Synthesis and Properties of NHC-Supported Palladium(I) Dimers with

Aug 29, 2013 - ... and the solid-state geometries, electronic structures, reactivity, and .... David Balcells , Gary W. Brudvig , Wei Dai , Louise M. ...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/Organometallics

Synthesis and Properties of NHC-Supported Palladium(I) Dimers with Bridging Allyl, Cyclopentadienyl, and Indenyl Ligands Wei Dai, Matthew J. Chalkley, Gary W. Brudvig, Nilay Hazari,* Patrick R. Melvin, Ravi Pokhrel, and Michael K. Takase The Department of Chemistry, Yale University, P.O. Box 208107, New Haven, Connecticut 06520, United States S Supporting Information *

ABSTRACT: The synthesis of a family of new Pd(I) dimers, (μ-All)(μ-Cp){Pd(IPr)}2 (All = C3H5, Cp = C5H5, IPr = 1,3bis(2,6-diisopropylphenyl)-1,3-dihydro-2H-imidazol-2-ylidene), (μ-All)(μ-Ind){Pd(IPr)}2 (Ind = C7H9), (μ-Cp)(μCp){Pd(IPr)}2, (μ-Cp)(μ-Ind){Pd(IPr)}2, and (μ-Ind)(μInd){Pd(IPr)}2, which contain a combination of bridging allyl, Cp, and indenyl ligands and are all supported by IPr as the ancillary ligand, is reported. All of these compounds are thermally stable at room temperature, and the solid-state geometries, electronic structures, reactivity, and redox chemistry of these new compounds have been compared with those of the dimer (μ-All)2{Pd(IPr)}2, which was previously reported. This work provides further evidence that bridging allyl, Cp, and indenyl ligands bind in a similar manner to Pd(I). However, it is demonstrated that there are notable differences between the IPr-supported species and related Pd(I) dimers with triethylphosphine ancillary ligands, which have been previously described.



INTRODUCTION In recent years there has been considerable interest in the use of Pd(I) dimers with bridging allyl ligands as both precatalysts and catalysts for cross-coupling and small-molecule activation.1 For example, it has been demonstrated that Pd(I) dimers with bridging allyl ligands are highly active precatalysts for Suzuki− Miyaura2 and Sonogashira couplings,3 Buchwald−Hartwig amination,4 α-arylation of carbonyls,4 and coupling of organosilanols with aryl halides.5 Furthermore, Pd(I) dimers can act as catalysts for the carboxylation of allylstannanes and allylboranes with CO26 and have been implicated as intermediates in allylic substitution reactions.7 In general, Pd(I) dimers with bridging allyl ligands can be divided into two classes. The first class consists of Pd(I) dimers with one bridging allyl ligand and one bridging halide, carboxylate, or thiol ligand.2,8 These species do not react with electrophiles such as HCl and CO2.6b The second class consists of species with two bridging allyl ligands (AllAll).8a−c,9 These complexes generally undergo facile reactions with HCl or CO2.6b In species with one bridging allyl and one bridging halide, carboxylate, or thiol ligand, the HOMO is localized on the Pd atoms, whereas in species with two bridging allyl ligands the HOMO is localized on the terminal carbon atoms of the bridging allyl ligands.6b,8o,10 It is proposed that this difference in the nature of the HOMO is responsible for some of the differences in reactivity between the two systems.6b,10 Another reason for the difference in reactivity is that, in systems which contain two bridging allyl ligands, two strongly donating allyl ligands are opposite each other.6b This weakens the Pd−allyl bonds and makes direct reactions at the bridging allyl ligand more likely. Given our recent work © 2013 American Chemical Society

investigating the catalytic conversion of CO2 to more valuable products,6a,11 we were interested in exploring the chemistry of species with two bridging allyl or related ligands. Cyclopentadienyl (Cp) and indenyl ligands are often considered to be analogous to allyl ligands,12 and a number of Pd(I) dimers with two bridging Cp and indenyl ligands have been prepared. For example, complexes with two bridging Cp ligands have been synthesized with ancillary phosphine ligands (CpCp),8d,13 while species with two bridging indenyl ligands have been prepared with supporting isocyanide ligands (IndInd).14 In general, complexes containing two different bridging ligands are less common. Although Werner synthesized a number of different complexes containing one bridging allyl and one bridging Cp ligand (AllCp),8d,15 for a long time complexes containing one bridging allyl and one bridging indenyl ligand (AllInd) were unknown. Similarly, species with one bridging Cp and one bridging indenyl ligand (CpInd) were also unknown. The full set of compounds that can be formed with only bridging allyl, Cp, and indenyl ligands is shown in Figure 1.16 It should be noted that there are a small number of examples of bridging Cp17 and indenyl17a,c,18 ligands binding to metals other than Pd; however, the vast majority of examples involve Pd(I) species. Recently we described the synthesis of AllCp, AllInd, and CpInd supported by ancillary triethylphosphine ligands19 and compared their solid-state and electronic structures to those of AllAll and CpCp, which had previously been prepared with Received: July 13, 2013 Published: August 29, 2013 5114

dx.doi.org/10.1021/om400687m | Organometallics 2013, 32, 5114−5127

Organometallics

Article

well as those of AllAll, which we previously prepared,6a are compared. This is the first time that all of the types of complexes described in Figure 1 have been synthesized with the same ancillary ligand.



RESULTS AND DISCUSSION Synthesis. The preparation of AllAll21 has already been described by our group, and the complex has been characterized by X-ray crystallography.6a As a result, our aim was to synthesize and crystallographically characterize the new complexes AllCp, AllInd, CpCp, CpInd, and IndInd. This would allow for a full comparison of complexes with different bridging ligands but the same ancillary ligand. Although we had prepared many of the analogous compounds with triethylphosphine as the supporting ligand,19 we found that these synthetic methods were not always translatable to the IPr ligand and new synthetic pathways had to be devised for many of the IPrsupported compounds. Synthesis of AllCp. Previously, it had been demonstrated that the reaction of (η5-Cp)(η3-All)Pd with triethylphosphine generated the triethylphosphine-supported AllCp dimer, via an intermediate of the type (η1-Cp)(η3-All)Pd(PEt3), which was characterized by NMR spectroscopy.19 The mechanism of dimer formation from (η1-Cp)(η3-All)Pd(PEt3) is proposed to involve reductive elimination of 5-allylcyclopenta-1,3-diene (or a related isomer) to generate Pd(PEt3), which is rapidly trapped by another equivalent of (η1-Cp)(η3-All)Pd(PEt3) to form the dimer.9d,19,22 An analogous series of reactions was used to prepare AllCp supported by IPr in 50% yield, as a thermally stable green solid, which could be stored indefinitely at room temperature (Scheme 1). Even at low temperature only one peak was observed in the 1H and 13C NMR spectra for the bridging Cp ligand in AllCp, indicating that the barrier to rotation of the Cp ligand is small. Similar NMR properties have been observed in monomeric systems with Cp ligands23 and in dimeric complexes with bridging Cp ligands.17a,19,22 The reaction of (η5-Cp)(η3-All)Pd with IPr led directly to the formation of AllCp, with no monomeric intermediate observed even at low temperature, although the organic byproduct 5allylcyclopenta-1,3-diene (and related isomers) was observed. Our inability to detect the intermediate (η1-Cp)(η3-All)Pd(IPr) was unexpected, as the related complex (η1-Cp)(η3-2methylallyl)Pd(IPr), which was stable at room temperature, has been isolated.24 The addition of the inductively electron donating methyl group presumably stabilizes the Pd(II)

Figure 1. Generic pictures of bridging allyl, Cp, and indenyl dimers.

triethylphopshine ligands.6a,8e The synthesis of a range of complexes with the same supporting ligand allowed us to confirm preliminary spectroscopic, crystallographic, and theoretical studies from Werner, which indicated that Pd(I) dimers with bridging Cp or indenyl ligands are similar to species with bridging allyl ligands.20 In all cases the bridging ligands bind through three carbon atoms to the two Pd atoms, with only the central carbon atom of the bridging group bound to both metal centers. Furthermore, a related combination of the π orbitals of the bridging allyl, Cp, or indenyl ligand is responsible for the predominant bonding interaction with the metal centers. However, there are also differences in the binding of bridging allyl, Cp, and indenyl ligands. In general there is less backbonding from the Pd to a bridging Cp ligand, in comparison to a bridging allyl or indenyl ligand, and the nature of the HOMO in complexes with bridging Cp ligands is different from that in species with bridging allyl or indenyl ligands. Although we were able to perform an extensive study using the triethylphosphine-supported complexes, we encountered two major problems: (i) All of the triethylphosphine-supported complexes except for AllAll and CpCp were thermally unstable at room temperature, which made it difficult to perform some types of spectroscopy, as well as reactivity studies. (ii) Despite repeated attempts, we were unable to synthesize IndInd supported by triethylphosphine ligands and as a result could not complete the series shown in Figure 1, with the same ancillary ligand. Here, we report the synthesis and X-ray structures of the thermally stable series of compounds AllCp, AllInd, CpInd, CpCp, and IndInd supported by the IPr ligand (IPr = 1,3-bis(2,6-diisopropylphenyl)-1,3-dihydro-2H-imidazol2-ylidene). The structures and reactivities of these species as Scheme 1. Synthesis of AllCp

5115

dx.doi.org/10.1021/om400687m | Organometallics 2013, 32, 5114−5127

Organometallics

Article

Scheme 2. Synthesis of AllInd

Scheme 3. Synthesis of IndInd

ligand, were observed at δ 0.09 ppm. We believe that the unusually low chemical shift for these protons is due to the ring current effect,27 as the solid-state structure of AllInd indicates one of the methyl groups is located under the phenyl ring of the bridging indenyl ligand (see the Supporting Information). Synthesis of IndInd. The complex IndInd was formed using a route similar to that described for AllInd (Scheme 3). Initially, (η3-Ind)Pd(IPr)Cl was formed through the reaction of IPr with {(η3-Ind)Pd(μ-Cl)}2.24 Subsequent treatment of (η3Ind)Pd(IPr)Cl with LiInd generated IndInd in 85% yield. This is the first example of a Pd(I) species with two bridging indenyl ligands, which does not have supporting isocyanide ligands.14 In this case the intermediate (η3-Ind)(η1-Ind)Pd(IPr), which we have previously characterized,24 was observed. This suggests that reductive elimination of biindene (or a related isomer) is more difficult than reductive elimination of either 5allylcyclopenta-1,3-diene or allylindene. This is consistent with previous results which suggest that reductive elimination of dicyclopentadienyl or biindene is difficult.19 As with AllInd, some of the methyl protons of the IPr ligand are shifted downfield in the 1H NMR spectrum due to the ring current effect.27 In fact, this phenomenon appears to be general to all complexes with bridging indenyl ligands studied in this work. Synthesis of CpInd and CpCp. Previously, we have demonstrated that the complexes (η5-Cp)(η1-Ind)Pd(IPr) and (η5-Cp)(η1-Cp)Pd(IPr) are relatively stable and do not rapidly

monomer and slows down reductive elimination to Pd(0). Furthermore, in this case the initial coordination of IPr to (η5Cp)(η3-All)Pd was slow, requiring 12 h at room temperature, whereas for triethylphosphine the reaction was complete in less than 1 min. This is most likely related to the greater steric bulk of IPr in comparison to triethylphosphine. Synthesis of AllInd. A synthetic route related to that described for AllCp was used for the preparation of AllInd (Scheme 2). However, in this case the order of the steps was reversed and the IPr ligand was installed prior to the last step, due to the thermal instability of (η5-Ind)(η3-All)Pd.25 Treatment of {(η3-All)Pd(μ-Cl)}2 with IPr resulted in the formation of the known air- and moisture-stable compound (η3All)Pd(IPr)Cl.26 Subsequent reaction of (η3-All)Pd(IPr)Cl with LiInd generated thermally stable AllInd in 78% yield, only the second example of a species with one bridging allyl and one bridging indenyl ligand.19 In a fashion similar to the synthesis of AllCp, the monomeric complex (η1-Ind)(η3All)Pd(IPr) is proposed to be an intermediate, although this species was not observed. The related complex (η1-Ind)(η3-2methylallyl)Pd(IPr) has been isolated; however, this species was only stable for short periods in solution.24 Our inability to observe (η1-Ind)(η3-All)Pd(IPr) is again consistent with the 2methyl substituent slowing down the rate of reductive elimination. In the 1H NMR spectrum of AllInd, six protons, which are assigned as two of the methyl groups of the IPr 5116

dx.doi.org/10.1021/om400687m | Organometallics 2013, 32, 5114−5127

Organometallics

Article

Scheme 4. Synthesis of CpInd

Figure 2. (a) ORTEP28 of AllCp at the 30% probability level (Pd atoms pink, N atoms blue, and C atoms gray). Hydrogen atoms and isopropyl groups of IPr are omitted for clarity. (b) Side-on view looking down the Pd−Pd bond of AllCp, showing only the Pd atoms and the bridging ligands.

Figure 3. (a) ORTEP28 of AllInd at the 30% probability level (Pd atoms pink, N atoms blue, and C atoms gray). Hydrogen atoms and isopropyl groups of IPr are omitted for clarity. (b) Side-on view looking down the Pd−Pd bond of AllInd, showing only the Pd atoms and the bridging ligands.

undergo reductive elimination.24 Therefore, our strategy for the synthesis of CpInd and CpCp involved using a sacrificial complex to generate a monoligated Pd(0) species which could be trapped by (η5-Cp)(η1-Ind)Pd(IPr) or (η5-Cp)(η1-Cp)Pd(IPr) to form the appropriate dimer (Scheme 4). In both cases, the complex (η3-Ind)(η1-Ind)Pd(IPr) was generated in situ

through the reaction of (η3-Ind)Pd(IPr)Cl with LiInd. Subsequent addition of (η5-Cp)(η1-Ind)Pd(IPr) or (η5-Cp)(η1-Cp)Pd(IPr) to the reaction mixture resulted in the formation of CpInd or CpCp, respectively, along with the organic byproduct biindene (or a related isomer). In the case of CpCp this synthetic route was extremely successful, with the 5117

dx.doi.org/10.1021/om400687m | Organometallics 2013, 32, 5114−5127

Organometallics

Article

Figure 4. (a) ORTEP28 of CpInd at the 30% probability level (Pd atoms pink, N atoms blue, and C atoms gray). Hydrogen atoms and isopropyl groups of IPr are omitted for clarity. (b) Side-on view looking down the Pd−Pd bond of CpInd, showing only the Pd atoms and the bridging ligands.

Figure 5. (a) ORTEP28 of CpCp at the 30% probability level (Pd atoms pink, N atoms blue, and C atoms gray). Hydrogen atoms and isopropyl groups of IPr are omitted for clarity. (b) Side-on view looking down the Pd−Pd bond of CpCp, showing only the Pd atoms and the bridging ligands.

Figure 6. (a) ORTEP28 of IndInd at the 30% probability level (Pd atoms pink, N atoms blue, and C atoms gray). Hydrogen atoms and isopropyl groups of IPr are omitted for clarity. (b) Side-on view looking down the Pd−Pd bond of IndInd, showing only the Pd atoms and the bridging ligands.

dimer being generated in 71% yield. The IndInd dimer was only observed in small quantities and could be separated by washing with pentane. This suggests that trapping of Pd(0) by (η5-Cp)(η1-Cp)Pd(IPr) is more efficient than trapping by (η3Ind)(η1-Ind)Pd(IPr). In the case of CpInd, both the CpCp and IndInd dimers were also formed, suggesting that the trapping rates of Pd(0) by (η5-Cp)(η1-Ind)Pd(IPr) and (η3-Ind)(η1Ind)Pd(IPr) are comparable and also that some ligand rearrangement is possible to generate the CpCp dimer.

Consistent with the hypothesis that ligand exchange is possible, benzene solutions of CpInd slowly formed CpCp and IndInd, along with some decomposition products at room temperature. Unfortunately, although we were able to crystallize the CpInd dimer, we were unable to generate a sample which did not contain small amounts of either CpCp or IndInd. Furthermore, for unknown reasons the ratio of CpCp and IndInd to CpInd was not reproducible between reactions. Spectroscopically, CpInd had many features in common with the other dimers; 5118

dx.doi.org/10.1021/om400687m | Organometallics 2013, 32, 5114−5127

Organometallics

Article

Table 1. Comparison of Pd−C and Pd−Pd Bond Lengths in AllAll,b AllCp, AllInd, CpCp, CpInd, and IndInda complex

bridging group

Pd(1)−Cterminal

Pd(2)−Cterminal

Pd(1)−Ccentral

Pd(2)−Ccentral

Pd−Pd

AllAllb,c AllCp

All All Cp All Ind Cp Cp Cp Ind Ind Ind

2.104(5) 2.037(13) 2.277(14) 2.128(4) 2.180(4) 2.1532(18) 2.1651(17) 2.04(2) 2.312(15) 2.1785(16) 2.1637(17)

2.147(5) 2.067(15) 2.258(24) 2.091(4) 2.226(4) 2.1882(18) 2.1651(17) 2.021(15) 2.336(19) 2.1785(16) 2.1637(17)

2.623(4) 2.517(19) 2.632(14) 2.424(4) 2.522(4) 2.618(2) 2.5086(17) 2.439(15) 2.621(12) 2.4451(17) 2.4681(17)

2.390(5) 2.407(16) 2.733(18) 2.580(4) 2.398(3) 2.4734(19) 2.5304(18) 2.490(8) 2.548(12) 2.4451(17) 2.4681(17)

2.72002(15) 2.7050(3)

AllInd CpCpc,d CpInd IndInde

2.7319(3) 2.6808(3) 2.6839(3) 2.6861(6) 2.6889(2)

a

All bond lengths are given in Å. bData taken from ref 6a. cThese stuctures contain an inversion center. dThere are two independent molecules present in the unit cell. eThis structure contains a mirror plane.

however, in the room-temperature 1H NMR spectrum the resonances associated with the isopropyl groups of the IPr ligand were broad, presumably due to restricted rotation caused by the increased steric bulk. At low temperature a well-resolved spectrum, which had features similar to those of the other dimers, was obtained. Solid-State Structures. As part of this work the complexes AllCp, AllInd, CpInd, CpCp, and IndInd were all characterized by X-ray crystallography. ORTEP drawings are shown in Figures 2−6, respectively, while important geometrical bond lengths are summarized in Table 1 and bond angles and dihedral angles are shown in Table 2. Data from the solid-state stucture of AllAll, which has previously been reported,19 are included in Tables 1 and 2 to allow for comparison with the new structures.

possibility of any bonding interaction. The C−C bond lengths in the bridging Cp ligand or the five-membered ring of the bridging indenyl ligand are similar to those found in monomeric η3-Cp and η3-indenyl ligands, respectively.29 For the bridging Cp ligands there are two long C−C bonds, two intermediate bonds, and one short bond, while for the fivemembered ring of the bridging indenyl ligands there are three intermediate bonds and two long bonds. However, in all cases the fold angles30 of the bridging Cp and indenyl ligands are significantly less than 10° and the bridging Cp and indenyl ligands are essentially planar. In contrast, in monomeric systems with η3-Cp or η3-indenyl ligands, the fold angles are generally greater than 10°.29 This provides further support for the hypothesis that bridging Cp or indenyl ligands are structurally distinct from η3-Cp or η3-indenyl ligands in monomeric complexes,17c,19 despite the fact that binding occurs through three carbon atoms to the metal centers in both cases. Previously, it has been proposed that folding occurs in monomeric systems due to the population of a metal−ligand antibonding orbital, whereas in the coordinatively unsaturated 16-electron dimers, this orbital is not populated.19 DFT calculations (vide infra) on the dimers studied in this work are consistent with this explanation. One of the main differences between the structural data for the triethylphosphine-supported dimers we reported previously19 and the IPr-supported complexes described here is that in this case there are no clear trends relating to the donor power of the bridging ligands. For triethylphosphine-supported species, a comparison of the respective bond lengths from the Pd centers to both the terminal carbons and the central carbons of the bridging ligands in AllCp, AllInd, and CpInd revealed that the Pd−bridging allyl bond distances were the shortest, followed by the Pd−bridging indenyl and Pd−bridging Cp bond distances.19 This suggested that the bridging allyl ligand binds the most tightly, followed by the bridging indenyl ligand and then the bridging Cp ligand. However, this trend is not clearly apparent in the IPr-supported compounds. Although the Pd−C bond lengths to the bridging allyl ligand in AllCp are considerably shorter than those in AllInd, suggesting that the bridging indenyl ligand is a more powerful donor than the bridging Cp ligand, the Pd−C bond lengths to the bridging Cp in CpInd are significantly shorter than the Pd−C bond lengths to the bridging indenyl ligand. Most likely, steric factors become more important in CpInd and cause the elongation in the Pd−C bond lengths to the bridging indenyl ligand;

Table 2. Experimental NHC−Pd−Pd Bond Angles and the Dihedral Angles Formed between the Allyl, Cp, or Indenyl Plane and the Plane Containing the Two Pd Centers and the Two Terminal C Atoms of the Bridging Allyl, Cp, or Indenyl Ligand in AllAll,a AllCp, AllInd, CpCp, CpInd, and IndIndb

complex a,c

AllAll AllCp AllInd CpCpc,d CpInd IndInde

NHC−Pd−Pd bond angle

θ

162.89(10) 175.87(5) 165.90(10), 163.49(10) 173.77(4), 178.94(4) 177.09(13) 173.93(5)

77.85 83.37 (allyl), 82.14 (Cp) 76.52 (allyl), 60.53 (indenyl) 78.02, 74.63 74.63 (Cp), 62.48 (indenyl) 65.35, 65.35

a Data taken from ref 6a. bθ is defined as shown in the diagram below the table title. All angles are in deg. cThese structures contain an inversion center. dThere are two independent molecules present in the unit cell. eThis structure contains a mirror plane.

In all of the structures described in this work the bridging ligands are coordinated in the manner previously described for bridging allyl, Cp, and indenyl ligands, with the central carbon bound to each Pd and the two adjacent carbons bound only to the nearest Pd.19,22 The long Pd−C distances (almost 3 Å) to the olefinic carbons in the bridging Cp ligands and the ringjunction carbons in the bridging indenyl ligands exclude the 5119

dx.doi.org/10.1021/om400687m | Organometallics 2013, 32, 5114−5127

Organometallics

Article

triethylphosphine ligand, the preference for a syn geometry from the bridging allyl ligand is greater than the preference for a anti geometry from the bridging Cp ligand and the molecule has a syn geometry. In contrast, for CpInd supported by a triethylphosphine ligand, the preference for an anti geometry from the bridging Cp ligand is greater than the preference for a syn geometry from the bridging indenyl ligand and the molecule has a anti geometry. The reverse trend is observed for IPrsupported species, and AllCp adopts an anti geometry (the bridging Cp ligand outcompetes the bridging allyl ligand), whereas CpInd adopts a syn geometry (the bridging indenyl ligand outcompetes the bridging Cp ligand). The exact reasons for this change in preference between IPr-supported species and triethylphosphine-supported species is unclear. The only exception to these observations is IndInd, which adopts an anti geometry, but the bridging indenyl ligand is the most sterically demanding. This is presumably responsible for the change from the anticipated syn geometry. It should be noted that the only other reported crystal structures of Pd(I) dimers with two indenyl ligands feature relatively sterically undemanding bridging isocyanide ligands and have a syn geometry.14 Similarly, calculations on (μ-indenyl)Pd2(PMe3)2 indicate that a syn geometry is preferred.19 Overall, the results of both this and previous studies indicate that, although broad trends are present in whether a bridging ligand prefers to adopt a syn or anti orientation, the difference in energy between the isomers is sufficiently small that the thermodynamically preferred orientation will depend on the nature of the ancillary ligand.6b,19,22 Given that the thermodynamic difference between the syn and anti isomers appears to be small, it is surprising that two isomers have never been conclusively observed either crystallographically or spectroscopically for the same compound. In systems with bridging Cp ligands, where only one resonance is observed by NMR spectroscopy for the bridging Cp ligand, it is clear that there is a low-energy pathway for interconversion of syn and anti isomers if they are in equilibrium. In systems with bridging allyl or indenyl ligands, it is less obvious that a low-energy pathway for interconversion exists and we believe that it is more likely that two isomers will be observed for the same compound in these cases. Electronic Structures of Dimers and Redox Properties. In order to probe the electronic structures of the dimers, DFT calculations were performed. All calculations used the TPSSH functional, which we have previously shown is able to accurately model the geometry of Pd(I) dimers,19 and featured the simplified NHC ligand 1,3-bis(dimethyl)-1,3-dihydro-2Himidazol-2-ylidene, instead of IPr, for computational simplicity. The calculated complexes have a prime after the compound name (for example AllAll′) to indicate that a modified ancillary ligand was used. Although we were unable to accurately predict the preferred geometry, syn or anti, for the complexes, the bond lengths and angles showed good agreement with the experimental structures. Presumably, the geometries could not be accurately modeled due to the use of the simplified ligand, which does not capture the full steric bulk of the experimental system. For the purposes of our discussion we have only considered the experimentally observed isomer for each dimer, rather than the lowest energy structure. The HOMOs of the six dimers studied in this work are shown in Figure 7. In general, the NHC-supported species have electronic structures similar to those of the corresponding triethylphos-

however, it is not possible to differentiate between electronic and steric effects from the X-ray structure. An observation that is consistent with the hypothesis that steric factors affect the bonding of dimers supported by the IPr ligands is the variation in the Pd−NHC bond lengths between the dimers. The shortest bond length is in AllAll (Pd−Ccarbene = 2.029(3) Å), while the next shortest Pd−NHC bond length is in AllInd and is approximately 0.02 Å longer (Pd−Ccarbene = 2.052(4) Å). The longest Pd−NHC bond length is in IndInd (Pd−Ccarbene = 2.0609(16) Å), where the bridging ligand combination is the most sterically demanding. In addition, comparison of the dihedral angle formed between the allyl, Cp, or indenyl plane and the plane containing the two Pd centers and the two terminal carbon atoms of the bridging allyl, Cp, or indenyl ligand (θ in the diagram in Table 2), reveals acute angles of approximately 65° for bridging indenyl ligands. The dihedral angle is between 75 and 85° for IPr dimers with less sterically demanding bridging allyl and Cp ligands. The observation that the dihedral angle is less than 90° in all cases suggests that there is back-bonding from filled orbitals on Pd to the bridging ligand in all systems.6b,8o,p,19 One trend that was noted in triethylphosphine-supported systems is that, when less donating bridging ligands are present, the Pd−Pd bond length becomes shorter.19 Assuming our results from the triethylphosphine-supported systems on the donor power of the different bridging ligands are translatable to the IPr system, a similar trend is apparent for the IPr-supported species. The longest Pd−Pd bond lengths are observed in AllInd and AllAll and the shortest in CpCp. In all cases the Pd−Pd bond lengths are longer for the IPr-supported species in comparison to the triethylphosphine-supported complexes. In general, it appears that both IPr and triethylphosphine dimers have similar preferences for the orientation of the bridging ligands with respect to each other, although it is harder to draw conclusions on the basis of the IPr data (Table 3).19 In Table 3. Comparison of the Preferred Orientation of the Bridging Ligands in IPr- and Triethylphosphine-Supported Dimers bridging group orientation complex

IPr

PEt3

AllAll AllCp AllInd CpCp CpInd IndInd

syna anti syn anti syn anti

syna synb synb antib antib ndc

a

Data taken from ref 6a. bData taken from ref 19. cNot experimentally determined.

both cases, complexes with bridging allyl and indenyl ligands prefer a syn geometry, in which the central carbon atoms of the bridging ligands are eclipsed and are located on the same face of the molecule. For example, AllAll and AllInd adopt syn geometries. In contrast, complexes with bridging Cp ligands prefer an anti geometry, in which the central carbon atoms are on opposite faces of the molecules. For example, CpCp adopts an anti geometry. However, whereas in triethylphosphinesupported species a bridging allyl ligand has a stronger preference for a syn geometry than does a bridging indenyl ligand, it appears that the opposite is true for IPr-supported species. Therefore, in the case of AllCp supported by a 5120

dx.doi.org/10.1021/om400687m | Organometallics 2013, 32, 5114−5127

Organometallics

Article

Figure 7. Calculated HOMOs of the (a) AllAll′, (b) AllCp′, (c) AllInd′, (d) CpCp′, (e) CpInd′, and (f) IndInd′ dimers studied in this work.

phine-supported dimers.19 We have previously shown that changing the orientation of the bridging ligands of a dimer from syn to anti does not significantly affect the nature or energies of the molecular orbitals.6b This is supported by the fact that, even for NHC dimers that have the orientation of the bridging ligands opposite to that of the corresponding triethylphosphine-supported dimers, the molecular orbitals are analogous.19 Complexes which only contain bridging allyl or indenyl ligands have a ligand-centered HOMO, whereas species which contain a bridging Cp ligand have a greater contribution from the Pd centers to the HOMO. In species with a bridging Cp ligand, the interaction between the bridging Cp ligand and the metal centers in the HOMO is slightly antibonding, which raises the energy of the HOMO relative to species with bridging indenyl ligands. The HOMO in species with bridging indenyl ligands is stabilized due to delocalization of the electron density on the outermost carbon atoms bound to Pd with the aromatic sixmembered ring of the indenyl ligands. This stabilization cannot occur for bridging allyl ligands, and as a result the HOMO in AllAll′ is at an energy similar to that of CpCp′. We have previously established that the first step in the reaction of species with two bridging allyl ligands with CO2 is nucleophilic attack by one of the terminal carbons of the bridging allyl ligand on electrophilic CO2.6b The calculated HOMOs of the dimers suggest that this reaction will be the most facile for AllAll, as this is the dimer in which the ligand makes the largest contribution to the HOMO, suggesting that the ligand is the most nucleophilic in this system. The calculated HOMOs suggest that the dimers could undergo relatively facile oxidation, as an electron would be removed from either a nonbonding orbital or a metal−Cp

antibonding orbital. The cyclic voltammograms (CVs) of the complete series of IPr-supported species, including AllAll but excluding CpInd (due to the impurities present in the sample), were recorded and are shown in Figure 8. For all species, no reduction waves were observed within the solvent window. Given that the LUMOs of the dimers are mainly Pd−Pd antibonding in character, this is not surprising. In contrast, all dimers underwent initial oxidation at potentials between −0.8 and −0.2 V versus ferrocene (Table 4). In the case of AllInd and IndInd the oxidations were irreversible. A second irreversible oxidation was also observed for IndInd. Consistent with our calculations, it was more difficult to oxidize AllInd and IndInd than to oxidize AllAll, CpCp, and AllCp, which all showed quasi-reversible oxidations, at similar potentials. A second reversible oxidation was observed for CpCp at 0.10 V versus ferrocene. From the electrochemical data alone it is unclear whether the initial oxidation is ligand centered or metal centered. The nature of the oxidized species was probed using EPR spectroscopy. In these experiments, AllAll, CpCp, and AllCp, which all showed quasi-reversible oxidations, were exposed to [Fe(η5-C5H5)2][PF6] at low temperature. The EPR spectra of AllAll+ and CpCp+ are shown in Figure 9, with the experimental and calculated g values given in Table 5. Both spectra are axial, with gx = gy < gz for AllAll+ and gx = gy > gz for CpCp+. The oxidized products are extremely unstable and rapidly decompose in minutes at −60 °C (see the Supporting Information). In the case of AllAll+ a small amount of the decomposition product was even present in the spectrum recorded immediately after mixing at −60 °C (see the Supporting Information). Furthermore, for AllCp, no EPR 5121

dx.doi.org/10.1021/om400687m | Organometallics 2013, 32, 5114−5127

Organometallics

Article

Figure 8. Partial CVs of (a) AllAll, (b) AllCp, (c) AllInd, (d) CpCp, and (e) IndInd in a 0.10 M nBu4NPF6 solution of dichloromethane under N2 at room temperature. The working electrode was a 2.0 mm diameter platinum disk, the reference and counter electrodes were 0.8 mm platinum wires, and the scan rate was 0.50 V s−1.

the Pd centers and the bridging allyl ligand. The Pd−Pd bond length in the calculated structure of CpCp′+ (see the Supporting Information) is notably shorter than in CpCp′ (2.58 versus 2.68 Å). However, the most striking difference in the calculated structure of CpCp′+ in comparison to CpCp′ is the binding of the Cp ligands. In CpCp′+ the bridging Cp ligands are bound through four carbon atoms, and all of the C− C bond lengths within the Cp ligand are almost equivalent. As the bridging Cp ligands bind through only four carbon atoms, there is significant radical character on the fifth carbon. This is reflected in the SOMO of CpCp′+ (Figure 10), which also has contributions from a metal−Cp antibonding orbital. The SOMO of AllAll′ is also predominantly ligand-centered, although there is obviously some contribution from the metal

spectrum was obtained for an oxidized product and it is likely that the oxidized product was too unstable to be observed. The g values for AllAll+ and CpCp+ suggest that there is some metal character in the SOMO. DFT calculations were performed on AllAll′+ and CpCp′+ to assist in understanding the CVs and EPR spectra. There is good agreement between the calculated g values for AllAll′+ and CpCp′+ and those experimentally observed for AllAll+ and CpCp+, providing support for our computational model (Table 5). The optimized structure of AllAll′+ (see the Supporting Information) is similar to that observed for AllAll′, with the most significant difference being a contraction in the calculated Pd−Pd bond length from 2.76 Å in AllAll′ to 2.67 Å in AllAll′+. There is also a slight contraction in the bond lengths between 5122

dx.doi.org/10.1021/om400687m | Organometallics 2013, 32, 5114−5127

Organometallics

Article

Table 5. Experimental and Calculated g Values for AllAll+ and CpCp+

Table 4. Summary of Oxidation Potentials of IPr-Supported Pd Dimers in a 0.10 M nBu4NPF6 Solution of Dichloromethane under N2 at Room Temperaturea AllAll

AllCp

AllInd

oxidation potential, Vb IC/IAc

−0.72

−0.68

−0.41

ΔE (EA − EC), mVd

74

0.783

0.572 79

CpCp −0.79, 0.10 0.866, 0.934 127

IndInd

AllAll+ CpCp+

−0.37, −0.22

gx,y

gz

1.996 (gx = 2.019, gy = 2.026)a 2.064 (gx = 2.054, gy = 2.070)

2.345 (gz = 2.464) 2.046 (gz = 2.076)

a

Numbers in parentheses are calculated values for AllAll′+ and CpCp′+ from DFT.

79

contributions from both the bridging allyl and indenyl ligands, the protonation is not selective and a mixture of AllCl or IndCl is formed in approximately a 7:3 ratio. Protonation of CpCp or IndInd with 1 equiv of 2,6-lutidinium chloride results in the generation of (μ-Cp)(μ-Cl){Pd(IPr)}2 (CpCl) or IndCl, respectively, with no further reaction observed when additional equivalents of acid are added. This is consistent with previous results on the protonation of AllAll.6 Overall, our results suggest that the nature of the HOMO is a good indicator of the site of protonation.

a

The working electrode was a 2.0 mm diameter platinum disk, the reference and counter electrodes were 0.8 mm platinum wires, and the scan rate was 0.50 V s−1. bVersus the ferrocenium/ferrocene couple. c Ratio of peak cathodic current to peak anodic current. dPeak separation between the anodic and cathodic waves.

centers given the gz value (Figure 10). The calculated spin density populations on the atoms in the dimers are also consistent with a ligand-centered SOMO (see the Supporting Information). On this basis we believe that the first oxidations in the CVs of the dimers are primarily ligand-centered with some metal character. In the case of CpCp and IndInd, where a second oxidation is observed, this oxidation is most likely also ligand-centered and involves the removal of another electron from the π systems of the bridging Cp or indenyl ligand. Given that CpInd also has two aromatic bridging ligands, it would also be expected to display two oxidation waves. Reactivity with 2,6-Lutidinium Chloride. In triethylphosphine-supported Pd(I) dimers with one bridging Cp ligand, it was demonstrated that protonation with 2,6lutidinium chloride is selective at the bridging Cp ligand.19 This is also true for IPr-supported species. Treatment of AllCp or CpInd with 2,6-lutidinium chloride resulted in the quantitative formation of (μ-All)(μ-Cl){Pd(IPr)}2 (AllCl) or (μ-Ind)(μ-Cl){Pd(IPr)}2 (IndCl), respectively (Table 6). This is presumably because the HOMO of these species is mainly localized on the bridging Cp ligand and there is almost no contribution from the bridging indenyl or allyl ligand. In contrast, in AllInd, where the HOMO has significant



CONCLUSIONS We have prepared a family of Pd(I) dimers with different combinations of bridging allyl, Cp, and indenyl ligands, all supported by the IPr ligand. This is the first time a complete series of dimers has been prepared with all of the different combinations of the allyl, Cp, and indenyl bridging ligands and the same ancillary ligand. The IPr-supported dimers share many features with related triethylphosphine-supported complexes. For example, in all cases the bridging ligands bind through three carbon atoms to the two Pd centers, with the central carbon atom bound to both metal centers and the adjacent carbon atoms bound only to a single metal. The electronic structures of the IPr-supported species are also similar to those of triethylphosphine-supported complexes. In general, in species with a bridging Cp ligand the HOMO is a Pd−Cp antibonding orbital, whereas in complexes with exclusively bridging allyl or indenyl ligands the HOMO is predominantly localized on the bridging ligand. As a result, in mixed species which contain a bridging Cp ligand, the Cp ligand is selectively

Figure 9. EPR spectra of (a) AllAll+ and (b) CpCp+. Experimental spectra are shown in black, and simulated spectra are shown in blue. Simulation parameters for AllAll+ were gz = 2.345 (fwhm Gaussian broadening 110 MHz) and gx,y = 1.996 (fwhm Gaussian broadening 100 MHz). Simulation parameters for CpCp+ were gz = 2.046 (fwhm Gaussian broadening 130 MHz) and gx,y = 2.064 (fwhm Gaussian broadening 80 MHz). 5123

dx.doi.org/10.1021/om400687m | Organometallics 2013, 32, 5114−5127

Organometallics

Article

Figure 10. Calculated SOMO of (a) AllAll′+ and (b) CpCp′+.

Table 6. Summary of Reactivity of Species Prepared in This Work with 2,6-Lutidinium Chloridea

complex

productb

AllCp CpInd AllInd CpCp IndInd

AllCl IndCl AllCl (74%), IndCl (26%) CpCl IndCl

a

A representative reaction is shown below the table title. Conditions: C6D6, room temperature, 10 min. bProducts identified using 1H NMR spectroscopy. See the Experimental Section for more information. Ind)Pd(IPr)Cl, 24 (η1-Cp)(η5-Cp)Pd(IPr),24 (η1-Ind)(η5-Cp)Pd(IPr),24 AllAll6a, and 2,6-lutidinium chloride.34 X-ray Crystallography. Low-temperature diffraction data (ω scans) were collected on either a Rigaku R-AXIS RAPID diffractometer coupled to a R-AXIS RAPID imaging plate detector with Mo Kα radiation (λ = 0.71073 Å) for AllCp, CpCp, and IndInd or a Rigaku MicroMax-007HF diffractometer coupled to a Saturn994+ CCD detector with Cu Kα radiation (λ = 1.54178 Å) for AllInd and CpInd. All structures were solved by direct methods using SHELXS35 and refined against F2 on all data by full-matrix least squares with SHELXL-9736 using established refinement techniques.37 All nonhydrogen atoms were refined anisotropically. All hydrogen atoms, unless otherwise noted, were included in the model at geometrically calculated positions and refined using a riding model. The isotropic displacement parameters of all hydrogen atoms were fixed to 1.2 times the U value of the atoms to which they are linked (1.5 times for methyl groups). All disorders were refined with the help of similarity restraints on the 1,2- and 1,3-distances and displacement parameters as well as rigid bond restraints for anisotropic displacement parameters. Further details of the crystal and refinement data are given in the Supporting Information. Electrochemistry. Electrochemistry experiments were performed using an airtight three-electrode system, which was assembled in a nitrogen-filled glovebox. The working electrode was a 2 mm diameter platinum electrode. The reference and counter electrodes were 0.8 mm platinum wires. The electrolyte was 0.10 M nBu4NPF6 in dichloromethane, which was synthesized by the metathesis of nBu4NBr and HPF6, recrystallized from hot ethanol, and dried under vacuum overnight. The dichloromethane used in the experiment was HPLC grade and was dried before use. Ferrocene was used as an internal standard. Cyclic voltammetric data were measured with Princeton Applied Research VersaSTAT 4 potentiostatic instrumentation. Computational Details. All geometry optimizations were performed using Gaussian 09 Revision A.02,38 using the TPSSH functional,39 which has previously been used to accurately model the geometries of Pd(I) dimers.19 The LANL2DZ basis set40 was used for Pd, and the 6-31G++(d,p) basis set was used for all other atoms. The LANL2DZ pseudopotential was used for Pd. Frequency calculations were performed on all optimized structures to ensure that they were true minima. Calculated g values were determined using the

protonated. Furthermore, the HOMO is slightly raised in energy in species with bridging Cp and allyl ligands in comparison to species with bridging indenyl ligands. Therefore, it is easier to oxidize species with bridging Cp and allyl ligands. The oxidation of complexes with two bridging allyl or two bridging Cp ligands results in the formation of species with radical character predominantly on the bridging allyl or Cp ligands. Finally, in contrast to related triethylphosphinesupported complexes, the IPr-supported species are thermally stable at room temperature. The enhanced thermal stability of IPr-supported dimers should allow for their reactivity with a wide variety of small molecules to be probed, and further studies toward this goal are currently being performed in our laboratory.



EXPERIMENTAL SECTION

General Methods. Experiments were performed under a dinitrogen atmosphere in an MBraun drybox or using standard Schlenk techniques, unless otherwise noted. (Under standard glovebox conditions, purging was not performed between uses of pentane, diethyl ether, benzene, toluene, and THF; 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.) Moisture- and air-sensitive liquids were transferred by stainless steel cannula on a Schlenk line or in a drybox. Solvents were dried by passage through a column of activated alumina followed by storage under dinitrogen. All commercial chemicals were used as received, except where noted. Deuterated solvents were obtained from Cambridge Isotope Laboratories. C6D6 and d8-toluene were dried over sodium metal and vacuum-transferred prior to use. NMR spectra were recorded on Bruker AMX-400 and -500 spectrometers at ambient probe temperatures, unless otherwise noted. Chemical shifts are reported in ppm with respect to residual internal protio solvent for 1H and 13C{1H} NMR spectra; J values are given in Hz. Robertson Microlit Laboratories, Inc., performed the elemental analyses. Literature procedures were used to prepare the following compounds: (η5Cp)(η3-All)Pd,31 1,3-bis(2,6-diisopropylphenyl)-1,3-dihydro-2H-imidazol-2-ylidene (IPr),32 (η3-All)Pd(IPr)Cl,26 lithium indenyl,33 (η35124

dx.doi.org/10.1021/om400687m | Organometallics 2013, 32, 5114−5127

Organometallics

Article

Amsterdam Density Functional (Version ADF2012) package,41 with the geometry determined from the Gaussian optimization. The generalized gradient approximation was employed, using the local density approximation of Vosko, Wilk, and Nusair42 together with nonlocal exchange correction by Becke43 and nonlocal correlation corrections by Perdew.44 Scalar relativistic corrections were included via the ZORA (zero-order relativistic approximation) formalism.45 TZ2P basis sets were used with triple-ξ accuracy sets of Slater-type orbitals, with polarization functions added to all atoms. The ESR keyword46 as implemented in ADF2012 was used. Synthesis and Characterization of Compounds. AllCp. A solution of (η5-Cp)(η3-All)Pd (0.115 g, 0.542 mmo1) in 5 mL of toluene was added to a solution of IPr (0.211 mg, 0.542 mmol) in 4 mL of toluene at room temperature. The solution was stirred at room temperature for 12 h, during which time a small quantity of Pd black precipitated out of the reaction mixture. The Pd black was removed by quickly filtering the mixture through a Celite plug. The solvent was removed under reduced pressure to give a greenish yellow solid. The solid was washed four times with 5 mL portions of pentane and dried under vacuum to give AllCp as a forest green solid. Yield: 0.148 g, 50%. X-ray-quality crystals were grown in a saturated toluene and pentane mixture (V(toluene):V(pentane) = 2:1) at −35 °C. 1 H NMR (C6D6, 400 MHz): 7.16 (t, J = 7.6 Hz, 4H, para-H ArIPr), 7.07 (d, J = 7.6 Hz, 4H, meta-H ArIPr), 7.04 (d, J = 7.6 Hz, 4H, meta-H ArIPr), 6.69 (s, 4H, HCCH), 5.27 (s, 5H, Cp), 3.31−3.20 (m, 8H, (CH3)2CH), 1.95 (1H, central-allyl), 1.95 (2H, terminal-allyl), 1.24 (d, J = 6.7 Hz, 12H, (CH3)2CH), 1.21 (d, J = 6.3 Hz, 12H, (CH3)2CH), 1.06 (d, J = 6.0 Hz, 24H, (CH3)2CH), 0.07 (d, J = 10.4 Hz, 2H, terminal-allyl). 13C{1H} NMR (C6D6, 100 MHz): 193.5, 146.2, 137.7, 129.3, 124.0, 123.5, 89.6, 67.4, 28.9, 28.8, 26.3, 25.9, 23.2, 23.1. Anal. Calcd (found) for C62H82N4Pd2: C, 67.93 (67.68); H, 7.54 (7.40); N, 5.11 (4.92). AllInd. (η3-All)Pd(IPr)Cl (0.136 g, 0.238 mmo1) was dissolved in 10 mL of diethyl ether at −35 °C. The solution was added to lithium indenyl (0.029 g, 0.238 mmol) in 2 mL of diethyl ether at −35 °C. The above mixture was stirred at −35 °C for 30 min and at room temperature for 1 day during, which time a small quantity of Pd black precipitated out of the reaction mixture. The Pd black was removed by quickly filtering the mixture through a Celite plug. The solvent was removed under reduced pressure to give a light yellow solid. The solid was washed four times with 5 mL portions of pentane and dried under vacuum to give AllInd as a light yellow solid. Yield: 0.106 g, 78%. Xray-quality crystals were grown by slow diffusion of pentane into a saturated THF solution at −35 °C. 1 H NMR (C6D6, 400 MHz): 7.44 (d, J = 2.2 Hz, 2H, Ind), 6.84− 6.78 (m, 12H, meta-H and para-H ArIPr), 6.66 (s, 2H, HCCH), 6.63 (s, 2H, HCCH), 5.41 (d, J = 3.7 Hz, 2H, Ind), 3.83 (t, J = 3.5 Hz, 1H, Ind), 3.63−3.49 (m, 4H, (CH3)2CH), 3.24 (m, 2H, (CH3)2CH), 2.78 (1H, central-allyl), 2.67 (sept, J = 6.6 Hz, 2H, (CH3)2CH), 2.35 (d, J = 8.3 Hz, 2H, terminal-allyl), 1.55 (d, J = 6.5 Hz, 6H, (CH3)2CH), 1.42 (d, J = 6.6 Hz, 6H, (CH3)2CH), 1.20 (d, J = 6.4 Hz, 6H, (CH3)2CH), 1.10 (d, J = 6.9 Hz, 6H, (CH3)2CH), 1.08 (d, J = 7.0 Hz, 6H, (CH3)2CH), 1.01 (d, J = 7.2 Hz, 6H, (CH3)2CH), 0.98 (d, J = 7.5 Hz, 6H, (CH3)2CH), 0.09 (d, J = 6.2 Hz, 6H, (CH3)2CH), −0.47 (d, J = 12.6 Hz, 2H, terminal-allyl). 13C{1H} NMR (C6D6, 100 MHz): 192.6, 146.2, 145.4, 142.2, 137.9, 137.6, 125.7, 123.9, 120.6, 118.6, 94.2, 79.2, 48.1, 28.7, 26.9, 25.1, 23.8, 22.9, 21.5. Anal. Calcd (found) for C66H84N4Pd2: C, 69.16 (69.00); H, 7.39 (7.56); N, 4.89 (4.41). IndInd. (η3-Ind)Pd(IPr)Cl (0.110 g, 0.170 mmol) was dissolved in 10 mL of diethyl ether at −35 °C. The solution was added to lithium indenyl (0.021 g, 0.170 mmol) in 4 mL of diethyl ether at −35 °C and stirred at this temperature for 20 min. The precipitate was removed by quickly filtering the mixture through a Celite plug. The resulting mixture was stirred at −35 °C for 30 min and stirred at room temperature for 1 day. The solvent was removed under reduced pressure to give a yellow solid. The solid was washed three times with 5 mL portions of pentane and dried under vacuum to give IndInd as a bright yellow solid. Yield: 0.088 g, 85%. X-ray-quality crystals were grown by slow diffusion of pentane into a saturated THF solution at −35 °C.

1

H NMR (C6D6, 400 MHz): 7.25 (d, J = 2.4 Hz, 4H, Ind), 7.02 (d, J = 2.5 Hz, 4H, Ind), 6.89−6.96 (m, 12H, para-H and meta-H ArIPr), 6.80 (s, 4H, HCCH), 5.64 (d, J = 3.7 Hz, 4H, Ind), 3.78 (sept, J = 6.8 Hz, 4H, (CH3)2CH), 3.22 (sept, J = 6.7 Hz, 4H, (CH3)2CH), 3.15 (t, J = 3.7 Hz, 2H, Ind) 1.47 (d, J = 6.7 Hz, 12H, (CH3)2CH), 1.20 (d, J = 6.7 Hz, 12H, (CH3)2CH), 1.08 (d, J = 6.9 Hz, 12H, (CH3)2CH), 0.57 (d, J = 6.7 Hz, 12H, (CH3)2CH). The compound was too insoluble in all common solvents to record a 13C NMR spectrum. Anal. Calcd (found) for C72H86N4Pd2: C, 70.86 (70.77); H, 7.10 (7.35); N, 4.59 (4.47). CpCp. (η3-Ind)Pd(IPr)Cl (0.142 g, 0.220 mmol) was dissolved in 10 mL of diethyl ether at −35 °C. The solution was added to lithium indenyl (0.027 g, 0.220 mmol) in 2 mL of diethyl ether at −35 °C and stirred at this temperature for 20 min. The precipitate was removed by quickly filtering the mixture through a Celite plug. A solution of (η1Cp)(η5-Cp)Pd(IPr) (0.125 g, 0.200 mmol) in 4 mL of THF was added to the above mixture of (η3-Ind)Pd(IPr)Cl and lithium indenyl at −35 °C. The resulting mixture was stirred at −35 °C for 30 min and stirred at room temperature for 1 day. The Pd black was removed by quickly filtering the mixture through a Celite plug. The solvent was removed under reduced pressure to give a yellow solid. The solid was washed three times with 5 mL portions of pentane and dried under vacuum to give CpCp as a yellow solid. Yield: 0.159 g, 71%. X-rayquality crystals were grown by slow diffusion of pentane into a saturated THF solution at −35 °C. 1 H NMR (C6D6, 400 MHz): 7.02−7.08 (m, 12H, para-H and metaH ArIPr), 6.72 (s, 4H, HCCH), 4.50 (s, 10H, Cp), 3.42 (sept, 8H, (CH3)2CH), 1.28 (d, J = 6.8 Hz, 24H, (CH3)2CH), 1.09 (d, J = 6.8 Hz, 24H, (CH3)2CH). 13C{1H} NMR (C6D6, 100 MHz): 188.6, 146.0, 138.4, 129.6, 124.2, 123.5, 86.5, 28.9, 25.8, 23.2. Anal. Calcd (found) for C64H82N4Pd2: C, 68.62 (68.36); H, 7.38 (7.66); N, 5.00 (4.38). CpInd. (η3-Ind)Pd(IPr)Cl (0.066 g, 0.102 mmol) was dissolved in 10 mL of diethyl ether at −35 °C. The solution was added to lithium indenyl (0.013 g, 0.102 mmol) in 2 mL of diethyl ether at −35 °C and stirred at this temperature for 20 min. The precipitate was removed by quickly filtering the mixture through a Celite plug. A solution of (η1Ind)(η5-Cp)Pd(IPr) (0.069 g, 0.102 mmol) in 4 mL of THF was added to the above mixture of (η3-Ind)Pd(IPr)Cl and lithium indenyl at −35 °C. The resulting mixture was stirred at −35 °C for 30 min and stirred at room temperature for 1 day, during which time a small quantity of Pd black precipitated out of the reaction mixture. The Pd black was removed by quickly filtering the mixture through a Celite plug. The solvent was removed under reduced pressure to give a dark yellow solid. The solid was washed three times with 5 mL portions of pentane and dried under vacuum to give crude CpInd as a brown solid. X-ray-quality crystals were grown by slow diffusion of pentane into a saturated THF solution at −35 °C. No elemental analysis was obtained for this compound, due to slight CpCp and IndInd impurities. The 1H and 13C NMR spectra are shown in the Supporting Information. 1 H NMR (C7D8, 400 MHz, 223 K): 7.22−6.90 (12H, para-H and meta-H ArIPr, partially obscured by solvent), 6.82 (d, J = 7.3 Hz, 2H, Ind), 6.59 (s, 2H, HCCH), 6.51 (s, 2H, HCCH), 5.01 (d, J = 3.1 Hz, 2H, Ind), 4.65 (s, 5H, Cp), 4.00 (sept, J = 7.2 Hz, 2H, (CH3)2CH), 3.37 (sept, J = 4.2 Hz, 2H, (CH3)2CH), 3.32 (sept, J = 7.4 Hz, 2H, (CH3)2CH), 3.30 (t, J = 4.9 Hz, 1H, Ind), 2.82 (sept, J = 4.9 Hz, 2H, (CH3)2CH), 1.50 (d, J = 6.5 Hz, 6H, (CH3)2CH), 1.40 (d, J = 6.7 Hz, 6H, (CH3)2CH), 1.27 (d, J = 5.8 Hz, 6H, (CH3)2CH), 1.18 (d, J = 7.0 Hz, 6H, (CH3)2CH), 1.16 (d, J = 7.1 Hz, 6H, (CH3)2CH), 1.11 (d, J = 5.8 Hz, 6H, (CH3)2CH), 1.02 (d, J = 6.3 Hz, 6H, (CH3)2CH), 0.30 (d, J = 7.1 Hz, 6H, (CH3)2CH). 13C{1H} NMR (C6D6, 100 MHz): 188.1, 146.7, 137.8, 129.6, 124.4, 121.0, 120.0, 87.9, 83.7, 44.0, 29.0, 26.8, 23.9. Representative Procedure for Protonation Reactions. The desired Pd(I) dimer (10 mg) was dissolved in C6D6 (0.25 mL). A suspension of 1 equiv of 2,6-lutidinium chloride in C6D6 (0.25 mL) was added and the resulting suspension vigorously shaken until it was homogeneous. NMR spectra of the reaction mixture were then recorded at room temperature. The product formed, (μ-All)(μ5125

dx.doi.org/10.1021/om400687m | Organometallics 2013, 32, 5114−5127

Organometallics

Article

Cl){Pd(IPr)}2,6a (μ-Cp)(μ-Cl){Pd(IPr)}2,47 or (μ-Ind)(μ-Cl){Pd(IPr)}2,47 was identified by comparison of the 1H NMR spectra to those of authentic samples. No further reaction was observed when additional 2,6-lutidinium chloride was added. Representative Procedure for EPR Spectroscopy. In an EPR tube in the glovebox, 1 μmol of the desired dimer was dissolved in 100 μL of toluene and frozen. A 50 μL portion of toluene was subsequently layered on top of the dimer solution and frozen, followed by 50 μL of dichloromethane, which was also frozen. Finally, a solution containing 1 equiv of [Fe(η5-C5H5)2][PF6] in 100 μL of dichloromethane was added and frozen. The tube was removed from the glovebox, and just prior to being inserted into the EPR spectrometer, the solution was allowed to melt in a cold bath at −65 °C. The tube was rapidly shaken at room temperature to allow for mixing of the oxidant and dimer and then inserted into the EPR spectrometer. X-band EPR spectra were acquired on a Bruker ELEXSYS E500 EPR spectrometer equipped with an SHQ resonator and an Oxford ESR-900 helium-flow cryostat. EPR scans were acquired at 7 K with the following instrumental parameters: microwave frequency 9.39 GHz, modulation frequency 100 kHz, modulation amplitude 10 G, and microwave power 10 mW.



(5) (a) Denmark, S. E.; Baird, J. D. Org. Lett. 2006, 8, 793. (b) Denmark, S. E.; Baird, J. D.; Regens, C. S. J. Org. Chem. 2008, 73, 1440. (6) (a) Hruszkewycz, D. P.; Wu, J.; Hazari, N.; Incarvito, C. D. J. Am. Chem. Soc. 2011, 133, 3280. (b) Hruszkewycz, D. P.; Wu, J.; Green, J. C.; Hazari, N.; Schmeier, T. J. Organometallics 2012, 31, 470. (7) Markert, C.; Neuburger, M.; Kulicke, K.; Meuwly, M.; Pfaltz, A. Angew. Chem., Int. Ed. 2007, 46, 5892. (8) (a) Werner, H.; Kühn, A. Angew. Chem., Int. Ed. 1977, 16, 412. (b) Felkin, H.; Turner, G. K. J. Organomet. Chem. 1977, 129, 429. (c) Hayashi, Y.; Matsumoto, K.; Nakamura, Y.; Isobe, K. J. Chem. Soc., Dalton Trans. 1989, 1519. (d) Yamamoto, T.; Saito, O.; Yamamoto, A. J. Am. Chem. Soc. 1981, 103, 5600. (e) Werner, H.; Kraus, H.-J.; Schubert, U.; Ackermann, K. Chem. Ber. 1982, 115, 2905. (f) Kobayashi, Y.; Iitaka, Y.; Yamazaki, H. Acta Crystallogr., Sect. B 1972, 28, 899. (g) Werner, H.; Kuhn, A. J. Organomet. Chem. 1979, 179, 439. (h) Yamamoto, T.; Akimoto, M.; Yamamoto, A. Chem. Lett. 1983, 1725. (i) Osakada, K.; Chiba, T.; Nakamura, Y.; Yamamoto, T.; Yamamoto, A. J. Chem. Soc., Chem. Commun. 1986, 1589. (j) Sieler, J.; Helms, M.; Gaube, W.; Svensson, A.; Lindqvist, O. J. Organomet. Chem. 1987, 320, 129. (k) Osakada, K.; Chiba, T.; Nakamura, Y.; Yamamoto, T.; Yamamoto, A. Organometallics 1989, 8, 2602. (l) Bogdanovic, B.; Goddard, R.; Rubach, M. Acta Crystallogr., Sect. C 1989, 45, 1511. (m) Osakada, K.; Ozawa, Y.; Yamamoto, A. J. Organomet. Chem. 1990, 399, 341. (n) Miyauchi, Y.; Watanabe, S.; Kuniyasu, H.; Kurosawa, H. Organometallics 1995, 14, 5450. (o) Kurosawa, H.; Hirako, K.; Natsume, S.; Ogoshi, S.; Kanehisa, N.; Kai, Y.; Sakaki, S.; Takeuchi, K. Organometallics 1996, 15, 2089. (p) Sakaki, S.; Takeuchi, K.; Sugimoto, M.; Kurosawa, H. Organometallics 1997, 16, 2995. (9) (a) Jolly, P. W.; Krueger, C.; Schick, K. P.; Wilke, G. Z. Naturforsch., B: Anorg. Chem., Org. Chem. 1980, 35B, 926. (b) Benn, R.; Jolly, P. W.; Mynott, R.; Raspel, B.; Schenker, G.; Schick, K. P.; Schroth, G. Organometallics 1985, 4, 1945. (c) Hayashi, Y.; Nakamura, Y.; Isobe, K. J. Chem. Soc., Chem. Commun. 1988, 403. (d) Krause, J.; Cestaric, G.; Haack, K.-J.; Seevogel, K.; Storm, W.; Pörschke, K.-R. J. Am. Chem. Soc. 1999, 121, 9807. (e) Krause, J.; Goddard, R.; Mynott, R.; Pörschke, K.-R. Organometallics 2001, 20, 1992. (10) Dau, P. D.; Hruszkewcyz, D. P.; Huang, D.-L.; Chalkley, M. J.; Liu, H.-T.; Green, J. C.; Hazari, N.; Wang, L. S. Organometallics 2012, 31, 8571. (11) (a) Wu, J.; Hazari, N. Chem. Commun. 2011, 47, 1069. (b) Schmeier, T. J.; Dobereiner, G. E.; Crabtree, R. H.; Hazari, N. J. Am. Chem. Soc. 2011, 133, 9274. (12) Crabtree, R. H. The Organometallic Chemistry of the Transition Metals, 5th ed.; Wiley: New York, 2009. (13) Werner, H.; Kraus, H.-J. Angew. Chem., Int. Ed. 1979, 18, 948. (14) (a) Tanase, T.; Nomura, T.; Yamamoto, Y.; Kobayashi, K. J. Organomet. Chem. 1991, 410, C25. (b) Tanase, T.; Nomura, T.; Fukushima, T.; Yamamoto, Y.; Kobayashi, K. Inorg. Chem. 1993, 32, 4578. (15) (a) Werner, H.; Tune, D.; Parker, G.; Krüger, C.; Brauer, D. J. Angew. Chem., Int. Ed. 1975, 14, 185. (b) Werner, H.; Kühn, A.; Tune, D. J.; Krueger, C.; Brauer, D. J.; Sekutowski, J. C.; Tsay, Y.-H. Chem. Ber. 1977, 110, 1763. (c) Kuhn, A.; Werner, H. J. Organomet. Chem. 1979, 179, 421. (d) Norton, D. M.; Mitchell, E. A.; Botros, N. R.; Jessop, P. G.; Baird, M. C. J. Org. Chem. 2009, 74, 6674. (16) In this paper the abbreviation AllAll means that a complex contains two bridging allyl ligands. The abbreviation Cp is used for a bridging cyclopentadienyl ligand, and Ind is used for a bridging indenyl ligand. Therefore, the abbreviation CpInd means that a complex contains one bridging cyclopentadienyl ligand and one bridging indenyl ligand. (17) (a) Beck, R.; Johnson, S. A. Organometallics 2013, 32, 2944. (b) Budzelaar, P. H. M.; Boersma, J.; Van der Kerk, G. J. M.; Spek, A. L. Organometallics 1984, 3, 1187. (c) Wu, J.; Brudvig, G. W.; Dai, W.; Guard, L. M.; Hazari, N.; Lin, P.-H.; Pokhrel, R.; Takase, M. K. Manuscript in preparation.

ASSOCIATED CONTENT

S Supporting Information *

Tables, figures, and CIF files giving X-ray information for AllCp, AllInd, CpCp, IndInd, and CpInd, selected NMR spectra, EPR spectra for AllAll+ and CpCp+, spin density plots for AllAll′+ and CpCp′+, and Cartesian coordinates and energies for optimized structures. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail for N.H.: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We gratefully acknowledge support through a Doctoral New Investigator Grant from the ACS Petroleum Research Fund (51009-DNI3) and the National Science Foundation through Grant CHE-1150826. The EPR spectroscopy work was supported by the Department of Energy, Office of Basic Energy Sciences, Division of Chemical Sciences, grant DEFG02-05ER15646 (R.P. and G.W.B.). W.D. thanks Yale University for an Anderson Postdoctoral Fellowship, and M.J.C. thanks Yale College for funding as part of the Yale Science Scholars Program. This work was supported in part by the facilities and staff of the Yale University Faculty of Arts and Sciences High Performance Computing Center and by the National Science Foundation under Grant CNS 08-21132 that partially funded acquisition of the facilities.



REFERENCES

(1) Hazari, N.; Hruszkewycz, D. P.; Wu, J. Synlett 2011, 1793. (2) Weissman, H.; Shimon, L. J. W.; Milstein, D. Organometallics 2004, 23, 3931. (3) Elliott, E. L.; Ray, C. R.; Kraft, S.; Atkins, J. R.; Moore, J. S. J. Org. Chem. 2006, 71, 5282. (4) Hill, L. L.; Crowell, J. L.; Tutwiler, S. L.; Massie, N. L.; Hines, C. C.; Griffin, S. T.; Rogers, R. D.; Shaughnessy, K. H.; Grasa, G. A.; Johansson Seechurn, C. C. C.; Li, H.; Colacot, T. J.; Chou, J.; Woltermann, C. J. J. Org. Chem. 2010, 75, 6477. 5126

dx.doi.org/10.1021/om400687m | Organometallics 2013, 32, 5114−5127

Organometallics

Article

Comput. Chem. 2001, 22, 931. (b) Fonseca Guerra, C.; Snijder, J. G.; te Velde, G.; Baerends, E. J. Theor. Chem. Acc. 1998, 99, 391. (c) ADF2012; SCM, Theoretical Chemistry, Vrije Universiteit, Amsterdam, The Netherlands; http://www.scm.com. (42) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200. (43) (a) Becke, A. D. Phys. Rev. A: At., Mol., Opt. Phys. 1988, 38, 3098. (b) Becke, A. D. J. Chem. Phys. 1988, 88, 1053. (44) Perdew, J. P. Phys. Rev. B 1986, 33, 8800. (45) (a) Vanlenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1993, 99, 4597. (b) Vanlenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1994, 101, 9783. (c) Vanlenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1996, 105, 6505. (d) Vanlenthe, E.; VanLeeuwen, R.; Baerends, E. J.; Snijders, J. G. Int. J. Quantum Chem. 1996, 57, 281. (e) Vanlenthe, E.; Ehlers, A.; Baerends, E. J. J. Chem. Phys. 1999, 110, 8943. (46) Kadantsev, E. S.; Ziegler, T. J. Phys. Chem. A 2009, 113, 1327. (47) Dai. W.; Hazari, N.; Melvin, P. R. Manuscript in preparation.

(18) (a) Heinemann, O.; Jolly, P. W.; Krüger, C.; Verhovnik, G. P. J. Organometallics 1996, 15, 5462. (b) Sieb, D.; Schuhen, K.; Morgen, M.; Herrmann, H.; Wadepohl, H.; Lucas, N. T.; Baker, R. W.; Enders, M. Organometallics 2012, 31, 356. (19) Chalkley, M. J.; Guard, L. M.; Hazari, N.; Hofmann, P.; Hruszkewycz, D. P.; Schmeier, T. J.; Takase, M. K. Organometallics 2013, 32, 4223. (20) Werner, H. Adv. Organomet. Chem. 1981, 19, 155. The bonding theory part of this review was taken from unpublished work by: Hofmann, P.; Hoffmann, R.; Dobosh, P. Private communication by these authors to H. Werner. (21) From this point forward it should be assumed that the supporting ligand on all complexes is IPr unless otherwise noted. For example, the abbrevation AllAll denotes (μ-All)2{Pd(IPr)}2 and the abbreviation CpInd denotes (μ-Cp)(μ-Ind){Pd(IPr)}2. (22) Werner, H. Adv. Organomet. Chem. 1981, 19, 155. (23) Sergeyev, N. M. Prog. Nucl. Magn. Reson. Spectrosc. 1975, 9, 71. (24) Bielinski, E. A.; Dai, W.; Guard, L. M.; Hazari, N.; Takase, M. K. Organometallics 2013, 32, 4025. (25) Ihara, E.; Maeno, Y.; Yasuda, H. Macromol. Chem. Phys. 2001, 202, 1518. (26) Viciu, M. S.; Germaneau, R. F.; Navarro-Fernandez, O.; Stevens, E. D.; Nolan, S. P. Organometallics 2002, 21, 5470. (27) Gomes, J. A. N. F.; Mallion, R. B. Chem. Rev. 2001, 101, 1349. (28) Farrugia, L. J. J. Appl. Crystallogr. 1997, 30, 565. (29) (a) Veiros, L. F. Organometallics 2000, 19, 5549. (b) Calhorda, M. J.; Felix, V.; Veiros, L. F. Coord. Chem. Rev. 2002, 230, 49. (30) For Cp systems the fold angle is the difference in the dihedral angles between the plane defined by the three allylic carbons and that formed by the carbon atoms of Cp except for the central bridging carbon atom, whereas for indenyl systems the fold angle is the difference in the dihedral angles between the plane defined by the three allylic carbons and that formed by the benzenoid carbons. (31) Tatsuno, Y.; Yoshida, T.; Otsuka, S. Inorg. Synth. 1979, 19, 220. (32) Arduengo, A. J., III; Krafczyk, R.; Schmutzler, R.; Craig, H. A.; Goerlich, J. R.; Marshall, W. J.; Unverzagt, M. Tetrahedron 1999, 55, 14523. (33) Christopher, J. N.; Diamond, G. M.; Jordan, R. F.; Petersen, J. L. Organometallics 1996, 15, 4038. (34) (a) Grönberg, K. L. C.; Henderson, R. A.; Oglieve, K. E. J. Chem. Soc., Dalton Trans. 1998, 3093. (b) Garratt, S. A.; Hughes, R. P.; Kovacik, I.; Ward, A. J.; Willemsen, S.; Zhang, D. J. Am. Chem. Soc. 2005, 127, 15585. (35) Sheldrick, G. M. Acta Crystallogr., Sect. A 1990, 46, 467. (36) Sheldrick, G. M. Acta Crystallogr., Sect. A 2008, 64, 112. (37) Müller, P. Crystallogr. Rev. 2009, 15, 57. (38) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, N. J.; Klene, M.; Knox, J. E.; 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.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö .; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, Revision A.02; Gaussian, Inc., Wallingford, CT, 2009. (39) Tao, J. M.; Perdew, J. P.; Staroverov, V. N.; Scuseria, G. E. Phys. Rev. Lett. 2003, 91, 146401. (40) (a) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270. (b) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299. (c) Wadt, W. R.; Hay, P. J. J. Chem. Phys. 1985, 82, 284. (41) (a) te Velde, G.; Bickelhaupt, F. M.; Baerends, E. J.; Fonseca Guerra, C.; Van Gisbergen, S. J. A.; Snijders, J. G.; Ziegler, T. J. 5127

dx.doi.org/10.1021/om400687m | Organometallics 2013, 32, 5114−5127