Tethered Aza-Wacker Cyclization Reactions with Unusual N-Alkoxy Carbamates

Abstract

Substrates with N-alkoxy carbamate “heads” and olefin “tails” can be productively cyclized into 1,3-oxazinan-2-one products bearing pendant alkenes. Our reaction protocol is operationally simple and involves heating the substrate with a mixture of Pd (II) and Cu (II) salts in CH₃CN under 1 atm of O₂. We hypothesize that an aza-Wacker mechanism is operative, based on prior art and our own experience with this reaction class. An array of substrates is compatible, and the heterocyclic products are amenable to further transformations.

Graphical Abstract

Classical aza-Wacker reactions continue to outnumber their tethered relatives.^1–6 In a 2020 review,⁷ we defined classical aza-Wacker reactions as those utilizing native amine and amide functional groups in oxidative cyclizations with pendant alkenes. Tethered reactions, in contrast, proceed after the attachment of a nitrogen-containing auxiliary.^8–12 This strategy is particularly powerful as it allows for the installation of a new C–N bond in the product without requiring a pre-existing C–N bond in the substrate. Inspired by literature precedent, we reported the first sulfamate-tethered^{13, 14} and phosphoramidate-tethered aza-Wacker reactions.¹⁵ We have used our sulfamate-tethered aza-Wacker reaction in numerous target-oriented syntheses, including for the preparation of galactosamines, kasugamines, (+)-kasugamycin, and (+)-kasuganobiosamine.^16–18

Given the continued sparsity of tethered aza-Wacker reactions, we have a programmatic focus¹⁹ on exploring new auxiliaries for this particularly efficient alkene amination modality. We have had some recent success²⁰ with a highly diastereoselective amino-trifluoroacetoxylation of alkenes utilizing unusual N-alkoxy carbamates.^21–23 Naturally, we wondered whether these auxiliaries would be competent for tethered aza-Wacker chemistry (Scheme 1). Carbamate tethers have been sparingly explored in aza-Wacker reactions,³⁰ and literature examples have focused on aminations with N-sulfonylcarbamates.^{24, 25, 31} Given that carbamates are ubiquitous in high-value targets and are themselves masked amino-alcohols, we hypothesized that further development of aza-Wacker technology with such moieties would be useful to the synthetic community.

Scheme 1.

Classical aza-Wacker cyclizations continue to predominate over their tethered cousins.

We began our exploration with trans-hex-3-en-1-yl methoxycarbamate, conveniently prepared from commercial trans-3-hexen-1-ol (Table 1).²⁶ Heating the substrate with catalytic Pd₂(dba)₃ and 1 equivalent of Cu(OAc)₂ in CH₃CN under 1 atm of O₂ gave desired cyclized product in a respectable yield of 40% (Table 1, Entry 1). Switching from Pd₂(dba)₃ to Pd(OAc)₂ improved the yield of product to 70% (Table 1, Entry 2). Mirroring what we have observed with previous tethered aza-Wacker reactions,¹⁵ an extended reaction time of 40 h was necessary as a ~10% reduction in yield was noted with a decreased time of 24 h (Table 1, Entry 3). Palladium salts other than Pd(OAc)₂ were much worse in promoting product formation (Table 1, Entries 4 – 6). Using DMSO, MeOH, and toluene, common solvents in Wacker-type reactions,^{27, 28} in place of CH₃CN was similarly deleterious (Table 1, Entries 7 – 9). Interestingly, product formation was excellent in DMA (Table 1, Entry 10). Only a portion of optimization reactions has been abstracted here, but a full treatment is present in the associated Supporting Information (Additional Optimization Section).

Table 1.

Optimization of reaction conditions.

Entry	[Pd]	Solvent	Time	Yield a
1	Pd2(dba)3 (10 mol%)	CH3CN	40 h	40%
2	Pd(OAc)2 (10 mol%)	CH 3 CN	40 h	70%
3	Pd(OAc)2 (10 mol%)	CH3CN	24 h	60%
4	Pd(TFA)2 (10 mol%)	CH3CN	40 h	11%
5	PdCl2 (10 mol%)	CH3CN	40 h	45%
6	Pd(acac)2 (10 mol%)	CH3CN	40 h	17%
7	Pd(OAc)2 (10 mol%)	DMSO	40 h	10%
8	Pd(OAc)2 (10 mol%)	MeOH	40 h	40%
9	Pd(OAc)2 (10 mol%)	Toluene	40 h	Trace
10	Pd(OAc)2 (10 mol%)	DMA	40 h	60%

^ayields estimated by ¹H NMR integration with an internal standard (p-nitrotoluene)

We next wished to explore the effect of various N-alkoxy substituents on reaction performance (Scheme 2). We were pleased to find that a range of N-alkoxy carbamate tethers was fully compatible with our optimized conditions. We were particularly impressed that sterically bulky substituents, such as isopropyl, isobutyl, cyclopropyl methyl, tert-butyl, and benzyl groups, did not preclude reactivity (Scheme 2, Entries 4 – 8). It was no surprise, however, that the yields dropped with tert-butyl and benzyl substrates (Scheme 2, Entries 7 – 8). Furthermore, we established that an N-alkoxy substituent was required for product formation, as the aza-Wacker cyclization failed with hydroxylamine substrate 17 and with ethyl carbamate substrate 18.

Scheme 2.

Structure-Reactivity Relationship with Carbamate Tethers.

Our optimized protocol was successful with a wide variety of linear alkenyl carbamates (Scheme 3). So far, we had focused on the preparation of products bearing di-substituted trans-alkenes. Thus, we were pleased to find that products with tri-substituted alkenes could be synthesized from appropriate substrates (Scheme 3, Entries 1, 8, and 10). Many interesting functional groups were tolerated by the reaction conditions, including aryl halides, pinacolborane, aryl ethers, alkyl ethers, and Boc-protected amines (Scheme 3, Entries 3, 5, 6, and 9). With a carbamate substrate prepared from a secondary alcohol, oxidative cyclization took place with excellent diastereoselectivity, furnishing an anti-1,3-oxazinan-2-one as the sole isomer detectable by ¹H NMR analysis (Scheme 3, Entry 7).

Scheme 3.

Substrate Scope Exploration.

There were a few substrates that unexpectedly failed to deliver products of standard oxidative cyclization (Scheme 4). The reaction performed best with di-substituted trans-alkenyl N-alkoxycarbamate substrates prepared from homoallylic alcohols. With di-substituted cis-alkene substrate 60, prepared from carbamoylation of commercial cis-3-hexen-1-ol, only trace product was observed. Substrates derived from allylic alcohols were similarly unsuccessful. Subjecting 61 to our standard protocol led to a complex mixture of products. With 62, starting material was largely recovered, and no discernible product was observed by ¹H NMR analysis of the unpurified reaction mixture. With tri-substituted alkene substrate 63, a complex mixture of products formed, including some N-N dimeric species.

Scheme 4.

Poor Performers

We were pleased that the reaction with 1 was scalable from ~0.2 mmol to 5.77 mmol (a 25-fold increase), supplying product 2 with an excellent yield of 61% (Scheme 5A). The alkenyl 1,3-oxazinan-2-one products of oxidative cyclization were quite amenable to further transformations (Scheme 5B). Heating 2 with LiAlH₄ and AlCl₃ in THF reduced the carbonyl, furnishing linear amino-alcohol 64.^{20, 29} The alkene of 2 could be cleanly hydrogenated with Pd/C without concomitant cleavage of the potentially labile N-methoxy group (Scheme 5B). Epoxide formation proceeded smoothly upon treatment with mCPBA (Scheme 5B). The N–methoxy linkage could be severed with SmI₂ solution (Scheme 5B).

Scheme 5.

(A) Scale up and (B) Applications.

In summary, we have demonstrated that substrates with N-alkoxy carbamate “heads” and olefin “tails” can be productively cyclized into 1,3-oxazinan-2-one products bearing pendant alkenes. Our reaction protocol is operationally simple and involves heating substrate with a mixture of Pd (II) and Cu (II) salts in CH₃CN under 1 atm of O₂. No rigorous exclusion of ambient moisture or air is required. We hypothesize that an aza-Wacker mechanism is operative, based on prior art and our own experience with this reaction class. An array of functional groups is well-tolerated and, where applicable, the diastereoselectivity of the reaction is excellent. The product 1,3-oxazinan-2-ones are amenable to further transformations, including as precursors for linear amino-alcohols. Given the importance of amination technology to a variety of chemical communities, we expect this reaction to be a welcome addition to existing methods.

Data Availability Statement

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