Visible light-mediated Alder-ene reaction of bicyclobutanes with cyclohexynes to access skipped dienes and skeletal remodeling to tetralines

Abstract

The strain-release driven reactions of bicyclo[1.1.0]butanes (BCBs) offer a powerful synthetic strategy for the construction of complex carbocyclic frameworks. Herein, we report the direct engagement of BCBs with transient cyclohexynes through a visible light-mediated dual-strain release manifold. By utilizing a photochemical peroxide precursor, this catalyst-free protocol bypasses the deleterious byproducts and super-stoichiometric fluoride sources typically required for cyclohexyne generation. This methodology allows the streamlined synthesis of skipped dienes bearing an all-carbon quaternary centre that are otherwise difficult to access via traditional means. The versatility of this reaction is underscored by its excellent functional group tolerance, scalability, and successfully extended to benzynes under mild and ambient conditions. Furthermore, the synthetic utility is demonstrated through the late-stage diversification of bioactive molecules and a synergistic S₈/DMSO-mediated skeletal remodeling of skipped dienes to functionalized tetralines.

A visible-light-mediated Alder-ene reaction between cyclohexynes and bicyclobutanes is presented, bypassing harsh conditions to couple two highly strained systems. This strategy provides efficient access to C(sp3)-rich skipped dienes and skeletally remodeled tetralins.

Introduction

As the geometrically constrained members of the fused bicyclic hydrocarbon family, bicyclo[1.1.0]butanes (BCBs) have emerged as pivotal motifs in contemporary molecular design. The extreme angular distortion inherent to the BCB framework manifests considerable strain energy of 66.3 kcal/mol. This energy is primarily localized within the highly destabilized central C-C bond, providing a substantial energetic bias that facilitates a broad spectrum of structural diversifications. This chemical spring allows BCBs to function as versatile scaffolds for the assembly of molecular architectures ranging from simple cyclobutanes to sophisticated bridged, fused, and spirocyclic systems (Fig. 1a)^1–9. These transformations follow three primary pathways: (i) the construction of complex bicyclic frameworks via divergent annulation pathways, (ii) the synthesis of functionalized cyclobutanes through strain-release bridgehead cleavage, and (iii) the generation of cyclobutene motifs. Bicyclic scaffolds are typically accessed through (3 + n) annulation strategies (n = 1–4)^10–43 and even higher order^44–46, leveraging the BCB unit as a potent three-carbon component. Alternatively, the cyclobutane^47–54 framework is forged through precise mono or difunctionalization and spirocyclization pathways^55–58, while unsaturated cyclobutene derivatives are efficiently realized through atom-economical isomerization^59–62 or Alder-ene-type rearrangements.

Fig. 1. Introduction and background to the present work.

a General reactivity of BCBs. b Previous reports on ene-type reactions of BCBs. c Comparison of Strain Energy. d Strategic Utility and Limitations of Strained Cyclohexynes. e Photochemical ene reaction of BCBs with Cyclohexynes. f Biologically relevant skipped diene. g Methodological extension.

The exploration of Alder-ene-type reactivity in BCBs traces back to the pioneering work of Blanchard and Cairncross⁶³, who demonstrated the reactivity of BCBs with the highly electron-poor alkene (Fig. 1b)^64–66. This was recently expanded by Bai, Chang and coworkers through a transition-metal-free, umpolung-type variant^67,68. In parallel, Pomerantz established the feasibility of ene reaction between BCBs and activated alkynes in the early 1980s⁶⁹. Building upon these foundations, the scope of this transformation has recently shifted toward strain-promoted systems; notably, Anderson and Biju have reported the engagement of BCBs with arynes and cyclopropenes for the construction of substituted cyclobutenes⁷⁰. Despite these advancements in harnessing reactive intermediates, a significant void remains regarding the utilization of cyclohexynes with BCBs. Such a transformation would be particularly compelling, as the coupling of these two highly strained species could facilitate a double strain-release manifold, providing efficient access to functionalized skipped diene moiety.

The synthetic utility of cyclohexynes^71–74 as transient intermediates is largely dictated by their significant angular distortion, which imparts a high degree of electrophilicity and facilitates a diverse array of dipolar and [4 + 2] cycloadditions^75–80. However, despite their established reactivity, reports of their engagement in Alder-ene-type transformations remain largely unexplored⁸¹, particularly under mild photochemical conditions. This distinction is evident upon comparison of the strain energies, while arynes⁸² and cyclopropenes⁸³ have extreme ring strain that forces them to react rapidly, cyclohexynes sit in a different energy category altogether (Fig. 1c).

Although cyclohexynes exhibit lower ring strain relative to their smaller-ring or benzenoid counterparts⁷⁴, they possess sufficient internal tension to undergo transformations that are inaccessible to unstrained, acyclic alkynes. In the absence of such geometric distortion, linear alkynes typically necessitate the presence of potent electron-withdrawing groups to achieve the necessary activation for ene-type reactivity. Despite their high reactivity and significant synthetic potential, the exploration of strained cyclohexynes has been hindered by several challenges, including their propensity for rapid decomposition and the requirement of excess fluoride sources or use of base during in situ generation (Fig. 1d)⁸⁴. Furthermore, achieving precise control over regio- and diastereoselectivity in these systems remains a persistent hurdle. Nevertheless, cyclohexynes provide high fraction of sp³ hybridized carbon atoms, facilitate the installation of multiple stereogenic elements, and maintain a high degree of saturation, which are increasingly prioritized in modern drug discovery⁸⁵.

Building upon our established strain-promoted ene reaction of BCBs with arynes/cyclopropenes⁷⁰, we sought to develop a photochemical strategy for the engagement of transient cyclohexynes with BCBs. The core of this strategy was finding a peroxide precursor, pioneered by Shi and co-workers⁸⁶, which facilitates the photochemical generation of cyclohexyne without generating deleterious byproducts. Herein, we disclose a catalyst- and fluoride-free ene reaction furnishing skipped dienes bearing an all-carbon quaternary center (Fig. 1e). Closely related skipped diene frameworks are prevalent in diverse natural products^87,88, including β-lumicolchicine⁸⁹, bakuchiol⁹⁰, and jerangolid A (Fig. 1f)⁹¹. Furthermore, this strategy was successfully extended to benzynes, providing a mild photochemical alternative to traditional thermal and fluoride-mediated conditions (Fig. 1g)⁷⁰. Finally, we also demonstrate the synthetic utility of these adducts through their conversion into skeletally remodeled tetralines, offering an atom-economical pathway to biologically relevant carbocycles⁹².

Results

Reaction optimization for Alder-ene reaction of BCBs with cyclohexynes

We began our investigation by examining the visible light-mediated ene reaction of bicyclobutane 1a with the peroxide-based cyclohexyne precursor 2a. Gratifyingly, irradiation of a dilute solution of 1a and 2a in MTBE (0.025 M) at 25 °C using 427 nm LEDs afforded the desired skipped diene 3a in 79% yield (Table 1, entry 1). The solvent concentration was found to play a crucial role in determining the efficiency of the transformation. Increasing the concentration to 0.05 M led to a significant erosion of yield (58%, entry 2), whereas further dilution to 0.0125 M maintained high efficiency (76%, entry 3). These observations are consistent with the involvement of a highly reactive, transient cyclohexyne intermediate, for which dilution likely suppresses competitive bimolecular decomposition pathways while favoring productive ene capture by the BCB. Reducing the amount of 2a from 3.0 to 2.0 equiv resulted in a pronounced decrease in reactivity (46%, entry 4). Shortening the reaction time from 48 h to 36 h led to incomplete conversion (66%, entry 5), while elevating the temperature to 50 °C proved detrimental, affording the desired product in only 36% yield (entry 6). A broad solvent screening revealed MTBE to be optimal for this photochemical ene reaction. Ethereal solvents such as Et₂O and DME afforded diminished yields (29–34%, entries 7 and 8), while more polar solvents including EtOAc, acetone, THF, and MeCN resulted in further attenuation of reactivity (entries 9–13). The use of molecular sieves was found to be beneficial as its omission led to a decrease in yield of 3a (63%, entry 14), indicating the sensitivity of either the cyclohexyne or BCB toward adventitious moisture. The reaction exhibited a marked dependence on the irradiation wavelength as the reaction performed at 365 nm or 440 nm enabled product formation in lower yields (entries 15, 16).

Table 1.

Optimization of the reaction conditions^a

Entry	Variation of the initial conditionsa	Yield of 3a (%)b
1	None	79(77)
2	0.05 M instead of 0.025 M	58
3	0.0125 M instead of 0.025 M	76
4	2 equiv of 2a instead of 3 equiv	46
5	36 h instead of 48 h	66
6c	50 °C instead of 25 °C	36
7	Et2O instead of MTBE	29
8	DME instead of MTBE	34
9	EtOAc instead of MTBE	37
10	Toluene instead of MTBE	31
11	Acetone instead of MTBE	23
12	THF instead of MTBE	41
13	CH3CN instead of MTBE	21
14	Without 4 Å MS	63
15	365 nm instead of 427 nm	51
16	440 nm instead of 427 nm	42

^aInitial conditions: 1a (0.1 mmol), 2a (3.0 equiv), MTBE (4.0 mL), 4 Å (25 mg), 25 °C and 48 h.

^bThe ¹H NMR yield of the crude products was determined using CH₂Br₂ as the internal standard and the isolated yield was given in parenthesis.

^cReaction mixture was stirred for 24 h.

Substrate scope studies

With the optimized conditions established, we investigated the generality of this visible light-mediated Alder-ene reaction across diverse BCBs and cyclohexyne precursors (Fig. 2). We first explored variations on the BCB aryl substituents. Substrates bearing electron-donating, electron-neutral, and halogen groups at the para- and meta- positions of aryl ring of BCB underwent smooth transformation, yielding the corresponding skipped dienes 3a–3k in moderate to good yields (52–77%). Steric effects were briefly examined with an ortho-substituted BCB, which successfully furnished 3l in 54% yield. Furthermore, disubstituted aryl rings were also well-tolerated, providing products 3m and 3n in 78% and 56% yields respectively. Beyond simple phenyl-substituted BCBs, the reaction tolerated broader structural complexity. For instance, the 2-naphthyl analogue afforded 3o (68% yield), while the vinyl-substituted BCB efficiently participated in the reaction to provide 3p (65% yield). A broad selection of ester groups, including phenyl, benzyl, isopropyl, tert-butyl, and cyclohexyl, proved compatible, delivering products 3q-3u with consistent efficiency. A cinnamyl-substituted ester BCB also participated effectively, yielding 3v in 57% yield. To showcase the synthetic utility for late-stage diversification, we subjected biologically relevant ester derivatives to the reaction conditions. BCB derived from geraniol, L-menthol, and gemfibrozil were successfully functionalized to afford 3w-3y (54–71%). Notably, in the case of the L-menthol-derived substrate, the corresponding product 3x was obtained as an inseparable mixture of diastereomers. Additionally, a pyrazole-derived amide and Weinreb amide derived BCB was successfully engaged in the present Alder-ene reaction to afford the skipped diene 3z and 3aa in 51% and 49% yield, demonstrating compatibility with amide and heterocyclic motifs. The scope of the methodology was further assessed using representative monosubstituted BCBs (1ac–1ae), however, these substrates were found to be incompatible under the present reaction conditions (see Supplementary Fig. 4).

Fig. 2. Substrate scope of visible light-mediated Alder-ene reaction of BCBs with cyclohexynes.

Reaction condition: 1 (0.1 mmol), 2 (0.3 mmol), 4 Å MS (25 mg), MTBE (4.0 mL), 25 °C. ^a The products were isolated as inseparable mixture of isomers, although the isomers were clearly visible in ¹H NMR, due to peaks overlapping, the unequivocal ratio determination was not done.

Next, substituted cyclohexyne partners were explored. The reaction of 1a performed with a series of monosubstituted cyclohexyne precursors readily furnished the desired products 3ab–3ae in 48–73% yields. These products were obtained as inseparable mixture of isomers inherent to the reactivity of unsymmetrical cyclohexynes. 3,3-Dimethyl cyclohexyne also proved to be an excellent partner, giving 3af and 3af′ in 71% yield. The method could be extended to a chiral cyclohexyne precursor, providing mixture of products 3ag and 3ag′ (42% yield) thus demonstrating the ability to handle highly complex structures. To further define the scope of the methodology, representative bridged cyclic alkyne 2h, seven-membered 2i and eight-membered 2j precursors were examined. However, no desired Alder-ene products were obtained under the optimized conditions (see Supplementary Fig. 4). Together, these results highlight the impressive range of functional groups this protocol can tolerate.

To demonstrate the versatility of this photochemical manifold, the methodology was extended to the generation and capture of benzynes. Although our group previously reported the similar transformation under thermal conditions, a photochemical variant remained an attractive goal for achieving greater operational mildness and efficiency. Initially, phthaloyl peroxide was explored as a potential benzyne source⁹³, analogous to the peroxide-based cyclohexyne generation; however, this precursor failed to deliver the desired ene product. Consequently, the focus was shifted to o-diazonium carboxylic acid 4a⁹⁴, which are known to undergo photochemical extrusion to generate benzyne intermediates upon 365 nm irradiation. Thus, irradiation of 1a with 4a in CH₃CN at 365 nm for 12 h at 25 °C afforded the aryl cyclobutene 5a in 96% yield (Fig. 3).

Fig. 3. Scope of the light-mediated ene reaction of BCBs with benzynes.

Reaction conditions: 1 (0.1 mmol), 4a (0.2 mmol), CH₃CN (3 mL), 365 nm, 25 °C.

This result prompted an evaluation of the BCB scope with respect to the benzyne partner. Differentially substituted BCBs bearing various aryl groups participated seamlessly, furnishing products 5a–5e in high yields. The reaction of pyrazole-derived BCB with benzyne generated from 4a afforded the cyclobutene 5f in 79% yield. Furthermore, the method was robust for the late-stage functionalization of complex molecules. For instance, BCBs derived from the drug gemfibrozil afforded 5g in 81% yield. These results underscore the advantages of the photochemical approach, generating benzyne under mild and fluoride-free conditions, offering nearly quantitative conversion and broad functional group compatibility.

To showcase the synthetic utility of the skipped dienes, their potential as precursors for the construction of complex carbocyclic frameworks were explored. We hypothesized that the cyclobutene moiety within this adduct could undergo a thermally induced 4π-electrocyclic ring-opening to generate a hexatriene species, followed by an intramolecular cyclization and a subsequent aromatization to form tetralines (see Supplementary fig. 5). In an initial experiment, treatment of 3a with elemental sulfur in DMSO at 140 °C in open air conditions afforded the desired tetraline 6a in 86% yield (Fig. 4). This transition-metal-free, oxidative skeletal remodeling protocol utilizes sulfur in catalytic quantities as the terminal oxidant, with DMSO functioning as both the reaction medium and an auxiliary oxygen source^95–97. This synergistic system bypasses the requirement for toxic or expensive metal-based reagents, offering a sustainable alternative for oxidative aromatization.

Fig. 4. Substrate scope for tetraline synthesis via skeletal remodeling.

Reaction conditions: 3 (0.1 mmol), S₈ (0.03 mmol), DMSO (1 mL), 140 °C.

To evaluate the generality of this remodeling protocol, we examined a series of structurally diverse skipped dienes. Gratifyingly, the deuterium-labeled analogue 3a′ underwent the same transformation to furnish the corresponding deuterium-labeled analogue 6a′, a biologically relevant scaffold, in excellent yield with complete retention of deuterium. Variations on the para- and meta-positions of the aryl ring at the cyclobutene were well-tolerated, delivering the corresponding substituted tetralines 6b–6h in excellent yields (88–97%). Moreover, replacement of phenyl group with 2-naphthyl moiety successfully provided the β-naphthyl-derived tetralin 6i in 89% yield. The versatility of the ester moiety was demonstrated by the benzyl-ester substrate, which furnished tetralin 6j in 96% yield. Notably, the cascade proved effective for complex molecular architectures, as evidenced by the successful transformation of a gemfibrozil-derived skipped diene into product 6k in 85% yield. To demonstrate the applicability of the remodeling strategy for the synthesis of tetralines bearing stereogenic centers in the saturated ring, regioisomeric pairs 3ab/3ab′ were subjected to the standard conditions. Gratifyingly, the corresponding tetralin 6l/6l′ was obtained in 81% albeit as mixtures of regioisomers.

Discussion

To gain insight into the reaction mechanism, a series of control experiments were conducted. First, the necessity of photochemical activation was confirmed, as no formation of product 3a was observed in the absence of light, with the starting materials remaining unreacted (Fig. 5a). Likewise, performing the reaction at 50 °C under otherwise identical conditions in the absence of light did not afford 3a, ruling out a thermally driven pathway operating in the present case. Notably, the transformation proceeds under low-energy blue-light irradiation (427 nm) in a significantly more diluted reaction medium than previously reported for the photochemical generation of cyclohexynes and therefore requires a longer reaction time (48 h)⁸⁶. These observations suggest that the slower generation of cyclohexyne under the present conditions suppresses competing decomposition pathways, thereby enabling its efficient trapping by BCB. Furthermore, irradiation of BCB 1a in the absence of peroxide precursor 2a did not result in the formation of the isomerized cyclobutene 7a (Fig. 5b), indicating that the BCB scaffold does not undergo independent homolytic or heterolytic bond cleavage under the reaction conditions. To investigate whether the transformation proceeds through a stepwise pathway, isolated cyclobutene 7a was subjected to the optimized reaction conditions in the presence of 2a. However, no formation of 3a was observed, effectively excluding 7a as a viable intermediate in the reaction pathway (Fig. 5c).

Fig. 5. Mechanistic experiments.

a Reaction in the absence of light source. b Reaction in the absence of 2a. c Reaction with cyclobutene 7a instead of 1a. d Radical Quenching Experiment. e Reaction in presence of D₂O. f Reaction with d₄ - BCB. g Reaction of d₄ - BCB in presence of H₂O. h Light On/Off Experiment. i Competition experiments.

To get further insight into the mechanism, we performed radical quenching studies using TEMPO and BHT as the radical quenchers (RQ). The addition of TEMPO completely inhibited the formation of 3a, likely due to a direct interaction with the in situ generated cyclohexyne (confirmed by HRMS) thereby inhibiting the reaction (Fig. 5d). Similarly, the reaction performed in the presence of BHT led to a significant yield depletion and 3a was formed in 33% yield. To distinguish between radical and concerted pathways, the reaction was conducted in the presence of D₂O. This transformation proceeded with the formation of 3a in 60% yield with no detectable deuterium incorporation (Fig. 5e).

To further probe the proposed Alder-ene pathway, a deuterium-labeling experiment was performed using d₄-labeled BCB 1a′, which possessed 92% deuterium incorporation. Under the standard reaction conditions, the corresponding product 3a′ was obtained in 70% yield with 92% deuterium incorporation at the cyclohexene position, indicating essentially complete retention of the deuterium label during the transformation (Fig. 5f). To further examine the possibility of proton/deuterium exchange, the reaction was repeated in the presence of 2.0 equiv of water. Notably, the deuterium incorporation in 3a′ remained unchanged (92%), indicating that no significant H/D exchange occurred under the reaction conditions. (Fig. 5g). Collectively, these observations are consistent with a concerted Alder-ene process involving direct deuterium transfer. To confirm the photochemical requirement of the transformation, an intermittent light on/off experiment was performed under the standard reaction conditions. Product formation proceeded efficiently during the irradiation periods, whereas only minimal additional conversion was observed during the dark intervals. Upon re-irradiation, product formation resumed, demonstrating that continuous light irradiation is required for efficient reaction progression. (Fig. 5h).

The role of strain as a primary driving force was further elucidated through a few competition experiments. When BCB 1a was exposed to a mixture of the in situ generated cyclohexyne and the electronically activated but unstrained alkyne 8, the Alder-ene product 3a was formed exclusively in 62% yield (Fig. 5i). The absence of a DMAD-derived product 9 underscores that strain alone is the decisive factor in the present strategy. Finally, a competition experiment between the in situ generated cyclohexyne and benzyne revealed a preference for the aryne-derived product 5a. Under 427 nm irradiation, a 1:2 ratio of 3a to 5a was observed, which further shifted to 1:5 at 365 nm. These observations suggest that although both species are highly reactive, the higher strain energy and superior electrophilicity of the benzyne intermediate lead to faster interceptive kinetics with the BCB.

Synthetic applications

To demonstrate the scalability and synthetic utility of the present method, a scale-up experiment was conducted, which successfully provided 3a in 74% yield (Fig. 6). The robustness of the synthesized product 3a was further highlighted through a series of diverse chemical transformations. Reduction of 3a with LiAlH₄ smoothly afforded the corresponding primary alcohol 10a in 89% yield, while nucleophilic addition using MeMgBr yielded the tertiary alcohol 11a (62% yield). Hydrolysis of the ester moiety provided the carboxylic acid 12a in 81% yield. The structure of 12a was unambiguously confirmed by the X-ray crystallographic analysis. Furthermore, the alkene motifs within the product were subjected to oxidative and reductive transformations. Epoxidation with m-CPBA furnished the epoxide 13a in 55% yield as a mixture of diastereomers in 2.2:1 ratio. Further exhaustive hydrogenation over Pd-C yielded the saturated cyclobutane derivative 14a in 78% yield and 1.1:1 diastereomeric ratio. The utility of the scaffold was further shown through the ozonolysis of 3a, which led to a productive deconstruction of the molecule. This process involved the cleavage of both double bonds followed by decarboxylation to provide product 15a in 70% yield. Furthermore, hydrogenation of 3a under mild conditions in presence of [Rh(cod)Cl)]₂ and PtO₂ afforded fully saturated trisubstituted cyclobutane 16a in 71% yield (dr = 1.3:1). Collectively, these transformations highlight the synthetic versatility of the skipped diene products and further underscore the utility of the present methodology for accessing structurally diverse and synthetically valuable molecular scaffolds.

Fig. 6. Synthetic applications.

Scale-up reaction and product functionalization.

In conclusion, we have established a catalyst- and fluoride-free photochemical manifold leveraging the dual-strain release of BCBs and transient cyclohexynes. By employing visible light active precursors, this protocol bypasses harsh reagents to forge C(sp³)-rich skipped dienes containing all-carbon quaternary center. Mechanistic studies are consistent with a concerted Alder-ene pathway. The utility of the strategy is underscored by its scalability, compatibility with benzynes, and late-stage diversification of bioactive motifs. Furthermore, a synergistic S₈/DMSO-mediated remodeling provides a transition-metal-free gateway to complex tetralines, offering a powerful toolkit for constructing architecturally dense molecular space.

Methods

Procedure for synthesis of skipped dienes via Alder-ene reaction of BCBs with cyclohexynes

An oven-dried Schlenk tube equipped with a magnetic stir bar was taken the BCB 1 (0.10 mmol, 1.0 equiv), cyclohexyne precursor 2 (0.30 mmol, 3.0 equiv) and 4 Å molecular sieves (25 mg). This was followed by the addition of MTBE (4.0 mL) under a nitrogen atmosphere at 25 °C. The reaction vessel was then placed in an EvoluChem PhotoRedOx Box and exposed to light at a wavelength of 427 nm for 48 h. Upon completion, the reaction mixture was filtered through a short pad of silica gel, eluting with EtOAc (10 mL). The filtrate was concentrated under reduced pressure, and the resulting residue was purified by flash column chromatography on silica gel to afford the corresponding skipped dienes.

Author contributions

S.S. and A.T.B. designed the project. S.S. optimized the reaction conditions. S.S. examined the substrate scope and performed the mechanistic experiments. S.S. wrote the first draft of the manuscript and A.T.B. edited the manuscript. All authors have given approval to the final version of the manuscript. A. T. B. directed the research.

Peer review

Peer review information

Nature Communications thanks Jin-Shun Lin, Hai-Jun Zhang and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Funding

This work was financially supported by the SERB, Government of India, (File Number: SCP/2022/000837). S.S. thanks UGC for Junior Research Fellowship.

Data availability

The data generated in this study has been provided within the article and in the Supplementary Information (PDF). All data supporting the findings of this manuscript are also available from the corresponding author upon request. Crystallographic data for the structure reported in this article have been deposited at the Cambridge Crystallographic Data Center, with the deposition number CCDC 2539293 (12a). These data can be obtained free of charge from Data Center via www.ccdc.cam.ac.uk/data_request/cif.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41467-026-77468-2.