Exploration of One-Pot, Tandem Sulfamoylation and Aza-Michael Cyclization Reactions for the Syntheses of Oxathiazinane Dioxide Heterocycles

Abstract

We show the first examples of one-pot, tandem sulfamoylation/aza-Michael reactions for the preparation of oxathiazinane dioxide heterocycles from linear alkenyl alcohol precursors. Our optimized protocols are tolerant of a variety of functional groups and provide products which are amenable for further transformations. The reactions scale well, and no special precautions are required to exclude air or ambient moisture.

Graphical Abstract

Introduction

Oxathiazinanes are now established synthetic intermediates for a variety of valuable targets.^1–4 The most common protocols to synthesize oxathiazinane heterocycles involve two-step, two-pot sequences (Scheme 1). Traditionally, a linear alcohol is converted into its corresponding sulfamate ester by base-promoted condensation with a sulfamoyl chloride.⁵ Milder methods have also been developed which allow for the sulfamoylation of functional-group rich, sensitive substrates.^6–11 The resulting sulfamates are then cyclized using a variety of creative methods, which include C–H amination,^{12, 13} aza-Wacker,^14–18 aza-Michael,¹⁹ and strain-release reactions.^{20, 21} In sharp contrast, we have found almost no examples of one-pot syntheses of oxathiazinanes from linear alcohol starting materials. Our laboratory has a programmatic focus on elevating the synthetic potential of the sulfamate functional group.¹ Here, we show the first examples of one-pot, tandem sulfamoylation/aza-Michael reaction cascades for the convenient preparation of oxathiazinane heterocycles from linear alkenyl alcohol substrates.

Scheme 1.

Oxathiazinane heterocycles are valuable but generally require two-step, two-pot protocols to prepare. One-pot reactions for their syntheses are rare.

Results and Discussion

We chose (E)-5-hydroxy-1-phenylpent-2-en-1-one as a convenient test-substrate to optimize a potential one-pot sulfamoylation/aza-Michael reaction sequence. With ~2 equiv. of ClSO₂NH₂ (freshly prepared by the reaction of HCO₂H with ClSO₂NCO) and 2 equiv. of DABCO, we were pleased to observe desired product in a 16% yield (Table 1, Entry 1). The yield of product did not improve by replacing DABCO with TBAF (Table 1, Entry 2). Using 1,1,3,3-tetramethylguanidine, triethylamine, or pyridine in place of TBAF (Table 1, Entries 3–5) was far better. With 3 equivalents of pyridine and 3 equivalents of ClSO₂NH₂, the yield of desired product improved to 70% (Table 1, Entry 6).

Table 1.

Select Optimization Conditions.

	Eqa	Baseb	Timec	Yield of B
1	1.8	DABCO [2]	15 h	16%
2	1.8	TBAF [2]	15 h	10%
3	1.8	TMGd [2]	24 h	40%
4	2	NEt3 [2]	6 h	43%
5	2	CsHsN [2]	15 h	48%
6	3	CsHsN [3]	15 h	70%
7	4	CsHsN [4]	15 h	77%
8	2	NaHCO3 e	6 h	70%

^aEq = equivalents

^bequivalents in parentheses

^ctime shown in hours

^dTMG = 1,1,3,3-tetramethylguanidine

^eIn this reaction, dimethylacetamide [DMA] was used as a co-solvent with MeCN. Ratio of DMA:MeCN = 1:2. A NaHCO₃ [sat. aq.) quench was essential for product formation.

Further increasing the equivalents of pyridine and ClSO₂NH₂ did not lead to marked improvement (Table 1, Entry 7). Using a solvent mixture of DMA and MeCN also gave product in a respectable yield of 70% (Table 1, Entry 8). With this reaction (Table 1, Entry 8), quenching with saturated, aqueous NaHCO₃ solution and stirring this mixture for at least 10 minutes was essential for product formation. In trial runs where this quench was eliminated, only trace product was observed, suggesting the important role of NaHCO₃ in promoting cyclization.

We were able to confirm product identity by X-ray diffraction analysis (Scheme 2, Entry 1). We wondered if tandem sulfamoylation/aza-Michael reactions were possible with N-substituted sulfamoyl chlorides. We found that one-pot preparation of N-Me (Scheme 2, Entry 2), N-Et (Scheme 2, Entry 3), and N-aryl (Scheme 2, Entries 4–6) oxathiazinanes was possible. One-pot reactions failed with N-hexyl sulfamoyl chloride and N-cyclohexyl sulfamoyl chloride. Here, we hypothesize that the increased steric bulk of these alkyl substituents precluded aza-Michael cyclization. Aryl groups are also bulky, but the increased nucleophilicity of the aniline nitrogen compensated for this.

Scheme 2.

Structure-Reactivity Relationship with Diverse Sulfamate Esters.

Our optimized protocols were compatible with several α,β-unsaturated ketones (Scheme 3 and Scheme 4). A variety of para-substituted aryl alkyl ketones reacted nicely (Scheme 3, Entries 1 – 4). We hypothesize that the strong electron donating ability of p-NMe₂ precluded efficient aza-Michael cyclization (Scheme 3, Entry 5). We were particularly pleased that several heteroarenes were compatible with our optimized conditions (Scheme 3, Entries 7, 8, and 11). Dialkyl ketones also reacted well (Scheme 3, Entry 10 and Scheme 4, Entry 7). In addition to α,β-unsaturated ketones, one-pot cyclization reactions with nitroalkenes were also possible (Scheme 4, Entries 4,6, and 8). Where relevant, the cyclizations proceeded with synthetically useful diastereocontrol, and the reaction diastereoselectivities ranged from 3:1 to 5:1 (Scheme 4, Entries 1, 5, and 6). Oxathiazinanes with substituents in a 1,2-anti configuration (Scheme 4, Entry 1) and a 1,3-syn configuration (Scheme 4, Entries 5 and 6) were favored. In each case, the major diastereomer has a lower energy chair conformation relative to the minor diastereomer, and it may be possible to use straightforward conformational analysis to predict product outcomes for substrates not depicted here.

Scheme 3.

Structure-Reactivity Relationship with Various α, α-Unsaturated Ketones.

Scheme 4.

Examining Stereoselectivity and Functional Group Compatibility.

While sulfamoylation of chalcones was very efficient, a one-pot aza-Michael cyclization did not occur in any substrate tested (Scheme 5). For these substrates, we found that treatment with 1,1,3,3-tetramethylguanidine in PhCl allowed for efficient cyclization, albeit in a second pot. We note that Mannich reactions have been previously used to synthesize similar scaffolds.^22-24

Scheme 5.

Chalcone substrates require a two-pot protcoi for efficient cyclizations.

Our optimized protocol was amenable to scaling up (Scheme 6A), and the products were convenient intermediates for further transformations (Scheme 6B). For example, the imidazole in 29 could be activated by methylation and then displaced with methanol to give ester 56. The oxathiazinane ring of 2 could be activated by appending a Cbz group and then opened with KOAc to give protected amino-alcohol 58. The nitro group of 37 could be reduced to a primary amine using a combination of NiCl₂•6H₂O and NaBH₄ and then derivatized with CbzCl in one pot to give differentially protected diamine 59.

Scheme 6.

(A) Scale-up and (B) Applications.

In summary, we show the first examples of one-pot, tandem sulfamoylation/aza-Michael reactions for the preparation of oxathiazinane heterocycles from linear alkenyl alcohol precursors. Our optimized protocols are tolerant of a variety of functional groups and provide products which are amenable for further transformations. The reactions scale well, and no special precautions need be taken to exclude air or ambient moisture. We expect this technology to be a valuable addition to existing methods for preparing densely functionalized heterocycles.²⁵

Experimental Section

Please see the associated Supporting Information for full experimental details.

Data Availability Statement

The data underlying this study are available in the published article and its Supporting Information.