有機ヒドリドの光再生を利用する非遷移金属触媒系CO2光還元反応の開発

概要

Introduction:
The ever-increasing demand for fossi I fuels has exacerbated dramatic energy shortages and massive greenhouse gas emissions. Especially, CO2, the most major greenhouse gas at the surface of the Earth's atmosphere has a serious impact on climate change which causes global warming and rise in sea-levels.1 In face of such critical situation, carbon capture and storage (CCS) seems to be an effective physical tactic to cope with carbon dioxide emission.2 Besides, Compared with CCS, directly chemical conversion of CO2 into useful chemicals and value-added fuels is more appealing using thermochemical, electrochemical and photochemical.2 Conspicuously, using renewable solar light as an energy source to photo-catalytically convert carbon dioxide into value-added fuels is an ideal approach to address the current predicament. As a thermodynamicaly stable gas, carbon dioxide, its reduction requires high energy as inputs.3 Besides, the selectivity of the formed products deserves consideration. Thus, developing an efficient photocatalyst system is an indispensable key to achieving photocatalytic reduction of carbon dioxide and practical solar energy storage.

Experiments and results:
Prior to the demonstration of photocatalytic reduction of CO2, we first examined the photochemical reduction of BJ+(J-) to determine whether BIH could be photochemically recycled. The photocatalytic reduction o「BJ+(J-)was converted to its reduced-state BIH in 44% yield with 54% BJ+(I-)remaining using PC 1 as a photocatalyst and ascorbic acid (H2A) as a sacrificial reductant under visible light irradiation (400 nm) (Scheme 1). The results fully suggested that regeneration of BIH from Bi+(I-) was viable and BJ+(J-) and BIH both held sufficient stability to be competent for long-term catalytic capability.

To date, using organic photocatalytic cycle to provide electrons for regeneration of BIH from oxidized state BI+(1-）as a feasible strategy. We have successfuly designed and demonstrated an excellent method CO2 was photo-catalytically converted to value-added formate comprising carbazole moiety as a visible light-driven photosensitizer, benzimidazoline-based organohydride as co-catalyst, and ascorbic acid as a sacrificial reductant using visible light irradiation at 400 nm. In this study, 13C labeling experiment further confirmed the carbon source of formate ptoduct derived from CO2 through the photocatalytic reduction of CO2. During experiment optimization, Photocatalytic reduction of CO2 generated formate in a 143% of yield under standard condition (Scheme 2). Reduced state BlH (without iodine sources) was employed as co-catalyst in the current reaction system. As a result, the yield of formate product using BIH as co-catalyst produced the same result as that of Bl+(J-), revealing that iodide anion was not involved in the photocatalytic cycle. The use of other inorganic bases instead of potassium carbonate in the current reaction system implied that the type of counter cations and basic anions had a slight impact on the photoinduced reaction to certain extent. In addition, other sacrificial reductants were discovered as an alternative solution to ascorbic acid, such as sodium sulfite and sodium hydrogen sulfite. Eventualy, a few of control experiments revealed that photocatalyst, light, sacrificial electron donor and base were all necessary for photoinduced CO2 reduction. However, photocatalytic reduction of CO2 still provided 8% of formate product in the absence of BI+(I・), prompting its meaningful per「ormanceas a co-catalyst.

As the current reaction system was conditioned under 400 nm irradiation condition, some weak chemical bonds from photosensitizer and co-catalyst might be at risk of being broken to decrease their stability and activity. Initial purpose for design and synthesis of integrated-form catalysts (27, 29 and 31) obtained electron donor-acceptor (EDA) complex to enable the reaction to occur under red-shifted visible light irradiation condition (Figure I). Although the formation of EDA complex was not demonstrated by their adsorption regions compared with that of PC 1, the photochemical reduction of CO2 using incorporated catalyst was conceptually viable. Photocatalytic reduction of CO2 using I mol% of incorporated catalyst 27, 29, and 31 provided formate product with 68, 72 and 56% yields in order. Nevertheless, we also noticed that their photo-activity was slightly inferior to that of dual catalysts available for photochemical reduction of CO2, ascribing to speedily intramolecular back electron transfer. The formation of the EDA complex from photocatalyst and co-catalyst will continue study further.

Next, potential photosensitizers were also investigated. We found that carbazole-based organic PCs with electron-withdrawing groups could not function as active photocatalysts to participate in reduction of CO2. Among these organic PCs, Nphenyl substituted carbazole moieties exhibited stronger durability in photocatalytic reduction of CO2. Additionally, we noticed that Jr(ppy)3 complex showed a productive efficiency in a 132% of yield which was roughly equivalent to performance of carbazole moiety; and formate product was not yielded from Ir(ppy)3 without BJ+(I-). Thereby, it suggested that the occurrence of lr(ppy)3-photocatalyzed CO2 reduction was through BI+(I-）that involved a mechanism similar to the reaction using PC 1. Thus, the combination of catalysts lr(ppy)J and BI+(J-) in a current study was unprecedented and novel.

Besides, the durability and activity of photocatalysts and cocatalyst were further explored. Results implied that the catalytic amount of PC was reduced by a dramatic order of magnitude up to 0.0 I mo I%. Most ideally, using organic PC 7, its TON and TOF reached 8820 and 2205/h, respectively (Table I, Entry 2). In addition, we further investigated TON and TOF which were calculated based on co-catalyst. When catalytic amount of PC 5 was set at 3 mol%, the amount of co-catalyst BI+(I-) was decreased to 0.0 I mol%. Its TON and TOF also exhibited dramatical increment up to 6070 and 1520/h (Table I, Entry 4). These results fully suggested that the employed PCs and BJ+(J-) had high durability and activity for photocatalytic reduction of CO2. Besides, the catalyst loadings of both PC 5 and BJ+(I-)could be simultaneously reduced to 0.1 mo)% in a 71 % yield (Table I, Entry 5), while further reduction of both catalyst loadings resulted in a much lower yield.

Knowingly, metal-catalyzed CO2 reduction in which CO and I Ii are o「ten generated becomes a side reaction. Indispensably, we need to determine whether CO and H2 gaseous are yiclded in current reaction system. Therefore, a quantitative analysis of evolved gas in the headspace of photochemical reactor by GC was carried out. Calculation result rcvcaled that H2 and CO evolution were below the detection limit(< 0.1 % yield for出 and< 0.8% yield for CO). Additionally, formaldehyde and oxalate as potential evolved products were undetected using 13C-labeling experiment. Therefore, photocatal:>-tic reduction of CO2 showed exclusive selectivity to yield formate as a sole product in current reaction system.

During the photocatalytic CO2 reduction, ascorbic acid was used as sacrificial electron donor. Thc actual yield based on the loading of ascorbic acid (two electrons provided per molecule) exceeded theoretical yield. We suspected that ascorbic acid might further undergo two-time redox to supply extra electrons and allow extra formation of formate (Figure 2). To testify the hypothesis, ascorbic acid was substituted with a commercially available dehydroascorbic acid in the photoinduced reduction of CO2 to corroborate formate was still produced in 74% yield. This revealed that dehydroascorbic acid was also competent as a sacrificial reductant since intermediate 34 had a structural pattern of enediol comparable to HA-(Figure 2). Therefore, ascorbic acid as sacrificial reductant in the photocatalytic reduction of CO2 fully corroborated that a two-time oxidation of ascorbic acid (four electrons provided per molecule) via intermediate dehydroascorbic acid was involved.

To further confirm the relationship between reaction mechanism and photocatalytic electron transfer, we investigated the quenching rate constants (kq) based on the interactions of PC 1 as photocatalyst with various concentrations of potential quenchers (Bl+(J-), HA-, H2A and BIH) according to the Stern-Volmer equation. As expected, Bl+(J-) exhibited stronger quenching effect, whereas HA-, H2A and BIH exhibited much weaker effects. Thus, these results confirmed that interaction of singlet excited state PC 1 with BJ+(I-) by single electron transfer was existent in the current reaction system. The electronic and optical properties of BJ+(J-) and PC 5 were further studied to provide important information that energy transfer from 35* to BJ+(I．)was not possible and single electron transfers from 15* to Bl+(J-) and/or from 35* to BJ+(I-) were both possible. Additionally, the magic blue experiment and time-resolved optical absorption spectrometer analysis revealed that s•+ was captured and energy transfer and/or electron transfer from 35* to BJ+(I.)pathways were not major mechanisms for the interaction of PC 5 with BI+（「）． Factually,single electron transfer from 15* to BJ+(I-) pathway was a major mechanism for interaction of PC 5 with BJ+(I-). Besides, via related experiments, we further demonstrated that treatment of BIH with CO2 (I atm) could not directly proceed, but BIH reacted with CO2 (I atm) to provide HCoo・ product in the presence of photocatalyst under irradiation condition. It was further revealed that the possible mechanism: Apart from carbazole moiety as a photocatalyst in photochemical regeneration of BIH from oxidized state BJ+(I-)，carbazole based photocatalyst was irradiated under irradiation of 400 nm condition again and produced excited state photocatalyst, which transferred one electron to CO2 to provide resulting CO2•- and then interaction of BIH with CO2•—produced HCoo・ via hydrogen atom transfer (Figure 3).

a Standard enthalpy change for the reaction (Br+ HA•• BIB + A) was calculated by the density functional theory (OFT) method with (U)しC-BLYP/6-31++G(d)/CPCM basis sets. b Estimated value by the OFT calculation. Standard redox potentials are referred to saturated calomel electrode (SCE). SET = single electron transfer; lSC = intersystem crossing; HAT= hydrogen atom transfer.

Conclusion:
We have developed a feasible strategy that transition metal-free photocatalytic reduction of CO2 was converted to formate with a dramatically high TON and TOF. Current reaction system for photocatalytic CO2 reduction reaction showed exclusive selectivity to yield formate as a sole product, while formaldehyde, oxalate, CO, and H2 as potential evolved products were not detected by 13C-label ing experiment and GC-based quantitative analysis of evolved gases experiment. Finally, based on the experimental observations and results, Single electron and hydrogen atom transfer-based photocatalytic mechanism of CO2 reduction was proposed.