No abstract

The synthesis of a series of C-10 trifluoromethyl ethers of artemisinin has been achieved from key bromide 8, itself carried out in two steps from artemisinin. The substitution of 8 with methanol, ethanol, or succinic acid allowed the access of C-10 CF(3) analogues of beta-artemether, beta-arteether, or artesunate, respectively, in good yields (up to 89%). The presence of the CF(3) group at C-10 of artemisinin clearly increased the chemical stability under simulated stomach acid conditions. For example, the CF(3) analogue of arteether was found to be around 45 times more stable than arteether itself. The influence of the CF(3) moiety on biological activity was also highlighted. CF(3) analogues of artemether and arteether exhibited a high in vivo antimalarial activity on mice infected with Plasmodium berghei NK173, with a complete clearance of the parasitemia during the entire observation period (25 days).

A straightforward route to trifluoromethyl analogs of piperidines is described. These syntheses involve a Barbier-type allylation reaction of trifluoroacetaldimines, followed by Nallylation (one-pot), and ring-closing metathesis. An efficient asymmetric version is also reported (Ͼ98 % de). FunctionPiperidines are an important class of compounds, and this heterocycle pattern is found in many therapeutic agents.[1] In this frame, the introduction of novel substituents on these heterocycles has been the focus of many investigations.[2] Among them, trifluoromethyl groups can greatly modify the biological properties of the target molecule.[3] The building-block strategy is a general approach to trifluoromethylated target compounds. Thus, aldimines derived from fluoral [4] in cycloaddition reactions have been shown to be a rich source of nitrogen heterocycles bearing a CF 3 group, especially β-lactams, [5] aziridines, [6] quinolines, and piperidinones. [7] An alternative route to the piperidine pattern involves ring-closing metathesis (RCM) [8] of doubly unsaturated amines. Recently, Billard and Langlois have successfully applied RCM to the preparation of nonchiral, CF 3 -substituted piperidines.[9] However, a limitation of the methodology stemmed from the protection/deprotection steps required to access the substrates. In this context, a fast and general methodology to give the starting doubly unsaturated trifluoromethylamines would be highly desirable. Along these lines, we report here our investigations on a simple, one-pot methodology leading to doubly unsaturated trifluoromethylated amines, and their subsequent cyclization in RCM and Pauson-Khand reactions. The asymmetric synthesis of a nonracemic CF 3 -containing piperidine (Ͼ98 % de) is also described. Results and DiscussionIn a previous communication we disclosed the efficient synthesis of homoallyl trifluoromethylamines 4 and 5 by [a] Barbier-type allylation of trifluoromethyl aldimines 1 and 2, respectively (Scheme 1). [10,11] This reaction proceeds (75-97 % yields) under mild conditions (at room temperature in DMF, or at reflux in THF) with allyl bromides as reaction partners and activated zinc (Zn*) as promoter.[12] Moreover, starting from the enantiopure aldimine 3, the corresponding homoallylamines 4-6 were obtained with high de (up to 98 %).[13] We also found that the propargylation reaction was successful under the same conditions (4d, 5d).[14] Consequently, this approach allows the synthesis of a great diversity of substituted homoallyl (and homopropargyl) amines.In order to access to doubly unsaturated amines, we studied the N-allylation reaction of homoallylamines 4a-d and 5a. Previous examples of N-allylation of α-trifluoromethylamines concerned only substrates with the nitrogen atom activated by an acyl group.[9] Nevertheless, the direct N-allylation of 4a-d and 5a was investigated under simple conditions: NaHCO 3 and allyl bromides, in refluxing acetonitrile, in the presence of a catalytic amount of KI (Table 1). Despite the fact that very long ...

No abstract