Photocatalyzed Direct α-Alkylation of Esters Using Styrenes

Abstract

A mild photocatalyzed approach to achieve the α-alkylation of esters via formation of an α -radical is disclosed here. Cesium enolates of esters were generated in situ using Cs₂CO₃ as a base. A subsequent photocatalyzed oxidation at the α-carbon of these enolates produced an α-radical that was added into activated alkenes. This is the first example accessing the α-carbon radical of esters in photoredox catalyed transformations.

The carbonyl group is central to the synthesis of many organic molecules. Carbonyl functional groups exist widely in synthetic precursors as well as bioactive compounds and enable transformations to access a vast of array of other important functional groups. Most importantly, carbonyl groups are a chief moiety in many pharmaceutical drugs. For example, approximately 80% of the highest grossing small molecule pharmaceutical drugs in 2021 contain at least one carbonyl group.^[1] Ketones, amides, and esters are the most common moieties that include a carbonyl group, and diverse functionalities are often found at the carbonyl α-position, highlighted in these pharmaceuticals^[2] (Figure 1A).

Figure 1.

Pharmaceutical Drugs Containing Carbonyl Group. Previous and Current Approaches for the α-Functionalization of Carbonyl Compounds.

Compounds featuring carbonyl groups are traditionally electrophilic at the carbonyl carbon and nucleophilic at the α-carbon. Considering the inherent reactivity of the carbonyl group, countless chemical transformations at the α-position of carbonyl compounds have been developed. Classical approaches include the aldol reaction,^[3] Claisen condensation,^[4] and Mannich reaction^[5] (Figure 1B). The transition metal (TM)-catalyzed α-arylation of carbonyl compounds has also been developed employing aryl halides and aldehydes.^[6] Silyl enol ethers and enamines serve as stabilized nucleophiles that can undergo nucleophilic addition with a carbonyl group.^[7] More recently, photocatalysts have been utilized as powerful tools to construct C—C bonds at the α-position of carbonyl compounds. These transformations often begin with pre-activation of the carbonyl group to the corresponding silyl enol ethers^[8] or enamines^[9] (Figure 1C). Early evidence for silyl enol ether radical cation formation was reported by Gassman in 1987, when it was reported that photoinduced single-electron oxidation of silyl enol ethers produced a radical cation species via excitation of 1-cyanonapthalene using high energy 300 nm light.^[8a] The combination of enamine catalysis and photoredox catalysis for the α-functionalization of aldehydes was first reported by MacMillan and co-workers in 2008 using a base and a catalytic amine.^[9a] Inspired by these reports, Gianetti and co-workers described a method for the α-arylation of cyclic ketones using low energy, green light-mediated photoredox catalysis in 2022.^[9e] While numerous approaches for the α-functionalization of carbonyl compounds have been developed, each method features restrictions which constrain their utility in organic synthesis. For example, enolate chemistry typically involves strongly basic conditions that are incompatible with many functional groups, limiting its use in the late-stage functionalization of complex small molecules.Under photocatalytic conditions, enamine chemistry requires multiple additives and has various scope limitations. The use of silyl enol ethers necessitates pre-functionalization of the corresponding carbonyl compound, which restricts the overall potential of these methods. To our knowledge, no general or mild conditions for the direct α-functionalization of esters currently are available.

Given our recent work combining NHC and photoredox catalysis for the synthesis of cycloalkanones,^[10] wherein mechanistic studies indicated the formation of an α-radical of a cesium enolate^[11] in situ, we sought to develop a mild and broad strategy to achieve the direct α-functionalization of carbonyl compounds. Reported herein is a mild photocatalyzed approach for the α-alkylation of esters (Figure 1D). This process features the formation of a cesium enolate in situ and subsequent photocatalyzed single-electron oxidation to an α-radical intermediate. The newly formed α-radical intermediate is then added into abundant styrene derivatives.

We began our investigation of this α-alkylation reaction by setting up high-throughput screening experiments to evaluate the effectiveness of different bases, solvents and photocatalysts with ethyl 2-phenylacetate 1 a and styrene 2 a (see the Supporting Information). In the presence of base Cs₂CO₃ and organophotocatalyst 3DPAFIPN in degassed acetonitrile under 456 nm blue LEDs, the desired product 3 a was obtained in 76% NMR yield. The excess styrene was easily recovered after chromatographic purification (Table 1, entry 1). Iridium (Ir-1 to Ir-3), ruthenium [Ru(bpy)₃] and various organophotocatalysts^[12] (4-CzIPN and Mes—Acr—Ph) were also examined under the same conditions, but no product was observed (entries 2–7). To increase solubility of the base in acetonitrile, we replaced Cs₂CO₃ with CsOAc^[13] and DBU, but these bases did not lead to an improvement in reaction yield (entries 8–9). Alkylation product 3 a was obtained in only 22% NMR yield when N,N-dimethylformamide (DMF) was used instead of acetonitrile (entry 10). Additionally, increasing the photocatalyst loading and varying the reaction concentration did not improve the reaction efficiency (entries 11–14). Adjusting the light wavelength to 370 nm and 390 nm failed to produce the desired product, and utilizing a blue light at 470 nm afforded 3 a in only 61% NMR yield (entries 15–17). Decreasing the amount of Cs₂CO₃ and styrene 2 a engendered a sluggish transformation, elongating the reaction time to 72 hours or more (see Supporting Information). Similarly, it was found that high levels of oxygen were detrimental to the reaction efficiency, leading to a dramatic reduction in reaction yield when the reaction was sparged with air (see Supporting Information).

Table 1.

Optimization of Reaction Conditions.

entrya	deviation from standard	Yield(%)b
1	none	76
2	Ir-1 instead of 3DPAFIPN	0
3	Ir-2 instead of 3DPAFIPN	0
4	Ir-3 instead of 3DPAFIPN	0
5	4-CzIPN insetad of 3DPAFIPN	0
6	Mes-Acr-Ph instead of 3DPAFIPN	0
7	Ru(bpy)3 instead of 3DPAFIPN	0
8	CsOAc instead of Cs2CO3	0
9	DBU instead of Cs2CO3	0
10	DMF instead of CH3CN	22
11	3 mol% 3DPAFIPN instead of 1 mol%	46
12	0.05 M instead of 0.1 M	5
13	0.15 M instead of 0.1 M	59
14	0.2 M instead of 0.1 M	25
15	370 nm instead of 456 nm	0
16	390 nm instead of 456 nm	0
17	470 nm instead of 456 nm	61

^aSee the Supporting Information for details. The reaction was carried out with 0.20 mmol of 1 a and 1.0 mmol of 2 a.

^bNMR yield was determined by ¹H NMR spectroscopy with 1,3,5-trimethoxybenzene as an internal standard.

With optimized reaction conditions, the scope of the α-alkylation of esters was investigated. It is worth noting that the yield of 3 a was increased to 86% when the reaction was performed at a 4-fold larger scale, demonstrating the scalability of this transformation. Different substitutions on the phenyl moiety of esters were initially examined (Scheme 1). These reactions conditions afforded 3 b and 3 c with ortho-methoxy and para-methyl substitution in 60% and 40% yield, respectively. Various halogen-substituted phenyl groups were well-tolerated, as demonstrated by the successful preparation of 3 d–f. An ester substrate substituted with a bulkier and electron-donating para-phenyl substitution generated 3 g in 36% yield. The reaction with an electron-deficient 2,4-difluoride substitution furnished 3 h in 43% yield, showing evidence for the reaction feasibility with both electron-rich and electron-deficient substrates.

Scheme 1.

Reaction Scope for the Photocatalyzed α-Alkylation Reaction.^{[a,b] [a]} Isolated yields are reported. ^[b] Reaction conditions: ester 1 (0.25 mmol, 1.0 equiv.), styrene-derivative 2 (1.25 mmol, 5.0 equiv.), Cs₂CO₃ (0.30 mmol, 1.2 equiv.), 3DPAFIPN (2.5 mmol, 0.01 equiv.), anhydrous acetonitrile (0.1 M), 456 nm blue LED, room temperature and 20 h. ^[c] Reaction performed at a 1.0 mmol scale. ^[d] Reaction performed at a 0.25 mmol scale. ^[e] Reaction performed at 0.15 M. ^[f] dr value was determined by ¹H NMR spectroscopy.

Pyridine and its derivatives are some of the most important heterocycle ring structures in medicinal chemistry and the pharmaceutical industry.^[14] This fundamental building block of most drugs on the market today plays a predominant role in creating molecular reactivity and metabolic stability. To our delight, the reaction conditions reported herein produced meta-substituted pyridine 3 i in 67% yield. Diverse styrene derivatives were subsequently explored. Para-methyl, meta-methoxy, para-bromide and meta-fluoride substituted products 3 j–m were obtained in moderate yields. Significantly, the reaction of 1 a and 4-vinylpyridine afforded 3 n in 58% yield.

Lastly, several variants of esters were explored. Replacing the ethyl group of the ester starting material with a methyl group afforded product 3 o in 74% yield. Terminal alkyne moieties were also well-tolerated, as product 3 p was obtained in 69% yield. We also investigated the feasibility of a phenylethyl group in our ester starting material, in order to compare the reactivity of two benzylic positions that we hypothesized could generate a radical. To our surprise, the remote benzylic position did not engage in the radical process, and product 3 q was synthesized in 55% yield without observing other adducts, thus indicating that the α-position of esters selectively forms an α-radical species under these reaction conditions. An ester featuring a cyclohexane motif yielded product 3 r in 60% yield using a longer reaction time (36 h). This reaction recovered a small amount of starting material, which we expect to have stemmed from intramolecular hydrogen atom transfer (HAT) between the hydrogen atom on the adjacent cyclohexane carbon and the α-radical of the ester. Furthermore, alkylation of (R)-1-phenylethyl-2-phenylacetate with styrene afforded 3 s in 71% yield with 1.2:1 dr.

To further explore the reactivity, 2-substituted furan and thiophene substrates gave a trace amount of the corresponding products 3 t and 3 u that might stem from these photoreactive heterocyclic ring. The phenyl group was replaced with an aliphatic chain, but the reaction failed to produce the product 3 v, suggesting that the aryl group is essential to stabilize the α-carbon radical of esters. Alternatively, a diphenyl substituted ester cannot afford the target product 3 w, indicating that the tertiary stabilized α-radical of esters does not engage in the subsequent alkylation. In addition to the formation of α-radical of esters, the current conditions produced α-alkylation ketone products 3 x and 3 y in moderate yields. The attempts on amide substrates failed to generate 3 z, suggesting incompatibility between the photoactive acyl imidazole and esters in the radical generation process. We pursued a late-stage modification of native compounds like cholecalciferol, however, the reaction process was sluggish and accompanied by unknown side products. We suspected the extra alkenes in cholecalciferol might interfere with the radical addition into the alkene, but it still remains elusive.

To examine and compare the feasibility of classic approaches versus our newly developed method to construct α-alkylated product 3 a, the alkylation of ethyl phenylacetate was performed in the presence of LiHMDS and 2-phenylethyl bromide (Scheme 2). The transformation was sluggish with formation of many unknown products. Ultimately, 3 a was obtained in 32% yield. Another reported method to access the metalloenolate of α-alkylation of phenylacetic esters^[15] is the use of sodium amide as base, but this method requires handling with sodium and consuming excess amount of ammonia which limits its practicality and potential applications.

Scheme 2.

Classic Methods for α-Alkylation of Phenylacetic Esters via Formation of Enolates.

To investigate the reaction pathway of the overall process, several mechanistic studies were performed (Scheme 3). Control experiments in the absence of Cs₂CO₃, 3DPAFIPN, or light did not afford the desired product 3 a, suggesting a photocatalyzed and base-promoted reaction. The addition of water to the reaction mixture under the standard conditions also proved detrimental to the desired transformation. Various hydrogen atom transfer (HAT) reagents were examined under the same conditions, but these additives decreased the reaction yield (see Supporting Information), indicating that the photocatalyzed oxidative α-radical formation in oxidative processes is not compatible with a HAT process. Potassium tert-butoxide (KOtBu) has been reported to catalyze the addition of ketones to styrene by Knochel and co-workers in 2000.^[16] Unlike this report, the addition is not described as a radical process in the KOtBu-catalyzed transformation. To confirm the radical-based α-alkylation process under our conditions, the reaction was carried out in the presence of KOtBu as the sole base. In contrast to Knochel’s work, no alkylation product 3 a was observed. Additionally, radical trapping experiments were performed using (2,2,6,6-tetramethylpiperidin-1-oxyl) radical (TEMPO). TEMPO trapping adduct 3 a-1 was observed from UPLC-MS without formation of the desired product 3 a in the presence of 2.0 equiv. of TEMPO, supporting the formation of an α-radical intermediate in the current α-alkylation process. The reaction was further carried out with a substoichiometric amount of TEMPO (0.2 equiv.). TEMPO-adducts 3 a-1 and 3 a-2, starting material 1 a and product 2 a were detected, indicating the benzylic radical formation after the α-radical’s addition to alkenes.

Scheme 3.

Mechanistic Studies: Negative Controls, Radical Trapping, Deuterium Labeling Experiments and Proposed Catalytic Cycle.

To evaluate the formation of a cesium enolate intermediate, ester 1 a was irradiated alongside base and photocatalyst in deuterated acetonitrile. Analysis of the ¹H NMR spectra revealed significant deuterium incorporation at the α-position of 1 a, likely due to the α-proton acidity. 12% deuterium incorporation of 1 a-D and 87% of 1 a-D₂ was identified, suggesting the equilibrium and formation of a cesium enolate.

Based on these mechanistic studies, the following catalytic cycle is proposed (Scheme 3). Under irradiation of light (456 nm), photocatalyst 3DPAFIPN is activated to its excited state. The cesium enolate is formed in situ upon treatment of base, Cs₂CO₃. Subsequent single electron transfer (SET) of the cesium enolate to 3DPAFIPN* furnishes an α-radical ester species along with the radical anion of the photocatalyst. The coupling of II with styrene produces intermediate radical species III. The desired product is then formed by reduction of the radical intermediate with the radical anion photocatalyst and subsequent protonation, presumably from the hydrogen carbonate formed from the initial deprotonation step. Unfortunately, additional Stern-Volmer experiments have been difficult due to the heterogenous nature of the reaction mixtures.

In summary, a mild photocatalyzed α-alkylation of esters with alkenes has been developed. To the best of our knowledge, this is the first transformation accessing the α-radical of esters under such simple and mild reaction conditions. Applications of this methodology to other carbonyl compounds with various radical coupling partners are ongoing in our laboratory.

Experimental Section

General Synthetic Procedure (1) for Ester Precursors

Method A: Sulfuric acid (1.0 mL, 30.0 mmol) was slowly added to a solution of the carboxylic acid (30.0 mmol) in ethanol (38 mL) previously cooled to 0 °C on an ice bath. After warming to room temperature, the mixture was heated to reflux and allowed to stir for 16 h. On cooling to room temperature, the mixture was concentrated by rotary evaporation at room temperature. The residue was then dissolved in EtOAc. The organic layer was washed three times with half the volume of saturated NaHCO₃ solution. The organic layer was then washed with brine and dried over MgSO₄. The organic layer was then evaporated to dryness to afford the desired ester.

Method B: A solution of the carboxylic acid (30.0 mmol), the alcohol (33.0 mmol) and DMAP (3.0 mmol) in anhydrous CH₂Cl₂ (100 mL) was treated with EDC•2HCl (33.0 mmol) at room temperature. The resulting mixture was stirred at room temperature for 16 h. The reaction mixture was filtered and washed with anhydrous Et₂O. The filtrate was concentrated and purified by chromatography (Hexanes/ethyl acetate).

General Synthetic Procedure (2) for α-Alkylation

All reactions were set up inside of a glovebox under N₂ atmosphere. The respective ester/ketone (0.250 mmol,1.0 equiv.), the respective alkene (1.250 mmol, 5.0 equiv.), 3DPAFIPN (2.50 mmol, 0.01 equiv.) and cesium carbonate (0.300 mmol. 1.2 equiv.) were added to an oven-dried 2-dram vial containing a stir bar. Acetonitrile (0.1 M) was added, and the reaction was capped. The resulting vials were removed out of the glovebox and parafilm was wrapped around the cap to prevent air from entering. The reaction mixture was stirred and irradiated under 456 nm LEDs. The reaction progress was monitored by GCMS. When complete consumption of ester was observed, the reaction was concentrated under reduced pressure and purified by column chromatography.