Direct and Asymmetric Nickel(II) Catalyzed Construction of Carbon–Carbon Bonds from N -Acyl Thiazinanethiones

: A wide array of new N -acyl thiazinanethiones are employed in a number of direct and enantioselective carbon–carbon bond forming reactions catalyzed by nickel(II) complexes. The electrophilic species are mostly prepared in situ from ortho esters, methyl ethers, acetals, and ketals, which makes the overall process highly efficient and experimentally straightforward. Theoretical calculations indicate that the reactions proceed through an open transition state in a S N 1-like mechanism. The utility of this novel procedure has been demonstrated by the asymmetric preparation of syntheti-cally useful intermediates and the total synthesis of peperomin D. Theoretical calculations were carried out using QM/MM algo- rithm methodology to localise the transition states. These show that one of the phosphine hinders the approach of said carbocation to the Si face of the enolate ( ΔΔ G 2.7 kcal mol ) and thus deter-mines the π –face selectivity of the addition. For details of theoretical calculations, see Supporting Information.

The stereocontrolled construction of carbon-carbon bonds from metal enolates holds a prominent position among carbon backbone forming methods in asymmetric synthesis. 1 Unfortunately, most of the reported methods hinge on the stoichiometric generation of the enolate and subsequent reaction with the chosen electrophile, so they do not meet the current demands for economy in synthesis. 2 Organocatalysis does meet such challenges, 3 but the source of nucleophiles is often restricted to aldehydes and a few privileged compounds. 4 Hence, there is a lack of direct, catalytic, and asymmetric transformations based on metal enolates from non-activated carboxylic derivatives. In this context, pioneering studies underlined the benefits of working with easily removable scaffolds attached to the carboxylic moiety. [5][6][7] This led Kobayashi to use amides in highly enantioselective aldol and Michael additions, 8 similarly Evans described aldol reactions and orthoester alkylations from N-acyl thiazolidinethiones, 9 Kumagai and Shibasaki also reported a number of reactions based on 7-azaindoline amides. 10, 11 Inspired by such precedents and taking advantage of our own experience in S N 1-like stereoselective transformations 12 and the isolobal principle, 13 we envisaged that treatment of N-acyl thioimides with easy to handle chiral nickel(II) complexes might catalytically produce the corresponding metal enolates. These would then be capable of taking part in asymmetric carbon-carbon bond forming reactions with cationic intermediates. According to such ideas, we herein report that the direct TESOTf-mediated addition of N-acyl thiazinanethiones to a wide array of electrophiles catalyzed by chiral nickel(II) complexes and the ensuing removal of the thiazinanethione scaffold provides enantiomerically pure compounds in high yields and in a straightforward manner (Scheme1).

Scheme 1. Direct, Asymmetric, and Catalytic C-C Bond Forming Reactions
We were aware from the very beginning that such a challenging process called for: (a) the catalytic formation of an enolate possessing the necessary chiral environment in parallel to (b) the generation of the required electrophile for (c) the installation of up to two TESOTf, 2,6-lutidine new stereocenters whilst (d) minimizing undesired reactions. Therefore, we carried out a careful examination of all the species involved in such a process.
Exploratory studies on the addition of N-propanoyl derivatives 1a-5a to 4,4'-dimethoxybenzhydryl methyl ether in the presence of commercially available (Me 3 P) 2 NiCl 2 demonstrated the crucial role of the exocyclic C=S bond (Scheme 2). Indeed, oxazolidinone 1a and thiazolidinone 2a did not react at all, whereas thiazolidinethione 4a and thiazinanethione 5a were converted into the alkylated products quantitatively.

Scheme 2. Assessment of the Scaffold
Therefore, we focused our attention on the alkylation of 4a and 5a promoted by a few chiral complexes (L*NiCl 2 in Table 1), easily prepared by simple heating of mixtures of NiCl 2 and the corresponding diphosphines in CH 3 CN. 9b Initial screening of the reaction conditions revealed that both substrates were appropriate platforms to carry out such alkylations. Thereby, treatment of thiazolidinethione 4a with 4,4'dimethoxybenzhydryl methyl ether, TESOTf, and 2,6lutidine in the presence of 5 mol % L*NiCl 2 at -20 °C for 15 h produced the quantitative and enantioselective (ee up to 96%) conversion into the alkylated adduct 9a (entries 1-3 in Table 1). Even better, parallel reactions from thiazinanethione 5a afforded adduct 10a as a single enantiomer (entries 4-6 in Table 1). Importantly, the temperature could be raised to 0 °C without any detrimental effect, which enabled us to dramatically reduce the reaction time and to scale down the catalyst loading (compare entries 6-9 in Table 1). Eventually, the alkylation of 5a with a mere 1 mol % of [(R)-DTBM-SEGPHOS]NiCl 2 took place at 0 °C in just 10 min and gave 10a with a 98% ee and a 96% yield after chromatographic purification (entry 8 in Table 1).
The optimized conditions were then applied to a broad array of N-acyl thiazinanethiones 5 ( Table 2). The reaction proved to be sensitive to the bulk of the acyl group, so the catalyst loading had to be increased to 10 mol % for the sterically hindered (R: i-Pr) thiazinanethione 5d (compare entries 1-4 in Table 2). Otherwise, it tolerated the presence of common functional groups as alkenes, alkynes, and carboxylic esters as well as Cα-benzyl or phenyl ethers, in most cases with an outstanding enantiocontrol (ee up to 98%) and yields from 78 to 96% (entries 5-9 in Table 2). Unfortunately, the synthesis of the azidoacetyl thiazinanethione counterpart proved troublesome, but a parallel alkylation reaction was carried out successfully with the N-azidoacetyl thiazolidinethione 4j Table 1   (entry 10 in Table 2). Significantly, the results for 10i and 9j make this alkylation a new approach to the asymmetric synthesis of α-hydroxy and α-amino acids respectively (entries 9 and 10 in Table 2).

. Initial Trials on the Direct and Asymmetric Reactions Catalyzed by Chiral Nickel(II) Complexes
The thiazinanethione scaffold of the products 10 was easily removed to release alkylated products (Scheme 3). 14-16 Indeed, reduction of 10a with NaBH 4 led to alcohol 11a with a yield of 87%, whereas treatment of 10a with methanol afforded ester 12a with a 96% yield. In turn, (S)-α-methylbenzylamine and morpholine reacted smoothly with 10a to produce amides 13a and 14a respectively in yields up to 96%. At this point, absolute configuration of adducts 10 was firmly established by chemical correlation of 11a-12a and X-ray analysis of amide 13a. 17 Interestingly, thiazinanethione may also act as a coupling reagent and permitted us to obtain diastereomerically pure N-acyl amino acid 15f by simple addition of methyl (S) leucinate to adduct 10f with an 89% yield.
Once the feasibility of the catalytic and asymmetric alkylation of 5 with 4,4'-dimethoxybenzhydryl methyl ether was established, we examined the synthetic potential of such a transformation through the use of other electrophiles represented in Scheme 1. The reactions with trimethyl orthoformate, a dimethyl ketal, and a tropylium salt, which involve the installation of a single stereocenter, proceeded smoothly and led to enantiomerically pure adducts 16a-18a (ee ≥97%) after slight adjustments of the former experimental conditions (Scheme 4). More concretely, the reaction with trimethyl orthoformate was carried out at -40 °C to suppress the competitive alkylation of the exocyclic C=S bond, whereas the addition to the dimethyl ketal lasted for six hours, probably because of the steric bulk of the oxo carbenium intermediate. Besides these results, it is important to highlight that the related reaction with 4-MeOPhCH(OMe) 2 , which involves the simultaneous construction of two new stereocenters, was also satisfactory and afforded the syn adduct 19a (dr 75:25, 98% ee for the syn diastereomer) and with a high overall yield (Scheme 4).

Scheme 4. Reactions with other electrophiles
An S N 1-like mechanism based on the approach of cationic reagents to the Re face of a putative chiral nickel enolate accounts for all these results. [(R)-DTBM-SEGPHOS]NiCl 2 is a robust and bench stable nickel(II) complex with a distorted square planar geometry which can be seen in the crystal structure obtained; 17 it does not catalyse the alkylation reaction but it is easily activated in situ with TESOTf to produce the true catalyst containing two triflate ligands. 18 Coordination of this species to the thioimide moiety enhances the acidity of 5 and facilitates the deprotonation of the Cα position. At the same time, the TESOTf reacts with the benzhydryl methyl ether and produces the corresponding carbocation, which in turn adds to the nickel(II) enolate. 19 Once the carbon-carbon bond is formed, the alkylated adduct 10a is released and the nickel(II) complex may start a new catalytic cycle (Scheme 5).
Eventually, we considered the synthesis of peperomin D (20 in Scheme 6), a secolignan metabolite isolated from Peperomia glabielle. 20 Featuring a five membered lactone with a benzhydryl appendage at the β position, we envisaged that it might be synthesized through alkylation of thiazinanethione 5k with the appropriate benzhydryl methyl ether in the presence of [(S)-DTBM-SEGPHOS]NiCl 2 . 21 Indeed, quenching the reaction mixture with LiBH 4 gave chemoselectivily the hydroxy ester 21, which was then treated with Amberlyst resin to obtain the desired lactone 22 with an excellent stereocontrol (95% ee) and a 51% yield. Remarkably, just a single chromatographic purification was required. The installation of the α-stereocenter was next accomplished by substrate-controlled alkylation of 22 with MeI, which allowed us the isolation of enantiomerically pure peperomin D 20 with an overall yield of 42% over three steps.

Scheme 6. Synthesis of Peperomin D
In summary, we have demonstrated the utility of Nacyl thiazinanethiones in a number of direct, chemo-and enantioselective carbon-carbon bond forming reactions usually promoted by 1-5 mol % of [DTBM-SEGPHOS]NiCl 2 . The thiazinanethione scaffold can be smoothly released from the resulting adducts to provide a broad array of enantiomerically pure intermediates. Theoretical studies suggest that these transformations proceed through an open transition state in an S N 1-like mechanism, in which the configuration of the αstereocenter is absolutely controlled by the chiral biphosphine. The efficiency of such an alkylation has been proved in the total synthesis of peperomin D, a five membered lactone containing two stereocenters.