One-Pot, Telescoped Alkenylation of Amides via Stable Tetrahedral Intermediates as Lithium Enolate Precursors

A mild and efficient telescoped procedure for the stereoselective alkenylation of simple, non-activated amides using LiCH2SiMe3 and carbonyl compounds as surrogates of alkenyllithium reagents is reported. Our methodology relies on the formation of stable tetrahedral intermediates, which, upon collapse into highly reactive lithium enolates in a solvent-dependent fashion, allows for the assembly of α,β-unsaturated ketones in a single synthetic operation with high stereoselectivity.

A mides are excellent substrates for the chemoselective assembly of ketones by direct 1,2-nucleophilic addition. 1 To this aim, several chemoselective strategies have been developed to overcome the inertness of the amide C−N bond toward nucleophiles. 2 However, there is still an urgent need to develop new bench-stable reagents for the high chemoselective transformation of amides using cheap and readily available synthons. In this context, the conversion of amides into α,βunsaturated ketones by direct nucleophilic addition of unsaturated organometallic reagents has received scant attention, and only a few examples of alkenylation of amides using non-stabilized vinyllithium reagents have been described. 3 Additionally, among the several methods developed for the preparation of chalcones (including Claisen−Schmidt condensation, 4 cross-coupling approaches, 5 Friedel−Crafts acylation, 6 photo-Fries rearrangement, 7 and catalytic one-pot synthesis from alcohols and ketones 8 ), the transformation of benzamides into chalcones by nucleophilic addition of βstyrenyllithiums remains hitherto unexplored. One possible explanation for this is the cumbersome methods often required for the preparation of alkenyllithiums. Although these reagents are stable and could be conveniently handled at room temperature, 9 their generation from alkenes by metalation, X/Li exchange, or reductive lithiation is often performed at low temperatures to avoid competitive elimination processes or configurational instability issues, using lithium metal or highly pyrophoric alkyllithiums under strictly controlled Schlenk conditions. 10 To overcome these issues, one-step methods based on the direct addition of α-alkoxyvinyllithiums 11 ( Figure  1A, i), Grignard reagents ( Figure 1A, ii) to both activated 12 and non-activated 13 amides, or weaker nucleophiles (alkenylcerium 14 and olefins 15 ) to secondary amides upon electrophilic activation ( Figure 1A, iii) as well as a lithium diisopropylamide (LDA)-promoted Claisen−Schmidt condensation approach on salicylamides ( Figure 1A, iv) 16 have been developed. The formal alkenylation of amides using organometallic reagents could however be realized resorting to canonical multistep approaches proceeding via isolated ketone intermediates. 17 In the course of our studies on the reactivity of s-block polar organometallic reagents under bench-type aerobic conditions, 18 we reported a general chemoselective route to ketones from amides using non-activated N-acylpyrrolidines as privileged acylating agents of organolithiums. 19 The notorious overaddition reaction was effectively suppressed, owing to the stabilizing effect of the reaction medium [cyclopentyl methyl ether (CPME)] on the dimeric tetrahedral intermediates. On these grounds, we envisioned that the addition of α-silylated organolithium to a simple amide could mediate the generation of a stable tetrahedral intermediate in CPME, which could be exploited as a transient nucleophile upon collapse to promote a Claisen−Schmidt-type olefination process in the presence of a carbonyl source.
We thus herein report a systematic study on the synergic combination of LiCH 2 SiMe 3 (a canonical synthon for olefination reactions) 20 and carbonyl compounds to telescope the transformation of simple, non-activated amides into α,β-unsaturated ketones ( Figure 1B). This protocol allows for the high (E)-stereoselective alkenylation of amides avoiding the preparation/use of unsaturated organolithiums, working under mild reaction conditions and in the absence of additives typically required to suppress the formation of disproportionation byproducts. 21 We started our preliminary investigations using amide 1a as a model substrate. On the basis of our previous results, a solution of compound 1a (0.2 mmol, 0.04 M) in CPME was reacted with a commercially available solution of LiCH 2 SiMe 3 (0.7 M in hexanes, 1.0 equiv) at room temperature (RT) (entry 1 in Table 1). After 30 min, the reaction was quenched with benzaldehyde (1.2 equiv), releasing in 1 h the desired chalcone 2a in 35% yield and complete (E) stereoselectivity. Pleasingly, increasing the amount of LiCH 2 SiMe 3 significantly improved the yield of compound 2a without affecting the stereoselectivity of the condensation step (entries 2 and 3), with optimal results using 1.5 equiv of organolithium (entry 2). Less satisfactory results were obtained increasing the amount of benzaldehyde (entry 4), the reaction time (entry 5), or the temperature (entry 6). Performing the alkenylation reaction in the presence of quinuclidine or LiCl as additives releases the target chalcone 2a in 57 and 53% yields, respectively (Table S1 of the Supporting Information). The use of solvents with higher coordinating ability, 22 such as tetrahydrofuran (THF) (entry 7) and its greener alternative 2-MeTHF (entry 8), was slightly less effective in promoting the reaction. As expected, comparable results in THF and 2-MeTHF were obtained when the analogous Weinreb amide of compound 1a (N,4dimethoxy-N-methylbenzamide) was chosen as the substrate (entries 10 and 11), thus confirming our previous findings on the efficacy of non-chelating N-acylpyrrolidines as chemoselective acylating agents. 19 Noteworthy, the title alkenylation reaction could be performed under aerobic conditions (entry 12), however with slightly lower yields, owing to the nonnegligible competitive protonolysis of organolithium occurring over prolonged reaction times. In this case, the use of highly hydrophobic CPME is essential to prevent the moistureinduced protonolysis process. The use of more hygroscopic solvents (CPME < 2-MeTHF ≪ THF) led to a progressive decrease of the reaction yield (Table S1 of the Supporting  Information).
With satisfactory conditions in place, the scope and limitations of this transformation were evaluated for a series of functionalized amides (Scheme 1). The alkenylation of Nacylpyrrolidines 1 proceeded smoothly en route to a variety of substituted chalcones bearing electron-donating (2c−2e), fluorinated (2f and 2g), naphthalene (2h), and heteroaromatic (2l and 2m) groups with moderate to good yields (28−80%). Our methodology also allowed (a) the chemoselective preparation of highly conjugated chalcones 2i and 2j and (b) the simultaneous alkenylation of two amide groups (2k) by simply increasing the amount of LiCH 2 SiMe 3 and carbonyl compound in a single synthetic operation. Remarkably, the preparation of chalcone 2b has been easily scaled up to 5.7 mmol of compound 1b (1 g) with comparable efficiency in terms of the yield and selectivity (69% versus 80% on a small scale). However, other sensitive functional groups, such as cyano-, nitro-, diazo-, bromine, and hydroxyl, were incompatible with the reaction conditions, affording complex reaction mixtures or recovery of the starting material after workup (see Table S2 of the Supporting Information).
We next investigated the aldehyde scope of the reaction (Scheme 2). The methodology well tolerates the use of several electron-donating group (EDG)-substituted (2p−2r and 2v)  Owing to the large potential of the chalcone scaffold in drug discovery, 4 we then applied our alkenylation conditions for the preparation of selected chalcones with prominent pharmacological applications (Scheme 3). Pleasingly, a series of chalcones with potential biological activity for the treatment of cancer (2ad and 2af), 23 microbial infections (2ag), 24 chronic myeloid leukemia (2ah), 25 and inflammations (2ai) 26 or possessing enhanced fluorescent properties for bioimaging purposes (2ae) 27 have been obtained starting from the properly substituted amides 1 and aldehydes in satisfactory yields (50−95%).
To gain more mechanistic insights into the nucleophilic acyl substitution (S N Ac)/condensation sequence, additional electrophilic quenching experiments were performed (Scheme 4A). As expected, treatment of amide 1a with LiCH 2 SiMe 3 (1.5 equiv) in CPME under optimized reaction conditions, followed by quenching with water (5 equiv), led to the exclusive formation of the α-silyl ketone 3a (entry 1).
Interestingly, performing the S N Ac step in THF led to the sole formation of acetophenone derivative 4a lacking the SiMe 3 group upon aqueous quenching (entry 2). Deuterium-labeling experiments afforded (a) α-deuterated α-SiMe 3 ketone 3a-D (40% D incorporation) when the reaction was performed in CPME (entry 3) and (b) deuterated acetophenone 4a-D as a mixture of isotopomers with an overall 130% D incorporation using THF as the reaction medium (entry 4). 28 These findings suggest the solvent-dependent formation of two different lithium enolates upon the addition of the electrophile to the reaction mixture, which can act as nucleophiles in a Claisen− Schmidt-type condensation in the presence of a carbonyl compound. To confirm that the formation of ketones 3a−4a occurs only upon electrophilic quench, we next investigated the stability of the tetrahedral intermediate by 13 C NMR analysis (Scheme 4B, i−iv). After 30 min from the addition of LiCH 2 SiMe 3 (1.5 equiv) to a 0.12 M solution of 13 C-labeled amide 1b-13 C (0.07 mmol, 1.0 equiv) in dry CPME under nitrogen, neither starting material nor ketones 3b-13 C or 4b- 13 C were detected in the 13 C NMR spectra. Evidence of the formation of the tetrahedral intermediate tetr-1b-13 C, stable under these conditions up to 3 h, was assessed by a significant upfield shift of amide carbonyl to 89.6 ppm. The addition of a stoichiometric amount of water induced the rapid conversion

Organic Letters pubs.acs.org/OrgLett
Letter of tetr-1b-13 C into α-silyl ketone 3b-13 C (δ CO = 195.6 ppm), alongside a negligible amount of acetophenone 4b-13 C. While the formation of the tetrahedral intermediate in the S N Ac step is undeniable, its solvent-dependent collapse into two different lithium enolates remains however unclear. Preliminary DFT calculations on the addition of LiCH 2 SiMe 3 to amide 1b revealed that tetr-1b exists as two conformations in which the SiMe 3 group arranges in an anti position (anti-tetr-1b) or between a gauche and syn conformation (syn-tetr-1b) with respect to OLi (Scheme 4C). Interestingly, the relative stability of the syn conformation is higher in THF (−3.6 kcal mol −1 versus anti-tetr-1b), whereas the anti conformation is more stable in less coordinating CPME (−1.3 kcal mol −1 versus syntetr-1b). Hence, we are inclined to propose the initial LiCH 2 SiMe 3 addition to amide to form the stable tetrahedral intermediate tetr-1b, which can equilibrate to the anti or syn conformation depending upon the reaction media. In CPME, the collapse of anti-tetr-1b upon electrophilic quench (path a) affords the α-SiMe 3 ketone, which can be further deprotonated to corresponding C-silylated lithium enolate by the excess of LiCH 2 SiMe 3 or the lithium amide leaving group, releasing a reactive nucleophile for a Claisen−Schmidt-type olefination process. In addition, the endergonic heterolytic dissociation of the C−N bond in the tetr-1b intermediate has slightly higher energy in THF than in CPME (19.7 and 17.0 kcal mol −1 , respectively). In THF, an intramolecular elimination of N-SiMe 3 pyrrolidine from the predominant syn-tetr-1b conformer might occur (path b), leading to the formation of the corresponding lithium enolate intermediate. 29 In conclusion, we have developed an efficient one-pot, telescoped procedure for the stereoselective alkenylation of simple, non-activated amides using LiCH 2 SiMe 3 and carbonyl compounds as surrogates of β-alkenyllithium reagents. Our strategy relies on the preliminary formation of stable tetrahedral intermediates, which, upon collapse in a solventdependent fashion, efficiently release highly reactive lithium enolates for in situ Claisen−Schmidt-type condensations. Our methodology allows for the assembly of substituted chalcones in good yields in a single synthetic operation with high stereoselectivity. Furthermore, bench-type aerobic conditions could also be employed using highly hydrophobic CPME as sustainable reaction media. The development of other electrophilic quenching strategies for the chemo-and stereoselective one-pot functionalization of lithium enolates, and complete DFT calculations aimed at clarifying the whole reaction mechanism and evaluating the energy barriers involved are under investigation and will be reported in due course.

■ ASSOCIATED CONTENT Data Availability Statement
The data underlying this study are available in the published article and its Supporting Information.  In situ 13 C NMR monitoring of the S N Ac reaction on labeled 1b-13 C in dry CPME. (C) Proposed reaction mechanism based on experimental data and reaction free energies (kcal mol −1 , at 298 K) estimated by preliminary density functional theory (DFT) calculations [M06-2X/6-311+G(d) level; see the Supporting Information for details]. Dimeric aggregates in solution were used in the computations. 19 For clarity, structures are represented as monomers and hydrogen atoms have been omitted.