|Home | About | Journals | Submit | Contact Us | Français|
The synthesis of cyclic sulfamides via enantioselective Pd-catalyzed alkene carboamination reactions between N-allyl sulfamides and aryl or alkenyl bromides is described. High levels of asymmetric induction (up to 95:5 er) are achieved using a catalyst composed of Pd2(dba)3 and (S)-Siphos-PE. Deuterium labelling studies indicate the reactions proceed via syn-aminopalladation of the alkene and suggest that control of syn- vs. anti-aminopalladation pathways is important for asymmetric induction.
The enantioselective Pd-catalyzed coupling of aryl or alkenyl bromides with N-allyl sulfamides affords substituted cyclic sulfamides in good yield with high levels of asymmetric induction. The reactions proceed via stereoselective syn-aminopalladation of the alkene, which is important for high enantioselectivity.
Cyclic sulfamides are present in a wide variety of natural products and pharmaceuticals. In addition, the sulfonyl group can easily be cleaved from these compounds to afford synthetically useful 1,2-diamines.[2, 3] Although a number of approaches have been developed for the synthesis of racemic cyclic sulfamides, few methods exist for the enantioselective generation of these compounds. Enantiomerically enriched cyclic sulfamides are often prepared from amino acids via multistep routes, and only one asymmetric metal-catalyzed reaction for the synthesis of cyclic sulfamides has been previously reported.
Our group recently developed a new method for the synthesis of racemic cyclic sulfamides via Pd-catalyzed alkene carboamination reactions between N-allylsulfamides and aryl halides or triflates.[7,8] We reasoned that an asymmetric variant of this transformation could provide straightforward access to enantiomerically enriched cyclic sulfamides. However, our prior studies had illustrated that the sulfamide-forming carboamination reactions may proceed via either syn- or anti-aminopalladation of the alkene depending on catalyst structure and reaction conditions. Since the two different C–N bond forming (and potentially enantiodetermining) aminopalladation pathways proceed through very different transition states, it seemed likely that achieving high selectivity for one pathway over the other would be critical for asymmetric induction.[9,10]
We have previously found that palladium catalysts supported by the chiral ligand (S)-Siphos-PE provide good to excellent results in Pd-catalyzed asymmetric carboamination reactions that afford pyrrolidines, cyclic ureas, or benzo-fused six-membered nitrogen heterocycles. All of these transformations were demonstrated to proceed via syn-aminopalladation of the alkene. As such, we elected to focus on reaction conditions that would promote syn-aminopalladation over anti-aminopalladation: use of aryl bromide electrophiles, xylenes as a nonpolar solvent, and NaOtBu as a base. In our previous studies on asymmetric carboamination reactions of ureas we discovered that the nature of the protecting group on the cyclizing nitrogen atom has a large influence on enantioselectivity. The best results were obtained with substrates bearing electron-poor aryl groups (e.g., p-nitrophenyl or p-chlorophenyl) on the cyclizing nitrogen atom. Thus, we initially examined the reactivity of substrates 1a–b bearing an N-aryl group on the cyclizing nitrogen atom and a tert-butyl group on the other nitrogen atom. Unfortunately, these reactions failed to provide satisfactory results; products 2a–b were generated in low yield and modest er (Table 1, entries 1–2). However, we subsequently discovered that substrates 1c–e with an N-benzyl group (or substituted N-benzyl group) on the cyclizing N-atom were transformed to the desired products 2c–e in good yields with useful levels of enantioselectivity (entries 3–5). Further exploration of substrate structure indicated the presence of the tert-butyl group on the non-cyclizing nitrogen atom was essential for reactivity, as substrates 1f–i were not transformed to the desired cyclic sulfamides (entries 6–9).
Having identified satisfactory reaction conditions and a set of nitrogen-protecting groups that provided acceptable enantioselectivities, we set out to explore the scope of this transformation. Thus, substrate 1c was treated with a range of aryl and alkenyl halides using the optimized conditions (Table 2). In general, transformations of electron-neutral or electron-rich aryl bromides provided the highest levels of asymmetric induction (entries 1–7), whereas use of electron-poor substrates resulted in slightly lower selectivity (entries 8–10). Transformations of substrates bearing m- or p-substituents proceeded smoothly, but o-substituted aryl bromides were converted to the desired products with dramatically lower enantioselectivities and yields (entries 12–13). Reactions of alkenyl bromides proceeded with high enantioselectivity (>93:7 er) but low yields were obtained due to competing Heck alkenylation of the N-allylsulfamide starting material (entries 14–15).
In our previous investigations of asymmetric Pd-catalyzed carboamination reactions of ureas, we discovered that the addition of 2 equivalents of water improved enantioselectivities in transformations of electron-poor aryl bromides.[12a] For example, the use of 4-bromobenzotrifluoride as an electrophile in the asymmetric carboamination of 1-allyl-1-methyl-3-(4-nitrophenylurea) under standard conditions resulted in the formation of the desired cyclic urea product with only 77% ee. However, when this reaction was conducted in the presence of water the enantioselectivity improved to 95% ee.[12a] In our present studies on transformations of N-allylsulfamides we found that addition of 2 equivalents of water led to slight improvements in enantioselectivity with some electron-deficient electrophiles (Table 2, entries 8, 10, and 13), but did not have an impact on asymmetric induction with most electron-neutral or –rich aryl or alkenyl halides. However, we were surprised to discover that in almost all cases, water exhibited a significant positive influence on the chemical yield of these reactions (an increase of 10–20% for most electrophiles and a 45% increase with bromobenzene). As observed with urea substrates, use of an aryl iodide rather than an aryl bromide led to a significant decrease in both yield and enantioselectivity (entry 3). The addition of water to this reaction improved the yield of the product, but interestingly the enantioselectivity was diminished. Substrates 4a–b that contain substituents at the internal alkene position proved to be unreactive even when the reaction temperature was increased to 135 °C [Eq. (1)].
To further illustrate the synthetic utility of this method we examined the deprotection and desulfonylation of products 3a and 3b. As shown in [Eq. (2–3)], the tert-butyl group is easily and selectively cleaved by trifluoracetic acid to afford 6 in near quantitative yield with no loss of optical purity. In contrast, subjection of 3b to more strongly acidic conditions (HBr/phenol) leads to formation of chiral 1,2-diamine 7 (in high yield and er) by removal of both the tert-butyl and SO2 groups. Thus, overall this method provides straightforward access to enantiomerically enriched chiral diamines, which are useful building blocks for organic synthesis.
We had initially postulated the mechanism of these transformations should proceed via syn-aminopalladation under our chosen reaction conditions. In order to directly probe the alkene addition stereochemistry and test this hypothesis we prepared deuterated substrate 8 and investigated its coupling with several different aryl bromide electrophiles under our standard conditions (Table 3). In all cases examined the major product stereoisomer resulted from syn-addition of the N-atom and the aryl group to the alkene, confirming the syn-aminopalladation pathway was operational. However, both the alkene addition stereoselectivity and the enantioselectivity changed depending on whether the electrophile was electron rich or electron poor. For example, the use of 4-bromo-tert-butylbenzene as the electrophile led to formation of the desired product 9a in 17:1 dr and 94:6 er (entry 1) whereas the analogous reaction with 4-bromobenzotrifluoride provided product 9d in 7:1 dr and 88:12 er (entry 4). The diminished diastereoselectivity likely results from the competing anti-aminopalladation mechanistic pathway, and this competing pathway also is likely responsible for the decreased enantioselectivity in reactions of electron-poor aryl halides. Addition of water to the reaction between 8 and 4-bromobenzotrifluoride led to a significant improvement in yield, but little change to diastereoselectivity or enantioselectivity.[18,19]
In conclusion, we have developed an enantioselective Pd-catalyzed alkene carboamination between N-allyl sulfamides and aryl or alkenyl bromides that affords cyclic sulfamide products in good yield with high enantioselectivity. The sulfamide starting material is easily prepared in one step from the corresponding allylic amine. Deprotection and/or desulfonylation of the products can be achieved in good yield, thus providing access to enantiomerically enriched substituted 1,2-diamines. These transformations constitute rare examples of asymmetric metal-catalyzed reactions that afford cyclic sulfamides. These studies also illustrate the importance of controlling syn- vs. anti- alkene aminopalladation pathways in asymmetric Pd-catalyzed alkene carboamination reactions.
The authors acknowledge the NIH-NIGMS (GM098314) for financial support of this work. The authors thank Mr. Ryan M. Fornwald for conducting preliminary experiments in this area.
Supporting information for this article is given via a link at the end of the document.