Photo-organocatalytic Enantioselective Perfluoroalkylation of β-Ketoesters.

The visible-light-driven, phase-transfer-catalyzed, enantioselective perfluoroalkylation and trifluoromethylation of cyclic β-ketoesters is described. The photo-organocatalytic process, which occurs at ambient temperature and under visible light illumination, is triggered by the photochemical activity of in situ-generated electron donor-acceptor complexes, arising from the association of chiral enolates and perfluoroalkyl iodides. Preliminary mechanistic studies are reported.

Visible-light-driven enantioselective catalytic processes hold great potential for the sustainable preparation of chiral molecules.[1] Their development, however, is challenging.[2] A central theme of modern stereoselective chemistry is the identification of strategies for exploring the untapped potential of enantioselective photocatalysis.[3,4] In this context, our laboratory recently introduced a unique approach[5] based on the ability of chiral enamines, key intermediates in thermal organocatalytic asymmetric processes,[6] to actively participate in the photoexcitation of substrates while inducing the stereocontrolled formation of chiral products. The photo-organocatalytic strategy, which did not require external photosensitizers, relied upon the formation of photoactive electron donor–acceptor (EDA) complexes,[7] arising from the ground-state association of the electron-rich enamine I with electron-deficient alkyl bromides II (Figure 1a). Visible light irradiation of the colored EDA complex III induced a single electron transfer (SET), allowing access to radical species under mild conditions. This reactivity enabled the development of a light-driven stereoselective α-alkylation of carbonyl compounds,[5] a process which could not be realized under thermal activation.

Figure 1

EDA complex activation strategy for designing light-driven enantioselective catalytic reactions: (a) chiral enamines as the donor; (b) chiral enolates as the donor. The gray circles represent the chiral organocatalyst scaffold; PTC = phase transfer catalysis.

In this Communication, we further advance the EDA complex activation concept to develop a photochemical enantioselective perfluoroalkylation of β-ketoesters 1 (Figure 1b). Conceptually, this study demonstrates that chiral enolates IV, generated upon deprotonation of 1, can serve as suitable donors for EDA complex formation. The chemistry occurs at ambient temperature and requires visible light irradiation to proceed. It provides straightforward access to highly valuable products 3 bearing an RF-containing quaternary stereocenter[8] (RF indicates the perfluoroalkyl fragment). Since fluorine-containing functional groups can greatly alter the intrinsic properties of organic compounds,[9] the catalytic production of perfluoralkyl-containing stereogenicity is a centrally important methodological goal.[10]

Our initial investigations were motivated by the desire to conceive novel and synthetically useful photo-organocatalytic asymmetric transformations. Specifically, we wondered if the EDA complex activation strategy could be expanded to include electron-rich chiral organocatalytic intermediates other than enamines I. Given the electronic similarities with I, in situ-generated enolates of type IV were considered as suitable donors.[11] Perfluoroalkyl iodides (RFI, 2) were selected as electron-accepting substrates, since a few literature precedents[12] qualify them as potential acceptors for facilitating EDA associations in the ground state. In addition, the electrophilic character of perfluoroalkyl radicals (RF•),[9c] emerging from the photoinduced SET, should facilitate trapping by the chiral enolate ion-pair IV.[13] The chemistry of chiral enolates has a rich history in enantioselective catalysis, with many different approaches available. One effective strategy relies on phase transfer catalysis (PTC),[14,15] where chiral quaternary ammonium salts can be used to generate a chiral ion-pair IV after deprotonation of β-ketoesters 1 by an inorganic base (Figure 1b). We anticipated IV to be electron-rich enough to form an EDA complex with RFI and then trap the resulting radical, forging a quaternary carbon stereocenter within product 3.

The feasibility of our plan was tested by reacting the methyl ester of the indanone-derivative 1a with perfluorohexyl iodide 2a in dichloromethane (DCM) under visible light irradiation by white light-emitting diodes (LEDs)[16] (Table 1). Performing the reaction in the presence of the commercially available cinchona-derived PTC catalyst 4a and Cs2CO3 (2 equiv), so as to form the corresponding chiral enolate of type IV, provided the product 3a in good chemical yield, albeit with low stereocontrol (entry 1). Irradiation by a compact fluorescent light (CFL) bulb resulted in reduced reactivity (entry 2). We noticed that, after mixing with the iodide 2a, the achromatic solution of the enolate IVa derived from 1a developed a yellow color (Figure 2a), while its optical absorption spectrum showed a bathochromic shift to the visible spectral region, diagnostic of an EDA complex (Figure 2b, blue line).

Table 1

Optimization of the Model Reactiona

entry	4	1	solvent	light	% yieldb	% eec
1	4a	1a	DCM	white LED	79	13
2	4a	1a	DCM	23 W CFL	34	15
3	4a	1a	DCM	no	0	–
4d	4a	1a	DCM	band-pass at 400 nm	53	13
5	4b	1a	DCM	white LED	61	35
6	5a	1a	DCM	white LED	65	65
7	5a	1a	C6H5Cl	white LED	12	86
8e	5b	1a	C6H5Cl	white LED	41	92
9e	5b	1a	C6H5Cl/C8F18	white LED	59	93
10e	5b	1b	C6H5Cl/C8F18	white LED	71	93
11e	5c	1b	C6H5Cl/C8F18	white LED	60	86

C8F18 = perfluorooctane. Reactions performed over 16 h on a 0.1 mmol scale using 1,2 equiv of 1a and a white LED strip to illuminate the reaction vessel.

Yield of 3a determined after isolation by chromatography.

Enantiomeric excess determined by HPLC analysis on a chiral stationary phase.

Using a 300 W xenon lamp.

Reaction time, 64 h; 3 equiv of 2a.

Figure 2

(a) Images showing the formation of the EDA complex (yellow) on the surface of Cs2CO3 (white solid) and its subsequent dispersion into the organic phase (chlorobenzene) upon addition of the PTC catalyst 4b. (b) Optical absorption spectra recorded in chlorobenzene in a 1 cm path quartz cuvettes using a Shimadzu 2401PC UV–visible spectrophotometer; [RF-I] = 15 μM, [1a] = 15 μM; [DBU] = 30 μM, [4b] = 15 μM. While the substrates 1a and 2a are achromatic, the resulting enolate IVa (formed upon mixing 1a with 2 equiv of DBU) showed a weak absorption at about 380 nm (red line); its combination with perfluorohexyl iodide 2a determines a strong bathochromic shift (blue line). The optical absorption spectrum of the reaction mixture under PTC conditions (recorded upon filtration of Cs2CO3) perfectly overlaid the blue line. DBU = 1,8-diazabicyclo[5.4.0]undec-7-ene.

Control experiments revealed how the exclusion of any of the reaction components, i.e., light (entry 3), PTC catalyst, and Cs2CO3, completely suppressed the process. Inhibition of the reactivity also occurred under an aerobic atmosphere or in the presence of TEMPO (1 equiv), the latter experiment being indicative of a radical mechanism. Additionally, an experiment using a 300 W xenon lamp, equipped with a band-pass filter at 400 nm so to exclude high-energy photons, did not significantly alter the reaction efficiency (entry 4). This result is mechanistically relevant since it excluded both possible homolytic cleavage of the C–I bond in 2a and direct photoexcitation of the enolate IVa (which is unable to absorb at 400 nm, red line in Figure 2) as pathways for RF• generation. All of these observations are consonant with the EDA complex-driven photochemical mechanism proposed in Figure 1b.

We next focused on identifying a chiral PTC organocatalyst that could infer high stereocontrol in the photochemical perfluoroalkylation of 1a. Representative results of our extensive investigations are listed in Table 1, with more details reported in the Supporting Information. We found that the substitution pattern of the benzyl moiety within the PTC catalyst had a direct influence on the stereoselectivity (catalyst 4b, entry 5). Gratifyingly, the pseudo-enantiomeric cinchonine derivative 5a inferred a higher level of stereocontrol (entry 6). Using chlorobenzene as solvent provided the product 3a with 86% ee, albeit with a greatly attenuated reactivity (entry 7). A final cycle of catalyst optimization revealed that the stereocontrol was sensitive to structural modifications at the 2′ position of the quinoline ring. Of the investigated catalysts, 5b provided the best results. Conducting the reaction with an excess of 2a (3 equiv) for a 64-h time period afforded the adduct 3a with 92% ee and a moderate yield (entry 8). Further evaluation of the reaction medium indicated that a chlorobenzene/perfluorooctane combination (in a 2:1 ratio) positively influenced the reactivity, without affecting the stereoselectivity of the process (3a formed in 59% yield and 93% ee, entry 9). Under the same conditions, the tert-butyl ketoester 1b was converted into the chiral adduct 3b with an improved chemical yield while retaining the stereoselectivity (71% yield, 93% ee, entry 10).

During the optimization studies, we noticed that the catalyst 5b was partially converted into the perfluorohexyl derivative 5c, due to an atom transfer radical addition[17] of 2a to the olefinic catalyst moiety followed by a net HI elimination. We found that 5c, isolated in 51% yield at the end of the reaction detailed in entry 10, was also a competent catalyst of the model reaction, providing only slightly inferior results than the progenitor 5b (compare entries 10 and 11).

We then evaluated the synthetic potential of the photo-organocatalytic asymmetric perfluoroalkylation strategy, reacting differently substituted indanone-derived β-ketoesters with perfluorohexyl iodide 2a under the catalysis of 5b.

As detailed in Figure 3b, a variety of electron-withdrawing substituents were well tolerated, independently of their position on the aromatic ring. The desired products 3b–g were obtained in good yields and high enantioselectivities. The presence of electron-donating substituents somewhat lowered the reactivity (products 3h and 3i). The process is amenable to scale up (1 mmol, product 3a), but with a slightly reduced yield, likely a consequence of a lower photon/mole ratio. Our efforts to react six-membered cyclic and linear substrates require further optimization, as only traces of the corresponding perfluoralkylated adducts could be obtained. Crystals of adduct 3f were suitable for X-ray crystallographic analysis, which established the stereochemical outcome of the photo-organocatalytic process.[18] We next found that the system is amenable to using other perfluoralkyl iodides (Figure 3c). Both shorter and longer perfluorinated chains could be installed in 1a in a good yield and with a high stereocontrol (ee ranging from 90% to 94%, product 3j–l). Notably, trifluoromethyl-containing quaternary stereocenters could be forged with high fidelity when reacting β-ketoesters with CF3I (Figure 3d, products 3m–o).

Figure 3

Photochemical enantioselective perfluoroalkylation of indanone-derived β-ketoesters under PTC conditions. (a) Reactions performed using the optimized conditions from entry 10 in Table 1. Yields are of isolated products 3. The X-ray structure of catalyst 5b is shown. (b) Scope of the β-ketoesters 1 using perfluorohexyl iodide 2b. (c) Scope of perfluoroalkylating agents. (d) Enantioselective trifluoromethylation. *1 mmol scale reaction.

As for the mechanism of this asymmetric photochemical perfluoroalkylation, we propose a radical chain propagation pathway, as depicted in Figure 4.[19] The chain reaction is initiated by the photochemical activity of the EDA complex of type V, formed upon the aggregation of the chiral enolate IV with RFI 2.[20] A visible-light-promoted electron transfer leads to the formation of the electron-deficient perfluoralkyl radical through the reductive cleavage of the C–I bond within 2. Consistent with a SET pathway, the model reaction was completely inhibited when performed in the presence of a redox trap such as 1,4-dinitrobenzene (0.2 equiv). The electrophilic perfluoroalkyl radical is next trapped by the chiral enolate IV in a stereocontrolled fashion. The resulting ketyl intermediate VI would then abstract an iodine atom from 2, thereby regenerating RF•.[21] The adduct VII is not stable and collapses to release the product 3 and the PTC catalyst 5b. At the present level of investigation, an alternative electron transfer process, where the ketyl intermediate VI reduces RFI to directly afford the final product 3, cannot be excluded.

Figure 4

Proposed mechanism: initiation, triggered by the photoactivity of the EDA complex, and radical chain propagation; X = I, Br.

In conclusion, we have developed a photochemical enantioselective perfluoroalkylation of cyclic β-ketoesters. The chemistry utilizes readily available substrates and proceeds at ambient temperature under visible light illumination. This study establishes the ability of chiral enolates, generated under PTC conditions, to be suitable donors in photoactive EDA complex while providing effective asymmetric induction in the trapping of the resulting radical species. Other applications of EDA complex as an activation strategy for the design of novel light-driven transformations are underway in our laboratory.