Mechanistic Study on Allylic Arylation in Water with Linear Polystyrene-Stabilized Pd and PdO Nanoparticles.

The catalytic cycle of allylic arylation in water catalyzed by linear polystyrene-stabilized Pd or PdO nanoparticles (PS-PdNPs or PS-PdONPs) was investigated. Stoichiometric stepwise reactions indicated that the reaction did not proceed stepwise on the surface of the catalyst. In the case of the reaction with PS-PdNPs, the leached Pd species is the catalytically active species and the reaction takes place through a similar reaction pathway accepted in the case of a complex catalyst. In contrast, allylic arylation using PS-PdONPs as a catalyst occurs via a Pd(II) catalytic cycle.

Introduction

Allylic arylation of allyl substrates with boron reagents is a powerful tool for carbon–carbon bond formation.[1−4] It is considered that the reaction proceeds in the following order: oxidative addition of the allyl substrate to Pd(0) to form a π-allyl palladium intermediate, transmetalation with the boron reagent, and reductive elimination to form the coupling product. Because of the formation of the π-allyl palladium intermediate in the first step, the use of either linear-type (e.g., cinnamyl esters) or branch-type (e.g., α-vinyl benzyl ester) substrates with arylboronic acid results in the same coupling product (e.g., 1,3-diarylpropene). On the other hand, a different mechanism for allylic arylation has been proposed in γ-selective substitution reactions using cationic palladium(II) species as a catalyst.[5]

Because transition-metal nanoparticles (NPs) are stable and exhibit high catalytic activities for various organic reactions in water, they have been considered to be green catalysts. Several kinds of Pd NPs that showed high catalytic activity for allylic arylation have been developed.[6] For example, Rhee et al. reported the allylic arylation of cinnamyl carbonates with arylboronic acids catalyzed by Pd NPs supported on a thermoresponsive polymer in water. The metalloenzyme-inspired polymeric imidazole Pd catalyst (MEPI-Pd) promoted the allylic arylation of allylic acetates and carbonates with tetraarylborates and arylboronic acids with a catalytic turn over number of 20 000–1 250 000. In this reaction system, 1,3-diphenylpropene was obtained from the reaction of either cinnamyl methylcarbonate (linear-type substrate) or phenyl vinyl carbinol carbonate (branch-type substrate) with phenylboronic acid, indicating that the reaction proceeded through the π-allyl palladium intermediate. The formation of 3,3-diarylpropene derivatives (branch-type product) has not been confirmed in most cases.

Although mechanistic studies on the metal NPs-catalyzed reaction had been focused on whether the reaction proceeds in solution or on the surface of the NPs, a new concept of a “Cocktail” of catalysts has been advocated recently.[7] In most cases, however, the reaction pathway is thought to be the same as that in the case of a complex catalyst. On the other hand, we discovered that polystyrene-stabilized Pd NPs (PS-PdNPs) did not produce coupling products in the reaction of aryl halides with aryltrimethoxysilanes (Hiyama coupling reaction), and Hiyama coupling reaction using polystyrene-stabilized PdO NPs (PS-PdONPs) as a catalyst proceeded through a different route from the commonly accepted one that starts from the oxidative addition of an aryl halide to a Pd(0) species.[8] This result suggests the reaction pathway or the reactivity of metal NPs is different from that in the case of a metal complex catalyst. Our continuing interest in the reactivity and mechanism of the catalytic reaction with metal NPs led us to examine allylic arylation in water using PS-PdNPs and PS-PdONPs as catalysts.

Results and Discussion

PS-PdNPs and PS-PdONPs were prepared using a previously reported method[8] and characterized by X-ray diffraction (XRD) and transmission electron microscopy.[9] When allylic arylation of cinnamyl methyl carbonate (linear-type substrate) with 4-methylphenylboronic acid was carried out in 1.5 mol/L K2CO3 aqueous solution at 50 °C for 5 h using PS-PdNPs as a catalyst, 3-(4-methylphenyl)-1-phenylpropene (linear-type product) and 3-(4-methylphenyl)-3-phenylpropene (branch-type product) were obtained in 68 and 31% yields, respectively (Scheme a). On the other hand, only 3-(4-methylphenyl)-1-phenylpropene was obtained in 72% yield from the reaction of methyl 1-phenyl-2-propenyl carbonate (branch-type substrate) with 4-methylphenylboronic acid (Scheme b). Furthermore, no coupling product was obtained in the reaction of cinnamyl acetate (linear-type substrate) with 4-methylphenylboronic acid (Scheme a), and the reaction using α-vinylbenzyl acetate (branch-type substrate) gave 3-(4-methylphenyl)-1-phenylpropene (linear-type product) in 97% yield (Scheme b). In addition, the reaction of cinnamyl acetate with 4-methylphenylboronic acid using PS-PdONPs as a catalyst took place smoothly to afford both linear- and branch-type products in 67 and 15% yields, respectively (Scheme c). These results suggest several routes may exist in allylic arylation in water using metal NPs as a catalyst.

Scheme 1

Allylic Arylation Using PS–Pd (or PdO) NPs in Water

To gain insight into the reaction mechanism of allylic arylation, stoichiometric stepwise reactions were performed (Scheme ). Solid (ReCat) and aqueous solution (Sol) were separated after the reaction of PS-PdNPs with 4-methylphenylboronic acid (or α-vinylbenzyl acetate). When ReCat was washed with water and diethyl ether, and then reacted with the other compound (α-vinylbenzyl acetate or 4-methylphenylboronic acid), no coupling product was obtained in either case. In contrast, 3-(4-methylphenyl)-1-phenylpropene was obtained after heating the mixture of Sol-1 and α-vinylbenzyl acetate (8% yield) or Sol-2 in the presence of α-vinylbenzyl acetate and 4-methylphenylboronic acid (68% yield).[10] These results are consistent with the fact that Pd species were detected in the solution phase after stirring the K2CO3 aqueous solution in the presence of PS-PdNPs with 4-methylphenylboronic acid or α-vinylbenzyl acetate (Table ). Indeed, a 99% yield of the coupling product was obtained from the reaction of α-vinylbenzyl acetate with 4-methylphenylboronic acid in the presence of a water-soluble Pd species, which was prepared by dissolving Pd(OAc)2 in 1.5 mol/L K2CO3 aqueous solution (Scheme ).[11] However, no Pd species was detected in the solution phase at middle of the reaction and after the catalytic reaction, indicating that the Pd species were restabilized by linear polystyrene immediately.[12] If Pd leaching occurred after oxidative addition of α-vinylbenzyl acetate, the reaction of the leached Pd species with 4-methylphenylboronic acid would give the coupling product because the allyl moiety is on Pd. However, no coupling product was obtained from the mixture of Sol-2 and 4-methylphenylboronic acid (Scheme ), suggesting that the allylic substrate does not involve in Pd leaching.

Scheme 2

Stepwise Reaction Using Stoichiometric Amounts of PS-PdNPs

Table 1

Leaching Test

entry	reagent	amount of Pd leached (%)a
1	4-methylphenylboronic acid	0.7
2	α-vinylbenzylacetate	2.9
3	α-vinylbenzylacetate 4-methylphenylboronic acid	<0.1

Average weight of Pd atoms detected by ICP-AES in the supernatant liquid after exposure of PS-PdNPs to different reagents in K2CO3 aqueous solution at 50 °C for 5 h.

Scheme 3

Reactivity of Water-Soluble Pd Species

Because no correlation between the yield of the coupling product and the reaction time was observed (Figure S6) when the reaction was monitored by 1H NMR, the reaction profile was investigated after preheating at 50 °C for 1 h (Figure ). Little change in the yield of the product was observed in allylic arylation using different amounts of α-vinylbenzyl acetate and 4-methylphenylboronic acid, suggesting that leaching of the Pd species into the reaction solution is the rate-determining step. In addition, little increase of the product yield was confirmed in the hot filtration test after preheating.[13] This result is consistent with the assumption that the leached Pd species are restabilized by linear polystyrene immediately.

Figure 1

Reaction profile for allylic arylation after preheating.

According to these results, a reaction mechanism is the same as a “Cocktail” of catalysts and that in the case of a complex catalyst,[14] that is, Pd species leach into the reaction medium, and then the allylic arylation takes place in the solution phase. After the corresponding product form via reductive elimination, leached Pd species are restabilized on the surface of the catalyst. Indeed, when the reaction of (2S,3E)-4-phenyl-3-buten-2-yl acetate with 4-methylphenylboronic acid was performed in the presence of PS-PdNPs as a catalyst, (3R,1E)-3-(4-methylphenyl)-1-phenyl-1-butene was obtained in 75% yield with 91% enantioselectivity (eq ). This result is consistent with the plausible reaction pathway shown in Scheme .

Scheme 4

Plausible Mechanism for Allylic Arylation Using PS-PdNPs as a Catalyst

As shown in Scheme a, no coupling product was obtained from the reaction of cinnamyl acetate with 4-methylphenylboronic acid using PS-PdNPs as a catalyst, although the reaction took place in the case of cinnamyl methyl carbonate or α-vinylbenzyl acetate. These differences would be attributable to the reactivity of these substrates.[15] On the other hand, the reaction of cinnamyl acetate with 4-methylphenylboronic acid using PS-PdONPs as a catalyst (Scheme c) afforded the coupling product, indicating the existence of another reaction pathway. In order to confirm the reaction route using PS-PdONPs as a catalyst, stoichiometric stepwise reactions were performed (Scheme ). The catalyst was separated from the solution after the reaction of PS-PdONPs with 4-methylphenylboronic acid (or cinnamyl acetate) and then reacted with the other compound (cinnamyl acetate or 4-methylphenylboronic acid).[16] No formation of the coupling product was confirmed in either case, suggesting that the reaction did not proceed stepwise on the surface of the catalyst. Although the coupling product was not obtained after heating the filtrate (Sol-3) in the presence of cinnamyl acetate, the reaction of cinnamyl acetate with 4-methylphenylboronic acid in the filtrate (Sol-4) gave 3-(4-methylphenyl)-1-phenylpropene (linear-type product) and 3-(4-methylphenyl)-3-phenylpropene (branch-type product) in 7 and 3% yields, respectively. When using a water-soluble Pd species as a catalyst, the yield of the coupling product was very low (Scheme ), and no coupling product was obtained in the hot filtration test (Scheme ). These results suggest that the leached Pd species is not the catalytically active species in this reaction system.

Scheme 5

Stepwise Reaction Using Stoichiometric Amounts of PS-PdONPs

Scheme 6

Reactivity of Water-Soluble Pd Species

Scheme 7

Hot Filtration Test

When the reaction profile was monitored using various amounts of cinnamyl acetate, the product yield was decreased with an increase in the amounts of cinnamyl acetate at the beginning of the reaction (Figure ). In contrast, the product yield was increased with increasing amount of 4-methylphenylboronic acid (Figure ). The final yield of the product was 62% when using 1.0 mmol of 4-methylphenylboronic acid, likely due to the formation of Pd(0) species on the surface of the catalyst.[17]

Figure 2

Reaction profile for allylic arylation with various amounts of cinnamyl acetate using PS-PdONPs as a catalyst.

Figure 3

Reaction profile for allylic arylation with various amounts of 4-methylphenylboronic acid using PS-PdONPs as a catalyst.

From the results of the preheating experiments (Scheme ), it was confirmed that the catalyst was not activated by 4-methylphenylboronic acid. Electron-rich cinnamyl acetate derivatives reacted more efficiently than electron-deficient ones (Table , entries 1–4).The existence of an electron-donating substituent on the aryl ring of the arylboronic acid accelerated the reaction because of the increase in nucleophilicity (entries 2 and 5–7). When the catalyst was reacted with cinnamyl acetate after the phenyl group was modified on the surface of the catalyst by the treatment with phenyl(trimethoxy)silane,[8a] no coupling product was observed (Scheme ). This result is consistent with the assumption that the reaction does not proceed stepwise on the surface of the catalyst. In addition, considering the results in Schemes and 6, the reaction takes place via a concerted mechanism on the surface of the catalyst.

Scheme 8

Preheating Experiment

Table 2

Effect of Substituents

entry	Ar1	Ar2	yield (l/b)a
1	4-MeOC6H4–	4-MeC6H4–	82% (94/6)
2	Ph	4-MeC6H4–	78% (83/17)
3	4-FC6H4–	4-MeC6H4–	46% (76/24)
4	4-NO2C6H4–	4-MeC6H4–	21% (86/14)
5	Ph	4-MeOC6H4–	93% (86/14)
6	Ph	Ph	68% (72/28)
7	Ph	4-CF3C6H4–	0% (-/-)

NMR yield.

Scheme 9

Stepwise Reaction on the Surface of the Catalyst

When the reaction profiles were investigated using the variable-time normalization analysis graphical methods[18] to gain further information on the mechanism of this reaction, an order of 1.0 for Pd and an order of 2.5 for arylboronic acid were obtained (Schemes S10 and S11).

A plausible reaction pathway is shown in Scheme . After coordination of the substrate onto the Pd(II) surface, nucleophilic attack of arylboronic acid (SN2 or SN2′) occurs to give the coupling product. Arylboronic acid would also function as a Lewis-acid catalyst to activate allylic acetate.[19] The order over two in arylboronic acid is consistent with the hypothesis that arylboronic acid acts as both of a nucleophile and a Lewis acid. In addition, the reason why the product yield was decreased with an increase in the amounts of cinnamyl acetate is arylboronic acid would act only as a Lewis-acid catalyst when more amount of substrate was used. The preheating experiment in the presence of arylboronic acid gave lower yield than the experiment in which only the catalyst was preheated (Scheme ), suggesting transmetalation step do not involve in this reaction route. When the reaction of (2S,3E)-4-phenyl-3-buten-2-yl acetate with 4-methylphenylboronic acid was performed in the presence of PS-PdONPs as a catalyst at 50 °C, the coupling product was obtained in 91% yield with 80% ee (eq ).[20] Although the enantioselectivity was slightly lower (81%), this result is consistent with the proposed reaction pathway.

Scheme 10

Plausible Reaction Pathway for Allylic Arylation Using PS-PdONPs as a Catalyst

Conclusions

In summary, the reaction mechanism for allylic arylation in water using polystyrene-stabilized Pd or PdO NPs as the catalyst was investigated. In the case of substrates that have higher reactivity for oxidative addition to Pd(0) species, allylic arylation took place smoothly using PS-PdNPs as a catalyst. On the other hand, allylic substrates with lower reactivity for oxidative addition reacted with arylboronic acid in the presence of PS-PdONPs. Stoichiometric stepwise reactions indicated that the reaction did not proceed stepwise on the surface of the catalyst. In the case of the reaction with PS-PdNPs, leaching of Pd species into the reaction solution was the rate-determining step, and the reaction took place through the commonly accepted route starting from the oxidative addition of the allylic substrate. In contrast, allylic arylation using PS-PdONPs as a catalyst occurred by nucleophilic attack of arylboronic acid after coordination of the Pd(II) species to the alkene moiety. Further investigations on the reaction mechanism are underway in our laboratory.

Experimental Section

A 400 MHz NMR spectrometer (ECZ400S, JEOL Ltd., Tokyo, Japan) was used for 1H NMR spectra in CDCl3 with TMS (δ = 0) as an internal standard. Inductively coupled plasma-atomic emission spectroscopy (ICP-AES) was performed using ICPS-8100 (Shimadzu Co., Kyoto, Japan). A JEM 2100F transmission electron microscope (JEOL Ltd., Tokyo, Japan) was used for investigation of Pd NPs. The samples were prepared by placing a drop of the solution on carbon-coated copper grids and allowed to dry in air. A Rigaku RINT 2500 diffractometer (Cu Kα radiation) equipped with a monochromator was used for powder XRDs. Polystyrene of narrow molecular weight distribution standards and Pd(OAc)2 were purchased from Tosoh Co., Ltd. (Tokyo, Japan) and Sigma-Aldrich Co. (Missouri, USA), respectively.

Preparation of PS-PdNPs[8b]

A stirring bar, 13 mg of polystyrene (0.13 mmol of styrene unit), Pd(OAc)2 8.4 mg (37 μmol), 4-methylphenylboronic acid 0.012 g (88 μmol), and 1.5 mol·L–1 aqueous KOH solution (3 mL) were added to a screw-capped vial. The aqueous solution was removed after stirring at 90 °C for 5 h. Subsequently, the polystyrene stabilized Pd NPs were washed with water (5 × 3.0 mL), MeOH (1 × 3.0 mL), and Et2O (5 × 3.0 mL). The loading of Pd (2.51 mmol/g) was confirmed by ICP-AES analysis.

Preparation of PS-PdONPs[8a]

A stirring bar, 13 mg of polystyrene (0.13 mmol of styrene unit), Pd(OAc)2 8.4 mg (37 μmol), and 1.5 mol·L–1 aqueous K2CO3 solution (2 mL) were added to a screw-capped vial. The aqueous solution was removed after stirring at 90 °C for 5 h. Subsequently, the polystyrene stabilized PdO NPs were washed with water (5 × 3.0 mL), MeOH (1 × 3.0 mL), and Et2O (5 × 3.0 mL). The loading of Pd (2.48 mmol/g) was confirmed by ICP-AES analysis.

Typical Procedures for Stepwise Experiments[8a]

A stirring bar, PS-PdNPs (200 mg, 0.50 mmol of Pd), 4-methylphenylboronic acid (68 mg, 0.50 mmol), and 1.5 mol·L–1 aqueous K2CO3 solution (2 mL) were added to a screw-capped vial. After the mixture was stirred at 50 °C for 5 h, the aqueous phase (Sol-1) was transferred to the other screw-capped vial. In order to remove the residual 4-methylphenylboronic acid, the recovered catalyst (ReCat-1) was washed with H2O (3 × 3.0 mL) and diethyl ether (3 × 3.0 mL). α-Vinylbenzyl acetate (88 mg, 0.50 mmol) and 1.5 mol·L–1 aqueous K2CO3 solution (2 mL) were added to a screw-capped vial including ReCat-1; the mixture was stirred at 50 °C for 22.5 h. After separating the catalyst and the aqueous phase, the recovered catalyst was washed with H2O (3 × 3.0 mL) and diethyl ether (3 × 3.0 mL), which were then added to the aqueous phase. The aqueous phase was extracted with diethyl ether (8 × 5.0 mL). The combined organic extracts were dried over MgSO4 and concentrated with an evaporator. On the other hand, the aqueous phase (Sol-1) was filtered with a membrane filter and stirred at 50 °C for 18 h after addition of α-vinylbenzyl acetate (88 mg, 0.50 mmol). After the mixture was extracted with diethyl ether (8 × 5.0 mL), the organic extracts were dried over MgSO4 and concentrated with an evaporator.

Typical Procedures for Leaching Tests[8a]

A stirring bar, PS-PdNPs (3.0 mg, 7.5 μmol of Pd), α-vinylbenzyl acetate (88 mg, 0.50 mmol), and 1.5 mol·L–1 aqueous K2CO3 solution (1 mL) were added to a screw-capped vial. After stirring at 50 °C for 5 h, the aqueous phase was filtered with a membrane filter and adjusted to 10 mL by hydrochloric acid (1.0 mol·L–1). The amount of Pd metal (2.3 ppm) was confirmed by ICP-AES analysis.

General Procedure for Allylic Arylation

A stirring bar, α-vinylbenzyl acetate (88 mg, 0.5 mmol), 4-methylphenylboronic acid (102 mg, 0.75 mmol), PS-PdNPs (1.0 mg, 0.5 mol % of Pd), and 1.5 mol·L–1 aqueous K2CO3 solution (1 mL) were added to a screw-capped vial. After the mixture was stirred at 50 °C for 5 h, the catalyst and the aqueous phase were separated. The recovered catalyst was washed with H2O (5 × 3.0 mL) and diethyl ether (5 × 3.0 mL), which were then added to the aqueous phase. The aqueous phase was extracted with diethyl ether (8 × 5.0 mL). The combined organic extracts were dried over MgSO4 and concentrated with an evaporator. 3-(4-Methylphenyl)-1-phenylpropene was obtained after silica-gel column chromatography in 97% yield.