Palladium(II)-Catalyzed Dehydrogenative Alkenylation of Cyclic Enaminones via the Fujiwara-Moritani Reaction

Abstract

A new Pd(II)-catalyzed dehydrogenative alkenylation reaction involving two alkenes was developed. A variety of nonaromatic, cyclic enaminones were successfully coupled to primary and secondary alkenes yielding a series of unique 1,3-dienes. The generality of this transformation presents a useful strategy for directly cross-coupling alkenes and offers an attractive new approach to functionalize enaminones.

Direct C–H functionalization chemistry has seen significant progress during the last decade. ¹ Cross dehydrogenative coupling reactions that use two C–H bonds to form a new C–C bond are highly sought after, because these processes do not require pre-funtionalization and as a result have high atom economy.² The Fujiwara-Moritani reaction is a process by which aromatic substrates undergo an intermolecular dehydrogenative alkenylation. This reaction was initially developed using stoichiometric amounts of Pd(II) (Figure 1a), ³ but soon thereafter was shown to also take place with catalytic amounts of Pd(II) (Figure 1b).⁴ In recent years, the scope of the Fujiwara-Moritani has been expanded to a plethora of aromatic substrates.⁵ However, only few cases of alkenylation reactions involving two alkene C–H donors (Figure 1c) are known. They are limited to select substrates, require high Pd loading, and need long reaction time. ⁶ Herein, we report the development of an efficient Pd(II)-catalyzed Fujiwara-Moritani reaction, featuring nonaromatic cyclic enaminones that react with a variety of alkenes to furnish 5-alkenylated reaction products.

Figure 1.

Development of the Fujiwara-Moritani reaction.

In our quest to generate libraries for biological screening, we were particularly interested in functionalizing the cyclic enaminone nucleus I (Figure 2) because of its unique chemical and biological properties.⁷ Recently, we reported a Pd(II)-catalyzed direct arylation of cyclic enaminones with aryl trifluoroborates (Figure 2a). ⁸ Alkenyl trifluoroborates, however, did not furnish the desired alkenylated products. We speculated that the rate of transmetallation with the alkenyl reagents surpassed that of C5-palladation and as a result homocoupling depleted the alkenyl precursors and reoxidants.

Figure 2.

Pd(II)-cross coupling reactions of cyclic enaminones.

To address this problem, we examined the feasibility of a Fujiwara-Moritani reaction (Figure 2b) to access 5-alkenyl enaminone derivatives. Key to this strategy would be the use of alkenes with complementary reactivity. We envisioned that the palladated enaminone II could be orthogonally intercepted by an electron-deficient alkene to furnish product VI.

We selected enaminone 1 as the substrate and examined its reaction with tert-butyl acrylate (2a) (Table 1). Pd(OAc)₂ (10 mol %) was initially selected because of its well-established reactivity.^{8, 9} Next, DMF was identified as the best solvent (entries 1–4). Cu(OAc)₂ proved to be the most effective and economical reoxidant (entries 4–7). The incorporation of additives, KTFA specifically, notably increased the yield of the reactions (entries 8–12). Furthermore, the highest yields were observed when the reaction was run at 80 °C (entries 12– 14), and the reaction time was reduced from 24 h to 3 h (entries 12 and 15).

Table 1.

Optimization of the reaction conditions^a

entry	solvent	reoxidant	additive	temp (°C)	3a (%)b
1	tBuOH	Cu(OAc)2	--	80	55
2	DMSO	Cu(OAc)2	--	80	53
3	DMA	Cu(OAc)2	--	80	70
4	DMF	Cu(OAc)2	--	80	78
5	DMF	CuCl2	--	80	0
6	DMF	AgOAc	--	80	43
7	DMF	PhCO3tBuc	--	80	59
8	DMF	Cu(OAc)2	LiBF4	80	80
9	DMF	Cu(OAc)2	BiCl3	80	39
10	DMF	Cu(OAc)2	CsOAc	80	77
11	DMF	Cu(OAc)2	K2CO3	80	68
12	DMF	Cu(OAc)2	KTFA	80	85
13	DMF	Cu(OAc)2	KTFA	50	73
14	DMF	Cu(OAc)2	KTFA	110	69
15	DMF	Cu(OAc)2	KTFA	80	87d

^aReaction conditions unless otherwise specified: 1 (0.2 M), 2a (2 equiv), Pd(OAc)₂ (10 mol %), reoxidant (2 equiv), additive (1 equiv) under N₂ at 80 °C in 24 h. (PMP=para-methoxyphenyl)

^b1H NMR yield vs. Ph₃SiMe (1 equiv) as the internal standard.

^c1 equiv.

^dCompleted in 3 h. (Detailed optimization is available in the Supporting Information.)

With these optimized conditions, we were pleased to find that this direct alkenylation reaction was amenable to a variety of alkenes (Scheme 1). Acrylate esters and vinyl ketones readily reacted with 1, providing the desired dienes (3a–3e) in excellent yields. Notably, the highest yield (95%) was observed with N,N-dimethylacrylamide (3f). Phosphonates, styrene and sulfones were also viable coupling partners, yet furnished the products (3g–3i) in slightly lower yields. Acrylic acid and vinyl ethers, however, failed to afford the desired products 3j or 3k. Methyl crotonate, as a multi-substituted alkene, could also be coupled to 1 to produce a single isomer (3l), whereas α-methylene-γ-lactones afforded both the conjugated and unconjugated diene products with a greater preference for the latter (3ma:3mb=1:2.5). This preference for unconjugated dienes has been previously noted.^{5k, 10} Interestingly, alkenylation of cyclohexene yielded two inseparable, unconjugated dienes (3na/3nb). Mechanistically, we speculated that 3nb was generated from 3na through Pd–H insertion and immediate β-H elimination.¹¹

Scheme 1.

Scope of alkenes^a

^a Conditions: 1 (0.2 M), 2 (2 equiv), Pd(OAc)₂ (10 mol %), Cu(OAc)₂ (2 equiv), KTFA (1 equiv) in DMF under N₂ at 80 °C for 3 h. Isolated yield. ^b Isolated ratio. ^{c 1}H NMR ratio.

We next assessed a series of enaminones (Scheme 2). We found that this reaction could be extended to mono-and bicyclic, electronically unattenuated enaminones (5a–5g). Importantly, alkenylation of the diastereomeric substrates (5a/5b) took place without epimerization of the stereocenters and with virtually the same yields. The introduction of a C6-substituent to the cyclic enaminone, however, significantly decreased the yield (of 5h). N-H and N-Cbz enaminones also showed poor reactivity (for products 5i and 5j) which is consistent with our previous findings.⁸ 4-Pyridone was found to furnish a mono-coupling product 5k albeit in only 7% yield. An E-enaminone was also tested, but only a trace amount of product 5l was observed.

Scheme 2.

Scope of enaminones^a

^a Conditions: 4 (0.2 M), 2a (2 equiv), Pd(OAc)₂ (10 mol %), Cu(OAc)₂ (2 equiv), KTFA (1 equiv) in DMF under N₂ at 80 °C for 3 h. Isolated yield.

To elucidate the alkenylation process, a mechanistic analysis of the initial interaction of Pd(II) with enaminone 1 was carried out by ¹H NMR (in DMSO-d₆ at room temperature, Figure 3). 50 mol % of Pd(OAc)₂ with 1 (Figure 3b) furnished intermediate 6 at room temperature along with the same amount of acetic acid and unreacted 1. With 100 mol % of Pd(OAc)₂ (Figure 3c), a complete conversion of 1 was observed after only 20 min, yielding only 6 and acetic acid (See the Supporting Information for the full spectra). 20% of product 3a was subsequently furnished from intermediate 6 when heated with acrylate 2a at an elevated temperature (140 °C) (Scheme 3).¹² This suggests that the C–C bond formation is the rate-limiting step. It is worth noting that intermediate 6, however, was not detected when DMF-d₇ was used as the solvent, presumably because DMF does not stabilize 6 as well as DMSO. ¹³ The discrepancy of their stabilizing effect as solvents was also reflected in the yield of 3a, where 78% was produced in DMF compared to 53% in DMSO (Table 1, entries 2 and 4).

Figure 3.

Detection of palladated intermediate 6 by ¹H NMR: (a) Pure 1; (b) 1 with 50 mol % of Pd(OAc)₂; (c) 1 with 100 mol % of Pd(OAc)₂.

Scheme 3.

Reaction of 1 with acrylate 2a in DMSO-d₆

Hence, we suggest the following mechanism (Figure 4).^5a An electrophilic attack of Pd(II) on enaminone A followed by deprotonation forms palladated B, which then undergoes alkene insertion to afford C. Subsequent β-H elimination delivers product D. Reductive elimination and reoxidation by Cu(II) regenerates Pd(II).

Figure 4.

Suggested mechanism of dehydrogenative alkenylation.

In summary, we have developed a direct, convenient and highly atom-economic approach for the dehydrogenative coupling of nonaromatic, cyclic enaminones and simple alkenes. The generality of this transformation presents a useful strategy for directly cross-coupling alkenes and offers an attractive new approach to prepare functionalize enaminones.