The Effect of a Fourth Binding Site on the Stabilization of Cationic SPS

Mar 20, 2009 - ACS eBooks; C&EN Global Enterprise .... The Effect of a Fourth Binding Site on the Stabilization of Cationic SPS Pincer Palladium Compl...
0 downloads 0 Views 522KB Size
Download PDF
2020

Organometallics 2009, 28, 2020–2027

The Effect of a Fourth Binding Site on the Stabilization of Cationic SPS Pincer Palladium Complexes: Experimental, DFT, and Mass Spectrometric Studies Matthias Blug,† Marjolaine Doux,† Xavier Le Goff,† Philippe Maıˆtre,‡ Franc¸ois Ribot,§ Pascal Le Floch,†,* and Nicolas Me´zailles*,† Laboratoire He´te´roe´le´ments et Coordination, Ecole Polytechnique, CNRS, 91128 Palaiseau, France, Laboratoire de Chimie Physique, UniVersite´ Paris-Sud 11, CNRS, 91405 Orsay, France, and Chimie de la Matie`re Condense´e de Paris, UPMC-UniVersite´ Paris 6, CNRS, 75252 Paris, France ReceiVed July 21, 2008

A potentially tetradentate ligand was synthesized by the addition of R-Li-picoline to the 2,6bis(diphenylphosphine sulfide)-3,5-diphenylphosphinine (SPS) ligand. The corresponding Pd complex (3) was obtained by the reaction with [(COD)PdCl2]. An air-stable cationic Pd(II) complex (4) was obtained after abstraction of Cl- with silver salt. The role of the additional binding site in the stabilization of the cationic complex was explored using NMR spectroscopic and tandem ESI-MS studies combined with DFT calculations. Additionally the picoline moiety was used to tune the electronic properties of the cationic complex 7 to modify the activity in the Lewis acid-catalyzed allylation of aldehydes. Introduction After the seminal report by Shaw in the 1970s, the chemistry of pincer ligands has evolved to maturity.1 The structures that have been studied early on, [2,6-(LCH2)2C6H3]- (L∼C∼L), where L is a two-electron donor and C is an anionic aryl carbon atom, have uncovered the peculiarities of these tridentate ligands.2 Of these, N∼C∼N, P∼C∼P, and S∼C∼S have then found numerous applications.3 In particular, the “P∼C∼P-metal” fragments have found applications in fundamental processes such as C-H activation as well as in several catalytic processes.4 For example, highly efficient Pd catalysts for cross-coupling processes were developed by the groups of Milstein, Morales-Morales, Jensen, and others.5 For a few years now, we have been involved in the development * Corresponding author. E-mail: [email protected]. Fax: +33 169 33 44 40. † Ecole Polytechnique, CNRS. ‡ Universite Paris-Sud 11. § UPMC-Univ. Paris 6. (1) Moulton, C. J.; Shaw, B. L. J. Chem. Soc., Dalton Trans. 1976, 1020. (2) Morales-Morales, D. J. C. M. The Chemistry of Pincer Compounds; Elsevier: 2007. (3) (a) Albrecht, M.; van Koten, G. Angew. Chem., Int. Ed. 2001, 40, 3750. (b) Cantat, T.; Jaroschik, F.; Nief, F.; Ricard, L.; Me´zailles, N.; Le Floch, P. Chem. Commun. 2005, 5178. (c) Cantat, T.; Ricard, L.; Me´zailles, N.; Le Floch, P. Organometallics 2006, 25, 6030. (d) Friggeri, A.; van Manen, H. J.; Auletta, T.; Li, X. M.; Zapotoczny, S.; Schonherr, H.; Vancso, G. J.; Huskens, J.; van Veggel, F.; Reinhoudt, D. N. J. Am. Chem. Soc. 2001, 123, 6388. (e) Dani, P.; Karlen, T.; Gossage, R. A.; Gladiali, S.; van Koten, C. Angew. Chem., Int. Ed. 2000, 39, 743. (f) Rietveld, M. H. P.; Grove, D. M.; van Koten, G. New J. Chem. 1997, 21, 751. (g) Steenwinkel, P.; James, S. L.; Grove, D. M.; Veldman, N.; Spek, A. L.; van Koten, G. Chem.-Eur. J. 1996, 2, 1440. (h) Gozin, M.; Weisman, A.; Bendavid, Y.; Milstein, D. Nature 1993, 364, 699. (4) (a) Gupta, M.; Hagen, C.; Kaska, W. C.; Cramer, R. E.; Jensen, C. M. J. Am. Chem. Soc. 1997, 119, 840. (b) Salem, H.; Ben-David, Y.; Shimon, L. J. W.; Milstein, D. Organometallics 2006, 25, 2292. (c) GomezBenitez, V.; Baldovino-Pantaleon, O.; Herrera-Alvarez, C.; Toscano, R. A.; Morales-Morales, D. Tetrahedron Lett. 2006, 47, 5059. (5) (a) Ohff, M.; Ohff, A.; vanderBoom, M. E.; Milstein, D. J. Am. Chem. Soc. 1997, 119, 11687. (b) Morales-Morales, D.; Grause, C.; Kasaoka, K.; Redon, R.; Cramer, R. E.; Jensen, C. M. Inorg. Chim. Acta 2000, 300, 958. (c) Morales-Morales, D.; Redon, R.; Yung, C.; Jensen, C. M. Chem. Commun. 2000, 1619.

of anionic S∼P∼S ligands based on phosphinine derivatives, the phosphorus analogues of pyridines. We found that complexes such as [(S∼P∼SR)PdCl] proved to be very robust and could efficiently catalyze the Miyaura coupling process, which allows the synthesis of boronic esters from halogenoarenes and pinacolborane.6 In 2003, Szabo and co-workers reported on interesting advances in the promotion of new allylation processes by using pincer palladium complexes in the catalyzed conversion of aldehydes into homoallyl alcohols through reaction with allylstannanes.7 In their studies, these authors have tested several ligand systems and showed that Se∼C∼Se tridentate pincer ligands have the optimal electronic properties for the desired transformation. They proposed the involvement of η1-allyl species based on the strong coordination of the pincer ligand, which occupies three of the four coordination sites of a square-planar Pd(II) complex.8 Interested by these results, we performed a combined experimental and theoretical study on this catalytic process, with the in-house designed S∼P∼SR and S∼C∼S pincer ligands. In terms of reactivity, the two Pd complexes were as efficient as the Se∼C∼Se-based system. Notably, we could show that the activity was increased by the addition of silver salts. The DFT study allowed us to rule out the involvement of the η1-allyl intermediates because of much higher energetic demand compared to the Lewis acid-promoted process. In practice, the two in situ synthesized [(S∼X∼S)Pd]+ (X ) C or P) systems were found to be excellent Lewis acids capable of activating the aldehyde toward nucleophilic attack by the allylstannane derivative (Scheme 1).9 The purpose of the present study is threefold. First we wished to isolate cationic Pd derivatives bearing SPS-based tridentate ligands in order to probe their reactivity in Lewis acid-catalyzed transformations. Second, we envisioned the introduction of a potentially fourth binding moiety, which would be able to stabilize the highly electrophilic cationic Pd species during a catalytic process. A geometrical requirement for this fourth moiety is that it should be connected to the ligand backbone in a manner that prevents the formation of a highly stable squareplanar Pd(II) complex, which would hamper any catalytic process. Finally, it was desirable to be able to tune the Lewis

10.1021/om800690t CCC: $40.75  2009 American Chemical Society Publication on Web 03/20/2009

Cationic SPS Pincer Palladium Complexes

Organometallics, Vol. 28, No. 7, 2009 2021 Scheme 1

acidity of the Pd center via the fourth ligand, which was therefore selected as a pyridine moiety. The results of this study are presented herein. In the course of these endeavors, density functional theory (DFT) calculations combined with electrospray ionization tandem mass spectrometry (ESI-MS/MS) techniques were used to uncover the geometry of a cationic Pd complex.

Results and Discussion As mentioned above, the [(S∼P∼SBu)PdCl] complex 1 can be activated to a postulated [(S∼P∼SBu)Pd]+ cationic species upon addition of silver salts. When this reaction was performed stoichiometrically, in a noncoordinating solvent, the disappearance of the starting complex and the formation of species that rapidly decomposed were observed, proving again the high Lewis acid character of the cationic complex. On the other hand, when the reaction was performed in the presence of a coordinating species, such as a few equivalents of acetonitrile, a single new complex was obtained (eq 1).

This complex, 2, was isolated in 85% yield and characterized by NMR spectroscopy and elemental analyses. In particular the 31 P NMR spectrum shows an AX2 spin system pattern with a doublet at δ 49.7 ppm (2JPP ) 69.6 Hz) and triplet at δ 64.6 ppm (2JPP ) 69.6 Hz), contrasting with the AB2 spin system of the starting complex (δ 49.2 and 55.4 ppm with 2JPP ) 87.0 Hz). The activity of complex 2 was tested in the nucleophilic allylation of aldehydes. Not surprisingly, this complex appeared as a poor catalyst because of the high stability of the squareplanar Pd(II) complex. Indeed, as shown in our proposed mechanism, aldehydes have to coordinate the electrophilic Pd center prior to attack by the nucleophile. In this case, acetonitrile (6) Doux, M.; Me´zailles, N.; Melaimi, M.; Ricard, L.; Le Floch, P. Chem. Commun. 2002, 1566. (7) Solin, N.; Kjellgren, J.; Szabo, K. J. Angew. Chem., Int. Ed. 2003, 42, 3656. (8) (a) Szabo, K. J. Chem.-Eur. J. 2000, 6, 4413. (b) Wallner, O. A.; Szabo, K. J. J. Org. Chem. 2003, 68, 2934. (c) Szabo, K. J. Chem.-Eur. J. 2004, 10, 5269. (d) Nakamura, H.; Bao, M.; Yamamoto, Y. Angew. Chem., Int. Ed. 2001, 40, 3208. (9) Piechaczyk, O.; Cantat, T.; Me´zailles, N.; Le Floch, P. J. Org. Chem. 2007, 72, 4228.

is clearly a better ligand than aldehydes, preventing the catalytic process from occurring. It was therefore highly desirable at this point to synthesize complexes that would be (1) more electrophilic than the chloride derivative 1, allowing the direct incorporation in the catalytic cycle; (2) more stable than the “unstabilized” cationic Pd derivative; (3) yet more reactive than the acetonitrile cationic derivative 2; (4) finally, if possible, allowing fine-tuning of the catalytic activity. Point 2 clearly requires the presence of a labile two-electron donor, and point 3 implies a weaker two-electron donor than a nitrile ligand. We thus envisioned the incorporation of an R-picoline moiety on the phosphorus center of the SPS ligand, in a way to prevent the formation of a stable squareplanar complex. The synthesis of this ligand has been reported recently,10 and its neutral Pd complex 3 synthesized readily (eq 2) and isolated in 74% yield. This complex was characterized by NMR spectroscopy, elemental analyses, and X-ray crystallography (Figure 1).

In the 31P NMR spectrum, an AB2 system (δ 48.4 and 51.9 ppm with 2JPP ) 89.0 Hz) is seen as for 1. In the 1H NMR spectrum the H4 proton is found at rather high field, δ 5.46 ppm in CDCl3, compared to the starting anion, and at a similar chemical shift than in complex 2, and resonates as a triplet because of the coupling with the two phosphorus atoms of the Ph2PdS groups. As expected for a Pd(II) complex, its geometry (10) Doux, M.; Thuery, P.; Blug, M.; Ricard, L.; Le Floch, P.; Arliguie, T.; Me´zailles, N. Organometallics 2007, 26, 5643.

2022 Organometallics, Vol. 28, No. 7, 2009

is square planar with the pyridine moiety not in interaction with the metal center. The chloride abstraction was then attempted on this complex with 1 equiv of AgBF4 in methylene chloride (Scheme 1). Instantaneous precipitation of the expected AgCl was observed. The 31P{1H} NMR spectrum showed the complete transformation of 3 into a new complex, 4, characterized by a complex spin system between δ 45.4 and 49.8 ppm. These chemical shifts, similar to the ones observed for complex 2, are in agreement with the expected formation of a cationic species. In the 1H NMR spectrum in CD2Cl2 this complex also shows a triplet for the H4 proton at δ 6.08 ppm (4JHP(B) ) 4.7 Hz). This complex, unlike the “SPSBu” analogue, is stable in solution for extended periods of time, which clearly points to the coordination of the pyridine fragment to the Pd center. Despite many efforts, we have been unable to crystallize this species. Two possibilities have thus to be envisaged for the geometry of this complex: either monomeric or dimeric. In the first case, the geometry at the Pd center would be highly distorted from square planar, because it is obvious from the structure of 3 that the pyridine moiety cannot reach over to be located trans to the phosphorus center. In the second case, the pyridine fragment of one ligand coordinates a second metal center to form a square-planar arrangement at the Pd center. A combined DFT and ESI tandem MS study was then performed to shed light on the precise nature of this complex (Vide supra). Further experimental proof of the cationic character of complex 4 was obtained by adding a few equivalents of pyridine (Scheme 1). The new complex is 5, characterized in the 31P NMR spectrum in CD2Cl2 by an AB2 system at δ 45.7 and 47.6 ppm (2JPP ) 79.5 Hz). The formulation of 5 was confirmed by elemental analyses as well as X-ray diffraction analysis (Figure 2). The complete formation of this complex is quite interesting in the sense that it shows that the pyridine fragment coordinated on a Pd center in complex 4 is readily displaced by another external pyridine ligand. This points to a weak stabilization, yet efficient in terms of complex stability in time, of the cationic Pd center.

Figure 1. ORTEP plot (50% thermal ellipsoids) of the X-ray crystal structure of compound 3. Ph groups were omitted for clarity. Selected bond lengths [Å] and angles [deg]: Pd(1)-P(1) 2.1783(8), Pd(1)-S(1) 2.3207(8), Pd(1)-S(2) 2.332(1), Pd(1)-Cl(1) 2.3894(8), S(1)-P(2) 2.033(1), S(2)-P(3) 2.043(1), P(1)-C(1) 1.761(3), P(1)-C(5) 1.760(3), P(1)-C(6) 1.839(3), P(1)-Pd(1)-S(1) 87.93(3), P(1)-Pd(1)-S(2) 87.17(3), S(1)-Pd(1)-S(2), 172.76(3), P(1)Pd(1)-Cl(1) 170.25(3).

Blug et al.

Figure 2. ORTEP plot (50% thermal ellipsoids) of the X-ray crystal structure of cation of compound 5. Ph groups and BF4- were omitted for clarity. Selected bond lengths [Å] and angles [deg]: Pd(1)-P(1) 2.2040(6), Pd(1)-S(1) 2.3126(6), Pd(1)-S(2) 2.3359(6), S(1)-P(2) 2.0313(8), S(2)-P(3) 2.036(1), P(1)-C(1) 1.776(2), P(1)-C(5) 1.771(2), P(1)-C(6) 1.841(2), P(1)-Pd(1)-S(1) 87.49(2), P(1)-Pd(1)-S(2) 86.83(2), S(1)-Pd(1)-S(2), 173.78(2), P(1)Pd(1)-N(2) 175.43(5).

In parallel, in order to probe the possibility of tuning the catalytic activity, quaternarization of the lone pair of pyridine was attempted. Indeed, the overall increase of the charge of the complex should in turn increase the Lewis acidity of the Pd center. In practice, addition of a few equivalents of HCl to a solution of complex 3 led to the instantaneous formation of complex 6, which was isolated quantitatively. The electronic influence of the protonation on the whole ligand backbone was clearly observed in the 1H NMR spectrum. Indeed, the CH2 group of the picoline experiences a downfield shift by 0.46 ppm (from δ 3.83 ppm in 3 to δ 4.29 ppm in 6). This effect is also observed on the shift of H4 on the phosphinine ring (from δ 5.46 ppm in 3 to δ 5.72 ppm in 6). The ORTEP plot of complex 6 is presented in Figure 3. As for 3, the formation of a stable cationic Pd center was then attempted. Two equivalents of AgBF4 were added to a solution of 6 in dichloromethane, leading to the precipitation of AgCl. In the 31 P NMR spectrum, the second-order spectrum of 6 disappeared rapidly in favor of a very broad singlet centered at δ 47.4 ppm. The fact that this species, 7, is unique was evidenced by addition of 10 equiv of acetonitrile. Indeed, it resulted in the clean formation of a single complex, 8, characterized again by an AB2 spin system at δ 44.7 and 47.7 ppm (2JPP ) 80.4 Hz). This complex was isolated in 75% yield and fully characterized by NMR techniques. Interestingly, exhaustive drying of this complex under vacuum resulted in the partial loss of acetonitrile to re-form 7. This process is reversible and shows that this ligand is more weakly bound to the Pd center than in 2, which could be of interest for catalytic processes. Catalytic Tests. Having in hand several complexes, preliminary tests were performed to confirm our hypotheses, namely, the influence of pyridine as a fourth ligand, the use of isolated cationic species, and finally the increase of the overall charge by protonation of pyridine. The allylation of aldehydes was chosen, as complex 1 is known to perform well in this process and could therefore be used as benchmark. The reaction between

Cationic SPS Pincer Palladium Complexes

Organometallics, Vol. 28, No. 7, 2009 2023

Figure 3. ORTEP plot (50% thermal ellipsoids) of the X-ray crystal structure of compound 6. Ph groups were omitted for clarity. Selected bond lengths [Å] and angles [deg]: Pd(1)-P(1) 2.184(1), Pd(1)-S(1) 2.323(1), Pd(1)-S(2) 2.319(1), Pd(1)-Cl(1) 2.380(8), S(1)-P(2) 2.023(1), S(2)-P(3) 2.041(1), P(1)-C(1) 1.777(3), P(1)-C(5) 1.762(3), P(1)-C(6) 1.853(3), P(1)-Pd(1)-S(1) 88.39(3), P(1)-Pd(1)-S(2) 86.55(3), S(1)-Pd(1)-S(2), 174.37(3), P(1)Pd(1)-Cl(1) 173.92(4). Table 1. Catalytic Results entry

catalyst

yield (%)

1 2 3 4 5 6

1 1 + 1 AgBF4 3 4 6 6 + 1 AgBF4

18.2 37.8 10.4 18.9 15.3 30.2

Table 2. Relative Energies of Structures I-IVa relative energies (kcal · mol-1)

I

II

III

IV

∆E ∆EPCM ∆G ∆GPCM

0 0 0 0

1.7 -2.4 3.1 -1.1

-4.5 -10.4 5.2 -0.7

4.9 -1.7 11.7 5.1

a

With ∆GPCM ) ∆G + (∆EPCM - ∆E).

p-bromobenzaldehyde and allyltributyltin was conducted in CHCl3 using 1% of catalyst, at 70 °C for 4 h. The results are gathered in Table 1. Entries 1, 3, and 5 allow the direct comparison of “PdCl” fragments, whereas entries 2, 4, and 6 are related to the corresponding cationic complexes. As mentioned previously, the stoichiometric addition of a silver salt to a solution of complex 1 led to an increase in the yield of the product, resulting from the fast generation of the catalytically active species (entries 1 and 2). Entry 3 allows the direct assessment of the role of the pyridine ligand. Complex 3 is the least active precatalyst (entry 3). Reasonably postulating a kinetically similar formation of the active species as for 1, it is clear that the pyridine ligand has a significantly stabilizing influence, thereby reducing the Lewis acidity of the Pd center. Cationic complex 4 (monomer or dimer, Vide supra), obtained after chloride abstraction from 3 followed by isolation, performs almost twice as well as 3 (entry 4). This clearly shows that this complex is not only stable once isolated but also more efficient than the “PdCl” complex. Complex 6, the pyridine protonated version of complex 3, is also more efficient, which would result from the expected increase of the Lewis acidity of the Pd center (entry 5). However, its reactivity is similar to that of complex 1, which would point to a negligible influence of the overall charge. This complex also performs almost as well as complex 4. Eventually,

in situ chloride abstraction from 6 was done, and the resulting complex 7 directly engaged in the catalytic process (entry 6). As in the case of 1, it resulted in doubling the activity (30.2% vs 15.3%), but as mentioned above (activities of 1 and 6) the overall dicationic complex does not perform better than “1 + AgBF4” (30.2% vs 37.8%). Finally, it does not appear that protonation of the pyridine moiety is a positive factor on the activity of the complex. Overall, isolated cationic complex 4 provides the best compromise between catalytic performance (50% lower), kinetic stability, and ease of handling (no need for activation). It is obvious from these results that the pyridine moiety indeed brings stability to a cationic Pd center, yet does not suppress catalytic behavior. DFT Study. The question of the precise nature of complex 4 remained: monomer or dimer? A DFT study has been performed to investigate the possible conformations of the cationic compound [(S∼P∼SPic)Pd]+. All calculations were carried out using the Gaussian 03 set of programs11 with the B3PW91 functional,12 the 6-31G* basis set for all nonmetallic atoms (C, H, P, N, S), and the Hay-Wadt13 basis set for palladium with an additional f-polarization function.14 A model compound, replacing all the phenyl groups by protons, was used for this study in order to preserve calculation time (Note: this model has proven to be appropriate in theoretical studies reported before15). As we are dealing here with either monomeric-monocationic or dimeric-dicationic species, the effect of solvation had to be taken into account in the DFT studies. This was achieved by using the polarized continuum model (PCM),16 considering CH2Cl2 as the solvent. However, entropic factors are not incorporated in this PCM energy, and free energies were therefore also calculated. In turn, these relative free energies, ∆G, are related to the gas phase systems, and the extrapolation to the condensed phase is delicate and subject to intense debate.17,18 In particular, Maseras and co-workers proposed the use of a free energy ∆GPCM taking into account both the entropic and the solvation effects, which is given by the relation ∆GPCM ) ∆G + (∆EPCM - ∆E).17 Two monomeric species were envisaged: an S∼P∼S-bound monomeric complex, I, and an S∼P∼N-bound monomeric complex, II. Two dimeric structures were also envisaged: one in which the pyridine fragment of one ligand is coordinated to a second Pd center, complex III, and another in which sulfur atoms are bridging, complex IV. This latter coordination mode has not yet been observed in phosphinine-based SPS pincer complexes. Pictures of the optimized structures are shown in (11) Frisch, M. J.; et. al. Gaussian 03 (ReVision C.02); Gaussian Inc: Wallingford, CT, 2004. (12) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Perdew, J. P.; Wang, Y. Phys. ReV. B 1992, 45, 13244. (13) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299. (14) Ehlers, A. W.; Bohme, M.; Dapprich, S.; Gobbi, A.; Hollwarth, A.; Jonas, V.; Kohler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G. Chem. Phys. Lett. 1993, 208, 111. (15) (a) Doux, M.; Le Floch, P.; Jean, Y. Eur. J. Inorg. Chem. 2006, 2035. (b) Doux, M.; Ricard, L.; Le Floch, P.; Jean, Y. Organometallics 2006, 25, 1101. (c) Doux, M.; Le Floch, P.; Jean, Y. J Mol. Struct. (Theochem) 2005, 724, 73. (d) Doux, M.; Ricard, L.; Le Floch, P.; Jean, Y. Organometallics 2005, 24, 1608. (16) (a) Barone, V.; Improta, R.; Rega, N. Theor. Chem. Acc. 2004, 111, 237. (b) Cossi, M.; Scalmani, G.; Rega, N.; Barone, V. J. Chem. Phys. 2002, 117, 43. (c) Cossi, M.; Barone, V.; Cammi, R.; Tomasi, J. Chem. Phys. Lett. 1996, 255, 327. (d) Miertus, S.; Scrocco, E.; Tomasi, J. Chem. Phys. 1981, 55, 117. (17) Braga, A. A. C.; Ujaque, G.; Maseras, F. Organometallics 2006, 25, 3647. (18) (a) Sakaki, S.; Takayama, T.; Sumimoto, M.; Sugimoto, M. J. Am. Chem. Soc. 2004, 126, 3332. (b) Tamura, H.; Yamazaki, H.; Sato, H.; Sakaki, S. J. Am. Chem. Soc. 2003, 125, 16114.

2024 Organometallics, Vol. 28, No. 7, 2009

Blug et al.

Figure 4. Calculated structures I-IV for the conformation of cationic compound 4. Selected bond lengths [Å] and angles [deg] for I: Pd(31)-P(17): 2.201, Pd(31)-S(15): 2.393, Pd(31)-S(16): 2.402, Pd(31)-N(23): 3.013, P(17)-Pd(31)-S(15): 86.1, P(17)-Pd(31)-N(23): 76.7, N(23)-Pd(31)-S(15): 97.7 S(15)-P(31)-S(16): 176.0. For II: Pd(31)-P(17): 2.278, Pd(31)-S(15): 2.383, Pd(31)-S(16): 6.176, Pd(31)-N(23): 2.086, P(17)-Pd(31)-S(15): 96.3, P(17)-Pd(31)-N(23): 83.4, N(23)-Pd(31)-S(15): 167.6. For III: Pd(31)-P(17): 2.249, Pd(31)-S(15): 2.418, Pd(31)-S(16): 2.379, Pd(31)-N(32): 2.182, P(17)-Pd(31)-S(15): 83.6, P(17)-Pd(31)-N(32): 173.4, N(32)-Pd(31)-S(15): 90.7. For IV: Pd(31)-P(17): 2.239, Pd(31)-S(15): 2.393, Pd(31)-S(16): 2.414, Pd(31)-N(23): 4.943, Pd(31)-Pd(62): 3.777, Pd(31)-S(60): 2.558, P(17)-Pd(31)-S(15): 84.6, P(17)-Pd(31)-S(60): 151.6, S(15)-P(31)-S(16): 170.7.

Figure 4. Note that in complex I there is a weak interaction between the pyridine moiety and the Pd center, as shown by the long distance between these atoms: 3.013 Å. This distance is definitely longer than a true bond, yet the particular orientation of the pyridine pointing toward the Pd center is unambiguous. Structure I is used as a reference. Not considering the weak interaction of the pyridine ligand, the overall geometry around the Pd center is T shaped, with the two S-Pd-P angles of 86.1°. The geometry of complex II was optimized in distorted T shape, with the pending phosphino sulfide ligand pointing in the opposite direction. In this complex, the peculiar arrangement of the ligand leads to angles significantly different from the expected 90°: P(17)-Pd(31)-S(15) ) 96.3° and P(17)-Pd(31)N(23) ) 83.4°. Interestingly, this complex is lower in energy (∆EPCM) than I, by 2.4 kcal/mol, but the free energy of this complex is slightly higher (3.1 kcal/mol), showing that the phosphine sulfide is as good a ligand as pyridine for this Pd(II) cationic center. The potential energy, E, of compound III was found to be lower than that of the monomers, whereas compound IV was found at higher energy relative to the monomeric compounds. It should be noted that both dimeric structures were stabilized by ca. 6 kcal/mol, when the contribution of the solvation is added to the potential energy (∆EPCM), whereas the monomeric structures were only slightly stabilized (0-4.1 kcal/mol). However, dimerization is obviously entropically unfavorable, and the Gibbs free energies, ∆G, for the dimers III and IV are found higher than that of the monomers. Of these dimers, the sulfurbridged complex IV is less stable than complex III by 8.0 kcal/ mol. In complex IV, the coordination deviates significantly from the expected square planar, as shown by the P(17)-Pd(31)-S(60) angle of 151.6°. This probably results from geometrical constraints. The geometry of the two “(S∼P∼S)Pd” fragments in III is very similar to the one found in the monomeric complex I. In the dimer, the expected more favored square-planar geometry of Pd(II) is achieved by the coordination of a pyridine moiety of the opposite ligand. Finally, when ∆GPCM is calculated, compounds I, II, and III all lie within 1.1 kcal/mol and should therefore coexist in solution. In the end, taking only ∆EPCM into account, the dimericdicationic species III is predicted to be more stable and would be the only one to be observed in solution. On the other hand, ∆G predicts the monomeric-monocationic species I to be more stable and would be the only one to be observed in solution. Finally, the difference in energy ∆GPCM predicted the coexistence of the three complexes I, II, and III in solution. This case is therefore a complicated one, for which DFT calculations did

not provide a definitive answer. We thus had to rely on an experimental technique to quantify the amount of the different complexes in solution. TANDEM ESI-MS Study. Electrospray ionization tandem mass spectrometry (tandem ESI-MS) was used to further characterize the Pd complexes present in solution. Indeed, soft ionization tandem mass spectrometry appears to be a sensitive and rapid technique for the determination of molecular weight of organometallic compounds.19 Electrospray ionization allows for a soft transfer to the gas phase of pre-existing molecules or ions in solution. Short-lived reactive intermediates can also be mass analyzed, thus providing a method of choice for mechanistic studies.20 The present tandem ESI-MS study was performed using a 7 T Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometer. The aim of this investigation was to determine the chemical nature of the Pd complex 4, postulating that the ionization technique would not result in the splitting of the dimer if present in solution. The mass spectrum of the CH2Cl2 solution of the synthesized complexes is given in Figure 5, and a detailed view of the dominant signal at m/z 879.06 is provided in Figure 5b. The observed isotopic pattern of the dominant peak at m/z 879.06 nicely matches the theoretical isotopic patterns calculated for the dicationic dimer complex, modeled by III above (Figure 5b). A numerical fit of the experimental spectrum, however, shows that a small fraction (∼2.5%) of the monocationic monomeric complex (isotopic patterns given in Figure 5d) is also present under our experimental conditions. When used in conjunction with chemicalinduced dissociation (CID), ESI-MS/MS can provide additional insights. After mass selection of the ions at m/z 879.06 with an isolation width of 10 Da, CID was performed for 100 ms. The resulting mass spectrum suggests that the dicationic dimeric species [(S∼P∼SPic)Pd]22+ fragments into two monocationic monomers [(S∼P∼SPic)Pd]+. At the ion/argon collision energies used (20 V), loss of C6H6N from the monomeric species was also observed. In the end, the results of this electrospray ionization tandem mass spectrometry study very nicely complement the results of the DFT study. Indeed, the DFT calculations did not allow settling conclusively between monomers or dimer complex in solution. On the grounds of the ESI-MS study and the respective percentages of the species at room temperature, we can conclude that the dimer (type III) is more stable than the monomers by ca. 2 kcal/mol. (19) Chaplin, A. B.; Dyson, P. J. Organometallics 2007, 26, 2447. (20) Chen, P. Angew. Chem., Int. Ed. 2003, 42, 2832.

Cationic SPS Pincer Palladium Complexes

Organometallics, Vol. 28, No. 7, 2009 2025

the structure of the cationic complex 4. The increase of the calculated hydrodynamic radius (rH(4)/rH(3) ) 1.78) is in excellent agreement with the existence of a dimeric structure in solution. This final result clearly proves that the dimeric complex is present in solution and not formed during the ionization process.

Conclusions

Figure 5. Mass spectra of a sample solution of Pd complexes in CH2Cl2 at 10-5 mol · L-1. Full mass spectrum recorded without mass selection (a) and detailed view of the isotopic distribution of the dominant peak at m/z 879.06 (b). Theoretical isotopic patterns calculated for the dicationic dimer (c) and monocationic monomer complexes (d). Table 3. Diffusion Constants and Hydrodynamic Radii of Complexes 3 and 4 compound

diffusion constant, 10-10 m2 s-1

rH, Å

TMS complex 3 complex 4

25.1 9.5 5.4

2.95 7.76 13.80

1

H-DOSY-NMR Study. Nevertheless, it has been reported in specific cases that the dimerization of monocationic transitionmetal complexes bearing halide ions, leading to halide-bridged dimer complexes, could occur in a mass spectrometer21 using the electrospray ionization technique. To finally prove the dimeric structure of complex 4 in solution, a DOSY-NMR22 study using the bipolar pulse longitudinal eddy current delay (BPPLED) pulse sequence23 was performed. This technique allows measuring diffusion constants of molecules in solution. In recent studies it was shown that the hydrodynamic radii of these compounds can be estimated from the diffusion constants via the Stokes-Einstein equation.24

D ) (kBT)/(CπηrH)

(1a)

The diffusion constants of tetramethylsilane (TMS) and complexes 3 and 4, as well as the corresponding hydrodynamic radii, are listed in Table 3. In all cases the diffusion constant of TMS was determined to be DTMS ) 2.51 × 10-9 m2 s-1. This shows that the viscosity of the solutions is not influenced by complexes 3 and 4. As the size of the observed samples is relatively close to the size of the solvent molecules, the frictional constant C was set to 4.24 The difference of the diffusion constants of complexes 3 and 4 reveals an important change in (21) (a) Hofmann, P.; Volland, M. A. O.; Hansen, S. M.; Eisentrager, F.; Gross, J. H.; Stengel, K. J. Organomet. Chem. 2000, 606, 88. (b) Zanini, M. L.; Meneghetti, M. R.; Ebeling, G.; Livotto, P. R.; Rominger, F.; Dupont, J. Inorg. Chim. Acta 2003, 350, 527. (22) (a) Pregosin, P. S.; Kumar, P. G. A.; Fernandez, I. Chem. ReV. 2005, 105, 2977. (b) Johnson, C. S. Prog. Nucl. Magn. Reson. Spectrosc. 1999, 34, 203.

In this study, an anionic potentially tetradentate ligand, based on an in-house-developed S∼P∼S ligand, was coordinated to a Pd(II) center. Several cationic complexes have been isolated and fully characterized. In particular, complex 4, unlike other related cationic “(S∼P∼SR)Pd” species, is kinetically stable and isolable. Catalytic tests in the known allylation of aldehydes were performed. The stabilization of the Pd center in complex 4 results in lower catalytic activity than in situ generated cationic complexes from complex 1 or 6, yet this complex is easy to handle and does not need to be activated. It is obvious from these results that the pyridine moiety does indeed bring stability to a cationic Pd center, without suppressing catalytic behavior, as expected. The precise nature of complex 4 was not accessible by usual NMR spectroscopy, which prompted us to carry out a combined DFT and tandem ESI-MS spectroscopic study. The DFT study, taking into account the solvation effects on a model complex, predicts a dimeric species, III, to be more stable than a monomeric species by ca. 8 kcal/mol, which would imply the sole presence of the dimer in solution. Free energy calculations on the other hand favor the monomeric form I by ca. 5 kcal/mol. Finally, the use of a recently introduced free energy GPCM calculation, adding solvation effects, predicted the three complexes to lie within 1.1 kcal/mol (monomer II most stable, and, therefore, that they would coexist in solution. A precise quantification of the isomers in solution was obtained by the FT-ICR technique: about 97.5% to 2.5% (dimer III to monomers I and/or II). This corresponds to a difference in energy of ca. 2 kcal/mol in favor of the dimer. Our study points to an overestimation of the entropic factors by DFT calculations and confirms the importance of carrying out PCM calculations for charged species. In the case presented here, the recently proposed free energy ∆GPCM calculation also overestimated the entropic factors and did not lead to the accurate prediction of the experimental system. Finally, our study clearly shows the interest of using combined experimental, in this case mass spectrometric techniques, and theoretical studies when the precise nature of complexes cannot be solved by usual NMR techniques or X-ray crystallography.

Experimental Section General Methods. All reactions were routinely performed under an inert atmosphere of argon or nitrogen using Schlenk and glovebox techniques and dry deoxygenated solvents. Dry THF, hexanes, and diethyl ether were obtained by distillation from Na/ benzophenone and dry CH2Cl2 from P2O5. CDCl3 was dried from P2O5 and stored on 4 Å Linde molecular sieves. CD2Cl2 was used as purchased and stored in the glovebox. Nuclear magnetic resonance spectra were recorded on a Bruker Avance 300 spectrometer operating at 300.0 MHz for 1H, 75.5 MHz for 13C, and 121.5 MHz for 31P. Solvent peaks are used as an internal reference relative to Me4Si for 1H and 13C chemical shifts (ppm); 31P chemical shifts are relative to a 85% H3PO4 external reference. Coupling constants are given in hertz. The following abbreviations are used: s, singlet; d, doublet; t, triplet; q, quadruplet; p, pentuplet; m, multiplet; v, virtual; b, broad.

2026 Organometallics, Vol. 28, No. 7, 2009

Blug et al.

Table 4. Experimental Details for X-ray Crystal Structures of 3, 5, and 6 cryst size (mm) empirical formula mol wt space group a (Å) b (Å) c (Å) R (deg) β (deg) γ (deg) V (Å3) Z calcd density (g cm-3) abs coeff (cm-1) 2F max (deg) F(000) index ranges no. of collected/indep reflns no. of reflns used Rint abs corr no. of params refined no. of reflns/params final R1a/wR2 (I > 2σ(I))b goodness of fit on F2 diff peak/hole (e Å-3) a

3

5

6

0.22 × 0.20 × 0.12 C50H40Cl10NP3PdS2 1272.76 P21/c 9.311(1) 40.894(3) 14.666(1) 90.00 103.870(1) 90.00 5421.5(8) 4 1.559 1.037 27.48 2560 -12 to 12; -53 to 48; -18 to 19 19 701/11 969 9526 0.0216 multiscan; 0.8040 min, 0.8857 max 627 15 0.0479/0.1457 1.050 1.391(0.095)/ -1.030(0.095)

0.20 × 0.20 × 0.16 C52H42N2P3PdS2,BF4 1283.85 P1j 12.787(1) 13.547(1) 17.359(1) 111.820(1) 100.400(1) 93.630(1) 2717.6(3) 2 1.569 0.855 30.03 1296 -17 to 17; -15 to 19; -24 to 24 23 605/15 786 12 722 0.0183 multiscan; 0.8475 min, 0.8753 max 586 21 0.0466/0.1527 1.115 1.607(0.093)/ -1.121(0.093)

0.23 × 0.16 × 0.05 C47H38ClNP3PdS2, 2(CH2Cl2),Cl 1205.89 P1j 9.618(1) 13.260(1) 22.200(1) 82.918(1) 89.160(1) 68.868(1) 2619.4(4) 2 1.529 0.970 27.48 1220 -12 to 11; -17 to 17; -28 to 28 21 984/11 290 8108 0.0280 multiscan; 0.8077 min, 0.9531 max 562 14 0.0479/0.1442 1.062 1.201(0.089)/ -0.918(0.089)

R1 ) ∑|Fo| - |Fc|/∑|Fo|. b wR2 ) (∑w|Fo| - |Fc|2/∑w|Fo|2)1/2.

DOSY experiments were performed using the bipolar pulse longitudinal eddy current delay (BPPLED) pulse sequence.23 In all cases the lock signal was adjusted to CD2Cl2. The duration of the magnetic field pulse gradients and the diffusion times were optimized for each sample in order to obtain an attenuation of 95% of the signal with the maximum gradient strength. In each PFG NMR experiment a series of 32 spectra on 8 K data points were acquired. The pulse gradients were incremented from 1 to 95% of the maximum gradient strength in a linear ramp. The temperature control unit and the tube spinning were turned off to prevent convection. After Fourier transformation and baseline correction, the diffusion dimension was processed with the Bruker TopSpin software. A hybrid 7 T Fourier transform ion cyclotron resonance (FTICR) mass spectrometer was used for the mass spectroscopic study. Highresolution mass spectra, recorded in the 200-2000 m/z range in positive mode, were obtained by Fourier transformation of a timedomain transient signal averaged over 16 scans. Sample solutions of the Pd complexes (10-5 M in CH2Cl2) were infused via syringe pump at a flow rate of 140 µL/h. Standard ESI conditions were used, with a capillary voltage of 4 kV and a heater temperature of 200 °C. Our FT-ICR instrument (Bruker Apex IV Qh) features a quadrupole mass filter followed by a hexapole cell pressurized at ∼10-3 mbar with Ar for collision-induced dissociation (CID) experiments. CID experiments were performed at a collision energy of 20 V after selection of precursor ions (dimer of Pd complexes, with a small fraction of monomer) with an isolation width of 10 Da. 2,6-Bis(diphenylphosphine sulfide)-3,5-diphenylphosphinine (SPS)25 and R-Li-picoline26 were prepared according to reported procedures. (23) (a) Wu, D. H.; Chen, A. D.; Johnson, C. S. J. Magn. Reson. A 1995, 115, 260. (b) Gibbs, S. J.; Johnson, C. S. J. Magn. Reson. 1991, 93, 395. (24) Macchioni, A.; Ciancaleoni, G.; Zuccaccia, C.; Zuccaccia, D. Chem. Soc. ReV. 2008, 37, 479. (25) (a) Doux, M.; Bouet, C.; Me´zailles, N.; Ricard, L.; Le Floch, P. Organometallics 2002, 21, 2785. (b) Avarvari, N.; LeFloch, P.; Ricard, L.; Mathey, F. Organometallics 1997, 16, 4089.

[SPS(CH2Py)PdCl], 3. A freshly prepared solution of R-Li-picoline (500 µL, 0.31 mmol) was added to a stirred solution of SPS (200 mg, 0.29 mmol) in THF (5 mL) at -78 °C. The solution turned immediately deep red and was allowed to return to room temperature during 15 min. [(COD)PdCl2] (83.6 mg, 0.29 mmol) was added as a solid, and the resulting brick-red solution was stirred for 2 h at room temperature. The solvent was evaporated, and the resulting solid was rinsed with hexanes (3 × 10 mL) and diethyl ether (2 × 10 mL). The residue was dissolved in CH2Cl2 and filtered through a pad of Celite. After evaporation of the solvent the title compound was obtained as a brick-red powder. Single crystals suitable for an X-ray crystal structure analysis were grown from a concentrated solution of CHCl3. Yield: 200 mg (0.215 mmol, 74%). Anal. Calcd for [C47H37ClNP3PdS2] (914.7): C 61.71, H 4.08. Found: C 61.86, H 3.89. 1H NMR (CDCl3): δ 3.83 (d, 2JHP(A) ) 12.5 Hz, 2H, CH2), 5.46 (t, 4JHP(B) ) 4.3 Hz, 1H, H4), 6.74-7.52 (m, 31H, CH of Ph and CH of Py), 7.71-7.83 (m, 2H, CH of Py), 8.53 (d, JHH ) 3.9 Hz, 1H, CH of Py). 13C NMR (CDCl3): δ 47.7 (d, 1JCP ) 22.7 Hz, CH2), 72.3 (ddd, 1JCP ) 92.9, 1JCP ) 51.5, 3JCP ) 3.1 Hz, C2,6), 119.3 (pseudo q, 3JCP ) 11.3 Hz, C4H), 121.9 (d, JCP ) 3.0 Hz, CH of Py), 126.8 (d, JCP ) 3.0 Hz, H of Py), 127.8-129.0 (m, CH of Ph), 131.5 (dd, 1JCP ) 83.8, 3JCP ) 9.1 Hz of PPh2), 132.3-132.5 (m, CH of Ph), 132.9 (m, C of PPh2), 132.9-133.2 (m, CH of Ph), 136.6 (d, JCP ) 2.3, CH of Py), 140.0 (dt, 2JCP ) 7.6 Hz, 4JCP ) 3.0 Hz, C3,5), 149.6 (d, JCP ) 2.3 Hz, CH of Py), 153.4 (d, 2JCP ) 7.6 Hz, C of Py), 159.6 (s, C of Ph). 31 P NMR (CDCl3): δ 48.4 (AB2, m, 2JP(A)P(B) ) 89.0 Hz, PBPh2), 51.9 (AB2, m, 2JP(A)P(B) ) 89.0 Hz, PA). [SPS(CH2Py)Pd]2(BF4), 4. To a solution of complex 3 (82 mg, 0.09 mmol) in CH2Cl2 (2 mL) was added solid AgBF4 (20 mg, 0.10 mmol). The resulting orange-red solution was stirred for 2 h at room temperature. AgCl was separated by centrifugation. After evaporation of the solvent the title compound was obtained as a red powder. Yield: 75 mg (0.08 mmol, 86%). Anal. Calcd for [C47H37BF4NP3PdS2] (966.1): C 58.43, H 3.86. Found: C 58.24, H 3.81. 1H NMR (CD2Cl2): δ 4.08 (dd, 2JHH ) 13.6 Hz, 2JHP(A) ) (26) Fraser, R. R.; Mansour, T. S.; Savard, S. Can. J. Chem.-ReV. Can. Chim. 1985, 63, 3505.

Cationic SPS Pincer Palladium Complexes 8.9 Hz, 1H, CH2), 4.39 (pseudo t, 2JHP(A) ) 2JHH ) 13.6 Hz, 1H, CH2), 6.08 (t, 4JHP(B) ) 4.7 Hz, 1H, H4), 6.24 (dd, 2JHP(A) ) 14.0 Hz, 2JHH ) 7.3 Hz, 2H, CH of Ph) 6.77-7.79 (m, 30H, CH of Ph and CH of Py), 7.99 (td, JHH ) 7.8 Hz, JHH ) 1.3 Hz, 1H, CH of Py), 8.43 (d, JHH ) 5.7 Hz, 1H, CH of Py). 13C NMR (CD2Cl2): δ 44.3 (d, 1JCP ) 20.0 Hz, CH2), 65.6 (m, C2 or 6), 72.1 (m, C2 or 6), 119.9 (pseudo q, 3JCP ) 10.6 Hz, C4H), 123.3 (s, CH of Py), 127.0-132.8 (m, CH of Py, CH and C of Ph), 136.7 (m, ∑J ) 12.8 Hz, C3 or 5), 137.3 (m, ∑J ) 13.6 Hz, C3 or 5), 138.6 (s, CH of Py), 151.0 (d, 2JCP ) 6.0 Hz, C of Py), 153.0 (s, CH of Py), 159.7 (s, C of Ph), 160.0 (s, C of Ph). 31P NMR (CD2Cl2): δ 45.4-49.8 (ABC, m, PACH2, PBPh2, PCPh2). [SPS(CH2Py)Pd(Pyr)]BF4, 5. Pyridine (10 µL, 0.12 mmol) was added to a solution of complex 4 (98 mg, 0.10 mmol) in CH2Cl2 (2 mL). After evaporation of the solvent the title compound was obtained as a red powder. Yield: 94.6 mg (0.98 mmol, 98%). Anal. Calcd for [C52H42BF4N2P3PdS2] (1045.18): C 59.76, H 4.05. Found: C 59.54, H 4.09. 1H NMR (CDCl3): δ 3.94 (d, 2JHP(A) ) 11.5 Hz, 2H, CH2), 5.58 (t, 4JHP(B) ) 4.4 Hz, 1H, H4), 6.82-8.74 (m, 29H, CH of Py and CH of Ph). 13C NMR(CDCl3): δ 48.5 (s, CH2), 71.1 (s, C2, C6), 119.6 (s, C4H), 122.7 (s, CH of R-picoline), 124.2 (s, CH of Py), 126.1 (s, CH of R-picoline), 128.1-129.3 (s, 5 × CH of Ph), 130.1 (s, C of PPh2), 131.2 (s, C of PPh2), 131.9-133.4 (s, 4 × CH of Ph), 136.5 (s, CH of Py), 138.7 (s, CH of R-picoline), 139.0 (s, C3, C5), 150.2 (s, CH of Py), 150.6 (s, CH of R-picoline), 151.1 (s, C of R-picoline), 160.9 (s, C of Ph). 31 P NMR (CDCl3): δ 45.7 (AB2, m, 2JP(B)P(A) ) 79.5 Hz, PBPh2), 47.6 (AB2, m, 2J(PB-PA) ) 79.5, PA). [SPS(CH2PyH)PdCl]Cl, 6. A 2 M solution of HCl in diethyl ether (120 µL, 240 mmol) was added to a solution of 3 (200 mg, 0.22 mmol) in CH2Cl2 (5 mL). The solution turned immediately from deep red to light orange. The title compound was obtained by evaporation of the solvents as an orange powder. Single crystals suitable for X-ray crystal structure analysis were grown by the diffusion of hexanes into a concentrated solution of the product in CH2Cl2. Yield: 200 mg (0.21 mmol, 95%). Anal. Calcd for [C47H38Cl2NP3PdS2] (951.19): C 59.35, H 4.03. Found: C 59.30, H 4.07. 1H NMR (CD2Cl2): δ 4.29 (d 2J(P-H) ) 12.7 Hz, 2H, CH2), 5.72 (t, 4JPH) 4.6 Hz), 1H, H4), 6.70-7.1 (m, 10H, CH of Ph), 7.20-7.65 (m, 20H, CH of PPh2) 7.69, (br, 1H, CH of Py), 8.31 (br, 1H, CH of Py) 8.60 (br, 1H, CH of Py) 8.68 (br, 1H, CH of Py). 13C NMR (CD2Cl2): δ 42.2 (CH2), 72.3 (C2,C6) 119.9 (s, C4), 124.5 (CH of Py), 127.1 (CH of Ph), 127.6 (CH of Ph), 128.2 (CH of Ph and PPh2), 130.2 (CH of Py), 131.1 (CH of PPh2), 131.9 (CH of PPh2), 138.7 (C3,C5), 144.1 (CH of Py), 148.5 (C of Py), 160.0 (C of Ph). 31P NMR (CD2Cl2): δ 48.2 (m, AB2, JP(A)P(B) ) 91.0 Hz), 51.9 (m, AB2, JP(A)P(B) ) 91.0 Hz). [SPS(CH2PyH)Pd(MeCN)](BF4)2, 8. AgBF4 (8.2 mg, 0.04 mmol) was added as a solid to a solution of 6 (20 mg, 0.02 mmol) in CH2Cl2 (2 mL). The solution turned immediately deep red. AgCl was eliminated by centrifugation. Acetonitrile (10 µL, 0.2 mmol) was added via a microsyringe. Evaporation of the solvents led to the partial re-formation of the desolvated product. The NMR spectrum of 7 was fully restored after the addition of 1 equiv of MeCN-d6 (∼1 µL). Yield: 16.6 mg (0.015 mmol, 75%). 1H NMR (CD2Cl2): δ 2.03 (br, 3H, MeCN), 4.20 (d, 2JPH ) 12.3 Hz, 2H, CH2), 5.93 (t, 4JPH ) 4.7 Hz, 1H, H-4), 6.80-7.70 (m, 30H, CH of

Organometallics, Vol. 28, No. 7, 2009 2027 Ph and PPh2), 8.15 (m, 2H, CH of Py) 8.79 (t, 3JHH ) 7.5 Hz, 1H, CH of Py), 8.96 (br, 1H, CH of Py). 13C NMR(CD2Cl2): δ 1.39 (hep, MeCN), 42.9 (d, 1JPC ) 17.9 Hz, CH2), 70.8 (dd, 1JPC ) 57.1 Hz, 1JPC ) 93.0 Hz, C2,C6), 118.0 (br, MeCN), 120.7 (pseudo-q, ∑J ) 11.7 Hz), 126.2 (s, CH of Py), 128.1 (CH of Ph), 128.6 (CH of Ph), 128.9 (CH of Ph), 129.1 (CH of Ph), 129.2 (CH of Ph), 129.3 (CH of Ph), 130.2 (d, 1JCP ) 4.1 Hz, C of PPh2), 130.3 (CH of Py), 132.4 (CH of Ph), 132.5 (CH of Ph), 132.6 (CH of Ph), 133.2 (CH of Ph), 133.4 (CH of Ph), 137.8 (dt, C3,C5), 142.4 (s, CH of Py), 147.2 (d, C of Py), 147.5 (s, CH of Py), 161.2 (s, C of Ph). 31P NMR (CDCl3): 44.7 (m, AB2, JP(A)P(B) ) 80.4 Hz), 47.7 (m, AB2, JP(A)P(B) ) 80.4 Hz). Palladium-Catalyzed Allylation of Aldehydes. Representative procedure for the allylation of p-bromobenzaldehyde with allyltributyltin in the presence of complex 1: A dried Schlenk was charged with p-bromobenzaldehyde (92.5 mg, 0.5 mmol), complex 1 (4.4 mg, 0.0050 mmol, 1.0 mol %), and THF (0.5 mL) under nitrogen. Allyltributyltin (186 µL 0.6 mmol) was then added via a syringe. The resulting solution was stirred at the desired temperature for 24 h. Thereafter the reaction mixture was evaporated, diluted into water, and extracted with ether. The organic layer was separated, washed with brine, dried (Na2SO4), and concentrated. The crude product was then purified by flash column chromatography (hexanes/EtOAc) to give the homoallylic alcohol as a colorless oil. The NMR data obtained for the coupling products are in agreement with the corresponding literature.23 X-ray Crystallographic Study. Data were collected at 150.0(1) K on a Nonius Kappa CCD diffractometer using a Mo KR (λ ) 0.71070 Å) X-ray source and a graphite monochromator. All data were measured using phi and omega scans. Experimental details are described in Table 4. The crystal structures were solved using SIR 9727 and Shelxl-97.28 ORTEP drawings were made using ORTEP III for Windows.29 CCDC 695195-695197 contain the supplementary crystallographic data for this paper. These data can be obtained free of charge at www.ccdc.cam.ac.uk/conts/retrieving.html or from the Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: (int.) +44-1223/336033; e-mail: [email protected].

Acknowledgment. This work was supported by the CNRS and the Ecole Polytechnique. The authors would like to thank IDRIS (project no. 091616) for the allowance of computer time. Supporting Information Available: CIF files for 3, 5, and 6, complete ref 11, and computed Cartesian coordinates, SCF energies, thermochemistry, PCM enrgies, and three lower frequencies of the theoretical structures. This material is available free of charge via the Internet at http://pubs.acs.org. OM800690T (27) Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G.; Giacovazzo, C.; Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.; Spagna, R. SIR97, an integrated package of computer programs for the solution and refinement of crystal structures using single crystal data. (28) Sheldrick, G. M. SHELXL-97; Universita¨t Go¨ttingen: Go¨ttingen, Germany, 1997. (29) Farrugia, L. J. ORTEP-3; University of Glasgow.