Acid-Catalyzed Oxy-aminomethylation of Styrenes

We report a strong Brønsted acid-catalyzed three-component oxy-aminomethylation of styrenes with sym-trioxane and sulfonamides or carbamates. This transformation provides a variety of 1,3-oxazinanes in moderate to good yields under mild reaction conditions. The obtained heterocycles can be readily transformed into the corresponding 1,3-amino alcohols, which are useful building blocks for the synthesis of pharmaceutically relevant molecules. Mechanistic studies suggest the intermediacy of an in situ formed 1,3,5-dioxazinane and a subsequent reaction with the olefin.

O lefins are abundant and inexpensive starting materials for chemical synthesis.−9 From these type of transformations, the Prins reaction and the aza-version thereof stand out for enabling the direct synthesis of 1,3-difunctionalized moieties, which are widely present in pharmaceutically relevant molecules. 10The Prins reaction has been extensively documented, and several catalytic methodologies are available nowadays. 11,12Analogously, the aza-Prins reaction of an alkene, formaldehyde, and ammonia represents a straightforward approach to transform olefins into 1,3-amino alcohols.Formally, this would involve the formation of a formaldehydederived imine, which is subsequently attacked by the nucleophilic olefin, and the intermediate β-amino carbocation could be trapped with water to deliver the corresponding 1,3amino alcohol (Figure 1A).Surprisingly, despite the significant synthetic potential of this transformation, it remains hitherto underdeveloped with only a few reports in the literature to date, either with extremely limited scope 13,14 or biased substrates with tethered nucleophiles to trap highly reactive intermediates, 15 as well as with the use of preformed electrophiles that lead to a two-component process. 16We aimed, therefore, to contribute to closing this gap and present here our work on the Brønsted acid-catalyzed threecomponent reaction of aryl olefins, formaldehyde, and ammonia surrogates, such as sulfonamides or carbamates (Figure 1B).The proposed transformation not only contributes to the field of olefin functionalization but also benefits from the wide availability of sulfonamides and carbamates, which are common and widely used pharmacophores.
At the onset of our investigation, we studied the reaction of styrene (1a), sym-trioxane (2) as a formaldehyde source, and p-toluenesulfonamide (3a) (Table 1 and Table SI-1) by testing the performance of several common Brønsted acids as catalysts.When substrates 1a, 2, and 3a were mixed in a 1:1.5:3 ratio, the use of a weak Brønsted acid, such as acetic acid (AcOH), in catalytic amounts led to the formation of 1,3oxazinane 4a in poor yield after 24 h at 60 °C (Table 1, entry 1).
The use of either p-toluenesulfonic acid (p-TsOH) or trifluoromethanesulfonic acid (TfOH) resulted in higher conversions, thereby affording the product in 38% and 32% yield, respectively (entries 2 and 3).As expected, the choice of catalyst plays a crucial role since other established strong Brønsted acids were found to be less efficient.Bis-(trifluoromethane)sulfonimide (Tf 2 NH) as catalyst led to a complex mixture and gave only 10% of the desired product 4a (entry 4).The use of the strong Brønsted acid hexafluorophosphoric acid (HPF 6 ) in catalytic amounts led to the formation of 1,3-oxazinane 4a in 58% yield after 24 h at 60 °C (entry 5).In the absence of acid catalyst, no product was observed (entry 6).Upon screening several solvents, the use of chloroform proved beneficial and resulted in an increased yield (78%, entry 7).Remarkably, the HPF 6 -catalyzed reaction also proceeded at room temperature, although with a considerably reduced rate (entry 8).Satisfactorily, the reaction could be performed on a larger scale (10 mmol of olefin) to provide 2.2 g of product 4a (isolated yield of 68%, entry 7).sym-Trioxane proved to be an adequate source of formaldehyde given the significant decrease in yield when paraformaldehyde or formalin were used (entries 9 −10).
With these optimized conditions, we explored the scope of the transformation (Table 2A).Regarding the olefin component, we found that terminal styrenes containing either weakly electron-donating (Me, t-Bu, CH 2 Cl) or electronwithdrawing groups (F, Cl, Br) as para substituents, afforded the corresponding 1,3-oxazinanes in moderate to good yields (4b−4d and 4e−4g).More electron-deficient olefins, though, proved to be nonreactive under the tested conditions (see the Supporting Information for details).Styrenes with meta-or ortho-substituted rings were also suitable substrates, as is the case of 1h and 1i; remarkably, olefin 1j possessing a challenging double ortho substitution pattern could also be converted into the corresponding 1,3-oxazinane 4j in moderate yield.The methodology could also be applied to a naphthylsubstituted olefin, which formed product 4k in 38% yield.Cyclic and/or internal aryl olefins also proved to be suitable substrates for the transformation.For example, 1H-indene reacted to form product 4l exclusively as the cis diastereomer.Conversely, dialin reacted to produce 4m in 32% yield as a diastereomeric mixture (cis/trans ratio = 1:1).The internal noncyclic olefin 1n reacted to produce 1,3-oxazinane 4n as a mixture of trans and cis isomers in a 2:1 ratio, whereas the isomeric cis olefin was not converted under the same reaction conditions.Finally, our methodology proved to be highly selective for aryl olefins, as observed for substrate 1l, which contained both an alkyl olefin and an aryl olefin, where only product 4o was formed in 53% yield.Several alkyl olefins did not react under the reaction conditions (see the Supporting Information for details).
Subsequently, the scope with respect to various sulfonamides was explored considering the broad availability of these compounds.As shown in Table 2B, benzenesulfonamide itself, as well as derivatives thereof containing electron-donating (Me, t-Bu, OMe) substituents, represent suitable substrates (see products 4p−4s).Furthermore, a broad variety of benzenesulfonamides with electron-withdrawing groups (F, Cl, Br, NO 2 , CF 3 ) in a variety of substitution patterns on the aromatic ring could also be transformed to the corresponding 1,3-oxazinanes (4t−4ze) in moderate yields.Naphthalenederived sulfonamide also reacted to produce the corresponding heterocycle 4zf.The reaction also proceeded with aliphatic sulfonamides to provide the corresponding products (4zg− 4zj) in good yields.It is worth mentioning that the reaction with fluoromethanesulfonamide led to the formation of the expected oxazinane 4zk in 21% yield, along with 46% of 4phenyl-1,3-dioxane, the product of the Prins reaction between olefin and formaldehyde, probably because of the decreased nucleophilic character of the sulfonamide, which hinders its reaction with trioxane to form the oxy-aminomethylating species.−19 Using N,N-dimethylsulfamide, desired product 4zl could be isolated in 39% yield.Given the broad spectrum of sulfonamide drugs, 20,21 we wondered if our methodology could be applied for the derivatization of pharmaceutically active molecules containing that functional group as a late-stage modification approach.Zonisamide, a medication used to treat the symptoms of epilepsy and Parkinson's disease, 22 reacted to give product 4zm in 70% yield.Considering the commercial availability of enantiopure camphorsulfonic acid and derivatives and their extended use as chiral auxiliaries and resolving agents, 23−25 we decided to make use of (1S)-10-camphorsulfonamide in our three-component transformation.The reaction afforded product 4zn in 29% yield, although with negligible diastereoselectivity (dr = 1.1:1).
Importantly, carbamates proved also to be suitable substrates for the multicomponent transformation.When phenyl carbamate 5a was used as the ammonia surrogate, product 6a was obtained in 34% yield (Table 3).By virtue of the extended use of carbamates (Boc, Cbz, Fmoc, among others) as protecting groups in the synthesis of peptides, we considered the potential of N-carbamoyl-protected oxazinanes for the preparation of modified/unnatural peptide derivatives. 26,27Gratifyingly, both benzyl carbamate (5b) and 9fluorenylmethyl carbamate (5c) underwent the HPF 6 -catalyzed three-component reaction with sym-trioxane and styrene 1a to afford the corresponding N-Cbz-protected (6b) and the N-Fmoc-protected (6c) 1,3-oxazinanes (23% and 37% yield, respectively; see Table 3).Similarly, terminal styrenes 1b, 1f, 1h, and 1p also proved to be suitable substrates for this methodology.In our attempt to prepare an N-Boc-protected oxazinane, tert-butyl carbamate proved to be incompatible with the strongly acidic catalyst (see the Supporting Information for more details).The obtained oxy-aminomethylation products already display the connectivity of the desired 1,3-amino alcohols, which require only the removal of the N,O-acetal functionality and the nitrogen protecting group.Choosing appropriate conditions allowed the selective removal of these moieties.For example, by refluxing with HCl in methanol, 28,29 1,3-oxazinane Table 2. Olefin/Sulfonamide Scope a a All the reactions were conducted with 1 (0.2 mmol), 2 (1.5 equiv), 3 (3.0equiv), and aq HPF 6 55% (20 mol %) in CHCl 3 (2 mL) at 60 °C for 24 h.The data reported here corresponds to the isolated yields.
4a underwent a clean N,O-acetal ring opening to give access to N-Ts-protected amino alcohol 7a in 89% yield (Scheme 1a).
Treatment of 4a with Mg powder and sonication followed by the N,O-acetal ring opening procedure afforded the corresponding free 1,3-amino alcohol 7b in 75% yield over the two steps (Scheme 1b).Furthermore, 1,3-oxazinane 4a can be readily reduced to the corresponding N-methyl species, a characteristic feature of several pharmaceutically active substances, such as atomoxetine, fluoxetine and nisoxetine. 30hen refluxed with diisobutylaluminum hydride (DIBAL) in THF, heterocycle 4a underwent a clean reductive ring cleavage to produce N-Ts-protected amino alcohol 7c in 80% yield (Scheme 1c).Importantly, refluxing our reaction product 4a in the presence of LiAlH 4 enabled direct conversion into Nmethyl amino alcohol 7d in 85% yield (Scheme 1d).
Next, we dedicated our efforts to gain insight into the mechanistic pathway of the three-component olefin oxyaminomethylation.To shed light on the actual nature of the electrophile, we conducted several experiments in the absence of olefin.The acid-catalyzed reaction of sym-trioxane (2) and sulfonamide 3a results in the formation of several sulfonamideformaldehyde condensation products, such as the monosubstituted 1,3,5-dioxazinane 8a, the disubstituted 1,3,5-oxadiazinane 8b, and the trisubstituted 1,3,5-triazinane 8c.Also, in the presence of catalytic HPF 6 at 60 °C, both 8b and 8c reacted with sym-trioxane to form 8a, which indicates a dynamic equilibrium between all of these heterocycles in our threecomponent reaction (see Scheme 2 and the Supporting Information for further details).When each one of these condensation products was reacted with styrene 1a in the presence of catalytic amounts of HPF 6 , only 8a showed significant reactivity forming 4a in 70% yield, thereby suggesting 8a to be the actual reactive precursor in the annulation with styrene (Scheme 2).Table 3. Reaction Using Carbamates a a Conditions: 1 (0.5 mmol), 2 (1.5 equiv), 5 (1.5 equiv), and aq HPF 6 55% (10 mol %) in CHCl 3 (0.5 mL) at 60 °C for 24 h.

Scheme 2. Toward the Elucidation of the Actual Mechanism in the Three-Component Transformation
To gain further understanding of the reaction mechanism, we studied the three-component reaction using β-deuteriumlabeled styrenes (d-1a-cis and d-1a-trans, respectively) as substrates. 31The HPF 6 -catalyzed reaction of these olefins led in both cases to cis/trans mixtures of the corresponding 1,3oxazinane, which is consistent with the intermediacy of a benzylic cation.
On the basis of these results, we can now propose a mechanism for our three-component reaction (Scheme 3).
Accordingly, in the first part of the reaction, sym-trioxane 2 engages in an acid-catalyzed condensation with p-toluenesulfonamide 3a to form several cyclic products under a dynamic equilibrium from which 1,3,5-dioxazinane 8a is proposed to be the actual oxy-aminomethylating precursor.Subsequent protonation with the Brønsted acid catalyst produces highly reactive intermediate I-1 that, upon ring opening, forms an Nsulfonyl iminium ion.This highly electrophilic species undergoes an aza-Prins reaction with olefin 1a (see I-2) to generate benzylic carbocation I-3 followed by ring closure and release of formaldehyde to give the desired product 4a.
In conclusion, a Brønsted acid-catalyzed three-component oxy-aminomethylation of olefins was developed.1,3-Oxazinane derivatives are obtained from the reaction of a wide range of sulfonamides/carbamates with sym-trioxane and aryl olefins by using the strong Brønsted acid HPF 6 as catalyst.Preliminary mechanistic studies suggest the product from partial aldehyde/ sulfonamide condensation to be a key intermediate for the oxyaminomethylation reaction.The obtained 1,3-oxazinanes can be easily transformed into valuable 1,3-amino alcohols, which are highly demanded building blocks in synthesis.