Based on the hypothesis, we first examined the feasibility and properties of this proposed system. Further examination demonstrated the formation of a ground state EDA complex by UV/Vis spectroscopy (Figure 1). Equimolar amounts of pale yellow 1,2-dinitrobenzene and colorless tri-n-butylamine were combined in acetonitrile, resulting in a darker solution (Fig. 1a). Subsequent measurement of the UV-Vis spectrum indicated a significant red shift and increased absorption of visible light (Fig. 1b). Commercially available 1,2-dinitrobenzene might not be pure, leading to a darker solution that affects light absorption. To resolve this, conduct simple column chromatography to get a pure, pale-yellow reagent. The observed absorption of visible light corresponds to the charge-transfer absorption of the EDA complex. This phenomenon is consistent with the expected effects of intermolecular interactions. While ensuring a constant final concentration of the acceptor, as the proportion of the Donor increases, the absorbance of the entire system gradually increases, displaying an initially rapid and then slower rate of increase. Once the Donor and Acceptor reach an equimolar ratio, the absorbance no longer exhibits a significant increase (Fig. 1c).
It can be observed that the maximum increment in absorbance of the EDA complex occurs at 490 nm (Fig. 1d). Using this reference wavelength, a Job's Plot experiment was conducted to investigate the absorbance increment of the EDA complex at different ratios. Surprisingly, the optimal binding ratio between the Acceptor and Donor was found to be 4:1 (Fig. 1e), which deviated from the commonly reported 1:1 binding ratio observed in most Acceptor and Donor pairs.
Furthermore, the addition of water (600 μL) to the EDA complex solution resulted in a significant reduction in system absorption and weakened the previously observed red shift phenomenon (Fig. 1f). When considering the dilution effect caused by the water addition (as demonstrated by a slight red shift in visible absorption upon adding water to the acceptor solution due to increased solvent polarity in the π-π* transition of nitroaromatics) (Fig. 1g), it was speculated that water acted as both a strong hydrogen bond donor and acceptor, thereby disrupting the formation of the EDA complex. This observation also provided an explanation for why the addition of 600 μL of water completely suppressed the subsequent oxygenation reaction involving superoxide radical species.
To demonstrate the presence of superoxide radicals, a highly reactive form of reactive oxygen species (ROS), a validation experiment was conducted. To monitor the fluorescence signal, the ROS fluorescent probe 2′,7′-Dichlorodihydrofluorescein (DCFH)59 was introduced into the system, and an incremental increase in its positive fluorescence signal was observed with prolonged illumination time (Fig. 2a). This observation strongly supports the generation of ROS by the current dioxygen activation system. To further support our hypothesis, we utilized the superoxide-specific fluorescent probe dihydroethidium (DHE) to monitor the superoxide species, as previously reported in the literature60. Upon illumination in the present protocol, we clearly detected a positive fluorescence signal from the DHE probe (Fig. 2b), and the oxygenation product of DHE was successfully identified through mass spectrometry, providing conclusive evidence for the presence of superoxide species generated through dioxygen activation.
Having successfully demonstrated the compatibility and powerful capacity for the activation of oxygen to generate superoxide radical anions through this EDA complex photocatalytic system, we proceeded to employ organic boronic acid as substrates for the capture of the generated superoxide intermediates. This pivotal step provided further compelling evidence of the operability and practical value inherent in the entire oxygen activation process of this strategy. In contrast to the reported stoichiometric peroxide oxidants61-64, the novel EDA complex with 1,2-dinitrobenzene and tertiary amines was found very efficient to catalyzed the oxygenation process with O2 as the oxidant and oxygen-source, yielding corresponding phenol product (Table 1). Acetonitrile was the most effective solvent for promoting the oxidation-oxygenation reaction, while other commonly used solvents gave unfavorable results (entries 2-8). Solvents with strong hydrogen bond donating properties, like methanol, could hinder the formation of the EDA complex and deactivate the desired catalytic process. We further identified 1,2-dinitrobenzene as the most effective acceptor, and tri-n-butylamine as the optimal donor to avoid the undesirable SNAr reaction (entries 12-18)65-66. Control experiments indicated that the presence of light, dioxygen, donor, and acceptor was essential for this dioxygen activation methodology, providing preliminary confirmation of the rationality and feasibility of the designed mechanism (entries 19-23). It is noteworthy that only catalytic amount of electron acceptor (5 mol%) was employed for this highly efficient aerobic oxygenation (Table 1).
After determining the optimized reaction conditions, we carried out experiments to evaluate the process of dioxygen activation using various boronic acid substrates (Table 2). The optimized conditions enabled us to achieve high yields of the oxygenation products. Notably, the reactivity of the dioxygen activation system displayed minimal sensitivity to the positional and electronic properties of the functional groups. Both electron-donating and electron-withdrawing groups were effectively accommodated by the system (Table 2). Additionally, functional groups that are typically easily oxidized by oxidants, such as alkynes (2b), sulfide groups (2c), alkenes (2d), aldehydes (2f), and benzyl position (2y), could be preserved under the optimized reaction conditions. These strong oxidant sensitive functional groups exhibited excellent compatibility within this oxidation-oxygenation system, allowing their preservation without any oxidative damage. This mild oxidation system demonstrated a wide range of substrates, making it promising for selectively oxidizing boryl groups in complex drug molecules or bioactive compounds containing multiple sensitive functional groups. The high tolerance of the reaction system towards functional groups suggested a gradual and suitable activation of oxygen, resulting in the generation of superoxide anion radicals during photocatalysis.
Moreover, besides conventional aromatic boronic compounds, alkyl-boronic acids also underwent efficient oxygenation reactions in this system, yielding the corresponding alkyl alcohol products. It is noteworthy that, apart from conventional phenyl-boronic acid, boronic esters could also serve as substrates, leading to the desired oxygenation products with good efficiency (2p), thereby expanding the range of substrates. However, unprotected amino groups posed a challenge in obtaining the desired products by this protocol. This was due to the electron-rich nature of amino groups, which competed with the donor and hindered the formation of the EDA complex. Nevertheless, this issue could be easily resolved by implementing a simple functional group protection strategy.
Based on the aforementioned experimental results and relevant mechanistic validation experiments, the reaction mechanism is proposed in Scheme 2. Initially, tri-n-butylamine as the electron donor interacts with a catalytic amount of 1,2-dinitrobenzene, which served as the electron acceptor in solution, leading to the formation of the EDA complex. This EDA complex has a visible light absorption charge-transfer band. Under excitation by visible light, a single electron transfer (SET) process takes place, where an electron from the donor is transferred to the acceptor, resulting in the formation of a diradical cage in acetonitrile solution. The electron-accepting intermediate, 1,2-dinitrobenzene radical anion A, is stabilized by the presence of two strongly electron-withdrawing nitro groups, preventing unproductive back electron transfer (BET). Subsequently, the dioxygen molecule captures the electron from the dinitrobenzene radical anion A, generating a superoxide radical anion, which is a kind of classical reactive oxygen species (ROS) with strong oxidizing capability. Simultaneously, the acceptor is regenerated, completing the catalytic cycle. The superoxide anion radical is then trapped by boronic acid, forming a peroxide radical intermediate C, which undergoes hydrogen atom transfer (HAT) with the generated cationic amine radical intermediate B to form the peroxyl alcohol species D. Subsequently, the rearrangement of intermediate D occurs to produce intermediate E, which is further hydrolyzed to yield the final oxygenation product.