Benzannulation of isobenzopyryliums with electron-rich alkynes: a modular access to β-functionalized naphthalenes

Described here is a modular strategy for the rapid synthesis of β-functionalized electron-rich naphthalenes, a family of valuable molecules lacking general access previously. Our approach employs an intermolecular benzannulation of in situ generated isobenzopyrylium ions with various electron-rich alkynes, which were not well utilized for this type of reaction before. These reactions not only feature a broad scope, complete regioselectivity, and mild conditions, but also exhibit unusual product divergence depending on the substrate substitution pattern. This divergence allows further expansion of the product diversity. Control experiments provided preliminary insights into the reaction mechanism.

A substituent-controlled divergent benzannulation provides rapid access to various β-functionalized naphthalenes from electron-rich alkynes.

Functionalized naphthalenes are important structural motifs widely present in bioactive natural products, pharmaceuticals and functional materials.^1,2 They also serve as valuable intermediates in organic synthesis.³ Among them, naphthalenes bearing an electron-donating group (EDG) at the β-position (e.g., β-naphthols, β-naphthylamines, and β-naphthyl thioethers) are particularly versatile (Fig. 1).^2,3 For example, BINOL, which is derived from β-naphthol, is an extensively utilized synthetic precursor toward a wide range of privileged chiral ligands and catalysts.³ While various approaches have been developed for the synthesis of naphthalenes, efficient and selective de novo strategies toward these β-functionalized ones still remain in high demand. Moreover, to the best of our knowledge, a general and modular approach for rapid access to all these electron-rich β-functionalized naphthalenes remains unknown.⁴Open in a separate windowFig. 1Useful β-naphthol, β-naphthylamine, and β-naphthyl thioether units.Isobenzopyrylium ions are readily accessible and versatile intermediates in organic synthesis (Scheme 1).⁵ They are known to participate in cycloaddition reactions with diverse carbon–carbon double bonds or triple bonds for the synthesis of naphthalenes.^5,6 While extensive progress has been achieved in this topic, challenges still remain to be addressed. For example, the majority of these benzannulations with alkynes have to be executed at high temperature, except those intramolecular cases or with stoichiometric activators. Moreover, electron-rich alkynes have not been well explored as reaction partners for the benzannulations with isobenzopyrylium ions. Nevertheless, such reactions would provide expedient access to the valuable β-naphthol and β-napthylamine derivatives with diverse substitution patterns (Scheme 1). In this context, here we report our effort in achieving a general and modular strategy toward these electron-rich naphthalenes with high efficiency and regioselectivity under mild conditions. We also observed interesting selectivity divergence controlled by the substituent of the isobenzopyrylium substrates, giving rise to different types of naphthalene products (e.g., 2-hydroxy-1-naphthaldehydes, Scheme 1).Open in a separate windowScheme 1Benzannulation of isobenzopyryliums with electron-rich alkynes.We began our study with isochromene 1a as the isobenzopyrylium precursor. Siloxy alkyne 2a was first employed as the model electron-rich alkyne in view of the extraordinary versatility of this type of alkyne in various cyclization reactions, including benzannulation.^7–9 Various Brønsted and Lewis acids were evaluated as catalysts for this reaction. Unfortunately, most of them did not show obvious catalytic activity toward the formation of a naphthalene product (5dEvaluation of reaction conditions

Entry	Catalyst	Yielda (%)
1	MeSO3H	<5b
2	HNTf2	<5c
3	Sc(OTf)3	<5c
4	TiCl4	<5b
5	AgOTf	30c
6	BF3·OEt2	75
7	BF 3 ·OEt 2 (20 mol%)	85 (81) d
8	BF3·OEt2 (1.0 equiv.)	83
9	BF3·OEt2 (20 mol%), with 1a′	(75)d

Open in a separate window^aReaction scale: 1a (0.05 mmol), 2a (0.06 mmol), catalyst (10 mol%), and DCM (0.5 mL). Yield is based on the analysis of the ¹H NMR spectrum of the crude product using CH₂Br₂ as the internal standard.^bUnreacted substrates account for the major remainder of the mass balance.^cAn unidentifiable mixture accounts for the major remainder of the mass balance.^dYield in parentheses is isolated yield.With the optimized conditions, the generality of this process was examined (Scheme 2). A range of isochromenes 1 and siloxy alkynes 2 participated in this benzannulation to form the desired silylated β-naphthols 3 with moderate to good efficiency. Notably, these products were all generated as a single isomer. This is particularly noteworthy when compared with the cases using phthalazines, which gave a mixture of regioisomers if the phthalazine was unsymmetrically substituted.^8g The excellent regioselectivity observed in our reaction is likely attributed to the significant polarization of both cycloaddition partners. While alkyl-substituted siloxy alkynes reacted efficiently, unfortunately aryl-substituted ones led to low yield.Open in a separate windowScheme 2Reaction scope. Reaction scale: 1 (0.5 mmol), 2 (0.6 mmol), DCM (5 mL), 12 h, and isolated yield. ^a Run for 24 h.During the above scope study, we found that the reaction of 3-unsubstituted isochromene 1g and siloxy alkyne 2a did not form the desired product 3a. Instead, 2-hydroxy-1-naphthaldehyde 4a was formed as the major product (64% yield, eqn (1)). Careful analysis of this product structure indicated that the original leaving part in 1a (in the case of 3a) was not completely cleaved from the molecule. Instead, only the C–O bond cleavage took place, which led to a dangling aldehyde group. Another important observation is that the TIPS group was lost in the naphthalene product. It was proposed that the methoxide leaving group in 1a might help remove this silyl group and assist the above C–O bond cleavage. In contrast, in the formation of 3a, this methoxide group was a part of the whole leaving unit during rearomatization and thus unavailable for desilylation (vide infra).In fact, such 2-hydroxy-1-naphthaldehyde products, resembling the highly versatile salicylaldehyde family, are indeed a useful substructure of bioactive molecules, such as gossypol (Fig. 1).^2b They can also serve as key intermediates toward functional materials, such as molecular sensors (e.g., AHN, Fig. 1).^2e In view of these important applications, considerable efforts were next devoted to further improving the reaction efficiency (see the ESI for details^†). Finally, we found that the reaction yield of 4a could be improved to 85% with 2.0 equivalents of BF₃·OEt₂ and one equivalent of 2,4,6-collidine as the additive. We believe that the role of collidine is to help reversibly stabilize the isobenzopyrylium intermediate and prevent its decomposition during the reaction progress.^9b1The above protocol for the synthesis of 2-hydroxy-1-naphthaldehydes is general for a range of 3-substituted isobenzopyryliums (Scheme 3). The reaction efficiency was not affected by electron-donating or electron-withdrawing groups. Various siloxy alkynes were also suitable reaction partners, including aryl- and alkyl-substituted ones. Again, these products were all formed as a single regioisomer. While the majority of these reactions were highly chemoselective toward the formation of aldehydes 4, it is worth noting that ^tBu- and TBS-substituted alkynes resulted in a mixture of 3 and 4, with the former being major. This is likely due to the substantial steric hindrance in close proximity to the silyl group, whose cleavage was obstructed and thus the driving force for the subsequent C–O bond cleavage was weakened, thereby altering the pathway to preferentially form 3 (vide infra). Finally, the structures of 4a and 4b were unambiguously confirmed by X-ray crystallography.Open in a separate windowScheme 3Scope for the synthesis of 2-hydroxy-1-naphthaldehydes 4. Reaction scale: 1 (0.5 mmol), 2 (0.6 mmol), DCM (5 mL), 12 h, and isolated yield. ^a Run for 16 h. ^b Run for 19 h.The divergent reaction patterns observed with siloxy alkynes prompted us to explore other electron-rich alkynes. Thioalkyne 5a was next employed for the reaction with 1a. Indeed, direct extension with BF₃·OEt₂ as the catalyst successfully resulted in regioselective formation of β-naphthyl thioethers 6a in about 70% yield (†). While Brønsted acids led to a significant drop in the yield, we were pleased to find that the Lewis acid AgNTf₂ exhibited excellent catalytic activity, furnishing 6a in 89% yield (entry 6).Condition optimization^a

Entry	Catalyst	Yieldb (%)
1	BF3·OEt2	70
2	BF3·OEt2 (20 mol%)	72 (68)b
3	HNTf2	20
4	MeSO3H	7
5	AgOTf	80
6	AgNTf 2	89 (83) b

Open in a separate window^aReaction scale: 1a (0.05 mmol), 5a (0.06 mmol), catalyst (10 mol%), and DCM (0.5 mL). The yields are based on NMR analysis of the crude product using CH₂Br₂ as the internal standard.^bIsolated yield.With the above conditions, we carried out a brief examination of the reaction scope (Scheme 4). Interestingly, a similar selectivity divergence was also observed with thioalkynes. Indeed, under identical conditions, the 3-butyl-substituted and 3-unsubstituted isobenzopyryliums all reacted efficiently, with the former leading to only 2-naphthyl thioethers 6, but the latter preferentially to 2-sulfenyl-1-naphthaldehydes 7. It is worth noting that only one regioisomer was observed in all these products.Open in a separate windowScheme 4Scope for the synthesis of 2-naphthyl thioethers. Reaction scale: 1 (0.3 mmol), 5 (0.36 mmol), 12 h, and isolated yield. ^a Run for 37 h. ^b Run for 19 h. The unreacted alkyne accounted for the major mass balance.Finally, we were curious about the reactivity of ynamides in such benzannulations.¹⁰ To our delight, in the presence of 20 mol% of BF₃·OEt₂, the reaction between isochromene 1a and different ynamides 8 successfully afforded the desired 2-naphthylamine products 9 with excellent regioselectivity and good efficiency (Scheme 5). Different electron-withdrawing groups on the ynamides were all compatible with this protocol. However, to our surprise, the use of 3-unsubstituted isochromene 1g did not lead to the expected naphthaldehyde 10 at all. Instead, the same product 9a was obtained as the only product, even if two equivalents of BF₃·OEt₂ were used together with collidine as the additive (conditions in Scheme 3). This result is in dramatic contrast to the cases of siloxy alkynes and thioalkynes.Open in a separate windowScheme 5Scope of ynamides. Reaction scale: 1 (0.5 mmol), 8 (0.6 mmol), DCM (5 mL), 12 h, and isolated yield. ^a Run for 30 h. ^b Yield in parentheses is obtained with BF₃·OEt₂ (2.0 equiv.), 2,4,6-collidine (1.0 equiv.), 0 °C, and DCM.Next, further studies were performed to help understand the unusual selectivity divergence observed with siloxy alkynes and thioalkynes. It is obvious that the substituent at the 3-position of the isochromene substrates has a crucial impact on the product distribution. To further probe the possible steric and electronic effects, we incorporated various other substituents, such as different aryl groups and the bulky ^tBu group (