Report

Enantioselective Organocatalysis Using SOMO Activation

See allHide authors and affiliations

Science  27 Apr 2007:
Vol. 316, Issue 5824, pp. 582-585
DOI: 10.1126/science. 1142696

Abstract

The asymmetric α-addition of relatively nonpolar hydrocarbon substrates, such as allyl and aryl groups, to aldehydes and ketones remains a largely unsolved problem in organic synthesis, despite the wide potential utility of direct routes to such products. We reasoned that well-established chiral amine catalysis, which activates aldehydes toward electrophile addition by enamine formation, could be expanded to this important reaction class by applying a single-electron oxidant to create a transient radical species from the enamine. We demonstrated the concept of singly occupied molecular orbital (SOMO) activation with a highly selective α-allylation of aldehydes, and we here present preliminary results for enantioselective heteroarylations and cyclization/halogenation cascades.

Over the past four decades, the capacity to induce asymmetric transformations with enantioselective catalysts has remained a focal point for extensive research efforts in both industrial and academic settings. During this time, thousands of asymmetric catalytic reactions have been invented (1), in accord with the increasing need for enantiopure medicinal agents and the rapid advancement of the field of asymmetric synthesis. Most catalytic enantioinductive processes are derived from a small number of long-established activation modes. Activation modes such as Lewis acid catalysis (2), σ-bond insertion (3), π-bond insertion (4), atom transfer catalysis (5), and hydrogen bonding catalysis (6) have each spawned countless asymmetric reaction classes, thereby dramatically expanding the synthetic toolbox available to researchers in the physical and biological sciences. A necessary objective, therefore, for the continued advancement of the field of chemical synthesis is the design and implementation of distinct catalytic-activation modes that enable previously unknown transformations.

Over the past 8 years, our laboratory has been involved in the development of the field of organocatalysis, a research area that relies on the use of small organic molecules as catalysts for enantioselective transformations. As part of these studies, we introduced the concept of iminium catalysis (7): an enal or enone activation mode that lowers the energy of the substrate'slowest unoccupied molecular orbital, facilitating enantioselective C–C and C–N conjugate additions, cycloadditions, hydrogenations, and Friedel-Crafts alkylations (8). Simultaneously, Barbas and List (9) brought to fruition the concept of enamine catalysis (Fig. 1), which raises the energy of the highest occupied molecular orbital (HOMO) in aldehydes and ketones to promote enantioselective α-carbonyl functionalization with a large range of electrophiles (10). These two modes of catalyst activation (iminium and enamine) have provided, in total, more than 60 asymmetric methodologies over the past 7 years.

Fig. 1.

SOMO catalysis via single-electron oxidation of a transiently formed enamine. LUMO, lowest unoccupied molecular orbital; R, an arbitrary organic substituent.

Given the established capacity of enamines and iminium ions to rapidly interconvert via a redox process (enamine has four π electrons and iminium has two π electrons), we recently questioned whether it might be possible to interrupt this equilibrium chemically and thereby to access a mode of catalytic activation that electronically bisects enamine and iminium formation. More specifically, we hypothesized that a one-electron oxidation of a transient enamine species (Fig. 2A) should generate a three–π-electron radical cation with a singly occupied molecular orbital (SOMO) that is activated toward a range of enantioselective catalytic transformations not currently possible with established catalysis concepts.

Fig. 2.

(A) Catalytic chemical steps leading to formation of the SOMO-activated intermediate. Me, methyl; Ph, phenyl. (B) DFT–calculated three-dimensional structure of the enantio-differentiated radical cation. (C) Possible transformations arising from enantioselective organocatalytic SOMO catalysis.

From the outset, we identified three key design elements to substantiate this proposal. First, we recognized the mechanistic requirement that an equilibrium population of enamine must undergo selective oxidation in the presence of an amine catalyst, an aldehyde substrate, and an iminium ion precursor. Theoretical support for such a chemoselective pathway was derived from the ionization potentials (IPs) of 1-(but-1-enyl)pyrrolidine (IP = 7.2 eV), pyrrolidine (IP = 8.8 eV), and butanal (IP = 9.8 eV)—data that reveal the transient enamine component to be sufficiently more susceptible to oxidation than the accompanying reaction partners (11, 12).

Second, we understood that the widespread application of SOMO catalysis would require the identification of an amine catalyst that could generically enforce high levels of enantiocontrol in the coupling of the pivotal radical cation with a variety of π-rich nucleophiles. On the basis of density functional theory (DFT) calculations (13, 14), we proposed that the imidazolidinone catalyst 1 (8) should selectively form a SOMO-activated cation, DFT-2, that projects the three–π-electron system away from the bulky tert-butyl group, while the radical-centered carbon selectively populates an E configuration to minimize nonbonding interactions with the imidazolidinone ring (Fig. 2B). In terms of enantiofacial discrimination, the calculated DFT-2 structure also reveals that the benzyl group on the catalyst system should effectively shield the re face of the radical cation, leaving the si face exposed for enantioselective bond formation.

Third, we knew that the intrinsic value of this activation mode would be defined by its capacity to enable useful enantioselective reactions. Radical cations show great potential in this vein, because they already participate in many noncatalytic C–C, C–O, C–N, C–S, and C–X (where X is a halogen) bond formations (1519). Our analysis reveals the attractive prospect of applying asymmetric SOMO catalysis to important problems such as direct and enantioselective allylic alkylation, enolation, arylation, carbo-oxidation, vinylation, alkynylation, or intermolecular alkylation of aldehydes.

To test this activation concept, we selected the direct and enantioselective allylic alkylation of aldehydes as a representative transform (20, 21). We recognized that the accompanying allylation products have been established as important chiral synthons in chemical synthesis (22, 23). Experimentally, this allylation protocol was first examined in dimethoxy ethane (DME) solvent with octanal, imidazolidinone catalyst 1, ceric ammonium nitrate (CAN) as the stoichiometric oxidant, and allyltrimethylsilane as the SOMO nucleophile (Table 1) (24). Preliminary studies revealed the successful execution of our design ideas to provide (R)-2-allyl-octanal with excellent levels of enantioinduction and in good conversion [Table 1, entry 1; 81% yield, 91% enantiomeric excess (ee)]. Experiments that probed the scope of the aldehyde component in this reaction are summarized in Table 1, entries 1 to 6. There appears to be substantial latitude in the steric demand of the radical-cation substituent (compare entries 1 and 5, with the substituent being hexyl versus cyclohexyl), allowing access to a broad variety of 2-alkylsubstituted-4-pentenals (75 to 81% yield, 91 to 94% ee). Moreover, a variety of chemical functionalities appear to be inert under these mild oxidative conditions, including olefins, ketones, esters, and carbamates (entries 2 to 4 and 6; 70 to 75% yield, 87 to 95% ee).

Table 1.

Representative SOMO catalysis. Enantioselective aldehyde α-allylation is shown. Bz, benzoyl; Boc, tert-butyl carbamoyl; Et, ethyl.

Additionally, Table 1 reveals that a diverse array of π-rich olefinic silanes (25) will readily participate as allylic alkylating reagents in this catalytic protocol (entries 1 and 7 to 10). For example, methyl, phenyl, and 2-alkyl substituted allylsilanes can be tolerated without losses in reaction efficiency or enantiocontrol (entries 7 to 9; 77 to 88% yield, 88 to 91% ee). The electron-deficient olefin ethyl-2-(methyl-trimethylsilyl)-acrylate also functions effectively as a SOMO nucleophile to provide the corresponding alkylated adduct in 81% yield and 90% ee (entry 10). This last result provides circumstantial evidence for the generation and participation of a radical-cation species, given the capacity of ethyl-2-(methyl-trimethylsilyl)-acrylate to function as a viable SOMO nucleophile on account of the captodative effect (26), yet not as effectively as a HOMO nucleophile because of diminished π density at the olefin terminus. The sense of asymmetric induction observed in all cases (Table 1) is consistent with selective engagement of the allylsilane substrate with the si face of the SOMO-activated species 2, in complete accord with the calculated structure DFT-2.

A survey of reaction media for this organocatalytic allylation has revealed that a variety of solvents may be used without a substantial loss in reaction efficiency, provided that water is present as an addend (27). Although the use of DME provides optimal selectivity, reaction rate, and chemical yield (28), acetone can often be used as an alternative solvent without allylation of the bulk medium (29). Moreover, extended reaction times do not lead to product epimerization or the formation of α, α-diallylaldehydes or aldehyde dimerization adducts.

To highlight the anticipated broad scope of SOMO activation, we present preliminary results for the enantioselective α-heteroarylation of aldehydes (Fig. 3A). Specifically, exposure of octanal and N-tert-butyl carbamoyl pyrrole to our SOMO activation conditions enables formyl α-arylation with useful levels of enantioselectivity and excellent yields. Moreover, we have found that unsaturated aldehydes rapidly undergo enantioselective cyclization with trapping of an exogenous halide. As revealed in Fig. 3B, activation of cis-6-nonenal with our SOMO protocol in the presence of LiCl leads to the formation of a stereochemically dense cyclopentyl ring system with excellent stereocontrol [85% yield, ≥10:1 diastereomeric ratio (dr), 95% ee]. This latter result again provides circumstantial evidence for the generation and participation of a radical-cation species, given the propensity of radicals to undergo cyclization with unactivated olefins.

Fig. 3.

(A) Enantioselective α-heteroarylation of aldehydes via SOMO catalysis. CAN, ceric ammonium nitrate. (B) Enantioselective olefin cyclization via SOMO catalysis. THF, tetrahydrofuran. (C) Mechanistic investigation to determine the intermediacy of a radical cation versus a carbocation.

We have undertaken studies to more definitively investigate the participation of the putative radical-cation intermediate in this catalytic process. Specifically, aldehyde activation-addition experiments were performed with the vinyl cyclopropane substrate 3, an established radical clock that was designed by Newcomb et al. (30) to differentiate between radical-mediated pathways and cationic mechanisms (Fig. 3C). Addition of the activated aldehyde intermediate to the olefin 3 occurs with subsequent scission of the benzylic cyclopropyl bond and not the α-methoxy cyclopropyl bond. This result is in complete accord with a radical pathway and not a cationic addition mechanism.

These results highlight the substantial scope of SOMO activation for useful transformations in organic synthesis. We anticipate a wide range of applications for pharmaceutically important compounds and intermediates. After our initial submission of this work, a similar approach was also applied to aldehyde α-oxidation (31).

Supporting Online Material

www.sciencemag.org/cgi/content/full/1142696/DC1

Materials and Methods

References and Notes

References and Notes

View Abstract

Navigate This Article