Visible-Light-Mediated Catalyst-Free [2+2] Cycloaddition Reaction for Dihydrocyclobuta[b]naphthalene-3,8-diones Synthesis under Mild Conditions

A facile and efficient visible-light-mediated method for directly converting 1,4-naphthoquinones into dihydrocyclo-buta[b]naphthalene-3,8-diones (DHCBNDOs) under mild and clean conditions without using any photocatalysts is reported. This approach exhibited favorable compatibility with functional groups and afforded a series of DHCBNDOs with excellent regioselectivity and high yields. Moreover, detailed mechanism studies were carried out both experimentally and theoretically. The readily accessible, low-cost and ecofriendly nature of the developed strategy will endow it with attractive applications in organic and medicinal chemistry.


Introduction
The development of green, concise and mild synthetic methodologies for extensively used organic compounds is one of the most essential tasks in the area of organic synthesis.Fused cyclobutene is a type of unique and intriguing structural motif in organic chemistry, which has attracted considerable attention in recent years because it is luxuriant in nature and shows multiple biological activities [1][2][3][4][5][6][7][8][9].Furthermore, owing to the high chemical reactivity from the inherent and high ring strain, the fused cyclobutenes experience diverse transformations, ring-opening and ring-expansion reactions in particular, providing distinctive approaches to the construction of complex organic molecules [10][11][12][13].Dihydrocyclobuta [b] naphthalene-3,8-diones (DHCBNDOs) containing fused cyclobutene structures are highly valued scaffolds that occur in many bioactive compounds (Figure 1) and, in particular, operate as lead compounds for SARS-CoV-19 management, [14] miR-1 inhibitors [15], and microRNAs activators [16].A great deal of attention has been devoted to developing better strategies for synthesizing DHCBNDOs and their analogs, of which [2+2] cycloaddition is generally considered to be one of the most forthright and efficient methods [17][18][19][20][21].The DHCBNDOs, an interesting scaffold for developing new drugs for SARS-CoV-19, coronary artery diseases, heart attacks and miRNA-based tumors, are underrepresented in medicinal chemistry [14][15][16], and this potentially results from deficiencies in extant synthetic methodologies.
tumors, are under-represented in medicinal chemistry [14][15][16], and thi from deficiencies in extant synthetic methodologies.As shown in Scheme 1, the DHCBNDO system has only recentl by formal [2+2] cycloaddition under reflux condition, as reported by and coworkers [22][23][24].Some great improvements via [2+2] photocyclo made by Zhang's (high-pressure mercury lamp) [15] and Shah's (vis [2+2] cycloaddition) groups [25].However, the reported methods seve unfriendly conditions, such as heating, high-pressure mercury lamp, a idants, which created some drawbacks in these strategies in terms of tection, safety, economy, reaction selectivity and functional group t significant progress has been achieved in the synthesis of DHCBNDO vious decades, the development of approaches for preparing DHCBN green and facile way is an ongoing challenge that urgently needs to b To supplement our initial research in the synthesis of tricyclic organic compounds using photocatalytic reactions [26][27][28][29], we wou green, facile and visible-light-mediated [2+2] cycloaddition reaction fo thesis at ambient temperature.We believe this approach not only repr and economical method for the preparation of DHCBNDOs, but also prospects in synthetic chemistry, such as constructing organic molecul tures.As shown in Scheme 1, the DHCBNDO system has only recently been synthesized by formal [2+2] cycloaddition under reflux condition, as reported by Diederich, Morita and coworkers [22][23][24].Some great improvements via [2+2] photocycloaddition have been made by Zhang's (high-pressure mercury lamp) [15] and Shah's (visible-light-mediated [2+2] cycloaddition) groups [25].However, the reported methods severely relied on some unfriendly conditions, such as heating, high-pressure mercury lamp, acid catalyst and oxidants, which created some drawbacks in these strategies in terms of environmental protection, safety, economy, reaction selectivity and functional group tolerance.Although significant progress has been achieved in the synthesis of DHCBNDO systems over previous decades, the development of approaches for preparing DHCBNDOs in an efficient, green and facile way is an ongoing challenge that urgently needs to be addressed.As shown in Scheme 1, the DHCBNDO system has only recently by formal [2+2] cycloaddition under reflux condition, as reported by D and coworkers [22][23][24].Some great improvements via [2+2] photocycload made by Zhang's (high-pressure mercury lamp) [15] and Shah's (visib [2+2] cycloaddition) groups [25].However, the reported methods severe unfriendly conditions, such as heating, high-pressure mercury lamp, acid idants, which created some drawbacks in these strategies in terms of en tection, safety, economy, reaction selectivity and functional group tole significant progress has been achieved in the synthesis of DHCBNDO s vious decades, the development of approaches for preparing DHCBNDO green and facile way is an ongoing challenge that urgently needs to be a To supplement our initial research in the synthesis of tricyclic or organic compounds using photocatalytic reactions [26][27][28][29], we would green, facile and visible-light-mediated [2+2] cycloaddition reaction for D thesis at ambient temperature.We believe this approach not only repres and economical method for the preparation of DHCBNDOs, but also ex prospects in synthetic chemistry, such as constructing organic molecules tures.

Scheme 1.
[2+2] cycloaddition reaction for the synthesis of DHCBNDOs.To supplement our initial research in the synthesis of tricyclic or polycyclic fused organic compounds using photocatalytic reactions [26][27][28][29], we would like to report a green, facile and visible-light-mediated [2+2] cycloaddition reaction for DHCBNDOs synthesis at ambient temperature.We believe this approach not only represents a mild, clean and economical method for the preparation of DHCBNDOs, but also exhibits promising prospects in synthetic chemistry, such as constructing organic molecules of complex structures.

Results and Discussion
In our pilot experiment, menadione 1a (1 mmol) and phenylacetylene 2a (1 mmol) were chosen as model substrates in solvet MeCN, which was irradiated under blue LEDs (460 nm) for 3 h.The target tricyclic framework compound 8a-methyl-1-phenyl-2a,8adihydrocyclobuta[b]naphthalene-3,8-dione 3aa was obtained in 81% yield without any catalyst (Table 1, entry 1).Furthermore, the use of several different solvents, including DCM, acetone, dioxane, chlorobenzene, MeOH, THF, toluene, and DCE, did not give a better result than MeCN (Table 1, entries 220139, respectively).However, when a solvent of MeCN was used with irradiation for 4 h, the yield of 3aa was improved to 86%.(Table 1, entry 10).Increasing the reaction time to 5h or increasing the equivalent of 2a to 1.4eq failed to give any better results (Table 1, entries 11-14, respectively).And the control experiments of no light and different wavelength (365 nm) did not give better results (Table 1, entries 15-16, respectively).Therefore, the optimized reaction conditions for this [2+2] cycloaddition reaction turned out to be those obtained using MeCN as the solvent (0.1 mmol/L) under blue LEDs (460 nm) irradiation for 4h.

Results and Discussion
In our pilot experiment, menadione 1a (1 mmol) and phenylacetylene 2a (1 mmol) were chosen as model substrates in solvet MeCN, which was irradiated under blue LEDs (460 nm) for 3 h.The target tricyclic framework compound 8a-methyl-1-phenyl-2a,8a-dihydrocyclobuta[b]naphthalene-3,8-dione 3aa was obtained in 81% yield without any catalyst (Table 1, entry 1).Furthermore, the use of several different solvents, including DCM, acetone, dioxane, chlorobenzene, MeOH, THF, toluene, and DCE, did not give a better result than MeCN (Table 1, entries 220139, respectively).However, when a solvent of MeCN was used with irradiation for 4 h, the yield of 3aa was improved to 86%.(Table 1, entry 10).Increasing the reaction time to 5h or increasing the equivalent of 2a to 1.4eq failed to give any better results (Table 1, entries 11-14, respectively).And the control experiments of no light and different wavelength (365 nm) did not give better results (Table 1, entries 15-16, respectively).Therefore, the optimized reaction conditions for this [2+2] cycloaddition reaction turned out to be those obtained using MeCN as the solvent (0.1 mmol/L) under blue LEDs (460 nm) irradiation for 4h.With the optimized conditions in hand, a survey of the substrate scope was carried out by varying 1,4-naphthoquinone and alkyne compounds (Scheme 2).Phenylacetylene 2a provided the corresponding product 3aa in very good yields under optimized reaction conditions.And the variability of the phenylacetylene was checked using substituted alkyne 2. The phenylacetylene 2 tethered with an electron-donating group, such as -OMe, 3,5-OMe, -Et, -Pr, and -cyclohexyl, also underwent [2+2] cycloaddition reactions with very good yields (3ab-3af).Similarly, 2 with electron-withdrawing groups, such as -Br, -NO2, and -CN, also underwent [2+2] cycloaddition reactions with very good yields (3ag-3ai).With the optimized conditions in hand, a survey of the substrate scope was carried out by varying 1,4-naphthoquinone and alkyne compounds (Scheme 2).Phenylacetylene 2a provided the corresponding product 3aa in very good yields under optimized reaction conditions.And the variability of the phenylacetylene was checked using substituted alkyne 2. The phenylacetylene 2 tethered with an electron-donating group, such as -OMe, 3,5-OMe, -Et, -Pr, and -cyclohexyl, also underwent [2+2] cycloaddition reactions with very good yields (3ab-3af).Similarly, 2 with electron-withdrawing groups, such as -Br, -NO 2 , and -CN, also underwent [2+2] cycloaddition reactions with very good yields (3ag-3ai).Alkyne derivatives with heterocycle substituents were also found to be suitable for this [2+2] cycloaddition reaction (3aj and 3ak), and non-terminal alkynes with -CHO and -Ph groups were also suitable substrates, giving 3al and 3am in very good yields.Similarly, aliphatic alkynes also gave the corresponding [2+2] cycloaddition products (3an and 3ao) in excellent yield, and 1,4-naphthoquinones bearing electron-withdrawing substituents (-H, -Cl, and -2,3Cl) did the same for corresponding products (3ba-3da) in high yields (87-92%).These different groups did not significantly affect the yields, probably due to the weak influence of electronic effects.It is noteworthy that the reaction proceeded with excellent regioselectivity due to the trans-product being hard to access because of the ring strain, giving target cis-8a-methyl-1-phenyl-2a,8a-dihydrocyclobuta[b]naphthalene-3,8-diones 3. The proton NMR coupling signals (two doublets) from the two cyclobutene protons proved that -Me and -Ph groups were adjacent to each other (Supplementary Figure S1).The singlecrystal X-ray diffraction (CCDC-2288747) further confirmed the cis-structure (Figure 2).As a consequence, the NMR and X-ray analysis confirmed excellent regioselectivity of the visible-light-mediated [2+2] cycloaddition reaction.
Molecules 2023, 28, x FOR PEER REVIEW 4 of 12 in excellent yield, and 1,4-naphthoquinones bearing electron-withdrawing substituents (-H, -Cl, and -2,3Cl) did the same for corresponding products (3ba-3da) in high yields (87-92%).These different groups did not significantly affect the yields, probably due to the weak influence of electronic effects.It is noteworthy that the reaction proceeded with excellent regioselectivity due to the trans-product being hard to access because of the ring strain, giving target cis-8a-methyl-1-phenyl-2a,8a-dihydrocyclobuta[b]naphthalene-3,8diones 3. The proton NMR coupling signals (two doublets) from the two cyclobutene protons proved that -Me and -Ph groups were adjacent to each other (Supplementary Figure S1).The single-crystal X-ray diffraction (CCDC-2288747) further confirmed the cis-structure (Figure 2).As a consequence, the NMR and X-ray analysis confirmed excellent regioselectivity of the visible-light-mediated [2+2] cycloaddition reaction.To evaluate the synthetic utility of our visible light-mediated approach of 1a and 2a was carried out under sunlight irradiation (about 8 h per day), uct 3aa was given in 82% yield (Scheme 3A).In addition, a large-scale (8-mmo aration of 3aa under sunlight irradiation was successful, with only a sligh yield (71%), which confirmed the practicability of this strategy (Scheme 3B).After exploring the substrate scope and utility of this reaction, we fo study mechanism.In order to obtain full insight into this intermolecular [2+ tion reaction, some control experiments (Scheme 4) and theoretical calculat 5) were conducted by using 1a and 2a as the model reactants.No reaction oc dark environments (without blue LEDs) (Scheme 4B), which highlighted To evaluate the synthetic utility of our visible light-mediated approach, the reaction of 1a and 2a was carried out under sunlight irradiation (about 8 h per day), and the product 3aa was given in 82% yield (Scheme 3A).In addition, a large-scale (8-mmol scale) preparation of 3aa under sunlight irradiation was successful, with only a slightly decreased yield (71%), which confirmed the practicability of this strategy (Scheme 3B).To evaluate the synthetic utility of our visible light-mediated approach, the reaction of 1a and 2a was carried out under sunlight irradiation (about 8 h per day), and the product 3aa was given in 82% yield (Scheme 3A).In addition, a large-scale (8-mmol scale) preparation of 3aa under sunlight irradiation was successful, with only a slightly decreased yield (71%), which confirmed the practicability of this strategy (Scheme 3B).After exploring the substrate scope and utility of this reaction, we focused on the study mechanism.In order to obtain full insight into this intermolecular [2+2]-cycloaddition reaction, some control experiments (Scheme 4) and theoretical calculations (Scheme 5) were conducted by using 1a and 2a as the model reactants.No reaction occurred in the dark environments (without blue LEDs) (Scheme 4B), which highlighted that the [2+2] cycloaddition reaction need to be triggered by blue LEDs.In addition, 2,2,6,6-tetramethyl-1-piperidinyloxy (with TEMPO as a radical scavenger) was added in the reaction of Scheme 4C, and only a trace amount of compound 3aa was detected.After exploring the substrate scope and utility of this reaction, we focused on the study mechanism.In order to obtain full insight into this intermolecular [2+2]-cycloaddition reaction, some control experiments (Scheme 4) and theoretical calculations (Scheme 5) were conducted by using 1a and 2a as the model reactants.No reaction occurred in the dark environments (without blue LEDs) (Scheme 4B), which highlighted that the [2+2] cycloaddition reaction need to be triggered by blue LEDs.In addition, 2,2,6,6-tetramethyl-1-piperidinyloxy (with TEMPO as a radical scavenger) was added in the reaction of Scheme 4C, and only a trace amount of compound 3aa was detected.Based on the control experimental results, DFT-calculated results (see the computational details section) and previous studies of the [2+2] cycloaddition reaction [31], a possible mechanism for this visible-light-photocatalyzed [2+2] cycloaddition reaction was proposed, as shown in Scheme 5. First, the 1a 3,*s was populated by intersystem crossing (ISC) of the singlet excited state 1a * that resulted from visible light irradiation of ground state 1a.The excited state of 1a 3,* reacted with 2a to deliver the cyclopropane intermediate (IM1) through a transition-state structure TS (IRC scan results in ESI, Supplementary Figure S40), and its free energy barrier was calculated as 13.4 kcal mol −1 .The next step in the mechanistic pathway is the reaction of the ring cleavage of the cyclopropane with the possible formation of 1,4-biradical intermediate IM2.The intermediate IM2 subsequently underwent intramolecular radical recombination, giving target ring-closure product 3aa.

Experimental Section
All chemical agents were purchased from chemical manufacturers (Bide and Energy Chemical) and directly used without any further purification.Thin layer chromatography was carried out on glass silica gel GF254 plates.Melting points were detected by a digital melting point apparatus (XT-5A).The NMR tests (1D-NMR and 2D-NMR) were conducted on a Bruker spectrometer (Avance 400).HRMS data were obtained from a Thermo Fisher mass spectrometer (Q-Exactive) in ESI mode.

Computational Section
Geometrical optimization was carried out at the um062x 6-31g(d,p) theoretical level using the SCRF model (MeCN as solvent) [32][33][34].Frequency analysis and the thermodynamic correctional data were obtained at the same level.In addition, the IRC pathways [35] calculations were also performed for transition state (TS) to identify whether the transition state can connect the reactants and the key intermediate (IM1).All of the computational experiments were conducted by using the Gaussian 16 program package [36].

Figure 1 .
Figure 1.Representative examples with the skeleton of DHCBNDO.

Figure 1 .
Figure 1.Representative examples with the skeleton of DHCBNDO.

Figure 1 .
Figure 1.Representative examples with the skeleton of DHCBNDO.

Scheme 4 .
Scheme 4. Control experiments.(A) The reaction under standard conditions; (B) The reaction under standard conditions without LEDs; (C) The reaction under standard conditions with addition of TEMPO.

Scheme 4 . 12 Scheme 4 .
Scheme 4. Control experiments.(A) The reaction under standard conditions; (B) The reaction under standard conditions without LEDs; (C) The reaction under standard conditions with addition of TEMPO.
a The reactions were carried out on a 1 mmol scale in 10 mL of solvent under blue LEDs (460 nm); irradiation without any catalyst; b Isolated yields; c Without light; d Under irradiation of 365 nm.

Entry Solvent Time Eqn (2a) Yield b
b Isolated yields; c Without light; d Under irradiation of 365 nm.