Comparison of the catalytic activity for the Suzuki–Miyaura reaction of (η5-Cp)Pd(IPr)Cl with (η3-cinnamyl)Pd(IPr)(Cl) and (η3-1-t-Bu-indenyl)Pd(IPr)(Cl)

Complexes of the type (η3-allyl)Pd(L)(Cl) and (η3-indenyl)Pd(L)(Cl) are highly active precatalysts for the Suzuki–Miyaura reaction. Even though allyl and indenyl ligands are similar to cyclopentadienyl (Cp) ligands, there have been no detailed comparative studies exploring the activity of precatalysts of the type (η5-Cp)Pd(L)(Cl) for Suzuki–Miyaura reactions. Here, we compare the catalytic activity of (η5-Cp)Pd(IPr)(Cl) (IPr = 1,3-bis(2,6-diisopropylphenyl)-1,3-dihydro-2H-imidazol-2-ylidene, Cp) with two commercially available catalysts (η3-cinnamyl)Pd(IPr)(Cl) (Cin) and (η3-1-t-Bu-indenyl)Pd(IPr)(Cl) (tBuInd). We show that Cp gives slightly better catalytic activity than Cin, but significantly inferior activity than tBuInd. This order of activity is rationalized by comparing the rates at which the precatalysts are activated to the monoligated Pd(0) active species along with the tendency of the starting precatalysts to comproportionate with monoligated Pd(0) to form inactive Pd(I) dimers. As part of this work the Cp supported Pd(I) dimer (μ-Cp)(μ-Cl)Pd2(IPr)2 (CpDim) was synthesized and crystallographically characterized. It does not readily disproportionate to form monoligated Pd(0) and consequently CpDim is a poor catalyst for the Suzuki–Miyaura reaction.


Results and Discussion
Catalytic comparison of (η 5 -Cp)Pd(IPr)Cl, (η 3 -cinnamyl)Pd(IPr)Cl and (η 3 -1-t-Buindenyl)Pd(IPr)Cl The IPr supported precatalyst for the Suzuki-Miyaura reaction Cp was synthesized using a literature method starting from the commercially available Pd(II) dimer (μ-Cl) 2 Pd 2 (η 3 -allyl) 2 (Scheme 1) [18].It is notable that in this synthesis dimeric {(IPr)Pd(Cl)} 2 (μ-Cl) 2 is prepared as an intermediate, followed by treatment with two equivalents of NaCp to generate the monomer Cp.This synthesis makes rapid ligand screening using the Cp supported scaffold difficult as the Cp group is introduced after the ligand.In contrast, the syntheses of both Cin and tBu Ind involve the initial preparation of dimers of the form {(η 3 -cinnamyl)Pd} 2 (μ-Cl) 2 or {(η 3 -1-t-Buindenyl)Pd} 2 (μ-Cl) 2 , respectively [11,15], which can then be treated with a ligand to generate the ligated precatalyst.Despite repeated attempts we were unable to synthesize a related unligated Cp containing dimer, which could be used for ligand screening [21].
The catalytic activity of Cp for Suzuki-Miyaura reactions with different substrates under both strong (KOt-Bu) and weak (K 2 CO 3 ) base conditions is compared to Cin and tBu Ind in Figure 2 and Figure 3.In general, the performance of Cp is slightly better than Cin, but considerably worse than tBu Ind.At  [15].All yields were determined using GC and are the average of two runs.
times when reactions using the tBu Ind precatalyst are complete, between 10 and 60% conversion is achieved with Cp.If reactions catalyzed by Cp are left for longer periods of time complete conversion occurs, indicating that the difference in rates is not related to rapid catalyst decomposition in the case of Cp.These results confirm the previously reported high activity of Cp supported precatalysts [18].Although the difference in performance between Cin, Cp and tBu Ind varies depending on the specific substrate and catalyst loading, we are not able to discern any general trends in the data.For example, in some cases Cp gives better activity than Cin for the synthesis of di-ortho-substituted biaryls (products 1 and 7), whereas in another case Cp gives only slightly better activity (product 3).However, in general, the relative catalytic performance of the different precatalysts does not vary when the base is changed.
In order to understand the relative activity of Cp in comparison to Cin and tBu Ind, we measured both the rate at which it is activated to monoligated Pd(0) and its tendency to undergo comproportionation to a Pd(I) dimer.The rate of activation was measured using the same procedure that we have previously used for Cin and tBu Ind [16].Cp was treated with base in the presence of ten equivalents of 1,3-divinyl-1,1,3,3-tetramethyldisiloxane (dvds) under a variety of conditions which are relevant to the Suzuki-Miyaura coupling and the reaction followed using 1 H NMR spectroscopy (Table 1).The metal containing product of this reaction is the Pd(0) complex (IPr)Pd(dvds) [22].The rate of formation of (IPr)Pd(dvds) can be used as a model for the rate of Pd(0) formation in catalysis.In all cases Cp is  [15].All yields were determined using GC and are the average of two runs.activated slower than tBu Ind [16], consistent with its inferior catalytic performance.For example, under the reaction conditions used in Table 1, entry 4, the rate of activation for tBu Ind is 7.6 ± 0.1 × 10 −4 s −1 compared to 3.4 ± 0.1 × 10 −4 s −1 for Cp [16].In contrast, Cp is generally activated faster than Cin.The rate of activation for Cin under the conditions used in Table 1, entry 3 is 4.2 ± 0.1 × 10 −4 , less than half the rate of that observed for Cp.The conditions used in Table 1, entry 4 are the most relevant to the catalysis described above, but although in this case it appears that Cp is activated faster than Cin (3.4 ± 0.1 × 10 −4 s −1 vs 1.4 ± 0.2 × 10 −4 s −1 ), the relatively large error associated with these numbers makes a firm conclusion difficult.
The mechanism of activation of Cp appears to be analogous to that previously described for Cin and tBu Ind as the organic byproducts of Cp activation, cyclopentadiene and either acetone (in the case of reactions performed in iPrOH) or formaldehyde (in the case of reaction performed in MeOH), are consistent with the previously reported pathway (Scheme 2) [16].In this mechanism initial substitution of a Cl − ligand in Cp by the solvent gives rise to the alkoxide complex A. Subsequently, the η 5 -Cp ring can undergo slippage to form complex B, with an η 1 -Cp ligand.The η 1 -Cp ligand is nucleophilic and can abstract a β-hydrogen from the alkoxide ligand to generate a Pd(0) species with a coordinated cyclopentadiene ligand (C).In this step the formaldehyde or acetone byproduct originating from a Reaction conditions: 0.0087 mmol Cp, 0.087 mmol of base, 0.087 mmol of dvds in 500 μL of solvent.b All rates are the average of at least two runs and were measured using 1 H NMR spectroscopy.c 100 μL of THF-d 8 was added along with only 400 μL of iPrOH.d Two equivalents of 18-crown-6 (relative to K 2 CO 3 ) were added to solubilize the K 2 CO 3 .e 0.0087 mmol precatalyst, 0.087 mmol phenylboronic acid, 0.096 mmol base, 0.087 mmol dvds in 500 μL MeOH-d 4 .
Scheme 2: Proposed mechanism for the activation of Cp to monoligated Pd(0).
the solvent is released.Finally, dissociation of the olefin ligand from C generates the active monoligated Pd(0) species, which in catalysis undergoes oxidative addition with the aryl halide, but in the case of our activation experiments is trapped by dvds.
The considerably faster rate of activation for Cp compared to Cin, suggests that Cp should be a much better precatalyst than Cin, which is inconsistent with our catalytic results (Figure 2 and Figure 3).In our model for precatalyst performance, catalytic activity is also related to the ease at which the starting precatalyst undergoes comproportionation with monoligated Pd(0) to form a Pd(I) dimer [13].The reaction of Cp with a weak base, K 2 CO 3 , in an alcohol solvent (MeOH) provided the dimeric complex, (μ-Cp)(μ-Cl)Pd 2 (IPr) 2 (Cp Dim ), in excellent yield (82%, Scheme 3).This is the same procedure we previously described for the preparation of Pd(I) dimers with a bridging chloride ligand and one bridging allyl or indenyl ligand [13].Cp Dim was characterized by NMR spectroscopy and X-ray crystallography (see Figure 4).To determine if Cp Dim is catalytically relevant modified conditions were used to allow for the reaction to be monitored by 1 H NMR spectroscopy (Scheme 4).In order to observe the Pd containing species, an increased catalyst loading was used, 4 mol % Cp, compared to the loadings described in Figure 2 and Figure 3. Peaks consistent with the formation of Cp Dim are observed during catalysis, and approximately 40% of the Pd is in the form of Cp Dim upon completion of the catalytic reaction.
In contrast, for Cin under the same conditions, only a small amount of Pd was determined to be in the form of a Pd(I) dimer [13].This suggests that Cp is more likely to undergo dimerization than Cin.Cp Dim was confirmed to be a poor catalyst under the conditions employed in Figure 2 (Scheme 5).This result is indicative of Cp Dim as an off-cycle deactivation product, which reduces the amount of the active Pd(0) species in solution.
Previously, we have demonstrated that the comproportionation of Pd(0) and Pd(II) species to IPr supported Pd(I) dimers with one bridging allyl and one bridging chloride ligand is reversible [13,14].One method to measure the rate of disproportionation of Pd(I) dimers is to react these species with a trapping agent for Pd(0), such as dvds.This results in the formation of the Pd(0) species (IPr)Pd(dvds) and a Pd(II) species of the form (η 3 -allyl)Pd(IPr)(Cl).We examined the tendency of Cp Dim to undergo disproportionation in the presence of dvds.The disproportionation of Cp Dim is extremely difficult and at 60 °C the half-life for the formation of Cp and (IPr)Pd(dvds) is 60 minutes (Scheme 6).In contrast, in the presence of dvds (μ-cinnamyl)(μ-Cl)Pd 2 (IPr) 2 undergoes full disproportionation in approximately 40 minutes at 40 °C, while for (μ-allyl)(μ-Cl)Pd 2 (IPr) 2 the reaction is complete in less than 10 minutes at room temperature [13].Although these results show that disproportionation of Cp Dim is more difficult than related allyl species, they provide no information on whether this is related to thermodynamic or kinetic effects.
To probe the relative thermodynamic favorability of dimer formation between allyl and Cp systems we performed a crossover experiment (Scheme 7a).In this experiment (μ-allyl)(μ-Cl)Pd 2 (IPr) 2 was mixed with Cp.The products of crossover are Cp Dim and (η 3 -allyl)Pd(IPr)(Cl) and our experiments indicate that the equilibrium favors these species.The crossover reaction can be described as the combination of the disproportionation of the allyl dimer and the comproportionation of Cp with IPr-Pd(0) (Scheme 7b).From these results, we conclude that the comproportionation reaction to form Cp Dim

Conclusion
We have performed the first detailed comparative investigation of the catalytic activity for the Suzuki-Miyaura reaction of (η 5 -Cp)Pd(IPr)(Cl) (Cp), with the related commercially available catalysts (η 3 -cinnamyl)Pd(IPr)(Cl) (Cin) and (η 3 -1-t-Buindenyl)Pd(IPr)(Cl) ( tBu Ind).We found that Cp is a slightly more efficient catalyst than Cin, but significantly less active than tBu Ind.The low activity of Cp in comparison to tBu Ind is related both to its slower rate of activation to the monoligated Pd(0) active species and its tendency to form a significant amount of the inactive Pd(I) dimer (μ-Cp)(μ-Cl)Pd 2 (IPr) 2 (Cp Dim ) under catalytic conditions.The formation of this inactive dimer also explains why Cp is only a slightly more active precatalyst than Cin, which activates slower than Cp, but is less likely to form the corresponding inactive Pd(I) dimer.In principle, the addition of steric bulk to Cp could prevent the formation of a Pd(I) dimer and result in a more active precatalyst.However, an additional challenge that must be overcome if practical precatalyst scaffolds based on a Cp ligand are to be developed is that the synthetic routes to these species are currently not amenable to rapid ligand screening in an analogous fashion to Cin and tBu Ind.

Experimental General methods
As previously described in [13] and [15], experiments were performed under a dinitrogen atmosphere in an M-Braun dry box or using standard Schlenk techniques unless otherwise stated.Under standard glovebox conditions, purging was not performed between uses of pentane, benzene and toluene; 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.Stainless steel cannulas were used to transfer moisture-and air-sensitive liquids on a Schlenk line or in a dry box.THF, diethyl ether, and toluene were dried by passage through a column of activated alumina followed by storage under dinitrogen.All commercial chemicals were used as received; exceptions where noted.MeOH (J.T. Baker) and iPrOH (Macron Fine Chemicals) were not dried but were degassed by sparging with dinitrogen for one hour and stored under dinitrogen.Potassium tert-butoxide (99.99%, sublimed) was purchased from Aldrich.Potassium carbonate was purchased from Mallinckrodt and ground up with a mortar and pestle and stored in an oven at 130 °C prior to use.1,3-Divinyltetramethyldisiloxane was purchased from TCI. Deuterated solvents were obtained from Cambridge Isotope Laboratories.MeOH-d 4 and THF-d 8 were not dried but were degassed prior to use through three freeze-pump-thaw cycles.Agilent-400, -500 and -600 spectrometers were used to record NMR spectra at ambient probe temperatures.Gas chromatography analyses (GC) were performed on a Shimadzu GC-2010 Plus apparatus equipped with a flame ionization detector and a Shimadzu SHRXI-5MS column (30 m, 250 μm inner diameter, film: 0.25 μm).The following conditions were utilized for GC analyses: flow rate 1.23 mL/min constant flow, column temperature 50 °C (held for 5 min), 20 °C/min increase to 300 °C (held for 5 min), total time 22.5 min.Literature procedures were used to prepare the following compounds: (η 3 -cinnamyl)Pd(IPr)(Cl) (Cin) [11], (η 3 -1-t-Buindenyl)Pd(IPr)(Cl) ( tBu Ind) [15], (η 5 -Cp)Pd(IPr)(Cl) (Cp) [18] (μ-allyl)(μ-Cl)Pd 2 (IPr) 2 [13].

X-ray crystallography
X-ray diffraction experiments were carried out on a Rigaku MicroMax-007HF diffractometer coupled to a Saturn994+ CCD detector with Cu Kα radiation (λ = 1.54178Å) at −180 °C.The crystals were mounted on MiTeGen polyimide loops with immersion oil.The data frames were processed using Rigaku CrystalClear and corrected for Lorentz and polarization effects.
Using Olex2 [42] the structure was solved with the XS [43] structure solution program by Patterson methods and refined with the XL [43] refinement package using least-squares minimization.The non-hydrogen atoms were refined anisotropically.Hydrogen atoms were refined using the riding model unless otherwise stated.

KOt-Bu conditions
Reactions were performed under dinitrogen in a 1 dram vial containing a flea stir bar and sealed with a septum cap.To the vial was added 950 μL of a MeOH stock solution, containing 0.5263 M aryl chloride, 0.5525 M boronic acid, 0.5789 M KOt-Bu and 0.2632 M naphthalene.The vial was then heated using an aluminum block heater set to 25 °C.After thermal equilibration, the reaction was initiated via the addition of 50 μL of the appropriate precatalyst solution in THF (0.1 M [Pd]).Aliquots (≈50-100 μL) were removed at reaction times indicated.The aliquots were purified by filtration through pipet filters containing approximately 1 cm of silica and eluted with 1-1.2 mL of ethyl acetate directly into GC vials.Conversion was determined by comparison of the GC responses of product and the internal naphthalene standard.Biaryl products were initially synthesized using literature procedures [15], identified using NMR spectroscopy by comparison to the literature chemical shifts [11] and then these pure samples used to generate calibration plots for the GC.

K 2 CO 3 conditions
Potassium carbonate (0.75 mmol) was transferred on the benchtop into a 1 dram vial containing a flea stir bar.The vial was sealed with a septum cap, and placed under dinitrogen (by cycling three times between vacuum and dinitrogen) on a Schlenk line through a needle.To the vial was added 950 μL of a MeOH stock solution, containing 0.5263 M aryl chloride, 0.5525 M boronic acid and 0.2632 M naphthalene.The vial was then heated using an aluminum block heater set to 25 °C.After thermal equilibration, the reaction was initiated via the addition of 50 μL of the appropriate precatalyst solution in THF (0.1 M [Pd]).Aliquots (≈50-100 μL) were removed at reaction times indicated.The aliquots were purified by filtration through pipet filters containing approximately 1 cm of silica and eluted with 1-1.2 mL of ethyl acetate directly into GC vials.Conversion was determined by comparison of the GC responses of product and the internal naphthalene standard.Biaryl products were initially synthesized using literature procedures [15], identified using NMR spectroscopy by comparison to the literature chemical shifts [11] and then these pure samples used to generate calibration plots for the GC.
Experiments on activation of Pd(II) to Pd(0) Young NMR tube and the reaction was monitored by 1 H NMR spectroscopy for one hour at 25 °C.After this time, the solvent mixture was removed on a Schlenk line and benzene-d 6 was added.A final 1 H NMR spectrum was recorded to identify the Pd containing products of the reaction.Cp Dim was observed as the main Pd containing product, with a yield of 40% compared to the internal standard 2,6-dimethoxytoluene.
Catalysis using Cp Dim as precatalyst: 0.05 mmol of Cp Dim was transferred into a 1 mL volumetric flask in a glovebox.The precatalyst was dissolved in THF, and the solution was diluted to 1 mL.The solution was transferred to a flask with a Kontes valve.Reactions were performed under dinitrogen in a 1 dram vial containing a flea stir bar and sealed with a septum cap.To the vial was added 950 μL of the MeOH stock solution described above.The vial was then heated using an aluminum block heater set to 25 °C.After thermal equilibration, the reaction was initiated via the addition of 50 μL of the THF solution containing Cp Dim (0.1 M [Pd]).Aliquots (≈50-100 μL) were removed at 30 and 60 minutes.The aliquots were purified by filtration through pipet filters containing approximately 1 cm of silica and eluted with 1-1.2 mL of ethyl acetate directly into GC vials.Conversion was determined by comparison of the GC responses of product and the internal naphthalene standard.No conversion to the biphenyl product was observed at either time point.

Disproportionation of Cp Dim using dvds:
In a nitrogen filled glovebox, Cp Dim (5.5 mg, 0.005 mmol), dvds (9.2 mg, 0.05 mmol) and 2,6-dimethocytoluene (0.8 mg, 0.005 mmol) were added to a vial.MeOH-d 4 (300 μL) and deuterated benzene (200 μL) were added and the homogeneous mixture was transferred to a J. Young tube and sealed.The contents were heated at 60 °C for one hour, at which time an NMR spectrum was recorded.The methyl protons of the product (IPr)Pd(dvds) [22] were compared to the internal standard.At one hour, the reaction had reached 50% conversion.
The mixture was heated to 60 °C and allowed to equilibrate over 36 hours.At this time, an NMR spectrum was recorded at room temperature.The equilibrium constant was calculated by using relative integrations of the Cp protons from Cp and Cp Dim to yield a K of 1.1.

Figure 1 :
Figure 1: General depictions of allyl and related precatalysts that are highly active for the Suzuki-Miyaura reaction with NHC ligands.

Figure 2 :
Figure 2: Comparison of catalytic performance of Cin, Cp and tBu Ind for a series of Suzuki-Miyaura reactions using KOt-Bu as the base.Yields for Cin and tBu Ind are from previous literature results[15].All yields were determined using GC and are the average of two runs.

Figure 3 :
Figure 3: Comparison of catalytic performance of Cin, Cp and tBu Ind for a series of Suzuki-Miyaura reactions using K 2 CO 3 as the base.Yields for Cin and tBu Ind are from previous literature results[15].All yields were determined using GC and are the average of two runs.

Table 1 :
Rates of activation of Cp under different conditions in the presence of dvds.a of the bridging Cp ligand.Consistent with this pseudo η 3 -binding, the C-C bond distances relating to two long bonds, two bonds of intermediate length and one short bond in the bridging Cp ligand are similar to those observed in monomeric Pd centers are bound to three carbon atoms of the bridging Cp ligand.Two of the three carbon atoms are bound to only one Pd center, while the central carbon atom binds to both Pd centers.Pd-C bond distances of almost 3 Å clearly indicate that there is no interaction between the Pd centers and the other two carbonScheme 4: Observation of Cp Dim under modified catalytic conditions.Scheme 5: Cp Dim is not an active precatalyst for a Suzuki-Miyaura reaction at room temperature.atoms [22]rimental details for Table1: Rates of activation of Cp under different conditions in the presence of dvds Schlenk line, after which dinitrogen was introduced into the NMR tube.An array of 1 H NMR spectra was taken at 25 °C over the course of 3 hours.The strong -CH 2 peak from the 18-crown-6 ether was suppressed by presaturating its signal during the experiment.During this time, the growth of the methyl protons of the (IPr)Pd(dvds)[22]product were monitored.)wasdissolved in 100 μL of MeOH-d 4 .These solutions were combined in a J. Young NMR tube at -78 °C.The reaction mixture was degassed on a Schlenk line, after which dinitrogen was introduced into the NMR tube.An array of 1 H NMR spectra was taken at 25 °C over the course of 3 hours.During this time, the growth of the methyl protons of the (IPr)Pd(dvds)[22]product were monitored.