Electron Donor–Acceptor (EDA) Complex between a Triarylamine and B(C₆F₅)₃ for the Photocatalytic Dehydrogenative Cross-Coupling of Phenols

Abstract

In this article, an electron donor–acceptor (EDA) complex between a triarylamine and B(C₆F₅)₃ that catalyzes the dehydrogenative cross-coupling of phenols is described. We demonstrate, for the first time, that the use of both components of the radical ion pairs generated by the photoexcitation of the EDA complex as co-catalysts, and the triarylaminium radical cation (^+·NAr₃) successfully promotes dehydrogenative cross-coupling between electron-rich phenols and 2-naphthols to provide electron-rich biphenol motifs using molecular oxygen as a terminal oxidant.

Introduction

Triarylaminium radical cations (^+·NAr₃) are exceptionally stable persistent radicals that are used in organic synthesis, materials science, and organometallic chemistry as mild yet relatively strong single electron oxidants “magic blue”.¹⁾ They are generally prepared from the corresponding neutral triarylamines (NAr₃) with the assistance of a silver-based oxidant²⁾; however, from an economic and environmental perspective, it would be desirable to avoid the use of a stoichiometric amount of such an expensive metal reagent, and thus, approaches to generate the radical cation more efficiently have recently been explored^3–9) (Chart 1a). Liu and Klussmann have reported that triarylamines can be oxidized in the presence of a stoichiometric amount of benzoyl peroxide and that the resulting triarylaminium radical cations successfully promote the difunctionalization of styrenes.³⁾ Mejía and colleagues have demonstrated a phenazine-catalyzed dehydrogenative aza-Henry reaction in which the phenazine radical cation is likely to be generated as an active species under an O₂ atmosphere⁴⁾; however, these methods require heating, which might promote the oxidation of the triarylamines. Although an alternative approach employing a laser flash photolysis⁵⁾ or an electrochemical method⁶⁾ have recently been reported, the development of conceptually new methods to provide synthetically useful triarylaminium species under metal-free and mild ambient conditions is still required.

Chart 1. Generation of Triarylaminium Radical Cations

Against this background, we have investigated the catalytic utility of electron donor–acceptor (EDA) complexes between triarylamines and tris(pentafluorophenyl)borane (B(C₆F₅)₃). EDA complexes are supramolecular assemblies that consist of electron-donating and -accepting molecules, and are known to exhibit charge-transfer (CT) bands in their UV-vis spectra that correspond to the energy of their highest occupied molecular orbital (HOMO)–lowest unoccupied molecular orbital (LUMO) gap.¹⁰⁾ The photoexcitation of such a complex can provide a radical ion pair that can be used in an extended range of radical transformations.^10–12) In such EDA complexes, triarylamines have been reported to serve as potential electron donors. Indeed, Wang and colleagues and Slootweg and colleagues have reported the formation of triarylaminium species via EDA complexation and subsequent photoexcitation using B(C₆F₅)₃, even though there are no reported synthetic applications of these complexes^13,14) (Chart 1b). We envisioned that NAr₃/B(C₆F₅)₃ systems could enable the efficient in-situ formation of triarylaminium radical cations from the corresponding triarylamines in the oxidative transformation by simple irradiation with visible light. Moreover, both components of the complex, i.e., NAr₃ and B(C₆F₅)₃, could potentially be used as catalysts, considering recent reports on the catalytic use of B(C₆F₅)₃, in which B(C₆F₅)₃ is proposed to be regenerated from the boron radical anion (^−·B(C₆F₅)₃).^15,16) Based on this concept, we attempted the catalytic use of NAr₃/B(C₆F₅)₃ in the metal-free dehydrogenative cross-coupling of phenols to provide useful electron-rich biphenol derivatives, which have typically been synthesized using transition-metal catalysis^17–21) or strong oxidant-mediated reactions,^22,23) as a model transformation (Chart 1c).

Results and Discussion

First, we examined the cross-coupling of phenol (1) and 2-naphthol (2) to evaluate the catalytic activity of NAr₃/B(C₆F₅)₃ (Table 1). Via reaction optimization, we found that 1 and 2 are coupled in the presence of a catalytic amount of tris(p-bromophenyl)amine (N(p-BrC₆H₄)₃) and B(C₆F₅)₃ in 1,2-dichloroethane (1,2-DCE) under aerobic conditions under irradiation from blue light-emitting diodes (LEDs) to provide the desired biphenol (3) in 74% yield (entry 1). We then conducted several control experiments to demonstrate the importance of the NAr₃/B(C₆F₅)₃ system in this reaction, which revealed that both components of N(p-BrC₆H₄)₃ and B(C₆F₅)₃ are critical to obtain good yields, i.e., only much lower yields of 3 were observed without the catalysts, which is probably due to a background reaction mediated by molecular oxygen (entries 2–4).²⁴⁾ Importantly, the reaction did not proceed in the dark, which suggests that the photoexcitation of the EDA complex is of critical importance (entry 5). The reaction under an N₂ atmosphere resulted in a significantly decreased yield of 3, while conducting the reaction under an O₂ atmosphere provided 3 in 74% yield with more consumption of 1. Here, molecular oxygen is likely to act as a terminal oxidant by quenching ^−·B(C₆F₅)₃¹⁵⁾ (entries 6, 7). In contrast to the good results obtained using N(p-BrC₆H₄)₃, tris(p-tolyl)amine (N(p-tol)₃) did not promote the cross-coupling, despite a previous report on the formation of an EDA complex with B(C₆F₅)₃ (entry 8).¹⁴⁾ This is probably due to the insufficient oxidizing power of the corresponding radical cation ^+·N(p-tol)₃ (E(S/S⁺) = +0.47 V vs. Ag/Ag⁺),²⁵⁾ which supports the idea that ^+·NAr₃ is largely responsible for the progress of the reaction. Finally, we conducted the cross-coupling using the common iridium photocatalyst [Ir[dF(Me)ppy]₂(dtbbpy)][PF₆] instead of N(p-BrC₆H₄)₃/B(C₆F₅)₃, albeit that the yield of biphenol 3 was low (entry 9); it is well known that the efficiency of the desired reaction decreases in the presence of molecular oxygen because the excited iridium species are immediately quenched by molecular oxygen.^26,27) One potential advantage of the N(p-BrC₆H₄)₃/B(C₆F₅)₃ system might be that the reaction proceeds well in the presence of O₂.

Table 1. Investigation of the Reaction Conditions

Entry	Deviation from the above conditions	Yield [%]a)	Recovered 1 [%]a)
1	None	74	23
2	No N(p-BrC6H4)3	35	57
3	No B(C6F5)3	10	42
4	No N(p-BrC6H4)3 and B(C6F5)3	10	60
5	No light	0	100
6	Under N2 atmosphere	6	92
7	Under O2 atmosphere	74	12
8	N(p-tol)3 instead of N(p-BrC6H4)3	14	86
9	[Ir[dF(Me)ppy]2(dtbbpy)]PF6 (5 mol%) instead of N(p-BrC6H4)3 and B(C6F5)3	18	76

a) Yields were determined via ¹H-NMR spectroscopy using dimethylterephthalate as the internal standard.

With the optimal conditions in hand (Table 1, entry 1), we examined the substrate scope of the dehydrogenative cross-coupling reaction (Chart 2). Not only the reaction of highly activated 2,6-dimethoxyphenol provided product 3, but those of 2-methoxy-4-methylphenol and 4-methoxyphenol also gave the corresponding coupling products (4 and 5) in moderate yield. It is noteworthy that in the case of 4 and 5, the naphthyl group was introduced at the ortho position relative to the hydroxy group, and thus the reaction resulted in products with ortho/para orientation. Bromo-substituted naphthol afforded the coupled product (6), and the reaction of a 2,7-bishydroxynaphthalene derivative also proceeded successfully (7). In contrast to the good results obtained for 2-naphthol derivatives, the reaction of 4-t-butylphenol was less effective, i.e., the coupling product (8) was obtained in only 9% yield.

Chart 2. Substrate Scope^a)

a) Isolated yields are shown.

Subsequently, we conducted several experiments to gain insight into the reaction mechanism (Chart 3). We started by measuring the UV-vis spectra of catalysts used in this study to verify the formation of the EDA complex (Chart 3a). First, we found that N(p-BrC₆H₄)₃, and B(C₆F₅)₃ do not absorb light in the visible region by themselves. We then measured spectra for N(p-BrC₆H₄)₃ + B(C₆F₅)₃ mixture solution, and found that they showed a new absorption band (λ_max = 419 nm), which appears to be the CT band.¹⁴⁾ Moreover, for this mixture only, blue LED irradiation turned the solution blue, and an absorption band specific to ^+·N(p-BrC₆H₄)₃ (λ_max = 727 nm) was observed in the UV-vis spectrum.²⁸⁾ These results strongly suggest that N(p-BrC₆H₄)₃ interacts with B(C₆F₅)₃ and that photoexcitation of the complex provides ^+·N(p-BrC₆H₄)₃ in the reaction mixture. We also conducted the cross-coupling using a stoichiometric amount of the isolated aminium salt N(p-BrC₆H₄)₃[SbCl₆] instead of N(p-BrC₆H₄)₃/B(C₆F₅)₃, which furnished coupling product 3 in 50% yield even under an N₂ atmosphere in the dark; this result strongly supports that the aminium salt is the active species for the oxidative transformation of phenols.

Chart 3. Mechanistic Studies

Based on all the experimental results, a plausible reaction mechanism is shown in Chart 3c. First, the N(p-BrC₆H₄)₃/B(C₆F₅)₃ EDA complex is photoexcited under blue LED light to provide the highly reactive ^+·N(p-BrC₆H₄)₃ and ^−·B(C₆F₅)₃ radicals. ^+·N(p-BrC₆H₄)₃ (E(S/S⁺) = +0.87 V vs Ag/Ag⁺)²⁵⁾ oxidizes electron-rich phenol 1 (E(S/S⁺) = +0.98 V vs Ag/Ag⁺)²⁵⁾ while ^−·B(C₆F₅)₃ reduces molecular oxygen to regenerate both N(p-BrC₆H₄)₃ and B(C₆F₅)₃. The resulting phenol radical cation couples with 2-naphthol 2 to afford intermediate A, and the subsequent oxidation of intermediate A by a superoxide anion or ^+·N(p-BrC₆H₄)₃ provides the desired cross-coupling product (3).

Conclusion

In summary, we have discovered that the photoexcitation of electron donor–acceptor (EDA) complexes between triarylamines and B(C₆F₅)₃ represents an alternative approach to the catalytic generation of synthetically useful triarylaminium radical cations. Moreover, we have successfully demonstrated a strategy for the metal-free, visible light-mediated dehydrogenative cross-coupling of phenols using molecular oxygen as a terminal oxidant. This is, to the best of our knowledge, the first example to use both components of a radical ion pair generated by the photoexcitation of an EDA complex as co-catalysts. We are currently exploring further applications of this approach and investigating details of the transformation.

Experimental

In a glove box, a borosilicate glass tube was charged with a stirring bar, N(p-BrC₆H₄)₃ (9.64 mg, 0.02 mmol, 20 mol%), 1,2-DCE (1.0 mL), and B(C₆F₅)₃ (10.2 mg, 0.02 mmol, 20 mol%). The tube was removed from the glove box, and the atmosphere was substituted with air. The mixture was treated with 2,6-dimethoxyphenol (1; 15.4 mg, 0.100 mmol, 1.0 equivalent (equiv.)) and 2-naphthol (2; 14.4 mg, 0.100 mmol, 1.0 equiv.) at the benchtop, and the resulting mixture was stirred under irradiation from blue LEDs. After stirring for 12 h (the temperature of the reaction mixture was typically 35–40 °C), all volatiles were removed under reduced pressure. The solution was purified using column chromatography (SiO₂, 0→30% EtOAc/hexane), providing 3 (22.1 mg, 0.0746 mmol, 75% yield) as a white solid.

Acknowledgments

This work was supported by JSPS KAKENHI Grants JP22H02743 and JP21H04793.

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

This article contains supplementary materials.

References and Notes

• 1) Jia X., Synthesis, 48, 18–30 (2016).
• 2) Walter R. I., J. Am. Chem. Soc., 77, 5999–6002 (1955).
• 3) Liu S., Klussmann M., Org. Chem. Front., 8, 2932–2938 (2021).
• 4) Unglaube F., Hünemörder P., Guo X., Chen Z., Wang D., Mejía E., Helv. Chim. Acta, 103, e2000184 (2020).
• 5) Bonesi S. M., Protti S., Albini A., J. Org. Chem., 83, 8104–8113 (2018).
• 6) Wang Q.-Q., Jiang Y.-Y., Zeng C.-C., Sun B.-G., Chin. J. Chem., 37, 352–358 (2019).
• 7) In photoredox catalysis, triarylamines are sometimes used as reductive quenchers for photoexcited species to generate radical anions with higher reducing power. In such cases, the triarylaminium radical cation is also generated and serves in the subsequent oxidation step as a persistent radical.
• 8) Wang L., Byun J., Li R., Huang W., Zhang K. A. I., Adv. Synth. Catal., 360, 4312–4318 (2018).
• 9) Wang L., Li R., Zhang K. A. I., Macromol. Rapid Commun., 39, 1800466 (2018).
• 10) For a selected review on transformations involving EDA complexes, see: Crisenza G. E. M., Mazzarella D., Melchiorre P., J. Am. Chem. Soc., 142, 5461–5476 (2020).
• 11) Fu M.-C., Shang R., Zhao B., Wang B., Fu Y., Science, 363, 1429–1434 (2019).
• 12) Bosque I., Bach T., ACS Catal., 9, 9103–9109 (2019).
• 13) Zheng X., Wang X., Qiu Y., Li Y., Zhou C., Sui Y., Li Y., Ma J., Wang X., J. Am. Chem. Soc., 135, 14912–14915 (2013).
• 14) Holtrop F., Jupp A. R., van Leest N. P., Paradiz Dominguez M., Williams R. M., Brouwer A. M., de Bruin B., Ehlers A. W., Slootweg J. C., Chem. Eur. J., 26, 9005–9011 (2020).
• 15) Yuan W., Huang J., Xu X., Wang L., Tang X.-Y., Org. Lett., 23, 7139–7143 (2021).
• 16) Aramaki Y., Imaizumi N., Hotta M., Kumagai J., Ooi T., Chem. Sci., 11, 4305–4311 (2020).
• 17) Wallis P. J., Booth K. J., Patti A. F., Scott J. L., Green Chem., 8, 333–337 (2006).
• 18) Lee Y. E., Cao T., Torruellas C., Kozlowski M. C., J. Am. Chem. Soc., 136, 6782–6785 (2014).
• 19) Esguerra K. V. N., Fall Y., Petitjean L., Lumb J.-P., J. Am. Chem. Soc., 136, 7662–7668 (2014).
• 20) Libman A., Shalit H., Vainer Y., Narute S., Kozuch S., Pappo D., J. Am. Chem. Soc., 137, 11453–11460 (2015).
• 21) Shalit H., Libman A., Pappo D., J. Am. Chem. Soc., 139, 13404–13413 (2017).
• 22) Morimoto K., Sakamoto K., Ohshika T., Dohi T., Kita Y., Angew. Chem. Int. Ed., 55, 3652–3656 (2016).
• 23) More N. Y., Jeganmohan M., Org. Lett., 17, 3042–3045 (2015).
• 24) The UV-vis spectra of the mixtures of catalysts and substrates shows that the strong CT band was observed only for the mixture of 2 and B(C₆F₅)₃, while the spectra exhibit no absorption around 450 nm. These observations indicate that 2 was oxidized by the assistance of B(C₆F₅)₃ to promote the desired coupling reaction even in the absence of N(p-BrC₆H₄)₃, despite its lower efficiency compared to that in the presence of N(p-BrC₆H₄)₃, which is probably due to less absorption of light; for further details, see the supplementary materials.
• 25) The redox potential was determined by cyclic voltammetry (CV) using the Ag/Ag⁺ couple as a reference electrode in 1,2-DCE; for details, see the supplementary materials.
• 26) Li M., Gebremedhin K. H., Ma D., Pu Z., Xiong T., Xu Y., Kim J. S., Peng X., J. Am. Chem. Soc., 144, 163–173 (2022).
• 27) Halliwell B., Gutteridge J. M. C., “Free radicals in biology and medicine,” Oxford University Press, New York, 2015.
• 28) Vardhaman A. K., Lee Y.-M., Jung J., Ohkubo K., Nam W., Fukuzumi S., Angew. Chem. Int. Ed., 55, 3709–3713 (2016).