Rh‐Catalyzed [2 + 2 + 2] Cycloaddition of Linked Bis(Diynes) Ar‐C≡C‐C≡C‐(CH2)3‐C≡C‐C≡C‐Ar With Quinones: Synthesis and Photophysical Properties of Bis(Arylethynyl)‐Naphthoquinones and Anthraquinones

TL;DR

This paper describes a new method to synthesize complex quinone-based compounds using rhodium-catalyzed reactions and microwave radiation, and explores their light-related properties and reactions.

Contribution

A novel Rh-catalyzed [2 + 2 + 2] cycloaddition method for synthesizing bis(arylethynyl)quinones with microwave-enhanced yields.

Findings

Microwave irradiation significantly improved the yields of the synthesized quinone compounds.

The synthesized quinones lacked photoluminescence and singlet oxygen luminescence.

A fluorescent derivative underwent photooxidation under sunlight via sensitized 1O2.

Abstract

Herein, we report the Rh‐catalyzed [2 + 2 + 2] cycloaddition of linked bis(diynes) Ar‐C≡C‐C≡C‐(CH2)3‐C≡C‐C≡C‐Ar with quinones for the rapid and efficient assembly of rigid structures containing bis(arylethynyl)naphthoquinones or anthraquinones using microwave radiation. The photophysical properties of the products were investigated, including absorption, emission, transient absorption (TA) spectroscopy, and singlet oxygen luminescence studies. However, the quinones did not exhibit photoluminescence or detectable singlet oxygen luminescence. Computational analysis via density functional theory (DFT) was performed to elucidate the electronic and structural factors influencing their behavior. Moreover, reductive methylation of one of the bis(arylethynyl)naphthoquinones (3aa) generated the fluorescent bis(methoxy) naphthalene derivative 4aa (Φf
= 0.53). However, 4aa underwent photooxidation in air under sunlight via reaction with 1O2, which it sensitized, leading to ring opening of the bis(methoxy)arene ring.

We report the synthesis bis(arylethynyl)naphtho‐ and anthraquinones via Rh(I)‐catalyzed [2 + 2 + 2] cycloaddition reactions. The substrate scope of both quinones and linked bis‐diynes was explored. Microwave irradiation remarkably enhanced yields. A fluorescent bis(methoxy)‐bis(arylethynyl)naphthalene was synthesized, which underwent an unusual photooxidation in air under sunlight.

Figures

(A) [2 + 2 + 2] cycloaddition with benzoquinones to build complex systems including molecules with biological activity; (B) Construction of π‐extended systems using quinones for fluorescent applications; (C) Rh‐catalyzed [2 + 2 + 2] cycloadditions with α,ω‐tetraynes and quinones and application as building blocks for fluorescent compounds.

Substrate scope for the Rh‐catalyzed [2 + 2 + 2] cycloadditions with α,ω‐tetraynes and quinones. [a] Reaction conditions: 1a‐1 (0.2 mmol), 2a (2.0 equiv.), [Rh(cod)Cl)]2 (5 mol%), (R)‐tol‐BINAP (10 mol%), DCE 2 mL (0.1 M), 100°C (MW), 2 h. Isolated yields are shown.

Molecular structure of 3aa determined by single‐crystal X‐ray diffraction with displacement ellipsoids shown at the 50% probability level. CCDC number: 2338853.

Modification of the building block 3aa based on the reduction/alkylation sequence to afford the compound 4aa.

(a) UV‐vis absorption spectra of the compounds. (b) Emission spectra of the compounds, inset: 4aa in DCM under UV irradiation (365 nm). (c) Singlet oxygen luminescence of the compounds and the standard perinaphthenone (PeriNa).

Plots of the HOMOs and LUMOs of 3aa, 3ea, 3ag, 3ad, and 4aa. Contour values are ±0.03 (e/bohr3)1/2.

CAM‐B3LYP‐simulated UV‐visible absorption spectra of selected quinones and derivatives.

Density‐difference plot associated with the HOMO‐LUMO transition in 3aa and 4aa. The cyan and pink colors indicate an increase and decrease of density upon excitation, respectively [isocontour value = 0.002 e bohr−3].

Experimental (dashed line) and CAM‐B3LYP‐simulated (solid line) emission spectra of 4aa.

1H‐NMR NMR spectra (300 MHz) of fresh 4aa in C6D6 during photoaging under ambient conditions. The molecular structure of the photodegradation product 4aa‐ox determined by single‐crystal X‐ray diffraction is shown (top right), and more details are provided in the Supporting Information. CCDC number: 2484133.

Tables

TABLE 1. Optimization for the cycloadditions with bis‐diynes and quinones. a

TABLE 2. Photophysical data of the compounds.

Compound a	λ abs/nm a (ε/104M−1 cm−1)	λ em/nm	Stokes shift/cm−1	Φf	τ/ns	k r/107 s−1	k nr/107 s−1
3aa	353 (3.2), 371 (3.1)	− b	−
3ad	297 (8.0), 431 (7.2)	520 c	3900	0.05	3.3 c	1.5	28.6
3ae	376 (4.5), 390 (8.3)	−	−
3ag	342 (6.3), 360 (4.6)	−	−
3ea	273 (3.0), 356 (2.8), 376 (2.5)
4aa	277 (4.5), 321 (4.1),410 (4.1)	492 c	4000	0.53	5.92	9.0	7.9

Full text

Introduction

1

Transition‐metal‐catalyzed [2 + 2 + 2] cycloaddition reactions involving unsaturated systems such as alkynes and alkenes enable the construction of a wide range of six‐membered rings, increasing molecular complexity in a single step with high efficiency and atom economy. A range of transition metals, such as Co, Ni, Ru, Rh, Pd, and Ir, are capable of mediating or catalyzing such reactions. The generally accepted mechanism for this process involves the oxidative coupling of two alkynes to generate a key metallacyclopentadiene intermediate in which a third unsaturated moiety is inserted, affording the desired six‐membered ring after the reductive elimination of the metal [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19].

Amongst the available transition metals for this process, Rh(I)based complexes have been some of the most widely used as catalysts to mediate these kinds of reactions. In the pioneering work from Müller (1974), Wilkinson's complex, [RhCl(PPh_3_)3], was reacted with a series of 1,5‐ and 1,6‐diynes, forming the corresponding rhodacyclopentadiene intermediates. Subsequent reaction of these rhodium complexes with various alkynes yielded aromatic scaffolds through a [2 + 2 + 2] cycloaddition process [3].

Besides their role as synthetic key intermediates in [2 + 2 + 2] cycloaddition reactions, rhodacyclopentadienes show valuable luminescent properties. In 2001, Marder and coworkers developed a regiospecific, high‐yield, one‐pot synthesis of 2,5‐bis(arylethynyl) rhodacyclopentadienes by [2 + 2] reductive coupling of 1,4‐diarylbuta‐1,3‐diynes and rhodium complexes. Despite the presence of a heavy metal atom, the complexes revealed intense fluorescence [20]. During ongoing research, modification of the metal‐bound alkynyl group and at the para‐substituted 1,4‐diphenylbuta‐1,3‐diynes, has enabled fine‐tuning of their electronic and photophysical properties [21]. The further employment of α,ω‐tetraynes (i.e., linked bis‐butadiynes) leads to an additional rigidification of the rhodacycle structure, leading to improved fluorescence emission [22, 23, 24, 25, 26].

Quinones are molecules well‐known for their biological activity, playing essential roles in biochemical processes and showing pivotal pharmacological activities [27, 28, 29, 30, 31, 32, 33]. Exploiting the full potential of these compounds requires synthetic methods to access their structural frameworks in a simple and straightforward manner. However, functionalization at the benzenoid ring of these moieties is quite challenging due to its naturally low reactivity that makes it typically less prone to undergo conventional electrophilic arene functionalization [34]. In this sense, synthetic methods that can provide highly functionalized quinones are highly desirable. Tanaka and coworkers reported a rhodium(I)‐catalyzed [2 + 2 + 2] cycloaddition of biphenyl‐linked 1, 7‐diynes with 1, 4‐naphthoquinones and anthracene‐1, 4‐diones [35]. This protocol successfully led to dibenzotetracenedione derivatives. However, the use of benzoquinones as the third component in this protocol was not accomplished, probably due to problems related to (i) competing oligomerization of the diyne, (ii) secondary cycloaddition via the quinone unit of the product, and (iii) the reliability of the oxidation step, which are the most common issues [36, 37].

In 2020, the Bower and da Silva Júnior groups described the first examples of Rh‐catalyzed [2 + 2 + 2] cycloadditions between linked bis‐alkynes and benzoquinones for the construction of substituted naphthoquinones [38]. This strategy enables rapid access to the relevant polyketide framework present in complex substances such as γ‐rubromycin (inhibitor of human telomerase) and fredericamycin A (compound with antitumor activity). For example, using this methodology, the pentasubstituted benzenoid core of justicidone could be accessed in just three steps (vs. eight steps previously) (Scheme 1A).

(A) [2 + 2 + 2] cycloaddition with benzoquinones to build complex systems including molecules with biological activity; (B) Construction of π‐extended systems using quinones for fluorescent applications; (C) Rh‐catalyzed [2 + 2 + 2] cycloadditions with α,ω‐tetraynes and quinones and application as building blocks for fluorescent compounds.

In addition to quinones being used for the construction of molecules with biological activity, this class of compounds can also be used as π‐extended building blocks that have fluorescent properties. Kakiuchi and collaborators developed a regioselective C–H alkylation of the corresponding acenequinones for the synthesis of tetraalkylanthracenes and ‐pentacenes (Scheme 1B). It was suggested that a tetraalkylpentacene is stable under air in the dark and possesses an appropriate HOMO level as an active material for p‐type organic field‐effect transistors (OFETs) [39, 40].

In 2017, Kakiuchi and colleagues developed ruthenium‐catalyzed C─O arylation of 1,4‐ and 1,5‐dimethoxy‐anthraquinones for the syntheses of dibenzo‐[h,rst]pentaphenes and dibenzo[fg,qr]pentacenes [41]. The cross‐coupling reaction proceeded selectively at the C─O bonds in the anthraquinones to give the corresponding 1,4‐ and 1,5‐diarylated anthraquinones. Further transformations using the Corey‐Chaykovsky reaction in the carbonyl groups, followed by a Lewis acid‐catalyzed dehydrative aromatization, afforded derivatives of dibenzo[h,rst]pentaphenes and dibenzo‐[fg,qr]pentacenes. Some compounds, when incorporated into OFET devices with a bottom‐contact configuration, exhibited hole‐transporting characteristics (Scheme 1B).

Based on the extensive expertise of the Marder group in the synthesis of rhodacyclopentadienes and longstanding endeavors to access new substituted quinones in the da Silva Júnior group, we developed an efficient and reliable methodology based on the Rh‐catalyzed [2 + 2 + 2] cycloadditions between α,ω‐tetraynes and benzoquinones for the rapid assembly of highly conjugated structures containing naphthoquinones and anthraquinones. Based on our previous results with linked bis‐alkynes, the reaction involving the α,ω‐tetraynes is expected to proceed through a cycloaddition step, followed by an oxidation step mediated by an additional equivalent of the quinone, affording the fully conjugated system (Scheme 1C) [38]. It was hoped that the extended conjugation in the products would lead to distinct photophysical properties.

Furthermore, structural modification of a selected example generated a highly fluorescent naphthalene derivative, highlighting this work as providing access to new quinoidal compounds and for the preparation of new fluorescent bis(arylethynyl)‐naphthalene compounds (Scheme 1C).

Results and Discussion

2

Synthesis

2.1

Our initial attempts to perform the [2 + 2 + 2] cycloaddition reaction between the α,ω‐tetraynes (1a) and benzoquinone (2a) were carried in the absence of ligand using the following reaction conditions: 1a (0.2 mmol), 2a (2.0 equiv.), precatalyst–[Rh(cod)Cl)]2 (5.0 mol%), DCE 2 mL (0.1 M), 70°C, 18 h. Unfortunately, using these conditions, no reaction was observed, and the starting materials were fully recovered (Table 1, entry 1). Based on the review regarding the rhodium catalyst in [2 + 2 + 2] cycloaddition reactions by Pla‐Quintana and Roglans [42], several rhodium complexes were evaluated (Table 1, entries 2–6). However, in all cases the desired product 3aa, was not obtained. Next, we investigated the presence of phosphorus ligands. Tanaka and colleagues reported that cationic rhodium(I)/BINAP‐type bisphosphine complexes are versatile catalysts for highly chemo‐, regio‐, and enantioselective [2 + 2 + 2] cycloadditions [43]. In addition, it is well known that cationic rhodium(I) complexes containing phosphine ligands can be prepared using [Rh(diene)Cl]2, where diene = norbornadiene (NBD), 1,5‐cyclooctadiene (cod), or 2 coordinated cyclooctenes (coe) [44, 45]. Using [Rh(coe)2_Cl]2 as the catalyst precursor and phosphines containing electron‐withdrawing groups such as (4‐CF_3_C_6_H_4)3_P and (C_6_F_5)_3_P, no reaction was observed (Table 1, entries 7 and 8). Using Xantphos as ligand, the desired product was not obtained either (Table 1, entry 9). Interestingly, using [Rh(cod)Cl]2 as a catalyst precursor and (rac)‐BINAP as the ligand the desired product 3aa was obtained in 51% isolated yield (Table 1, entry 10). Further improvement was observed when (rac)‐BINAP was replaced by (R)‐tol‐BINAP, reaching a yield of 68% (Table 1, entry 11). In terms of solvent, using THF instead of DCE slightly reduced the yield to 47% yield (Table 1, entry 12).

TABLE 1: Optimization for the cycloadditions with bis‐diynes and quinones. a

To improve the catalytic performance, we investigated the presence of additives. Using AgBF_4_ as well as other additives, no significant improvements were observed (Table 1, entries 13–16). A significant improvement was observed when the conventional oil bath was replaced with microwave radiation. In only 2 h, the desired product 3aa was obtained in 94% yield (Table 1, entry 17). Notably, under reaction conditions with the use of microwave radiation, the presence of additives is not required.

Having established the optimized conditions, we first examined the scope of benzoquinones derivatives. As shown in Scheme 2, the Rh‐catalyzed [2 + 2 + 2] cycloadditions proceeded smoothly with different benzoquinone derivatives (2b‐2f). The product containing a methyl group (3ba) was obtained in 80%, while the product containing a methoxy group (3ca) gave a yield of 84%. The presence of a phenyl group gave a yield of 65% of the desired product 3da, while the fused benzene ring gave a yield of 96% of the isolated product 3ea. The product 3fa containing the acetate group in the fused benzene ring was formed in 78% yield.

Substrate scope for the Rh‐catalyzed [2 + 2 + 2] cycloadditions with α,ω‐tetraynes and quinones. [a] Reaction conditions: 1a‐1 (0.2 mmol), 2a (2.0 equiv.), [Rh(cod)Cl)]2 (5 mol%), (R)‐tol‐BINAP (10 mol%), DCE 2 mL (0.1 M), 100°C (MW), 2 h. Isolated yields are shown.

Encouraged by the success of the reaction with benzoquinone derivatives, we turned our attention to evaluating the scope of bis‐diynes (1b‐1g). Even in the presence of several functional groups, the methodology presented good to excellent yields (37%–80%). For example, the presence of electron‐donating groups such as those present in the products 3ab, 3ac, and 3ad was well tolerated and gave yields in a range of 37–62%. Interestingly, the product containing the thioether group (3ae) gave 80%. In our previous work, very electron‐poor benzoquinones were not suitable substrates [38]. Interestingly, this methodology showed good results even in the presence of electron‐withdrawing groups such as an ester (3af) or trifluoromethyl group (3ag), giving the products in 40 and 42% yield, respectively.

All products were characterized by ^1^H and ^13^C NMR spectroscopy as well as HRMS, FTIR, and melting point (see Supporting Information for details). Compound 3aa, which was obtained as single crystals, was subjected to X‐ray crystallographic analysis, with details provided in the Supporting Information (Figure 1).

Molecular structure of 3aa determined by single‐crystal X‐ray diffraction with displacement ellipsoids shown at the 50% probability level. CCDC number: 2338853.

In general, the fluorescence emissions of aromatic carbonyl compounds such as aldehydes and ketones are weak due to rapid intersystem crossing, rationalized by the El‐Sayed rule [46, 47, 48, 49, 50, 51, 52]. To enhance the fluorescence, we modified the carbonyl group to improve emission efficiency with strategies involving: i) the expansion of π‐conjugation system; ii) the introduction of σ–π conjugation system, and iii) the use of charge transfer emission induced by highly polarized groups [53, 54, 55, 56]. First, using the compound 3aa as a model substrate, we decided to transform it into a naphthalene derivative. Based on the works of Linker [57] and Barbasiewicz [58], the carbonyl groups were reduced using a SnCl_2_/MeOH/HCl mixture followed by alkylation with Me_2_SO_4_ under basic conditions to afford compound 4aa as yellow crystals with 40% yield over two steps (Scheme 3).

Modification of the building block 3aa based on the reduction/alkylation sequence to afford the compound 4aa.

Photophysical Properties

2.2

The photophysical properties of compound 4aa and its precursor 3aa were examined to elucidate their fluorescence behavior. As expected, compound 3aa exhibited negligible fluorescence emission. However, upon transformation into 4aa, a remarkable enhancement was observed, leading to a fluorescence quantum yield of 53% in dichloromethane. The emission spectrum of 4aa extends into the visible region, peaking at λ em = 498 nm, with a Stokes shift of 4000 cm^−1^ (Table 2). Additionally, the fluorescence lifetime of 4aa was determined to be 5.92 ns with k r = 9.0 ×10^−7^ s^−1^ and k nr = 7.9 ×10^−7^ s^−1^, highlighting its efficient emissive nature compared to its precursor 3aa (Figure 2).

(a) UV‐vis absorption spectra of the compounds. (b) Emission spectra of the compounds, inset: 4aa in DCM under UV irradiation (365 nm). (c) Singlet oxygen luminescence of the compounds and the standard perinaphthenone (PeriNa).

Moreover, only the quinone compound 3ad exhibited emission at 520 nm, with a low PL quantum yield of 0.05, whereas the other quinone compounds 3aa, 3ae, 3ag, and 3ea displayed negligible fluorescence. To investigate further their photophysical properties, singlet oxygen luminescence measurements were conducted. As shown in Figure 2c, none of the compounds exhibited detectable singlet oxygen phosphorescence in the NIR. As noted below (vide infra), we were also unable to detect the triplet states by ns‐transient absorption (TA) spectroscopy, so they are likely to have very short lifetimes.

Theoretical Studies

2.3

Density functional theory (DFT) and time‐dependent density functional theory (TD‐DFT) calculations were carried out at the CAM‐B3LYP/6‐31+G** level of theory (see the computational details in the Supporting Information) on a few representative naphthoquinones and derivatives to rationalize the observed spectroscopic properties of these molecules. First, compound 3aa was geometrically optimized and compared to the experimental structure determined by single‐crystal X‐ray diffraction (Table S1, Supporting Information). The two structures match well, giving confidence in the computed structural parameters in the other molecules for which no crystallographic data are available.

The electronic structures of these compounds (optimized with C s symmetry) were then analyzed. For quinones 3aa, 3ea, and 3ag, the highest occupied molecular orbital (HOMO) of a″ symmetry (π‐type) is heavily localized on the central benzene ring and the adjacent ethynyl groups, whereas the lowest unoccupied molecular orbital (LUMO), also of a″ symmetry (π‐type), is distributed over the quinone group (Figure 3). Similarly, the LUMO of 3ad is localized on the quinone group. On the other hand, its HOMO strongly differs from that in 3aa, 3ea, and 3ag, being heavily localized on the nitrogen atoms of the strong donor bis(methoxyphenyl)amino groups (Figure 3). In the case of 4aa, both HOMO and LUMO differ from those of the quinones, with the former distributed over the dimethoxybenzene group and to a lesser extent on the central benzene ring and the adjacent ethynyl groups, while the latter extends mostly on the central benzene ring and the adjacent ethynyl groups (Figure 3). A large HOMO‐LUMO energy gap (more than 4 eV at the CAM‐B3LYP level of theory) is computed for all compounds, indicating high kinetic and thermodynamic stability.

Plots of the HOMOs and LUMOs of 3aa, 3ea, 3ag, 3ad, and 4aa. Contour values are ±0.03 (e/bohr3)1/2.

The difference in energy and nodal properties of the HOMO and LUMO in 3ad and 4aa with respect to those of the quinones 3aa, 3ea, and 3ag should have some effect on the optical properties of these compounds, as observed experimentally (see above). Compared with experimental data, TD‐DFT computations were carried out using the optimized ground state geometries to simulate the optical absorption spectra of these molecules (Figure 4). Both the energy trend (absorption wavelength) and relative intensities are indeed well reproduced in the simulations (compare Figures 2b and 4). For example, compound 3ad shows the highest simulated absorption wavelength, with a maximum intensity around 424 nm and a shoulder at 495 nm. The experimental absorption bands are observed at 431 nm and approximately 510 nm, respectively. The second compound with the highest absorption wavelength, 4aa, shows a simulated maximum (both in intensity and energy) at 404 nm, which compares well with the experimental value of 410 nm. The absorption spectra of the other compounds, 3aa, 3ea, and 3ag, are also well simulated, with a good match in absorption wavelengths, which increase progressively from 3ag to 3aa to 3ea, as observed experimentally. As expected, the lowest energy absorption bands of 3ad and 4aa are somewhat red‐shifted with respect to those of 3ag, 3aa, and 3ea, due to the presence of the amino groups in the former and the methoxy groups in the latter.

CAM‐B3LYP‐simulated UV‐visible absorption spectra of selected quinones and derivatives.

It is possible to assign the bands of the absorption spectra, particularly the lowest energy ones, to specific electronic transitions. Interestingly, for all compounds except 4aa, the shoulder beyond the maximum absorption wavelength, located between 400 and 500 nm, is associated to a weakly allowed a′ transition from the HOMO to the LUMO. In contrast, for compound 4aa, the transition is still from the HOMO to the LUMO but is of a″ symmetry and has a significant oscillator strength, resulting in a strong absorption band at 404 nm. Indeed, these different HOMO‐LUMO electronic transitions are dominated by different charge transfers (see the density‐difference plot in Figure 5).

Density‐difference plot associated with the HOMO‐LUMO transition in 3aa and 4aa. The cyan and pink colors indicate an increase and decrease of density upon excitation, respectively [isocontour value = 0.002 e bohr−3].

As mentioned earlier, compound 4aa revealed a remarkably high fluorescence quantum yield (Table 2 and Figure 2c), whereas the related quinone compounds displayed negligible fluorescence emission. The first excited states of 3aa, 3ea, 3ag, 3ad, and 4aa were computationally investigated. As expected, compound 4aa, being a bis(methoxy)naphthalene rather than a naphthoquinone, behaves differently from the others. Indeed, while the compound 3 series exhibits small oscillator strengths (f < 0.25) for the de‐excitation of the singlet excited state S_1_ (fluorescence), compound 4aa shows an oscillator strength that is an order of magnitude larger (f = 1.45). The computed fluorescence emission of 4aa occurs at λ em = 499 nm with a Stokes shift of ca. 4700 cm^−1^, which compares well with the experimental values of 492 nm and 4065 cm^−1^, respectively (vide supra). Additionally, the computed fluorescence lifetime is 2.57 ns, which closely matches the experimental value of 5.92 ns. Finally, the strong asymmetric fluorescence band of compound 4aa is well reproduced in the simulation using the Adiabatic Hessian (AH) approach. As observed in Figure 6, the maximum occurs around 495 nm, with a long tail extending to 700 nm, which is accurately modelled. This behavior is clearly related to the strong vibronic coupling between the electronic and vibrational transitions.

Experimental (dashed line) and CAM‐B3LYP‐simulated (solid line) emission spectra of 4aa.

Stability Investigation

2.4

We attempted to gain further information about the excited states of representative compounds naphthoquinone 3aa, anthraquinone 3ae, and dimethoxynaphthalene 4aa by ns‐TA spectroscopy in deaerated THF and/or toluene solutions using ca. 5 ns pulses (see Supporting Information for details). However, only weak and short‐lived ground state bleaching (GSB) was observed, which was too fast to resolve with the laser setup employed. However, while 3ae was relatively stable, both 3aa and 4aa showed rapid photo‐induced decomposition, as observed by UV‐Vis spectroscopy (Figure S5). Whereas 3aa proved to be relatively stable in C_6_D_6_ solution under ambient conditions (light and air) by ^1^H NMR spectroscopy, with some decomposition being observed after ca. 1 week, 4aa proved to be much more prone to photooxidation.

Thus, during our investigation, we observed that the color of compound 4aa in a C_6_D_6_ solution in air changed upon exposure to sunlight, indicating that compound 4aa is not photostable, which motivated us to investigate this further. As shown in Figure 7, compound 4aa was dissolved in C_6_D_6_ and exposed to air under sunlight. The disappearance of the proton signals at 7.65, 6.98, and 6.61 ppm suggests the photodegradation of compound 4aa, while new proton signals at 9.90, 7.45, 6.85, and 5.83 ppm simultaneously increased as compound 4aa degraded. The photodegradation product 4aa‐ox was further confirmed by ^1^H‐^1^H COSY and ^13^C NMR analysis (Figure S6). Furthermore, single‐crystal analysis confirmed the structure of 4aa‐ox, proving that compound 4aa underwent a photooxidation process, the product having the formula of 4aa + O_2_. The transformation of compound 4aa to 4aa‐ox involves cleavage of the electron‐rich dimethoxyphenyl ring of the central naphthalene unit. This structural modification is characterized by the conversion of the arene ring into a carbomethoxy group and a methoxyvinyl aldehyde group, both being attached to the ortho positions of the remaining phenyl ring of the core.

1H‐NMR NMR spectra (300 MHz) of fresh 4aa in C6D6 during photoaging under ambient conditions. The molecular structure of the photodegradation product 4aa‐ox determined by single‐crystal X‐ray diffraction is shown (top right), and more details are provided in the Supporting Information. CCDC number: 2484133.

Such an intriguing transformation of a bis(methoxy)naphthalene was reported using photoexcited m‐chloronitrobenzene as the oxidant [59, 60]. A related reaction of ozone with naphthalene under various conditions resulted in ring‐opening and the formation of ortho‐aldehyde and vinylaldehyde groups [61, 62, 63].

In our case, the dimethoxynaphthalene derivative likely undergoes either [4 + 2] 1,4‐cycloaddition of ^1^O_2_ to the electron‐rich para‐carbons of the substrate, forming an endoperoxide [64] prior to rearrangement, or direct 1,2‐addition, either of which subsequently leads to ring opening. Although we were unable to observe emission from ^1^O_2_ directly, vide supra, this may well be because compound 4aa clearly sensitizes its formation but then reacts with it quite efficiently.

Conclusion

3

In conclusion, our investigation into the [2 + 2 + 2] cycloaddition reaction between linked bis‐diynes and benzoquinones leads to a new class of highly functionalized compounds with promising applications. Through systematic exploration of various reaction conditions, we achieved success with the combination of [Rh(cod)Cl]2 and R‐(tol)‐BINAP as the ligand, yielding the desired product in good yield. Further optimization using microwave radiation significantly enhanced the efficiency of the reaction. Subsequent exploration of the scope of benzoquinone derivatives and bis‐diynes demonstrated the versatility of the methodology, yielding a range of products with moderate to excellent yields. The comprehensive characterization of the products confirmed their structures, and the X‐ray crystallographic analysis provided additional insights. Inspired by the potential applications of the products, conversion of the carbonyl moiety of compound 3aa furnished the highly fluorescent bis(methoxy)‐bis(arylethynyl)naphthalene compound 4aa, which underwent facile and clean photooxidation in air under sunlight in a process involving ring opening of the bis(methoxy)arene moiety. Overall, our study not only establishes a robust methodology for [2 + 2 + 2] cycloaddition reactions but also opens avenues for further exploration and application of the synthesized compounds as potentially biological active compounds and also as building blocks for fluorescent compounds.

Experimental Section

4

Chemicals

4.1

Unless otherwise noted, all reactions were performed using standard Schlenk or glovebox (Innovative Technology Inc.) techniques under argon. HPLC‐grade solvents were argon saturated, dried using an Innovative Technology Inc. Pure‐Solv Solvent Purification System, and further deoxygenated by using the freeze‐pump‐thaw method.

The compounds 1,4‐benzoquinone and 1,4‐naphthoquinone were purchased from Sigma Aldrich and purified via reduced pressure sublimation using a cold finger sublimation apparatus (50°C, 0.9 mbar) and stored in a glovebox to prevent contact with moisture. All commercially available naphthoquinones and further commercial chemicals were purchased from Sigma Aldrich and TCI chemicals. The bis‐diynes 1a [65], 1b [66], 1c [21], 1d [67], 1e [24], 1f [24], and 1g [21] were obtained following a procedure previously described in the literature. [Rh(cod)Cl]2 was synthesized according to the literature procedure [45].

Materials and Instrumentation

4.2

Microwave heating was performed in a Biotage Initiator+ reactor. Automated flash chromatography was performed using a Biotage Isolera Four System on silica gel (Biotage SNAP cartridge KP‐Sil 10 g, KP‐Sil 25 g, and HP‐Sil 50 g). Commercially available, precoated TLC plates (Polygram Sil G/UV254 and Polygram Alox N/UV254) were purchased from Machery‐Nagel. The removal of the solvent was performed on a rotary evaporator in vacuo at a maximum temperature of 45°C.

High‐resolution mass spectra were obtained using a Thermo Fisher Scientific Exactive Plus Orbitrap MS System in APCI mode. All solution NMR spectra were acquired on a Bruker Avance 300 (^1^H: 300 MHz) or Bruker Avance 400 NMR spectrometer (^1^H: 400 MHz or 300 MHz, ^13^C: 100 MHz). The ^1^H and ^13^C NMR spectra were referenced to the appropriate residual solvent peak or TMS peak. Infrared spectra were recorded on a Perkin Elmer Spectrum One FTIR spectrometer solids compressed on a diamond plate. Melting points were determined using Stuart SMP30 melting point apparatus and are uncorrected.

General Procedure for [2 + 2 + 2] Cycloadditions of α,ω‐Tetraynes 1 With Quinones 2

4.3

An oven‐dried thick‐walled glass reaction tube, fitted with a magnetic stirring bar, was charged with the respective bis‐diynes (0.2 mmol), [Rh(cod)Cl]2 (5 mol %), (R)‐tol‐BINAP (10 mol %), and the appropriate quinone substrate 2 (200 mol%). DCE 2 mL (0.1 M) was added via syringe, and the tube was sealed under an inert atmosphere. The reaction mixture was stirred at 100°C for 2 h, in a microwave reactor, then cooled to room temperature and concentrated in vacuo. The crude reaction mixture was purified by automated flash column chromatography to yield the target quinone 3.

4,9‐bis(p‐Tolylethynyl)‐2,3‐Dihydro‐1H‐Cyclopenta[b]Naphthalene‐5,8‐Dione (3aa)

4.4

The general procedure was followed by using 1a and 2a as starting material. After flash chromatography on silica (n‐hexane/EtOAc, 9:1), the product 3aa was isolated as a red solid (80 mg, 94%). ** ^1^H NMR** (400 MHz, CDCl_3_) δ: 7.57 (d, J = 7.8 Hz, 4H), 7.20 (d, J = 7.9 Hz, 4H), 6.89 (s, 2H), 3.28 (t, J = 7.7 Hz, 4H), 2.39 (s, 6H), 2.21 (quint, J = 7.7 Hz, 2H). ** ^13^C NMR** (100 MHz, CDCl_3_) δ: 184.4, 153.5, 139.5, 138.2, 132.2, 131.4, 129.4, 120.4, 119.0, 100.8, 87.4, 34.5, 23.1, 21.8. HRMS (APCI^+^): Calcd for [C_31_H_23_O_2_]^+^ 427.1698 [M+H]^+^, found 427.1687. m.p. (°C) = 227–231. FTIR: 2917, 2189, 1732, 1653, 1612, 1434, 1391, 1348, 1322, 1290, 1262, 1174, 1154, 1090, 1031, 986, 943, 855, 707, 645, 571, 526, 465, 442, 424 cm^−1^. The structure of the product was also confirmed by X‐ray diffraction (CCDC number = 2338853).

Synthesis of Compound 4aa

4.5

This procedure was carried out through modifications of procedures present in the literature [57, 58]. To a solution of compound 3aa (42.6 mg; 0.1 mmol) in MeOH (400 mL) was added a solution of SnCl_2_·2H_2_O (78.9 mg, 0.35 mmol) in conc. HCl (1.2 mL). The reaction mixture was stirred at 65°C for 1 h, then poured into cold H_2_O (10 mL) and left at 0°C for 30 min while protected from light. The product was extracted with dichloromethane (10 mL), washed with H_2_O (2 × 10 mL), and left under suction to dry for 1 h while protected from light, affording the corresponding 1,4‐dihydroxynaphthalene (approx. 30 mg) as a brown solid. The 1,4‐dihydroxynaphthalene (30 mg, 0.07 mmol) obtained was dissolved in acetone (4 mL). To this solution, K_2_CO_3_ (29.0 mg, 0.21 mol), was added and the resulting suspension was stirred at room temperature for 15 min before being treated with dimethyl sulfate (17.0 mL, 0.175 mmol). The reaction mixture was heated under reflux for 2 h, then cooled to room temperature and concentrated in vacuo. The crude solid obtained was dissolved in CH_2_Cl_2_ (20 mL) and filtered to remove K_2_CO_3_. The filtrate was washed with H_2_O (2 × 10 mL) and brine (10 mL), then dried over Na_2_SO_4_ and concentrated in vacuo. Purification by flash column chromatography (20% EtOAc/Hex) on silica gel afforded compound 4aa (18.4 mg, 40% over two steps) as yellow crystals.

Crystallographic Data

4.6

X‐ray diffraction data for 3aa were collected on a Rigaku XtaLAB Synergy‐R diffractometer with an HPA area detector and multi‐layer mirror monochromated Cu‐K_a_ radiation. The structure was solved using the intrinsic phasing method [68], refined with the ShelXL program [69], and expanded using Fourier techniques. All nonhydrogen atoms were refined anisotropically. Deposition Numbers 2338853 (for 3aa), 2484133 (for 4aa‐ox) contain the supplementary crystallographic data for this paper. These data are provided free of charge by the joint Cambridge Crystallographic Data Centre and Fachinformationszentrum Karlsruhe Access Structures service.

Photophysical Properties of the Compounds

4.7

All measurements were performed in standard quartz cuvettes (1 cm x 1 cm cross‐section). For compound 4aa, UV‐visible absorption spectra were recorded using an Agilent 8453 diode array UV‐visible spectrophotometer. Emission spectra were recorded using an Edinburgh Instruments FLSP920 spectrometer equipped with a double monochromator for both excitation and emission, operating in right‐angle geometry mode, and all spectra were fully corrected for the spectral response of the instrument. The fluorescence quantum yields (Φ_F_) were measured using a calibrated integrating sphere (inner diameter: 150 mm) from Edinburgh Instruments combined with the FLSP920 spectrometer described above. For solution‐state measurements, the longest‐wavelength absorption maximum of the compound in the respective solvent was chosen as the excitation wavelength, unless stated otherwise. Fluorescence lifetimes were recorded using the time‐correlated single‐photon counting (TCSPC) method using an Edinburgh Instruments FLS980 spectrometer equipped with a high‐speed photomultiplier tube positioned after a single emission monochromator.

The luminescence of singlet oxygen was investigated using the Edinburgh Instruments FLSP920 spectrometer, which was equipped with a 450 W Xenon arc lamp, a double monochromator, a red‐sensitive photomultiplier (PMT‐R928P), and a near‐infrared (NIR) PMT. The measurements were conducted in dichloromethane at a concentration lower than 5 × 10^−6^ M. To determine the quantum yields for singlet oxygen formation, the emission intensity was measured in the wavelength range of 1230 to 1330 nm and then compared to the standard compound, perinaphthenone, which has a quantum yield of singlet oxygen formation close to unity (excited at 360 nm) [70, 71, 72].

Computational Details

4.8

Theoretical calculations were performed at the DFT level with the Gaussian 16 package [73] using the CAM‐B3LYP functional [74] with the 6–31+G** basis set. All structures, both in the ground (S_0_) and excited states (S_1_), were optimized and confirmed to be energy minima on the potential energy surface by diagonalizing their Hessian matrices. Electronic absorption spectra and excited‐state relaxations were computed using the time‐dependent (TD)‐DFT method. The luminescence spectra of the compounds studied were simulated via the Adiabatic Hessian (AH) approach within the Franck‐Condon approximation. The AH model incorporated mode mixing to provide an accurate description of the potential energy surfaces (PES) of both ground and excited states [75, 76]. Regarding vibronic coupling, we used the AH ansatz (Adiabatic Hessian), which allows that each electronic state (ground S_0_ and excited S_1_, typically) has its own optimized equilibrium geometry together with its respective vibrational modes. The vibrational structure is defined on the adiabatic potential energy surfaces of each electronic state. Calculations were performed within the Frank‐Condon approximation. The AH approach includes Duschinsky matrix (mode‐mixing), etc.

Concerning details of the calculations, the class‐based pre‐screening criteria were set to:

maxC1 = 70 (default = 20)

maxC2 = 70 (default = 13)

maxINT = 100 (default) that corresponds of maximum number of integrals to compute each class above C2

where the unit is million of integrals.

The lifetime was computed based on the methodology of Rega and coworkers [77].

Nanosecond Transient Absorption Spectroscopy (4aa)

4.9

Time resolved ns‐spectra were recorded using an Edinburgh LP 920 laser‐flash TA spectrometer with a 450 W Xe Arc flash lamp. A Q‐switched Ekspla NT340 Nd:YAG laser with integrated OPO and SFG was used as an excitation source. White light and the laser beam were oriented perpendicular to each other. The sample was prepared in a nitrogen‐filled glove box using quartz cuvettes (10×10 mm, Starna) equipped with a Young's tap. Dry solvent of spectroscopic quality was used, which was degassed by at least 5 freeze‐pump‐thaw cycles. All samples were excited with laser pulses (length ca. 5 ns, 10 Hz repetition rate) in the absorption maxima of the ground state or in the shoulder at 22200 cm^−1^ (450 nm). The sample stability was verified by recording UV‐Vis spectra before and after the measurements.

Conflicts of Interest

The authors declare no conflict of interest.

Supporting information

Experimental procedures and NMR spectra for the compounds, X‐ray crystallography data, photophysical and computational data are available in Supporting Information. Supporting File: chem70597‐sup‐0001‐SuppMat.pdf.