Radiation-induced aerobic oxidation via solvent-derived peroxyl radicals

Abstract

Oxidation is a fundamental transformation in synthesis. Developing facile and effective aerobic oxidation processes under ambient conditions is always in high demand. Benefiting from its high energy and good penetrability, ionizing radiation can readily produce various reactive species to trigger chemical reactions, offering another option for synthesis. Here, we report an ionizing radiation-induced aerobic oxidation strategy to synthesize oxygen-containing compounds. We discovered that molecular oxygen (O₂) could be activated by reactive particles generated from solvent radiolysis to produce solvent-derived peroxyl radicals (R_solOO·), which facilitated the selective oxidation of sulfides and phosphorus(iii) compounds at room temperature without catalysts. Density functional theory (DFT) calculations further revealed that multiple R_solOO· enable the oxidation reaction through an oxygen atom transfer process. This aerobic oxidation strategy broadens the research scope of radiation-induced chemical transformations while offering an opportunity to convert nuclear energy into chemical energy.

Solvent-derived peroxyl radicals (R_solOO·), a new group of reactive oxygen species (ROS), enable the radiation-induced aerobic oxidation strategy.

Introduction

Oxidation reactions are among the most widespread transformations in synthesis.¹ With the growing interest in green synthesis and sustainable chemistry, molecular oxygen (O₂) is regarded as an ideal oxidant due to its good natural abundance and high cost-effectiveness, making it a superior choice for synthesizing oxygen-containing compounds. Utilizing oxygen instead of classical oxidants not only eliminates the need for stoichiometric reagents, but also avoids harsh reaction conditions and the production of potentially toxic waste.² Emerging as an environmentally friendly approach, the photochemical aerobic oxidation strategy under ambient conditions has garnered extensive attention in recent years and continues to evolve.³ In this strategy, triplet oxygen with relatively low reactivity is usually activated by photosensitizers (PSs, e.g., transition metal complexes and π-conjugated aromatic dyes) or electron donor–acceptor (EDA) complexes to produce singlet oxygen (¹O₂) or superoxide anion radicals (O₂˙⁻) for subsequent reactions (Scheme 1a).⁴ Nonetheless, the diversity of reactive oxygen species (ROS) remains limited, and the incorporation of catalysts could compromise the cost-effectiveness and functional group compatibility.⁵ Therefore, developing facile and effective methods that enable aerobic oxidation is always in high demand.

Scheme 1. Solvent-derived peroxyl radicals enable radiation-induced aerobic oxidation. UV = ultraviolet, Sol. = solvent, and Sub. = substrate.

Ionizing radiation (e.g., γ-ray photons), the main form of nuclear energy utilization,⁶ has been widely applied in wastewater treatment,⁷ material fabrication,⁸ and radiotherapy.⁹ These high-energy particles can uniformly and effectively produce highly reactive species, which can overcome activation barriers for subsequent reactions under ambient conditions. Furthermore, ionizing radiation possesses good penetrability and industrial accessibility, holding promise for large-scale production.¹⁰ Recently, ionizing radiation has been harnessed to produce chemical feedstocks such as hydrogen,¹¹ methanol,¹² acetic acid,¹³ and ammonia,¹⁴ demonstrating the feasibility of nuclear energy-induced chemical transformations. To the best of our knowledge, radiation-induced aerobic oxidation has not yet been developed. Diverging from photochemical aerobic oxidation processes, in which photoresponsive groups harness ultraviolet-visible light energy to activate O₂, solvent molecules with the largest proportion predominantly absorb energy and trigger O₂ activation in radiation-induced aerobic oxidation reactions.¹⁵ Upon irradiation, stable solvent molecules can be converted into solvated electrons (e_sol⁻), hydrogen radicals (H·), and solvent-derived radicals (·R_sol) with a homogeneous distribution within tens of nanoseconds.¹⁶ In the presence of air, ·R_sol with a high radiolytic yield can rapidly activate O₂ (k > 10⁹ L mol⁻¹ s⁻¹) to generate solvent-derived peroxyl radicals (R_solOO·).¹⁷ As a group of ROS that have received little attention yet commonly exist in the radiolysis of various solvents, reactivity-adjustable R_solOO· are anticipated to enable a broad array of aerobic oxidation reactions (Scheme 1b).

Herein, we used γ-ray radiation from a ⁶⁰Co source to investigate whether ionizing radiation could induce aerobic oxidation processes. The results revealed that multiple R_solOO· could successfully promote the oxidation of sulfides and phosphorus(iii) compounds at ambient temperature and pressure without any catalysts. This aerobic oxidation strategy exhibits favorable functional group compatibility and is suitable for gram-scale synthesis, demonstrating that ionizing radiation can enable selective oxidation processes for synthesizing chemical feedstocks rather than unselectively degrading organics to carbon dioxide and water as in wastewater treatment. It provides an effective approach for organic synthesis and opens broad prospects for the valuable utilization of nuclear energy.

Results and discussion

Optimization of the reaction conditions

Sulfoxide is a ubiquitous structural building block widely applied in natural products, pharmaceuticals, agrochemicals, food additives, and other chemical feedstocks.¹⁸ Normally, sulfoxides are produced through the oxidation of the corresponding sulfides with the help of various oxidants, such as m-chloroperbenzoic acid, oxone, hydrogen peroxide, etc.¹⁹ To validate our hypothesis, we first selected methylphenyl sulfide (1a) as the model compound to investigate radiation-induced aerobic oxidation.

The initial reaction was carried out in different solvents under an air atmosphere and irradiated with γ-rays at room temperature with a dose rate of 110 Gy min⁻¹ (1 Gy = 1 J kg⁻¹) (Tables 1 and S1, ESI†). To our delight, acetonitrile (CH₃CN) as the solvent gave the highest yield of (methylsulfinyl)benzene (2a), reaching 95% after 15 h of irradiation (Table 1, entry 1). Methanol (CH₃OH), acetone, and ethyl acetate (EtOAc) could also facilitate the oxidation of 1a in moderate yields (Table 1, entries 2–4), while tetrahydrofuran (THF), cyclohexane, dimethyl sulfoxide (DMSO), N,N-dimethylformamide (DMF), toluene, and dichloromethane (CH₂Cl₂) were not conducive to the reaction (Table 1, entries 5–10). Considering that the addition of water can inhibit the conversion of sulfoxides to sulfones through hydrogen bond interactions, we then evaluated a mixed solvent of CH₃CN and water (Table 1, entries 11 and 12, and Table S2, ESI†).²⁰ However, with the increase in the proportion of water, the yield correspondingly decreased, which could be attributed to the decreased solubility of the substrate and the reduced concentration of the primary oxidizing species. In addition, hydroxyl radicals which are potent oxidants produced from water radiolysis tended to cause overoxidation of organics, thus reducing the yield. The dose rate had almost no effect on the reaction performance when the total absorbed dose was fixed (Table S3, ESI†). Of note, the reaction time played a vital role in determining the product distribution (Table S4, ESI†). We found that increasing the reaction time from 2 h to 15 h would raise the yield of 2a from 27% to 95%. However, when the reaction time exceeded 15 h, 2a started to convert to methylphenyl sulfone (3a), indicating that it is feasible to selectively obtain sulfoxides or sulfones by modulating the reaction time. When 2a was used as the substrate, 3a could also be obtained with the other reaction conditions unchanged, further demonstrating that sulfoxides served as necessary intermediates in the oxidation of sulfides to sulfones (Table S5, ESI†).

Table 1. Optimization of the reaction conditions^a.

Entry	Solvent	Yieldb (%)
1 c	CH 3 CN	95
2	CH3OH	87
3	EtOAc	62
4	Acetone	75
5	THF	18
6	Cyclohexane	22
7	DMSO	Trace
8	DMF	21
9	Toluene	24
10	CH2Cl2	n.d.
11	CH3CN/H2O (4/1)	72
12	CH3CN/H2O (1/4)	16

^aReaction conditions: 1a (0.3 mmol) in solvent (15 mL) with γ-ray radiation (dose rate: 110 Gy min⁻¹, measured using a Fricke dosimeter) at room temperature and under an air atmosphere for 15 h.

^bDetermined by gas chromatography-mass spectrometry (GC-MS) using 1,3,5-trimethoxybenzene as the internal standard.

^cThe optimized reaction condition. n.d. = no detected.

Substrate scope

Based on the optimized reaction conditions, we further evaluated the substrate scope of radiation-induced aerobic oxidation using various sulfides. Aryl methyl sulfide derivatives with different substituents were examined first. As shown in Fig. 1, both electron-donating groups and electron-withdrawing groups were tolerated. Aryl methyl sulfides with electron-donating groups (e.g., alkyl, alkoxy, and trifluoromethoxy) (1b–1d) were converted to the corresponding sulfoxides (2b–2d) in moderate yields. Similarly, aryl methyl sulfides with electron-withdrawing groups (e.g., halogen, nitro, cyano, acetyl, and trifluoromethyl) (1e–1s) could also be converted to sulfoxides (2e–2s) in satisfactory yields. High functional group compatibility was shown by the successful preservation of various groups in products, including trifluoromethoxy (2d), nitro (2n), cyano (2o), boronic pinacol ester (Bpin, 2p), and trifluoromethyl groups (2s). Additionally, the number and positional properties of the functional groups have a negligible effect on the reactivity (2i–2m). Next, we investigated the oxidation efficiency of naphthyl or heteroaryl methyl sulfides. To our satisfaction, sulfides bearing naphthalene (2t), thiophene (2u and 2v), pyridine (2w), pyrimidine (2x), thiazole (2y), or benzothiazole (2z) were also accommodated by the radiation-induced aerobic oxidation system. It is worth mentioning that the sulfur-containing heteroaromatic rings remained unoxidized, showing that this oxidation strategy is chemoselective.

Fig. 1. The substrate scope of radiation-induced aerobic oxidation of sulfides. Reaction conditions: 1 (0.3 mmol) in CH₃CN (15 mL) with γ-ray radiation (dose rate: 110 Gy min⁻¹, measured using a Fricke dosimeter) at room temperature and under an air atmosphere for 15 h. All yields are isolated. ^aReaction time: 30 h. ^bReaction time: 7 h.

Subsequently, we expanded the substrate scope to other aryl alkyl sulfides with different chain lengths, isomeric structures, and functional groups. They could all be oxidized to sulfoxides with alkyl motifs retained in moderate yields (2aa–2af). Apart from aryl alkyl sulfides, a series of diaryl sulfides and dialkyl sulfides were evaluated. The relatively low reactivity of diaryl sulfides is attributed to the conjugation effect of benzene rings (2ag–2ak). Considering that thianthrene contains two sulfur atoms, we found that only mono-oxygenation sulfoxide (2aj) could be obtained. In contrast to diaryl sulfides, dialkyl sulfides could be transformed into sulfoxides (2al and 2am) in excellent yields in a shorter reaction time. The above results indicate that electronic effects of the substituents linked to the sulfur atom play an important role in the conversion efficiency. The conversion of sulfides containing electron-donating groups was easier than those bearing electron-withdrawing groups. In addition, diaryl sulfides required a longer reaction time than aryl alkyl sulfides and dialkyl sulfides due to the decreased electron density on the sulfur atom.

Encouraged by the satisfactory results above, we further verified the applicability of this oxidation strategy for synthesizing bioactive molecules and pharmaceuticals. A potential anticancer drug sulforaphane (2an) was obtained in 71% yield with its sensitive isothiocyanate group preserved.²¹ Moreover, the corresponding sulfoxide of etoricoxib (2ao), a cyclo-oxygenase (COX)-2 inhibitor, could also be prepared in 52% yield.²²

Similar to the sulfur atoms in sulfides, the phosphorus atoms in phosphorus(iii) compounds also tend to accept oxygen atoms and be oxidized to form the corresponding phosphorus(v) compounds, which are crucial for the synthesis of organophosphorus compounds.²³ As shown in Fig. 2, aryl alkyl phosphine oxides with different alkyl motifs (5a–5f), triaryl phosphine oxides with different substituents (5g–5p), and tricyclohexylphosphine oxide (5q) could all be readily obtained with excellent yields in a much shorter reaction time. Alkene (5e), furan ring (5o), and pyridine ring (5p) were well preserved after the reaction. Of note, 5g was successfully obtained at a gram-scale level (91%, 1.26 g), suggesting that this protocol is promising for scale-up synthesis. Moreover, this radiation-induced aerobic oxidation strategy could also be applied to the synthesis of phosphinates, phosphonates, and phosphates (5r–5u).

Fig. 2. The substrate scope of radiation-induced aerobic oxidation of phosphorus(iii) compounds. Reaction conditions: 4 (0.3 mmol) in CH₃CN (15 mL) with γ-ray radiation (dose rate: 110 Gy min⁻¹, measured using a Fricke dosimeter) at room temperature and under an air atmosphere for 3 h. All yields are isolated. ^aReaction time: 100 min. ^bReaction time: 220 min. ^cGram-scale synthesis.

Mechanistic investigation

We next turned our attention to the mechanism of radiation-induced aerobic oxidation of sulfides and phosphorus(iii) compounds. All mechanistic studies were conducted with 1a as the model substrate and CH₃CN as the solvent. According to the results of control experiments, both γ-ray radiation and O₂ were essential for the success of this oxidation strategy (Fig. 3a). Then, we introduced the ROS fluorescent probe 2′,7′-dichlorodihydrofluorescein (DCFH) to the system and observed an increased fluorescence signal with prolonged irradiation time, supporting the generation of ROS in this oxidation process (Fig. S2, ESI†).

Fig. 3. Experiments for mechanistic investigation. (a) Control experiments for mechanistic studies. (b) Quenching experiments for investigating the key ROS. (c) High-resolution mass spectrometry of the adduct between TEMPO and ·CH₂CN. (d) High-resolution mass spectrometry of the adduct between BHT and ·CH₂CN. ^aDetermined by GC-MS using 1,3,5-trimethoxybenzene as the internal standard. ^bStandard conditions: 1a (0.3 mmol) in CH₃CN (15 mL) with γ-ray radiation (dose rate: 110 Gy min⁻¹, measured using a Fricke dosimeter) at room temperature and under an air atmosphere for 15 h. n.d. = no detected.

To further investigate the key ROS, a series of quenching experiments were performed by introducing various scavengers (Fig. 3b). When an electron quencher (e.g., AgNO₃) was added, there was no significant difference in the reaction performance, demonstrating that e_sol⁻ and O₂˙⁻ derived from e_sol⁻ were not the primary reactive species. Similarly, we utilized the O₂˙⁻-specific probe nitrotetrazolium blue chloride (NBT) to monitor the superoxide species, and the unabated absorbance at 260 nm revealed the absence of O₂˙⁻ (Fig. S3, ESI†). By introducing diphenyl sulfoxide (Ph₂SO), we could rule out the formation of a persulfoxide intermediate between sulfide and ¹O₂ in that the reaction performance remained essentially unchanged and only a trace amount of diphenyl sulfone (Ph₂SO₂) formed between persulfoxide and Ph₂SO (Fig. S4, ESI†).²⁴ Meanwhile, the reaction performance was hardly affected when deuterated acetonitrile (CD₃CN) was used as the solvent, where the lifetime of ¹O₂ is longer [τ_Δ(CD₃CN/CH₃CN) = 1625/82 μs] (Fig. S7, ESI†).²⁵ Therefore, we speculated that ¹O₂ also played a minor role in this oxidation process. The addition of 1,4-dimethoxybenzene (DMB) as a sulfide radical cation (R₂S˙⁺) scavenger also did not significantly suppress the reaction, suggesting that R₂S˙⁺ might not be as important as they are in photocatalytic aerobic oxidation driven by O₂˙⁻ or ¹O₂.²⁶ Hydroxyl radicals (·OH) were excluded by introducing 2-propanol (i-PrOH) as the scavenger. Furthermore, when 2,2,6,6-tetramethyl-1-piperinedinyloxy (TEMPO) or butylated hydroxytoluene (BHT) was added as the radical trapping reagent, the oxidation of 1a was severely suppressed, and the adduct with the cyanomethyl radical (·CH₂CN) 6 or 7 could be observed by high-resolution mass spectrometry (HRMS) (Fig. 3c, d, and S5, ESI†). These results indicated that a ·CH₂CN-mediated radical process may be involved.

Based on the above results, we propose that the radiation-induced aerobic oxidation process differs from most photocatalytic processes mediated by O₂˙⁻ or ¹O₂. Solvent-derived peroxyl radicals (e.g., cyanomethylperoxyl radicals, ·OOCH₂CN) generated from the rapid combination of O₂ and ·R_sol (e.g., ·CH₂CN) may enable this oxidation process.

To further unravel the mechanism, density functional theory (DFT) calculations were performed on the model substrate 1a with CH₃CN as the solvent. The proposed mechanism and calculated results are shown in Fig. 4. Upon irradiation with ⁶⁰Co γ-rays, CH₃CN is initially ionized to generate e⁻ and radical cations (CH₃CN˙⁺) along the tracks. After the subsequent rapid processes such as spur expansion, geminate recombination, and intra-track reaction, ·CH₂CN becomes the predominant reactive species in the solution within tens of nanoseconds.²⁷ Therefore, the cleavage of the C–H bond is not the rate-determining step, which is consistent with the findings of kinetic studies (Fig. S7, ESI†). According to previous studies and the calculated results, once ·CH₂CN is formed, it reacts with O₂ to produce ·OOCH₂CN smoothly (k = 1.3 × 10⁹ L mol⁻¹ s⁻¹), which is spontaneous (Fig. 4a and S8, ESI†).²⁸ The ·OOCH₂CN then undergoes the oxygen atom transfer (OAT) process with 1a to yield 2aviaTS_1a/2a with the Gibbs activation energy (ΔG^≠) and the Gibbs reaction energy (ΔG) being 23.3 kcal mol⁻¹ and −15.5 kcal mol⁻¹, respectively. By contrast, it is more difficult for ·OOCH₂CN to undergo the same process with 2a to produce 3aviaTS_2a/3a, where ΔG^≠ and ΔG are 26.6 kcal mol⁻¹ and −33.5 kcal mol⁻¹. These results reveal that sulfoxides are kinetic products and sulfones are thermodynamic products, which is consistent with the relationship between product distribution and reaction time (Table S4, ESI†). Moreover, the second oxygen atom of ·OOCH₂CN is unlikely to undergo the OAT process, because the values of ΔG^≠ for the reaction of cyanomethoxy radicals (·OCH₂CN) with 1a and 2a are as high as 35.3 kcal mol⁻¹ and 33.1 kcal mol⁻¹, respectively, indicating that the ·OCH₂CN pathway is unfavorable compared to the ·OOCH₂CN pathway (Fig. 4b). All of the transition state geometries were determined by intrinsic reaction coordinate (IRC) analysis (Fig. S9, ESI†).

Fig. 4. Mechanism proposal and DFT calculations. (a) The radiolysis of CH₃CN and the formation of ·OOCH₂CN. (b) Gibbs energy profile for the radiation-induced aerobic oxidation of 1a to 2a and 3avia ·OCH₂CN (left) or ·OOCH₂CN (right) was calculated at the B3LYP-D3/6-311+G(d,p)/SMD(acetonitrile) level of theory. (c) Distortion–interaction analysis for transition states in the Gibbs energy profile was calculated at the B3LYP-D3/6-311+G(d,p)/SMD(acetonitrile) level of theory.

Distortion–interaction analysis was also performed on these four transition states to explain why ·OOCH₂CN is more conducive to undergoing the OAT process (Fig. 4c and Table S8, ESI†).²⁹ Compared to the other three transition states, the total distortion energy (ΔE^≠_dist, the energy required to distort substrate and oxidizing species into the transition state geometry) of TS_1a/2a is 26.8 kcal mol⁻¹, which is higher than that of TS_2a/3a and . However, the interaction energy (ΔE^≠_int, the energy released when the two fragments form a complete transition state) of TS_1a/2a is down to −15.1 kcal mol⁻¹, which compensates for the relatively higher ΔE^≠_dist. Therefore, the activation energy (ΔE^≠, the sum of ΔE^≠_dist and ΔE^≠_int) of TS_1a/2a is the lowest (11.7 kcal mol⁻¹), suggesting that TS_1a/2a is relatively stable and the transformation from 1a to 2avia ·OOCH₂CN is more favorable.

Similar to ·CH₂CN, hydroxymethyl radicals (·CH₂OH), cyclohexyl radicals (·C₆H₁₁), and tetrahydrofuran-2-yl radicals (·C₄H₇O) generated from solvent radiolysis could also react with O₂ rapidly to produce the corresponding peroxyl radicals, which then promote the oxidation of 1a to 2a through the OAT process.³⁰ The results of control experiments and radical trapping experiments showed that γ-ray radiation, O₂, and ·R_sol also play important roles in radiation-induced aerobic oxidation mediated by other several solvents (Tables S6, S7 and Fig. S6, ESI†). As shown in Fig. 5a, as the peroxyl radicals change from ·OOCH₂CN, hydroxymethylperoxyl radicals (·OOCH₂OH), and cyclohexylperoxyl radicals (·OOCy) to tetrahydrofuran-2-peroxy radicals (·OOTHF), the value of ΔG^≠ increases from 23.3 kcal mol⁻¹ to 34.0 kcal mol⁻¹, which indicates that the OAT process becomes difficult to proceed. This trend is also consistent with the reaction performance shown in Table 1. The yield of 2a decreases with the increase in Gibbs activation energy, indicating that the oxygen atom transfer ability of R_solOO· is a key factor affecting the reaction performance. Distortion–interaction analysis reveals that the interaction between the substrate and R_solOO· is particularly important for the stabilization of the transition state (Fig. 5b and Table S8, ESI†). When the values of ΔE^≠_dist are nearly identical, the value of ΔE^≠_int in CH₃CN (−15.1 kcal mol⁻¹) or CH₃OH (−15.9 kcal mol⁻¹) is significantly lower compared to that in cyclohexane (−8.7 kcal mol⁻¹) or THF (−8.8 kcal mol⁻¹), contributing to a more stable transition state and a lower ΔE^≠ (11.7 kcal mol⁻¹). The unfavorable interaction between ·OOCy or ·OOTHF and 1a may be due to the steric hindrance brought by the cyclohexyl or tetrahydrofuran-2-yl group (Fig. 5c). Therefore, the efficacy of the radiation-induced aerobic oxidation strategy heavily relies on the solvent type and the reactive species generated from solvent radiolysis.

Fig. 5. Multiple solvent-derived peroxyl radicals enable the radiation-induced aerobic oxidation strategy. (a) Gibbs energy profile for the radiation-induced aerobic oxidation of 1a to 2avia different R_solOO· was calculated at the B3LYP-D3/6-311+G(d,p)/SMD level of theory. CyH = cyclohexane. (b) Distortion–interaction analysis for transition states in the Gibbs energy profile was calculated at the B3LYP-D3/6-311+G(d,p)/SMD level of theory. (c) Optimized geometries of transition states.

Conclusions

In summary, we presented an ionizing radiation-induced aerobic oxidation strategy that enables the selective oxidation of sulfides and phosphorus(iii) compounds at room temperature without catalysts. This strategy exhibited a wide substrate scope and favorable functional group compatibility. Through mechanistic experiments and DFT calculations, we established that highly active R_solOO· produced from solvent radiolysis, rather than O₂˙⁻, ¹O₂, and ·OH, serve as the primary oxidizing species and facilitate the oxidation through an OAT process. Notably, R_solOO· are commonly generated during the radiolysis of various solvents in the presence of O₂, making it easy to adjust the reaction performance by choosing the appropriate solvent. Combined with the increasing accessibility and strong penetrability of ionizing radiation, our aerobic oxidation strategy is expected to have a wider range of applications in various oxidation reactions and scale-up synthesis, presenting a fresh avenue for nuclear energy utilization.

Data availability

The ESI† includes details of optimization studies, substrate scope studies, mechanistic studies, and computational studies. Characterization data, NMR spectra, and Cartesian coordinates of all optimized geometries are included as well.

Author contributions

Z. L. conceived the study. Y. X., assisted by B. M., W. L. and J. L., performed optimization studies and substrate scope studies. Y. X., assisted by B. M. and Z. T., performed mechanistic studies. Y. X., assisted by B. M. and Z. S., performed computational studies. Y. X., B. M. and Z. L. analyzed the data. Z. L. wrote the paper with inputs from all authors. All authors discussed the results and commented on the paper.

Conflicts of interest

There are no conflicts to declare.