Palladium-Catalyzed Hydroarylation of 1,3-Dienes with Boronic Esters via Reductive Formation of π-Allyl Palladium Intermediates under Oxidative Conditions

Abstract

A palladium-catalyzed reductive cross coupling of 1,3-dienes with boronic esters is reported in which a π-allyl Pd-species is generated directly from a 1,3 diene via a Pd-catalyzed aerobic alcohol oxidation. Both the scope of the process and the origin of a highly selective 1,2-addition are discussed.

Since the first report in 1957,¹ the π-allyl complexes of palladium have played a central role in Pd-catalysis due to the ease in which these complexes undergo nucleophilic substitution.² Classically, π-allyl Pd-species are most commonly generated from an allyl fragment containing a leaving group reacting with Pd(0)^2c,3 or from nucleophilic addition to an 1,3-diene promoted by Pd(II).^2e,4 More recently, efforts have been reported to directly access these intermediates via allylic oxidation.⁵ Based on our programmatic focus on the development of Pd-catalyzed alkene hydrofunctionalization reactions coupled to the reduction of O₂,⁶ we became interested in accessing π-allyl Pd-species reductively directly from 1,3-dienes. These substrates have performed rather poorly in our previous reports.⁶ To accomplish this, we hypothesized that a Pd-hydride B formed from the oxidation of an alcohol could be intercepted by a 1,3-diene yielding the desired π-allyl intermediate C (Scheme 1).⁷ Herein we report the successful generation of these intermediates from dienes and an in situ formed Pd-hydride and subsequent cross coupling with aryl boronic esters to accomplish a highly regioselective diene hydroarylation reaction using O₂ as the terminal oxidant.

Scheme 1.

Proposed mechanism for hydroarylation of 1, 3-dienes.

Diene 1a and boronic ester 2a were selected as the substrates for optimization and initially evaluated under the conditions previously reported for styrene hydroarylation using [Pd(SiPr)Cl₂]₂ as the catalyst (Table 1, entry 1).⁶ This catalyst system performs poorly with little conversion of the substrate which may be due to the excellent stability of N-heterocyclic carbene allyl palladium species.⁸ Therefore, we explored the use of Pd[(−)-sparteine]Cl₂, whose π-allyl complexes have been previously structurally characterized and successfully used in nucleophilic addition chemistry.⁹ The use of this catalyst leads to a significant improvement in conversion and GC yield of the desired product. It is important to note that the ratio of the 1,2 addition to the 1,4 addition product is >99:1 based on GC analysis. Removal of exogenous (−)-sparteine leads to rapid decomposition of the catalyst. Therefore, the amount of excess sparteine was increased with an observed improvement in selectivity and yield of the desired product (entry 3). Decreasing the amount of ^tBuOK (entry 4) as well as raising the temperature (entry 5) leads to a noticeable increase in the product yield. Three equivalents of the boronic ester are required due to the consumption of this reagent through oxidative homocoupling and formation of phenol by hydrogen peroxide produced from the reduction of O₂.^6c It should be noted that ^tBuOK is required for adequate yields of product.

Table 1.

Optimization for the hydroarylation of 1a.

Entry	Pd source	x	y	T/°C	% Conv.a	%3ab	3a : 4a
1	[Pd(SiPr)Cl2]2	6	6	55	7	3	5 : 1
2	Pd[(−)-sparteine]Cl2	6	6	55	70	42	22 : 1
3	Pd[(−)-sparteine]Cl2	20	6	55	58	45	19 : 1
4	Pd[(−)-sparteine]Cl2	20	0.5	55	96	63	>99 : 1
5	Pd[(−)-sparteine]Cl2	20	0.5	75	>99	76	25 : 1

^aMeasured by GC using an internal standard.

^bGC yield using an internal standard.

With this optimized catalyst system in hand, the substrate scope was explored (Table 2). First, the nature of the arylboronic ester was evaluated using diene 1a (products 3a–3h). Isolated yields were generally good to excellent and all of the reactions were highly regioselective for the 1,2-addition product (>95:5 by ¹H NMR). Ortho-substitution is allowed (3b) while the electronic nature of the boronic ester has a minimal impact on the yield of the reaction in contrast to our previous reports (3c–3h).^6c This is highlighted by the incorporation of an ester (3e), nitrile (3g), and an acid sensitive acetal group (3h). Several substituted diene derivatives in the reaction with boronic ester 2a were also examined. Moderate to good yield is achieved with simple hydrocarbons (3i–3l). An electron rich aryl substituted diene gave the corresponding product in high yield (3m) and a TBS protected alcohol was tolerated using this catalyst system giving the product in excellent yield (3n). It should be noted that acceptable yields are found only when using arylboronic esters.

Table 2.

Substrate scope.

Yields are average isolated yields of at least two experiments. a >95 to 5 ratio of the 1,2 to 1,4 hydroarylation products was measured by ¹H NMR.

Classically, the addition to π-allyl palladium complexes leads to a mixture of α- (1,2 addition in our case) and γ-coupled products.¹⁰ The observation that high selectivity for the 1,2-addition product is found is not only synthetically attractive but also prompted us to investigate the intermediacy of a π-allyl Pd-species. Therefore, a series of reactions were carried out. First, both stereochemically pure Z and E diene isomers of 1a were submitted to the reaction conditions. Although the observed rate of conversion for the E-isomer is 1.6 times faster than that of Z-isomer, both lead to only a single isomeric product, suggesting a π-allyl Pd-intermediate. The Z-isomer does not appreciably form the E-isomer under the reaction conditions. Secondly, submission of a conjugated diene 1h or a skipped diene 1i both give the same hydroarylation product again consistent with a presumed π-allyl Pd-intermediate. Conversion of 1i to ~ 10% 1h was found by GC during the reaction consistent with the formation of a Pd-hydride. Finally, an experiment was carried out, where diene 1a was mixed with Pd[(−)-sparteine]Cl₂ in the presence of isopropanol for 1 hour at 60 °C wherein a composition consistent with the [(−)-sparteine]Pd^II-π-allyl(Cl) complex was confirmed by ESI-HRMS. ¹¹

To probe the nature of the high 1,2 selectivity, simple hydrocarbon substituted 1,3-dienes with differential steric impact were evaluated. Dienes with smaller substituents lead to a poorer ratio of 1,2 to 1,4 addition products although the formation of both is indicative of a π-allyl palladium intermediate (Figure 1). To confirm that the selectivity has a steric origin, the log of ratio of regioisomers was correlated to Charton steric values of the substituents on the dienes.¹² A linear free energy relationship is observed consistent with steric effects dominating the selectivity in this reaction (Figure 1). However, the substrate containing a cyclohexyl substituent clearly does not fit in the correlation and the reason for this is not obvious.

Figure 1.

The linear free energy relationship between Charton steric parameters and log of ratio of regioisomers (1,2/ 1,4-addition).

In summary, we have developed a novel approach for the hydroarylation of 1,3-dienes by accessing π-allyl intermediates directly using a coupled aerobic alcohol oxidation to access a Pd-hydride. The scope of the process shows wide tolerance of functional groups on the boronic ester and on the diene. Moreover, high selectivity is observed for the 1,2 addition product which has been shown to be of a steric origin. While an enantioenriched chiral ligand is used in this chemistry, only poor enantioselectivity can be achieved (<20 %ee). Therefore, future work is focused on identification of new ligand classes to promote this reaction in high enantioselectivity.

Scheme 2.

Reactions to probe the intermediacy of a π-allyl Pd-species.