CarbonâHeteroatom Coupling Using Pd-PEPPSI Complexes

Review pubs.acs.org/OPRD

Carbon−Heteroatom Coupling Using Pd-PEPPSI Complexes Cory Valente,† Matthew Pompeo,‡ Mahmoud Sayah,‡ and Michael G. Organ*,‡ †

The Dow Chemical Company, 400 Arcola Road, Collegeville, Pennsylvania, United States 19426 Department of Chemistry, York University, 4700 Keele Street, Toronto, Ontario M3J 1P3, Canada

‡

ABSTRACT: In this contribution, recent advances with the PEPPSI style of Pd−NHC catalysts in aryl aminations and aryl sulfinations are reviewed and summarized from both applications and mechanistic standpoints.

■

INTRODUCTION The ubiquitous presence of C−S and C−N bonds in organic and materials chemistry, coupled with the advent of relatively mild, selective, and highly efficient Pd-catalyzed methodologies, has made aminations and sulfinations of functionalized carbon backbones via cross-couplings one of the most popular transformations in organic synthesis over the past decade.1,2 Aiding the gain in popularity of these reactions has been the success in designing tailored ligands for Pd that enable historically difficult steps in the catalytic cycle. In effect, substrates that were inactive in cross-couplings and/or not compatible with the reactions conditions even 5 years ago now couple efficiently under increasingly mild reaction conditions. The summary of the Pd-catalysis community’s collective efforts is a wider substrate scope and a more efficient and applicable reaction. In this perspective, recent advances with the PEPPSI (pyridine-enhanced precatalyst preparation stabilization and initiation; see Figure 1 for general structure of the PEPPSI system)3 brand of Pd-catalysts in aryl aminations and aryl sulfinations are reviewed and summarized from both an applications and mechanistic standpoint.

optimization through iterative modifications is an obvious and productive means to fine-tune the steric topography around the Pd center and alter the kinetics of RE. Although bulky phosphines have been shown to enable facile RE, they typically require elevated temperatures to effect efficient catalysis.11 Moreover, when the steric topography around Pd is essential for catalyst turnover and limiting sulfur-poisoning pathways that lead to catalyst resting states, even the most transient dissociation of the ligand can be detrimental. Attenuating the basicity of the phosphine atom by modifying the substituents at phosphorus can certainly help minimize ligand dissociation; however, this is more readily accomplished through the use of the inherently more basic NHC ligand.12 To that end, the PEPPSI class of Pd−NHC complexes has been investigated and systematically modified to enable one of the most generally mild sulfinating protocols to date.13,14 Pd-PEPPSI-IPent (1, Figure 1) was found to be more effective at aryl sulfinations than its predecessors Pd-PEPPSI-IPr (2) and Pd-PEPPSI-IMes (3) due to the increased sterics of the NHC ligand.13 In the presence of LiOiPr and KOtBu, a range of (hetero)aryl bromides and chlorides were efficiently coupled with aryl and alkyl thiols at 40 °C in toluene (Table 1, 4−19).13 Primary, secondary, and tertiary alkyl thiols, which are known to be sluggish coupling partners,15 were routinely coupled (leading to 11−15) in high yields under the mildest reactions conditions reported to date. The coupling of triisopropylsilanethiol also proceeded well, furnishing silyl-protected arylthiols 16 and 17 in excellent yields. As reported by Hartwig,5 these substrates are useful precursors to free aryl thiols, which is of considerable value as there are a limited number of these derivatives that are commercially available. Reactions were conducted for 24 h; however, rate studies revealed that most reactions were complete within 4−5 h. This observation led to the development of a room temperature coupling protocol, wherein several thioethers (20−24) were prepared to illustrate the generality of these mild reaction conditions. This marked the first time that Pd-catalyzed aryl sulfinations could be carried out at ambient temperature.

1. ARYL SULFINATIONS WITH Pd-PEPPSI PRE-CATALYSTS Relative to Pd-catalyzed aryl amination chemistry, far less progress has been made on generating efficient aryl sulfination protocols, this despite the prevalence of thioethers in the pharmaceutical arena.4 Although recent strides in Pd-mediated catalysis are a significant improvement over the forcing conditions of nucleophilic aromatic substitution5 and reactions invoking elemental sulfur,6 there is still much headway to be made.7 Advanced phosphine ligands such as JosiPhos8 have been instrumental in bridging the gap between the forcing conditions of old and the efficiency and control that Pdmediated catalysis has become known. Historically, the siphoning of PdII intermediates into thiolatederived off-cycle resting states (Scheme 1, resting states A, B and C) has plagued Pd-catalyzed sulfinations and in turn limited widespread adoption of this transformation.9,10 One approach to this problem is to lower the activation barrier of reductive elimination (RE) so as to keep the catalyst engaged in the catalytic cycle and in effect reposition the equilibrium away from the nonproductive resting states of the higher order (resting state B) and thiolate-bridged (resting state C) PdII oxidative addition adducts (vide inf ra). Ligand design and © 2013 American Chemical Society

Special Issue: Transition Metal-Mediated Carbon-Heteroatom Coupling Reactions Received: October 3, 2013 Published: December 11, 2013 180

dx.doi.org/10.1021/op400278d | Org. Process Res. Dev. 2014, 18, 180−190

Organic Process Research & Development

Review

Figure 1. Structures of Pd-PEPPSI-IPent, Pd-PEPPSI-IPr and Pd-PEPPSI-IMes.

Table 1. Low-temperature aryl sulfinations with aryl-, alkyl-, and silylthiols using Pd-PEPPSI-IPent (1)

Scheme 1. General catalytic cycle for aryl sulfinations highlighting thiol-poisoning pathways that lead to catalyst resting states

Whereas precatalyst activation occurs routinely at or near room temperature in Negishi,16,17 Suzuki−Miyaura,17,18 Stille− Migita19 and Kumada−Tamao−Corriu20 cross-couplings and in aryl amination chemistry where the organometallic or alkyamine is sufficient to reduce PdII,21 the mechanism of reduction in aryl sulfinations is less clear. Although the aryl sulfinations themselves proceeded at 40 °C or below (Table 1), activation of the precatalyst required a preheating step in the presence of LiOiPr to reduce PdII to Pd0. Clearly, this additional preheating step complicates the procedure and lessens the practicality of the otherwise efficient and mild aryl sulfinations. Although catalyst activation can be achieved at room temperature with catalytic Bu2Mg or at 40 °C with catalytic morpholine, these approaches are similarly impractical. A significant improvement in these transformations would be achieved by the design of a catalyst that could be activated at low temperature while maintaining a high TON, once in cycle. To that end, a systematic evaluation of substituents on the pyridine ligand and NHC core of the precatalyst was conducted to determine their effect on catalyst activation using the coupling of thiophenol (26) with sterically hindered 2bromoxylene (25) as the model reaction (Table 2).14 LiOiPr was removed from the reaction conditions to simplify the study and limit the number of activation modes available to the precatalyst. As a start, Pd-PEPPSI-IPr (2) was found to be active at 80 °C but failed to provide coupled product at 70 °C. Surprisingly, unsubstituted pyridine facilitated precatalyst activation (28) at 10 °C lower than 3-chloropyridine, as did 2-picoline (precatalyst 29). Addition of a second ortho methyl group to 2-picoline (precatalyst 30) lowered the activation temperature another 10 °C. The original design of the PEPPSI series of catalysts (Figure 1) was predicated on the assumption that the electron-withdrawing effects of the meta chlorine substituent would facilitate ligand dissociation and thus

precatalyst activation.22 However, given the coupling results, it appears that for aryl sulfinations, ligand sterics may play a more pivotal role. In the absence of isopropoxide, double RS−/ Cl− anion exchange followed by RE to the disulfide is presumably the only mode of activation available to the precatalyst. Therefore, it stands to reason that increased steric bulk of the pyridine ligand would facilitate this initial RE to generate Pd0 and thus enable the catalytic cycle. X-ray crystallography reveals that the Pd−S bond is lengthened in the presence of 2,6-dimethylpyridine, which supports this 181



Review

Table 2. Optimization of NHC and pyridine ligand sterics in aryl sulfinations

Table 3. Room temperature aryl sulfinations mediated by pre-catalyst Pd-PEPPSI-IPentCl-picoline (35)

hypothesis.14 There does appear to be an upper limit on the steric topography around Pd as 2,6-diethylpyridine (precatalyst 31) only provides trace product at 70 °C, likely because the steric environment around Pd is now too encumbered to allow for the prerequisite RS−/Cl− ion exchange. The addition of two chlorine atoms to the NHC backbone (precatalyst 32) enables activation at 50 °C, again likely stemming from a favorable change in the steric topography around Pd for RE, as the Cl atoms serve to push the N-aryl groups inward toward Pd (vide inf ra).23 Given the observed role of sterics on catalyst activation, the NHC ligand was switched from IPr to IPent, and in combination with 2-picoline (precatalyst 34) near quantitative conversion to 27 was observed at 50 °C. The addition of two chlorine atoms to the NHC backbone (precatalyst 35) had a similar effect on the IPent catalyst, lowering the activation temperature by a further 10 °C enabling near quantitative conversion to 27 at only 40 °C. Using catalyst 35 under the optimized reaction conditions in Table 1, a range of (hetero)aryl sulfides were efficiently coupled with sterically congested aryl chlorides and bromides to provide thioethers 6, 27, and 36−44 (Table 3). The finely tuned steric topography of the Pd catalyst facilitated the formation of tetraortho-substituted thioethers (6, 36, and 39) and the coupling of electronically deactivated/heterocyclic sulfides (37, 38, 40, 41, and 43) in high yield under the mildest reaction conditions reported to date for this transformation. 1.1. Mechanistic Considerations of Aryl Sulfinations. The mechanism for catalyst activation, in particular the role that the additive LiOiPr serves in this process, was investigated in more detail to better understand the entryway of the active NHC−Pd0 species into the catalytic cycle.24 The results from this mechanistic study are summarized in Scheme 2 and elaborated on below. • The signature 3-chloropyridine ligand that is most common in the PEPPSI series of catalysts (represented

by D), which was believed to improve precatalyst activation kinetics relative to pyridine, actually undergoes Pd-mediated hydrodehalogenation under the reducing conditions of the high-temperature preactivation protocol to provide precatalyst intermediate E. Alternatively, removing this “precatalyst preactivation step” (i.e., step D → E) by replacing 3-chloropyridine with pyridine lowers the temperature required to produce active NHC−Pd0 catalyst (see Table 2). • Ligand exchange of sulfide for each chloride produces disulfide precatalyst intermediate F, itself a resting state, of which up to 50% disproportionates into the Pd-trimer G, yet another resting state. Empirical evidence suggests that the remaining fraction of F undergoes a monosubstitution of isopropoxide for sulfide to produce the unsymmetrical Pd adduct H. For each iteration of F → G, there is a concomitant loss of one NHC ligand, which effectively erodes the level of precatalyst available to the cross-coupling reaction. • There is a significant base and counterion effect that underpins the catalyst activation pathway. Whereas KOtBu, LiOtBu, and LiOiPr alone are not effective additives, KOiPr (or a mixture of KOtBu and LiOiPr) facilitates efficient reduction of NHC−PdII to NHC−Pd0 (I) and ensuing catalysis. The counterion effect appears to originate from the solubility difference between 182



Review

Scheme 2. Proposed activation pathway of Pd-PEPPSI catalysts under the reaction conditions for aryl sulfination

2. ARYL AMINATIONS WITH Pd-PEPPSI PRE-CATALYSTS The Pd-catalyzed amination of aryl halides has become one of the most widely used methods for the construction of aryl C− N bonds.25 The popularity of this reaction has benefited from the optimization of ancillary ligands that have served to improve the scope and efficiency of this transformation.2,3,26,27 Electron-rich, bulky tertiary phosphines have served well as a ligand platform due to their proven activity and relatively straightforward tunability.10,15,26,28,29 Some notable, highly active phosphine ligands (45−52) for aryl aminations are presented in Figure 2. Recently, much work has been focused on the use and optimization of the PEPPSI catalyst platform that relies on the unique electronic and steric properties that are inherent to NHC ligands.3 In 2008, Organ and co-workers disclosed that Pd-PEPPSI-IPr (2) was an effective precatalyst for the amination of aryl chlorides and bromides with anilines and secondary amines using KOtBu as the base.30 Despite the high reactivity of the reported reaction conditions, the strongly basic medium was not compatible with base-sensitive functionality. To improve the practicality of this methodology, and thus widen the substrate scope, a more mildly basic set of reaction conditions was developed wherein Cs2CO3 was shown to be a suitable replacement for KOtBu for the coupling of secondary amines with electron-deficient aryl halides.30,31 As oxidative addition (OA) was found not to be rate limiting, a mechanistic investigation was undertaken to determine why alkyl amines only couple effectively with electron-deficient aryl halides when employing a weak carbonate base, whereas no such dependence on the electronic nature of the aryl halide was observed when employing a stronger alkoxide base.21 Moreover, the type of

potassium and lithium thiolate salts. Whereas LiSAr salts are soluble in toluene, their corresponding potassium salts are not, which is essential to minimize thiolate concentration in solution so that the equilibrium favors intermediate H. The use of soluble lithium thiolates, however, repositions the equilibrium toward resting state F and in turn, G, essentially turning off the precatalyst activation pathway. The base effect appears to be one of sterics. Whereas MOtBu is too large to degrade the stable trimeric adduct G, MOiPr is small enough to facilitate the return of monomeric PdII complex H to the catalyst activation pathway. Although KOiPr is the preferred additive as it contains both the counterion and base required for efficient catalysis, the lack of commercial availability prompted the use of the KOtBu/LiOiPr mixed salt system throughout the sulfination work. Owing to the relatively strong reducing power of −OtBu and −OiPr, these bases also serve to prevent the accumulation of disulfide should it form in situ. This is imperative as disulfide can oxidatively add to Pd0, which will arrest the catalytic cycle (see Scheme 1), pushing the equilibrium toward thiol-derived catalyst resting states F and G. • Reduction of the unsymmetical Pd intermediate H can take place by one of two mechanisms. In the first, βhydride elimination of isopropoxide renders ArS−Pd−H that can undergo RE to provide ArSH, which in the presence of KOtBu ions will precipitate out of solution. Alternatively, deprotonation of the isopropoxide methine proton on the Pd adduct H by a second molecule of base would provide Pd0 directly with concomitant elimination of the insoluble ArSK salt. 183



Review

Table 5. Comparative study of Pd-PEPPSI-IPr (2) and PdPEPPSI-IPent (1) in aryl aminations with electron-deficient anilines

Figure 2. Selection of state-of-the-art phosphine ligands used in the aryl aminations.

base used in the analogous aryl amination with aniline derivatives appeared to be less dependent on the electronic nature of the electrophile.32 The results of this mechanistic study are provided at the end of this section. More recently, the bulkier Pd-PEPPSI-IPent precatalyst (1) was shown to greatly outperform Pd-PEPPSI-IPr (2) in the coupling of secondary amines with a wide variety of aryl chlorides at 80 °C using Cs2CO3 as the base (leading to products 53−57, Table 4).21,32 Notably, even sterically

the Pd−anilinium complex and in turn render this adduct more susceptible to carbonate-mediated deprotonation (Scheme 3). Scheme 3. Proposed NHC−Pd-catalyzed amination pathway for alkyl and aryl amines

Table 4. Comparative study of Pd-PEPPSI-IPr (2) and PdPEPPSI-IPent (1) in aryl aminations with dialkylamines

hindered and electronically deactivated substrates coupled in high yields (for example 56). Pd-PEPPSI-IPent (1) was also able to effectively promote the coupling of a diverse array of weakly nucleophilic anilines (leading to 58−62) with electron-rich aryl chlorides, the most challenging electronic combination for aryl aminations, compared to Pd-PEPPSI-IPr (2), which was almost completely inactive in these cases (Table 5). However, these reactions required high temperature (110 °C) to achieve modest yields (40−79%). In light of the modest reactivity of Pd-PEPPSI-IPent (1) in coupling electron-deficient anilines, backbone modification of the NHC ligand was investigated with the hypothesis that placing electron-withdrawing groups on the backbone would render the metal centre more electrophilic, similar in concept to the precatalyst development in the sulfination work.14 In doing so, the equilibrium of the amine coordination step would favor

In practice, the backbone-modified IPr ligands do generate more active catalysts in aryl aminations with anilines; however, the observed improvement in catalysis was essentially independent of the donating ability of the backbone substituent, suggesting that the enhancement is actually steric and not electronic in origin.33 Following this logic, a more rewarding path forward would be to modify the backbone of the IPent ligand, given its superior ancillary effects relative to the IPr ligand (vide supra).14 The IPentCl analogue (84, see Table 9) was synthesized and evaluated in the challenging coupling of electron-deficient 184



Review

anilines with electron-rich aryl chlorides (Table 6). Pd-PEPPSIIPentCl (84) was found to be significantly more active than the

aminations involving electron-deficient anilines and electronrich aryl chlorides. Although the answers to these questions are not yet clear, what is known is that the answers are not one and the same. For aryl aminations with alkylamines, the RDS is deprotonation of the NHC−Pd−ammonium complex (K → M). This is consistent with the maximum reaction rate being first order in carbonate and zeroth order in amine and aryl halide.21 In the above stated case, it appears then that the IPent−Pd−ammonium complex has a lower ef fective pKa relative to that of the analogous IPr−Pd−ammonium complex, which could account for the improved catalyst reactivity. This conjecture assumes that the ground state energy leading to the transition state of the rate-determining deprotonation step is higher for IPent than it is for IPr due to the increased steric environment of the former. Although a possibility, the mechanistic underpinnings for IPent-Pd vs IPr-Pd have not been conclusively fleshed out. In the case of aryl amination with anilines, the RDS is no longer deprotonation but rather RE (L → J). This is consistent with the maximum reaction rate being first order in aniline and carbonate and zeroth order in aryl halide.32 This is also reasonable as the pKa of anilines relative to that of amines is lower by approximately 10 logarithmic units. As such, deprotonation by carbonate base is now rapid and less dependent on the electronic structure of the OA partner. However, the OA partner does still play a role in the RE step, as the more electron-deficient derivatives improve the overall reaction kinetics. This may be due to the better leaving group ability of these arenes during RE. In contrast, we have observed that more nucleophilic anilines have overall higher reaction rates. This is thought to be a result of their better reducing power during the RE step, with the electronic ‘assist’ placing less impetus on ligand sterics (i.e., Pd-PEPPSI-IPr (2) is adequate for these couplings). In turn, when the aniline does not possess the required basicity for the reduction of PdII (i.e., Table 6), the catalyst must accommodate for this shortcoming through ligand sterics. Hence, the IPentCl ligand is superior to the IPent ligand in the coupling of electron-deficient anilines. Whatever the detailed mechanistic underpinnings are eventually elucidated to be, the empirical data reveals that PdPEPPSI-IPentCl (84) is one of the most active catalysts reported to date for aryl aminations. 2.2. Synthesis of Triarylamines by Way of Aryl Aminations. A previous result (see Table 6) wherein an aryl amination in the presence of Pd-PEPPSI-IPentCl (84) led to the formation of a small amount of triarylamine 67 by over-reaction prompted a follow-up study to assess the feasibility of a general coupling protocol for the synthesis of these difficult-to-form and valuable reaction products. Under a given set of amination reaction conditions involving primary anilines with a stoichiometric amount of aryl halide, the exclusive coupling product will almost always be the secondary aniline. The selectivity results because the diarylamine products are more challenging coupling partners than even the most deactivated primary anilines due to the presence of the second aryl ring, which severely curtails the nucleophilicity of the amine. However, as the results show, in the presence of a highly active catalyst and an excess of aryl halide, the preparation of symmetric triarylamines (67−76) is easily accessible (Table 7).35 In addition, one can prepare unsymmetrical triarylamines such as 78−80 through the reaction of a diarylamine (i.e., 77) with an aryl chloride (Table 8). In summary, under a standard

Table 6. Comparative study of Pd-PEPPSI-IPent (1) and PdPEPPSI-IPentCl (84) in aryl aminations with electrondeficient anilines and electron-rich aryl chlorides

unmodified IPent complex (1), effectively coupling both 3trifluoromethylaniline and 3,4,5-trifluoroaniline with 4-chloroanisole, furnishing 58 and 59 in 96 and 94% yield, respectively. Remarkably, even strongly deactivated pentafluoroaniline could be coupled with 4-chloroanisole in 58% yield (leading to 63), whereas no reaction was observed with Pd-PEPPSI-IPent (1). To our knowledge, this is the first example of Pd-catalyzed arylation of pentafluoroaniline, demonstrating the unprecedented reactivity of Pd-PEPPSI-IPentCl (84). 2.1. Mechanistic Considerations. It is not obvious how the increased steric topography around Pd for Pd-PEPPSIIPentCl (84) relative to Pd-PEPPSI-IPent (1) would have such a dramatic effect on aryl amination chemistry involving electronically unreactive arenes (i.e., electron-deficient anilines and electron-rich aryl chlorides). For aryl aminations involving primary and secondary alkyl amines, kinetic studies show that the rate-determining step (RDS) is deprotonation of the Pd− ammonium complex (K in Scheme 3).21,30,32 Whereas KOtBu is sufficiently basic to complete this step, Cs2CO3 relies more heavily upon the electronic nature of the aryl halide.29,34 This dependence exists because, once oxidative addition is complete, the electrophile itself becomes a ligand on Pd that can attenuate the Lewis acidity of Pd which in turn affects the Brønsted acidity of the Pd−ammonium complex (see intermediate K). Electron-withdrawing groups on the OA coupling partner lower the effective pKa of the Pd−ammonium adduct and thereby improve the kinetics of the RDS. Interestingly, in the presence of Pd-PEPPSI-IPent (1), both reactions reach near quantitative conversion, although the latter requires a longer reaction time to do so. Since deprotonation is the RDS and provided that the σ-donating ability of both IPr and IPent are not significantly different, why then does an increase in ligand sterics help facilitate the kinetics of deprotonation? Recall that this is similar to the question posed at the beginning of this section, only in that case the increased sterics of Pd-PEPPSI-IPentCl (84) over those of Pd-PEPPSI-IPent (1) have a dramatic effect in aryl 185



Review

a mild base and low reaction temperature offers unique opportunities for coupling substrates containing highly sensitive functional groups. Moreover, the developed methodology would be more amenable to process chemistry. In approaching this challenge, we took a page from our group’s recent work on the low-temperature Pd-PEPPSI-catalyzed sulfination chemistry in which the structure of the pyridine ligand in the precatalyst was found to have a considerable impact on catalyst performance (vide supra).14 Although the precatalyst activation pathways and catalytic cycles are somewhat different, it was reasoned that a similar dependence on the pyridine substitution might be manifested in low-temperature amination chemistry. To this end, a number of Pd-PEPPSI-IPentCl precatalysts with various pyridine ligands (35 and 84−87) were prepared and screened in the coupling of 4-chloroanisole (81) with 3,4,5trifluoroaniline (82) at room temperature to generate diarylamine 83an extremely challenging coupling for any catalyst even at 110 °C! (Table 11). The standard 3-chloropyridine catalyst Pd-PEPPSI-IPentCl (84) exhibited good reactivity at this temperature, furnishing 55% conversion to product after 24 h (see Table 9). The fact that Pd-PEPPSI-IPentCl (84) is as reactive at room temperature as Pd-PEPPSI-IPent (1) is at 80 °C is a testament to the sizable effect that backbone substitution has on modulating the performance of this NHC ligand. Switching to pyridine complex 85 resulted in a 15% jump in conversion to 70%. Placing a methyl or ethyl group in the ortho position of the pyridine resulted in another 10% improvement in conversion (35 and 86); however, employing precatalyst 87 with bulkier 2,6-dimethylpyridine resulted in only 39% conversion to product. Interestingly, 92% of 87 was recovered unreacted after 24 h; thus, only 8% of the 3 mol % of catalyst added was responsible for the observed catalysis. This implies that activation of 87 is slow, but once activated, 87 proceeds to catalyze product formation with a very high turnover frequency. From this study, Pd-PEPPSI-IPentCl-o-picoline (35) was identified as one of the most reactive precatalysts for room temperature aryl amination, which is coincidentally the same complex that was identified as the most active for room temperature aryl sulfination.14 The exact role of the pyridine ligand in improving catalyst reactivity in aryl aminations is currently under investigation. Unexpectedly, these reactions were found to be somewhat sensitive to the level of dissolved O2 in commercially available anhydrous DME. The initial pyridine optimization studies were

Table 7. Formation of symmetric triarylamines with PdPEPPSI-IPentCl (84) in the presence of excess aryl chloride and aniline derivatives

Table 8. Formation of asymmetric triarylamines with PdPEPPSI-IPentCl (84) using a diarylamine precursor

Table 9. Pyridine optimization of the IPentCl-Pd-based precatalysts in room temperature aryl aminations

set of reaction conditions, Pd-PEPPSI-IPentCl (84) facilitated the coupling of primary and secondary anilines with an aryl chloride, leading to the formation of symmetric (Table 7) and asymmetric (Table 8) triarylamines in good yield at 80 °C using either KOtBu or Cs2CO3 base. 2.3. Low-Temperature Aryl Aminations with PdPEPPSI-IPentCl. Given the superior activity of Pd-PEPPSIIPentCl (84) relative to Pd-PEPPSI-IPent (1) and all other evaluated NHC-based Pd catalysts, the question arose as to whether it would be possible to conduct these amination reactions at lower temperature. If successful, the combination of 186



Review

trifluoromethylaniline, forming aminated products 88 and 89 in excellent yields. Due to the limited solubility of the aryl chloride from which 89 was derived, this reaction was heated gently at 45 °C to ensure complete conversion to product. 2.4. Development of a Functional Group ‘Safe’ Alkoxide Base. As a consequence of the aggressive reactivity of alkoxide bases, the tolerance of various sensitive functional groups, such as ketones and esters, to the reaction conditions for aryl aminations is limited. To circumvent this drawback, the use of carbonate bases has gained in popularity (vide supra). Still, the amine-coordination/deprotonation step for many amination protocols shows a first-order dependence on base, and the use of metal alkoxides is required to obtain acceptable product yields in short reaction times. This is evident when comparing the relative pKa’s for the Pd−anilinium complex (∼8−10) with bicarbonate (∼10.3) and tBuOH (∼18). Moreover, the limited solubility of carbonate salts in organic solvent further exacerbates the already slow acid−base chemistry. In fact, deprotonation by Cs2CO3 has been shown to be a surface-mediated heterogeneous reaction, wherein the particle size and thus surface area of Cs2CO3 can affect the success of the amination protocol.37 Instead, a fully soluble, weakly nucleophilic base with a conjugate acid pKa between 11−15 would provide the ideal balance between basicity and functional group tolerance. To that end, potassium 2,2,5,7,8pentamethylchroman-6-oxide (92), a truncated version of αtocopherol (vitamin E), was developed and evaluated on the predication that the cyclic ether serves as a conformational lock to force one of the ether oxygen lone pairs into conjugation with the aromatic ring.38 In doing so, the pKa is raised from 10.2 (phenol) to 11.4. Indeed, the activity of 92 is significantly improved relative to that of the acyclic derivative 93 in the aryl amination of 4-chlorotoluene with morpholine in the presence of Pd-PEPPSI-IPent at 80 °C. To demonstrate functional group tolerance, 92 was evaluated directly against KOtBu in Pdcatalyzed aryl aminations leading to diarylamines (94−99) using base-sensitive coupling partners under two different sets of reaction conditions (Table 11). For the reactions involving KOtBu, the mass balance of starting material containing the sensitive functionality could not be assigned, suggesting decomposition in the presence of this reactive base.

successfully conducted using standard Schlenk techniques and commercial anhydrous DME. However, when switching to a new solvent bottle from the same supplier, the reactions essentially shut down. Reactivity could only be recovered if the solvent was degassed via three cycles of freeze−pump−thaw and the entire reaction was set up in an Ar-filled glovebox. Thus, to ensure consistently reproducible results, all room temperature aryl aminations were conducted under a strictly O2-free environment. This increased sensitivity might be due to competitive oxidative addition of (NHC)Pd(0)Ln towards molecular oxygen, forming unreactive (NHC)Pd(O2)L complexes which have been documented in the literature.36 For example, Fantasia and Nolan recently reported that peroxo complex (IPr)Pd(PPh3)(O2) forms rapidly and irreversibly from (IPr)Pd(PPh3) upon exposure to atmospheric O2 and is stable as such for many days without observable decomposition. The stability of these off-cycle intermediates at room temperature may account for the heightened O2 sensitivity of aminations conducted at room temperature relative to those conducted at elevated temperatures in which sufficient thermal energy is available to drive off the Pd-bound O2. To demonstrate the scope of this new catalyst at room temperature, a variety of functionalized, electron-rich aryl chlorides were coupled with a series of electron-deficient anilines (Table 10). Good to excellent yields were achieved Table 10. Examining the substrate scope of room temperature aryl aminations catalyzed by Pd-PEPPSIIPentCl-o-picoline (35)

3. COMPARING REACTIVITY OF Pd-PEPPSI PRECATALYSTS WITH OTHER Pd−NHC COMPLEXES Many catalysts have been developed and reported as being “highly active” in the cross-coupling literature. To those working in the field, a highly active catalyst would be defined as one that is able to conduct couplings that other catalysts simply cannot, or the ability to couple substrates together under signif icantly milder conditions than is the current state-of-theart in the field. Catalyst turnover numbers (TON) are sometimes used to compare the relative reactivity between different catalysts. However, TONs speak more to the stability of a catalyst and its ability to remain on cycle and is not a suitable metric to use in the relative ranking of the reactivity of a catalyst. Moreover, these TON comparisons rarely take into account the reaction conditions, which must be identical to hold any comparative value. It is more suitable to use the turnover frequency (TOF) of a catalyst, which is a better representation for the efficiency of a catalyst to generate product while it is active and on cycle, for drawing any comparative conclusions. For example, an infinitely stable

when coupling 4-chloroanisole with polyfluorinated anilines (leading to 58 and 91) and methyl 4-aminobenzoate (leading to 61). A chemoselective amination of 4-chlorophenylboronic acid pinacol ester was also achieved, yielding 91 in good yield without a trace of biaryl side products that might be observed at higher temperatures. Finally, higher-molecular weight aryl chlorides of potential medicinal interest were coupled with 3187



Review

Table 11. Aryl aminations with sensitive functional groups in the presence of optimized alkoxide base 92

Scheme 4. Direct comparison of pre-catalysts Pd-PEPPSIIPr, Pd-PEPPSI-SIPr and Pd-PEPPSI-IPr* in the coupling of morpholine with 4-chlorotoluene at room temperature

couplings but using different reaction conditions (Table 12).40 The reactivity of Pd-PEPPSI-IPr*OMe (101) was reported by the Table 12. Published comparison of Pd-PEPPSI-IPr (2) with Pd-PEPPSI-IPr*OMe (101) under different reaction conditions

a

Comparative yields obtained with KOtBu are provided. bComparison between catalysts under different reactions conditions carry little value.

catalyst may have a high TON, but if the TOF is low, then conversion to product will limit its practical use. Direct comparisons have been drawn between the reactivity of the Pd-PEPPSI-IPent precatalyst (1) and Pd(IPr*) derivatives (100−102) in both amination- and sulfination-type couplings (Figure 3). Nolan and co-workers compared the reactivity of a

When the same reaction was conducted in DME at 80 °C, 62% yield was obtained. bWhen the same reaction was conducted in DME at 80 °C using 85, 62% yield was obtained. a

Figure 3. Pd−NHC precatalysts reported by Nolan and co-workers.

authors as high and on par with that of Pd-PEPPSI-IPent (1) using the comparatively low catalyst loading as their justification. Given that different reaction conditions were used, comparisons in reactivity cannot be based on TON. Aryl amination reaction conditions employing soluble, highly aggressive tert-butoxide and its derivatives cannot be drawn into comparisons with reactions performed using a mild, insoluble carbonate base. For example, the reaction in Scheme 4 requires approximately 24 h using Pd-PEPPSI-IPr (2) to complete using carbonate base at 80 °C; the same reaction using KOtBu completes in just 15 s at RT!21 If the same catalyst performs vastly differently under different reaction conditions, any attempted comparison between catalysts under different reactions conditions holds little value. The [Pd(IPr*OMe)(cin)Cl)] (102) variant of IPr* has also very recently been applied to sulfination chemistry with aryl

series of PEPPSI precatalysts in the amination reaction shown in Scheme 4.39,40 They found that the rates of aryl amination for this particular substrate pairing were identical under the reaction conditions used (KOtBu base, DME solvent, RT, 1% catalyst load) for Pd-PEPPSI-IPr (1), Pd-PEPPSI-SIPr, and PdPEPPSI-IPr* (100), and that all three catalysts led to a similar level of conversion (∼80%) after 6 h.39 The conversion over time plots are suggestive of a similar TON and TOF among the three catalysts under these commonly employed aryl amination reaction conditions. In a related study, the same group reported the couplings of chloroanisole to a selection of aniline derivatives (leading to 58, 60, and 103) using modified IPr ligands, and comparisons were made to the performance of Pd-PEPPSI-IPent (1) in the same 188



Review

and alkyl sulfides.41 In all cases, the reactions required 110 °C in dioxane solvent with KOtBu base to achieve suitable catalyst TON. A wide sampling of the same couplings from this work have been achieved using Pd-PEPPSI-IPent (1) at RT. Further, as has been discussed above, an even wider group of substrates that are profoundly sterically and electronically deactivated can also now be routinely coupled at RT using Pd-PEPPSI-IPentClopicoline (35).14 In a direct comparison study for the coupling of 2,6dimethylchlorobenzene (104) with benzenethiol (26) under identical reaction conditions (Table 13), Pd-PEPPSI-IPent (1)

4.0. CONCLUSIONS A series of Pd-PEPPSI complexes featuring NHC- and pyridinemodified ligands were evaluated for activity in aryl sulfinations and aryl aminations. All NHC backbone-modified complexes exhibited enhanced reactivity relative to their unmodified counterparts, regardless of the electron-withdrawing or -releasing ability of the backbone substituents. This implies that the effect imparted by the NHC backbone substituents is primarily steric in origin and appears to corroborate the current mechanistic understandings for the respective catalytic cycles. The substitution on the pyridine ligand was also found to be important, with 2-picoline ligand being the most effective for both aryl aminations and sulfinations. A more thorough understanding of the role that the pyridine ligand plays in aryl sulfinations and aminations is being examined. Since their introduction in 2008, the PEPPSI style of Pd− NHC precatalysts have been demonstrated to be some of the most active catalysts available for cross-couplings. They are efficient catalysts under some of the mildest reaction conditions reported and have served well in widening the substrate scope to include functionalized electrophiles and nucleophiles that historically had proven very challenging to couple. Several PdPEPPSI catalysts are commercially available from Sigma Aldrich,42 and as of the date of this submission, they have found widespread application as evidenced by their use in 97 patents or patent applications.43

Table 13. Direct comparison between Pd-PEPPSI-IPr* (100) and IPent (1, 33) PEPPSI scaffolds under identical reaction conditions

entry

catalyst

conversion (%)a

1 2 3

Pd-PEPPSI-IPent (1) Pd-PEPPSI-IPr* (100) Pd-PEPPSI-IPent-py (33)

quant. traceb quant.c

■

a

Reactions were performed in a glovebox using degassed solvent. Conversions are based on 1H NMR spectroscopy. b2-Chloro-1,3dimethylbenzene was quantitatively reduced to o-xylene. cReaction was complete within 5 min.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] Notes

The authors declare the following competing financial interest(s): Some of the catalysts reported in this manuscript are distributed by Sigma-Aldrich, from which the PI receives a royalty payment.

furnished quantitative conversion to 27 (entry 1) while PdPEPPSI-IPr* (100) provided only trace product. The mass balance for the latter is predominately the reduced aryl halide (entry 2). Under the same reaction conditions, Pd-PEPPSIIPent-py (33) provided quantitative conversion to product in just 5 min (entry 3). These results clearly illustrate the effect that ligand structure can have on aryl sulfinations and the importance of using identical reaction conditions (an ‘applesto-apples’ approach) for adequately ranking the relative reactivity of a set of catalysts. On the basis of the proven ability of Pd−NHC complexes to effectively conduct both C−N and C−S coupling, we propose the catalyst performance rank order given in Figure 4. Again, this assignment is based on relative TOF for couplings that work for multiple catalysts, the relative demonstrated ease of the coupling (i.e., low temperature, mild base, etc.), and most importantly, the demonstrated ability in specific cases to effect couplings that other catalysts have not demonstrated the ability to do.

■

REFERENCES

(1) (a) Muci, A. R.; Buchwald, S. L. Top. Curr. Chem. 2002, 219, 131. (b) Surry, D. S.; Buchwald, S. L. Angew. Chem., Int. Ed. 2008, 47, 6338. (2) Hartwig, J. F. Acc. Chem. Res. 2008, 41, 1534. (3) Valente, C.; Ç alimsiz, S.; Hoi, K. H.; Mallik, D.; Sayah, M.; Organ, M. G. Angew. Chem., Int. Ed. 2012, 51, 3314. (4) (a) De Martino, G.; Edler, M. C.; La Regina, G.; Coluccia, A.; Barbera, M. C.; Barrow, D.; Nicholson, R. I.; Chiosis, G.; Brancale, A.; Hamel, E.; Artico, M.; Silvestri, R. J. Med. Chem. 2006, 49, 947. (b) Bagley, M. C.; Davis, T.; Dix, M. C.; Fusillo, V.; Pigeaux, M.; Rokicki, M. J.; Kipling, D. J. Org. Chem. 2009, 74, 8336. (5) Fernandez-Rodrigues, M. A.; Hartwig, J. F. Chem.Eur. J. 2010, 16, 2355. (6) Organic Chemistry of Sulfur, 1st ed.; Oae, S., Ed.; Plenum Press: New York, 1977.

Figure 4. Proposed ranking of reactivity of selected Pd−NHC catalysts in C−N and C−S cross-coupling. 189



Review

(7) Herrero, M. T.; SanMartin, R.; Domínguez, E. Tetrahedron 2009, 65, 1500. (8) Carril, M.; SanMartin, R.; Domínguez, E.; Tellitu, I. Chem.Eur. J. 2007, 13, 5100. (9) Fernández-Rodríguez, M. A.; Shen, Q.; Hartwig, J. F. Chem. Eur. J. 2006, 12, 7782. (10) Alvaro, E.; Hartwig, J. F. J. Am. Chem. Soc. 2009, 131, 7858. (11) Murata, M.; Buchwald, S. L. Tetrahedron 2004, 60, 7397. (12) Kantchev, E. A. B.; O’Brien, C. J.; Organ, M. G. Aldrichimica Acta 2006, 39, 97. Kantchev, E. A. B.; O’Brien, C. J.; Organ, M. G. Angew. Chem., Int. Ed. 2007, 46, 2768. (13) Sayah, M.; Organ, M. G. Chem.Eur. J. 2011, 17, 11719. (14) Sayah, M.; Lough, A. J.; Organ, M. G. Chem.Eur. J. 2013, 19, 2749. (15) Hartwig, J. F. Inorg. Chem. 2007, 46, 1936. (16) (a) Hadei, N.; Kantchev, E. A. B.; O’Brien, C. J.; Organ, M. G. Org. Lett. 2005, 7, 3805. (b) Ç alimsiz, S.; Sayah, M.; Mallik, D.; Organ, M. G. Angew. Chem., Int. Ed. 2010, 49, 2014. (c) Valente, C.; Belowich, M. E.; Hadei, N.; Organ, M. G. Eur. J. Org. Chem. 2010, 4343. (d) Ç alimsiz, S.; Organ, M. G. Chem. Commun. 2011, 47, 5181. (e) Hadei, N.; Kantchev, E. A. B.; O’Brien, C. J.; Organ, M. G. J. Org. Chem. 2005, 70, 8503. (f) Organ, M. G.; Avola, S.; Dubovyk, I.; Hadei, N.; Kantchev, E. A. B.; O’Brien, C. J.; Valente, C. Chem.Eur. J. 2006, 12, 4749. (g) Achonduh, G. T.; Hadei, N.; Valente, C.; Avola, S.; O’Brien, C. J.; Organ, M. G. Chem. Commun. 2010, 46, 4109. (17) Hadei, N.; Kantchev, E. A. B.; O’Brien, C. J.; Organ, M. G. Org. Lett. 2005, 7, 1991. (18) (a) O’Brien, C. J.; Kantchev, E. A. B.; Valente, C.; Hadei, N.; Chass, G. A.; Lough, A.; Hopkinson, A. C.; Organ, M. G. Chem.Eur. J. 2006, 12, 4743. (b) Organ, M. G.; Ç alimsiz, S.; Sayah, M.; Hoi, K. H.; Lough, A. J. Angew. Chem., Int. Ed. 2009, 48, 2383. (19) Dowlut, M.; Mallik, D.; Organ, M. G. Chem.Eur. J. 2010, 16, 4279. (20) Organ, M. G.; Abdel-Hadi, M.; Avola, S.; Hadei, N.; Nasielski, J.; O’Brien, C. J.; Valente, C. Chem.Eur. J. 2007, 13, 150. (21) Hoi, K. H.; Ç alimsiz, S.; Froese, R. D. J.; Hopkinson, A. C.; Organ, M. G. Chem.Eur. J. 2011, 17, 3086. (22) Love, J. A.; Morgan, J. P.; Trnka, T. M.; Grubbs, R. H. Angew. Chem., Int. Ed. 2002, 41, 4035. (23) Pompeo, M.; Froese, R. D. J.; Hadei, N.; Organ, M. G. Angew. Chem., Int. Ed. 2012, 51, 11354. (24) Sayah, M.; Organ, M. G. Chem. Eur. J. 2013, 19, 16196−16199. (25) (a) Decker, M.; Si, Y.-G.; Knapp, B. I.; Bidlack, J. M.; Neumever, J. L. J. Med. Chem. 2010, 53, 402. (b) Nicolaou, K. C.; Bulger, P. G.; Sarlah, D. Angew. Chem., Int. Ed. 2005, 44, 4442. (c) Torborg, C.; Beller, M. Adv. Synth. Catal. 2009, 351, 3027. (26) Martin, R.; Buchwald, S. L. Acc. Chem. Res. 2008, 41, 1461. (27) Glorius, F. Top. Organomet. Chem 2007, 21, 1. (28) (a) Littke, A. F.; Fu, G. C. Angew. Chem., Int. Ed. 1998, 37, 3387. (b) Yamamoto, T.; Nishiyama, M.; Koie, Y. Tetrahedron Lett. 1998, 39, 2367. (c) Urgaonkar, S.; Xu, J.-H.; Verkade, J. G. J. Org. Chem. 2003, 68, 8416. (d) Reddy, C. V.; Kingston, J. V.; Verkade, J. G. J. Org. Chem. 2008, 73, 3047. (e) Ratabaoul, F.; Zapf, A.; Jackstell, R.; Harkal, S.; Riermeier, T.; Monsees, A.; Dingerdissen, U.; Beller, M. Chem.Eur. J. 2004, 10, 2983. (f) So, C. M.; Zhou, Z.; Lau, C. P.; Kwong, F. Y. Angew. Chem., Int. Ed. 2008, 47, 6402. (g) Lundgren, R. J.; Peters, B. D.; Alsabeh, P. G.; Stradiotto, M. Angew. Chem., Int. Ed. 2010, 49, 4071. (29) Shen, Q.; Shekhar, S.; Stambuli, J. P.; Hartwig, J. F. Angew. Chem., Int. Ed. 2005, 44, 1371. (30) Organ, M. G.; Abdel-Hadi, M.; Avola, S.; Dubovyk, I.; Hadei, N.; Kantchev, E. A. B.; O’Brien, C. J.; Valente, C. Chem.Eur. J. 2008, 14, 2443. (31) Surry, D. S.; Buchwald, S. L. Chem. Sci. 2011, 2, 27. (32) Hoi, K. H.; Ç alimsiz, S.; Froese, R. D. J.; Hopkinson, A. C.; Organ, M. G. Chem.Eur. J. 2012, 18, 145. (33) Pompeo, M.; Organ, M. G. Unpublished results. (34) (a) Wolfe, J. P.; Tomari, H.; Sadighi, J. P.; Yin, J.; Buchwald, S. L. J. Org. Chem. 2000, 65, 1158. (b) Fors, B. P.; Krattiger, P.; Strieter,

E.; Buchwald, S. L. Org. Lett. 2008, 10, 3505. (c) Hamann, B. C.; Hartwig, J. F. J. Am. Chem. Soc. 1998, 120, 7369. (d) Old, D. W.; Wolfe, J. P.; Buchwald, S. L. J. Am. Chem. Soc. 1998, 120, 9722. (e) Shen, Q.; Ogata, T.; Hartwig, J. F. J. Am. Chem. Soc. 2008, 130, 6586. (35) Hoi, K. H.; Coggan, J. A.; Organ, M. G. Chem.Eur. J. 2013, 19, 843. (36) Fantasia, S.; Nolan, S. P. Chem.Eur. J. 2008, 14, 6987. (37) Meyers, C.; Maes, B. U. W.; Loones, K. T. J.; Bal, G.; Lemière, G. L. F.; Dommisee, R. A. J. Org. Chem. 2004, 69, 6010. (38) Hoi, K. H.; Organ, M. G. Chem.Eur. J. 2012, 18, 804. (39) Chartoire, A.; Frogneux, X.; Boreux, A.; Slawin, A. M. Z.; Nolan, S. P. Organometallics 2012, 31, 6947. (40) Meiries, S.; Speck, K.; Cordes, D. B.; Slawin, A. M. Z.; Nolan, S. P. Organometallics 2013, 32, 330. (41) Bastug, G.; Nolan, S. P. J. Org. Chem. 2013, 78 (18), 9303− 9308. (42) For commercially available catalysts, see: http://www. sigmaaldrich.com Pd-PEPPSI-IPr (Item #669032) Pd-PEPPSI-IPent (Item #732117) and http://totalsynthesis.ca/products for Pd-PEPPSIIPentCl, Pd-PEPPSI-IPentCl-o-picoline, and other related catalysts. (43) Using the Thomson Innovation patent search engine with “PEPPSI” as the search string reveals 97 patents or patent applications within which a Pd-PEPPSI catalyst is utilized. Date of search 08/15/13.

190


CarbonâHeteroatom Coupling Using Pd-PEPPSI Complexes

Recommend Documents

CarbonâHeteroatom Coupling Using Pd-PEPPSI Complexes