ZnBr2/Oxone-mediated ipso-cyclization of N-(3-phenylprop-2-yn-1-yl)aniline.

In this work, a selective synthetic strategy towards 1-azaspiro[4.5]deca-3,6,9-trien-8-ones from N-tosyl-N-(prop-2-yn-1-yl)aniline is developed. The transformation proceeds smoothly in a mixed solvent including acetonitrile and water when ZnBr2 and Oxone are employed. Mechanism studies show that the reaction proceeds in a regioselective manner via a radical ipso-cyclization pathway. This journal is © The Royal Society of Chemistry.

Quaternary carbons are ubiquitous in many useful architectures.[1] Consequently, tremendous effort has been made towards its synthetic methodology development.[2] Among these established achievements, ipso-cyclization of arene served as a powerful tool towards quaternary carbon-containing spirocycles.[3] Electrophilic ipso-cyclization was initially developed at the beginning of the new century.[4] To date, many unprecendented advances on electrophilic ipso-cyclization have already been witnessed, and a series of spirocyclic compounds were achieved accordingly. In the past years a particular emphasis of synthetic chemists was put on radical ipso-cyclization.[5] Compared to electrophilic ipso-cyclization, the radical way was able to be appliable for more substrates with a broader functional group tolerance.

Recently, a formal 6-endo radical cyclization of aryl propiolate was developed to construct substituent-shifted isocoumarin derivatives.[6] Mechanism studies suggested the above formal 6-endo-cyclization was constituted by radical α-addition of propiolate, radical ipso-cyclization, and ester migration, which caused the shift of the substituents in core of isocoumarin. To provide a hint on this ipso-cyclization, many elegant examples were developed on electrophilic and/or radical ipso-cyclization of propiolates, leading to spirocycles 1-oxaspiro[4.5]deca-3,6,9-triene-2,8-diones.[7]

As an important analog of propiolate, N-aryl propiolamides was reasonable to be investigated for ipso-cyclization.[8] Pleasingly, electrophilic/radical ipso-cyclization of N-aryl propiolamides was also reached so for. Particularly, for radical pathways, many radical precursors were recognized as efficient reaction partners. One example from our group indicated that bromo radical was also compatible for the ipso-cyclization of N-aryl propiolamides, and a series of bromo-containing 1-azaspiro[4.5]deca-3,6,9-triene-2,8-dione was observed (Scheme 1, eqn (a)).[9] As we know, bromo group served as a versatile building block for structural elaboration through these classical transformations with/without metal catalysis. Interestingly, bromo radical was in situ generated in the presence of Oxone and bromo anion. In the process, bromo anion occurred to be oxidized into bromo radical through a single electron oxidization, and triggered the radical ipso-cyclization of N-aryl propiolamides for the synthesis of 1-azaspiro[4.5]deca-3,6,9-triene-2,8-dione as desired.

Scheme 1

Proposed route for the synthesis of trienones.

Over the past years, our group was focusing intensively on developing a more efficient and more economic system to realize regioselective transformations of alkynes.[10] For instance, we recently reported a base-promoted α-addition of propiolamides, which led to various azetidin-2-ones with high efficiency and good reaction scope.[10] In addition, the ortho-cyclizative reaction of ortho-substituent-free N-tosyl-N-(prop-2-yn-1-yl)aniline was reported by our group under Oxone and tetrabutylammonium bromide (TBAB), where released a distinctive compound 3-bromo-1-tosyl-1,2-dihydroquinoline derivatives (Scheme 1, eqn (b)).[11] As our continuous interest on alkyne-based regioselective transformations and in light of the findings presented in Scheme 1a and b, we would like to disclose the radical ipso-cyclization of N-(prop-2-yn-1-yl)aniline 1 under mild conditions for the synthesis of 1-azaspiro[4.5]deca-3,6,9-trien-8-one 3 (Scheme 1, eqn (c)). Considering our findings on Oxone chemistry,[12] in this paper we envisioned the radical brominative ipso-cyclization of N-tosyl-N-(prop-2-yn-1-yl)aniline could be activated by Oxone and ZnBr2. To reduce the possibility of ortho-cyclization, the reaction of 2-iodo-N-tosyl-N-(3-arylprop-2-yn-1-yl)aniline 1a was herein employed as the model substrate. 2-Iodo-N-tosyl-N-(3-arylprop-2-yn-1-yl)aniline was often used as a dual-functional building block.[13]

To our delight, a preliminary result from the model reaction of 1a and ZnBr2 was favored to deliver the spirocyclic compound 1-azaspiro[4.5]deca-3,6,9-trien-8-one 3a as expected, and the reaction yield was 70% at room temperature.

Other reaction condition trials on temperature, solvent, and bromo source were optimized accordingly. The results were presented in Table 1. Based on the results, an increase of reaction temperature to 60 °C did not make positive impact on the reaction outcome, just providing the desired product 3a in 62% yield (entry 2, Table 1). A similar yield was observed when the loading of ZnBr2 was used (entry 3, Table 1). The reactions using other solvents, including DCE : H2O (v/v = 1 : 1) and THF : H2O (v/v = 1 : 1), gave rise to inferior yields (entries 4 and 5, Table 1). Changing the ratio between MeCN and H2O did not give better yields, suggesting MeCN : H2O (v/v = 4 : 1) being the best choice (entries 6 and 7, Table 1). The use of TBAB, KBr or NaBr as a replacement of ZnBr2 was not favourable for the reaction, resulting in lower yield of the desired product 3a (entries 8–10 Table 1). The use of 2.0 equiv. of TEMPO as a radical scavenger totally shut the model reaction down, probably suggesting the model reaction going through a radical pathway (entry 11, Table 1).

Initial studies for the reaction of regioselective spirocyclization of 2-iodo-N-tosyl-N-(3-arylprop-2-yn-1-yl)anilinea

Entry	Variation of standard conditions	Yield of 3aa,b (%)
1	—	70
2	60 °C	62
3	1.5 equiv. ZnBr2	70
4	DCE : H2O (v/v = 1 : 1)	61
5	THF : H2O (v/v = 1 : 1)	32
6	MeCN : H2O (v/v, 10 : 1)	67
7	MeCN : H2O (v/v, 1 : 1)	45
8	2.0 equiv. TBAB	47
9	2.0 equiv. KBr	41
10	2.0 equiv. NaBr	38
11	2.0 equiv. TEMPO	0

Isolated yield based on 1a.

Standard conditions: 1a (0.2 mmol), ZnBr2 (1 equiv.), Oxone (2.0 equiv.), solvent (2 mL), rt, overnight; TBAB = n-tetrabutyl ammonium bromide; Oxone = 2KHSO5·KHSO4·K2SO4.

With the optimized conditions in hand, we then explored the reaction generality. The results are illustrated in Table 2. As shown in Table 2, a series of substituted 1-azaspiro[4.5]deca-3,6,9-trien-8-one 3 was achieved in good yields. The substituent Ar effects of aryl alkynes were investigated. The results reveal that both electron-donating groups and electron-deficient groups were compatible for the reactions. For example, the reaction with the substrate containing a 4-cyanophenyl group on Ar group afforded the desired 1-azaspiro[4.5]deca-3,6,9-trien-8-one 3c in a 72% yield, while the substrate having a 4-methylphenyl gave the corresponding product 3g in a 77% yield. Other substituents including 4-acetylphenyl, 4-esterphenyl, 4-fluorophenyl, 4-bromophenyl, and 4-phenylphenyl groups were efficient reaction partners, leading to the corresponding products 3a–3h in 68–81% yields. However, the reaction of 2-iodo-N-tosyl-N-(but-2-yn-1-yl)aniline 1i became complex, and did not offer the desired product 3i. It was reasoned that methyl in alkyne took part in the reaction, causing the reaction complex.

Reaction scope: effect of Ar groupa

Isolated yield based on 1.

Subsequently, we explored the tolerance of substituents on ring of aniline (Table 3). Encouragingly, it seemed that R1 substituents could be replaced by methyl and chloro. The corresponding products 3j–3l were obtained in 50–65% yields. For instance, the reaction using the substrate with 5-methyl substituent produced a desired 1-azaspiro[4.5]deca-3,6,9-trien-8-one 3j in a 65% yield, while using the substrate substituted by 5-chlorogave rise to 3k in a 62% yield. Surprising, the reaction was suppressed by the presence of the 3-chloro group in substituent R1, resulting in a decreased yield of 3l in 50% yield. The substituent effect of R2 on the ring of aniline was also exploited. Beside iodo function, it was pleased to find the R2 could be also replaced by bromo and methyl, and the yield of corresponding products 3m and 3n were 75% and 68%, respectively. It was noteworthy that 1-azaspiro[4.5]deca-3,6,9-trien-8-ones 3b and 3n has been synthesized under the standard conditions of Scheme 1b.[11] Compared to our previous finding, this results were highlighted by a higher efficiency, indicating a broader reaction scope simultaneously.

Reaction scope: substituent effecta

Isolated yield based on 1.

Finally, various N-protecting groups were screened. Pleasingly, N-tosyl could be replaced with 4-bromophenylsulfonyl, 2-naphthalenesulfonyl, and 2-thiophenesulfonyl. The desired 1-azaspiro[4.5]deca-3,6,9-trien-8-ones 3o–3q were detected in 61–72% yields. However, when acetyl group or hydrogen atom attached on the nitrogen atom, no desired product was observed.

Enlightened by the above information and the previous results, a plausible mechanism was proposed in Scheme 2. As illustrated in Scheme 2, bromo anion was oxidized into bromo radical,[14] which occurred to undergo regioselective radical α-addition to form an intermediate A.[15] Radical ipso-cyclization of the intermediate A then took place. Followed by oxidation, the spiro intermediate B was converted into a spirocation intermediate C. Thanks to the use of water as a mixed solvent, herein the intermediate C was trapped by water to afford hydroxyl-linked species D. Oxidation again provided the final targeted molecules 3.

Scheme 2

Proposed mechanism.

In conclusion, we have developed a selective synthetic strategy towards 1-azaspiro[4.5]deca-3,6,9-trien-8-ones from N-tosyl-N-(prop-2-yn-1-yl)aniline. The transformation proceeded smoothly in a mixed solvent system of MeCN/H2O when ZnBr2 and Oxone was employed as the promoters. Mechanism studies showed that the reaction proceeded in a regioselective manner via a radical ipso-cyclization pathway. It was believed that the products were synthetically versatile since iodo and bromo groups in the product 1-azaspiro[4.5]deca-3,6,9-trien-8-ones was ready to be structurally elaborated through some classical transformations.

Conflicts of interest

There are no conflicts to declare.