTS This work was supported by the U.K. EPSRC (F.A.S., R.L.M., J.H.) and the Leverhulme Trust (D.S.W., R.J.L.). REFERENCES
(1) Layfield, R. A. Chem. Soc. Rev. 2008, 37, 1098. (2) (a) Layfield, R. A.; McAllister, J. A.; McPartlin, M.; Rawson, J. M.; Wright, D. S. Chem. Commun. 2001, 1956. (b) Soria-Á lvarez, C.; Bashall, A.; McInnes, E.; Layfield, R. A.; McPartlin, M.; Rawson, J. M.; Wright, D. S. Chem.Eur. J. 2006, 3053. (3) (a) Garcia-Á lvarez, J.; Kennedy, A. R.; Klett, J.; Mulvey, R. E. Angew. Chem., Int. Ed. 2007, 46, 1105. (b) Blair, V. L.; Clegg, W.; Conway, B.; Hevia, E.; Kennedy, A.; Klett, J.; Mulvey, R. E.; Russo, L. Chem.Eur. J. 2008, 14, 65. (c) Blair, V. L.; Clegg, W.; Mulvey, R. E.; Russo, L. Inorg. Chem. 2009, 48, 8863. (4) Mulvey, R. E.; Blair, V. L.; Clegg, W.; Kennedy, A. R.; Klett, J.; Russo, L. Nature Chem. 2010, 2, 588. (5) Soria-Á lvarez, C.; Boss, S. R.; Burley, J.; Humphrey, S. M.; Layfield, R. A.; Kowenicki, R. A.; McPartlin, M.; Rawson, J. M.; Wheatley, A. E. H.; Wood, P. T.; Wright, D. S. Dalton Trans. 2004, 3481. (6) Brinkmann, C.; García, F.; Morey, J. V.; McPartlin, M.; Singh, S.; Wheatley, A. E. H.; Wright, D. S. Dalton Trans. 2007, 1570. (7) (a) Soria-Á lvarez, C.; Bond, A. D.; Harron, E. A.; Layfield, R. A.; McAllister, J. A.; Pask, C. M.; Rawson, J. M.; Wright, D. S. Organometallics 2001, 20, 4135. (b) Soria-Á lvarez, C.; Cave, D.; Bond, A. D.; Harron, E. A.; Layfield, R. A.; Mosquera, M. E. G.; Pask, C. M.; McPartlin, M.; Rawson, J. M.; Wood, P.; Wright, D. S. Dalton Trans. 2003, 3002. (8) For relevant reviews see: (a) Greenberg, S.; Stephan, D. W. Chem. Soc. Rev. 2008, 37, 1482. (b) Waterman, R. Dalton Trans. 2009, 18. (9) Cp2Mn (102 mg, 0.55 mmol) was dissolved in THF (3 mL) and cooled to −78 °C. t-BuPH2 (0.07 mL, 0.55 mmol) was added by syringe and the reaction vessel warmed to room temperature. After it was stirred overnight, the resulting yellow-brown solution was stored at −30 °C, affording yellow crystals of {t-BuPH(Cp)Mn·(t-BuPH2)}2 (1). Yield: 101 mg (31%). Mp: 169−171 °C. Anal. Found: C, 52.00; H, 8.78; P, 19.66. Calcd for C26H52Mn2P4, C, 52.18; H, 8.76; P, 20.70. 1 H NMR spectroscopy (500.1 MHz, d8-THF): δ 4.20 (s, Cp, 10H), 3.76 (dm, HP, 2H, 1JHP = 301 Hz), 3.21 (dbr, H2P, 4H, 1JHP = 313 Hz), 2.80 (d, free H2P, 10H, 1JHP = 185 Hz), 1.53 (d, HPCMe3, 18H, 3 JHP = 9 Hz), 1.16 (d, free H2PCMe3, 45H, 3JHP = 11 Hz), 1.07 (d, H2PCMe3, 18H, 3JHP = 11 Hz). 13C{1H} NMR spectroscopy (125.8 MHz, d8-THF): δ 76.2 (Cp), 33.7 (HPCMe3), 32.8 (d, free H2PCMe3, 2 JCP = 12 Hz), 31.9 (H2PCMe3). 31P NMR spectroscopy (162.0 MHz, d8-THF): δ 237.1 (d, HP, 1JPH = 301 Hz), 75.2 (t, H2P, 1JHP = 313 Hz), −80.1 (br, free H2PCMe3). 31P{1H} NMR spectroscopy (162.0 MHz, d8-THF): δ 241.6 (s, H2P), 79.7 (s, H2PCMe3), −75.7 (br, free 26
dx.doi.org/10.1021/om200381p | Organometallics 2012, 31, 23−26
