Baowen
Zhou
,
Jinliang
Song
*,
Huacong
Zhou
,
Tianbin
Wu
and
Buxing
Han
*
Beijing National Laboratory for Molecular Sciences, CAS Key Laboratory of Colloid and Interface and Thermodynamics, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, China. E-mail: songjl@iccas.ac.cn; hanbx@iccas.ac.cn
First published on 9th October 2015
Direct utilization of the abundant hydrogen and oxygen in water for organic reactions is very attractive and challenging in chemistry. Herein, we report the first work on the utilization of the hydrogen in water for the hydrogenation of various organic compounds to form valuable chemicals and the oxygen for the oxidation of glucose, simultaneously by photocatalysis. It was discovered that various unsaturated compounds could be efficiently hydrogenated with high conversion and selectivity by the hydrogen from water splitting and glucose reforming over Pd/TiO2 under UV irradiation (350 nm). At the same time, glucose was oxidated by the hydroxyl radicals from water splitting and the holes caused by UV irradiation to form biomass-derived chemicals, such as arabinose, erythrose, formic acid, and hydroxyacetic acid. Thus, the hydrogen and oxygen were used ideally. This work presents a new and sustainable strategy for hydrogenation and biomass conversion by using the hydrogen and oxygen in water.
Lignocellulosic biomass is an abundant and renewable resource.4 Glucose can be obtained from the depolymerization of cellulose, which is a main component of lignocellulosic biomass.5 Conversion of glucose into various value-added chemicals and fuels has received much attention.6 Furthermore, H2 can also be generated from glucose by various routes, including photocatalytic processes,7 catalytic reforming,8 and enzymatic procedures.9 Up to now, attention has been mainly focused on the transformation of glucose into chemicals or H2. Exploration of new processes for glucose utilization is a long term task.
Hydrogenation is an important approach for producing various chemicals. Current hydrogenation processes usually suffer from harsh reaction conditions and lower selectivity. Moreover, the H2 used is mainly derived from fossil resources.10–12 Furthermore, the reduction processes that use stoichiometric reductants generally form large amounts of environmentally unfriendly waste.13–15 In recent years, photocatalytic reduction has been reported to be an alternative method for the traditional hydrogenation process with high selectivity under milder reaction conditions. Triethanolamine,16,17 isopropanol,18 and oxalic acid,19etc., are usually used as electron and proton donors, which makes the process less sustainable.
The use of cheap and abundant resources in the hydrogenation of organic compounds and transformation of biomass into value-added chemicals is of great importance. Herein, we report the first work of using water to provide hydrogen and oxygen for the hydrogenation of organic compounds and oxidation of biomass into value-added chemicals simultaneously by photocatalysis. In this approach, the hydrogen in water is utilized for the hydrogenation of different organic compounds, and at the same time the oxygen in water is involved in the oxidation of glucose to generate a series of biomass-derived valuable chemicals, as shown in Scheme 1, in which the main products are provided.
Scheme 1 Photocatalytic water splitting for hydrogenation of various organic compounds and conversion of glucose to valuable chemicals. |
Fig. 1d–f illustrates the influence of the molar ratio of glucose to substrate on the photocatalytic hydrogenation of 4-nitroacetophenone with different reaction times. It can be seen that the conversion of 4-nitroacetophenone (Fig. 1d), and the selectivity (Fig. 1e) and yield (Fig. 1f) of 4-aminoacetophenone increased with the increasing molar ratio of glucose to substrate. During the reaction process, glucose performed as the scavenger of photoexcited holes to inhibit the recombination of the electron–hole pairs on TiO2. More importantly, the photocatalytic reforming of glucose with the aid of hydroxyl radicals from water splitting could release protons, which could be reduced by photoinduced electrons to form the active H-PdNP species, this will be discussed in detail in subsequent sections. Therefore, increasing the molar ratio of glucose to substrate facilitated the elimination of the holes and promoted the formation of the H-PdNP species, which was beneficial to the photocatalytic reduction. Our control experiment showed that the hydrogenation reaction could be prohibited by the addition of 2,2′,6,6′-tetramethylpiperidine-N-oxyl (TEMPO, 100 mg), which could remove hydrogen atoms from the metal surfaces,24 indicating the formation of H-PdNPs. In addition, when the molar ratio of glucose to substrate was increased to 100/1, the reaction time for the complete conversion of 4-nitroacetophenone was reduced to 4 hours with a 4-aminoacetophenone selectivity of 98%, which further proved that a high molar ratio of glucose to substrate was beneficial for the reaction.
Water splitting accompanied with glucose reforming continuously proceeded over Pd/TiO2 under UV-light illumination, releasing protons and electrons for photocatalytic hydrogenation. As shown in Fig. 2 and Table 1, 0.13 mmol of glucose was consumed for the total conversion of 0.1 mmol of 4-nitroacetophenone to 4-aminoacetophenone, this was found from HPLC analysis with a refractive detector (Shimadzu RID-10A). Simultaneously, a series of biomass-derived chemicals, including arabinose, erythrose, formic acid, and hydroxyacetic acid were formed from the consecutive photooxidation of glucose by oxygen in the form of hydroxyl radicals from water and the holes on TiO2 caused by UV irradiation. In order to confirm that the hydrogen in water took part in the reaction, we carried out a control experiment using D2O to replace H2O with a reaction time of 12 hours and the other conditions were the same as those shown in Fig. 1b. The –NO2 was reduced to –NH2, –NDH, and –ND2 with a –NH2:–NDH:–ND2 ratio of 0.60:0.32:0.08, indicating that 24% of the hydrogen was from water in the reaction. This indicates that the hydrogen for the reduction was derived from both water splitting and glucose reforming. Our experiment showed that the reaction did not occur without glucose, indicating that to use the hydrogen in water for the hydrogenation reaction, the oxygen from water must be consumed in the oxidation of glucose simultaneously, this will be discussed in the following sections. In addition, hydrogenation did not occur in the dark or when the wavelength of the incident light was >420 nm. The reason for this was that the electron–hole separation resulting from the wide band gap of TiO2 (3.2 eV) could not be induced, which was the first key step for the subsequent photoredox reactions. Furthermore, only a very little amount (about 2%) of 4-aminoacetophenone was produced upon simulated sunlight because UV light energy is only a small proportion (about 4%) of solar energy. These results suggested that UV-light was the driving force of the reaction when Pd/TiO2 was used as the catalyst.
Fig. 2 The dependence of the amounts of the products from glucose reforming in water on time with the reaction conditions of Fig. 1b. |
Entry | Substrate | Time (h) | Nitroarenes con.b (%) | Anilines sel.c (%) | Glu. cons.d (mmol) |
---|---|---|---|---|---|
a Reaction conditions: substrates, 0.1 mmol; glucose, 0.5 mmol; water, 1 mL; temperature, 25 °C; 0.025 g Pd/TiO2 with 2 wt% Pd; UV-light (350 nm, 4 mW cm−2); illuminated area, 2 cm2. b Con. = conversion. c Sel. = selectivity. d Glu. cons. = glucose consumption. | |||||
1 | 12 | >99 | 99 | 0.13 | |
2 | 12 | 92 | 89 | 0.13 | |
3 | 12 | 98 | 96 | 0.12 | |
4 | 11 | >99 | 98 | 0.13 | |
5 | 10 | >99 | 99 | 0.11 | |
6 | 16 | 98 | 93 | 0.12 | |
7 | 18 | >99 | 96 | 0.12 | |
8 | 16 | 97 | 88 | 0.13 | |
9 | 16 | 95 | 90 | 0.13 | |
10 | 20 | 98 | 96 | 0.12 | |
11 | 48 | 95 | 97 | 0.20 |
Experiments were also conducted to examine the reusability of Pd/TiO2. In each cycle, Pd/TiO2 was recovered by centrifugation and washed with ethanol. After being dried at 60 °C under vacuum for 12 hours, the catalyst was reused for the next run. The conversion of 4-nitroacetophenone and the yield and selectivity of 4-aminoacetophenone are shown in Fig. 3. It was found that there was no considerable decrease in the conversion, selectivity, and yield after four cycles, indicating that Pd/TiO2 was stable in the reaction. Meanwhile, Pd/TiO2 recovered after being reused four times was characterized by TEM, and the results indicated that the size of the Pd particles was not changed obviously (Fig. S4†).
The photocatalytic reduction of nitrobenzenes with various substituted reducible groups was performed, and the results are presented in Table 1. It was found that all the nitrobenzenes examined could be converted to the corresponding anilines in the water–glucose system. The activity order of para-, meta- and ortho-substituted nitroacetophenones (entries 1–3, Table 1) was para- > meta- > ortho-, caused mainly because of steric hindrance. Meanwhile, 4-iodo-nitrobenzene (entry 8, Table 1) and 4-chloro-nitrobenzene (entry 9, Table 1) showed lower selectivity because of a dehalogenation reaction that occurred in the reaction process. In addition, nitroethane (entry 11, Table 1) could be reduced to ethylamine with a selectivity of 97% and a prolonged reaction time (48 h), indicating that our reaction system could also be used to reduce nitroalkanes.
From Table 1, it was found that the substituted nitrobenzenes with other reducible groups could be reduced to the corresponding anilines with good to excellent (88–99%) selectivity by the economic consumption of glucose. The high selectivity can be attributed mainly to two reasons. The first is that the nitro groups prefer to be adsorbed onto the catalysts in comparison with other reducible groups,10 which was beneficial to the selective reduction of the nitro group. A kinetics experiment was conducted to prove this. When a mixture of nitrobenzene (0.1 mmol) and benzaldehyde (0.1 mmol) was used as the reactants, it was found that nitrobenzene was preferentially hydrogenated in the process, while the reduction of benzaldehyde was inhibited completely in the presence of nitrobenzene (Fig. S2†). Secondly, the nitro group has a stronger electron-withdrawing ability than other substituted groups, exhibiting better electrophilicity,16 which is favorable for the reduction of the nitro group by the active H-PdNP species when compared with other reducible groups. We designed and carried out control experiments to find evidence to support this argument. In the experiment, the photocatalytic hydrogenation of nitrobenzene (0.1 mmol), benzaldehyde (0.1 mmol) and acetophenone (0.1 mmol) was performed separately in 2 mol L−1 glucose aqueous solution with a reaction time of 6 hours under UV-light (350 nm), and the results are illustrated in Fig. S3.† It was found that the nitro group had a higher reaction rate than the aldehyde and carbonyl groups. This suggests that the nitro group had a higher intrinsic catalytic activity than the other reducible groups.
Generally, there are two possible pathways for hydrogenating nitrobenzenes to anilines (Fig. 4). As shown by HPLC analysis, there were no azoxybenzene and azobenzene produced during the entire catalytic process. In contrast, nitrosobenzene and hydroxylamine were formed. These results indicated that the photocatalytic reduction of nitrobenzenes to the corresponding anilines was through pathway A in our reaction system (Fig. 4). On the basis of the experimental results and some related reports,10,25,26 we propose a possible mechanism about the photocatalytic selective reduction of nitrobenzenes to anilines in the water–glucose system over Pd/TiO2, which is shown in Fig. 5 and 6. Firstly, electron–hole separation on TiO2 occurred under UV-light irradiation (Fig. 5). Then, water splitting accompanied with the photocatalytic reforming of glucose was induced by the photogenerated holes to release protons and hydroxyl radicals (Fig. 5). The hydroxyl radicals were involved in the photocatalytic reforming of glucose. A series of biomass-derived chemicals were formed through C1–C2 cleavage from the continuous photooxidation of glucose by the hydroxyl radicals from water splitting and the holes on TiO2 caused by UV irradiation (Fig. 6). Subsequently, the photoinduced electrons could readily reduce the H+ from water splitting and glucose reforming on the PdNPs to form the active H-PdNP species (Fig. 5). Finally, the nitro group of the nitrobenzenes is preferably adsorbed onto the catalyst, and is predominantly reduced to the functionalized anilines by the H-PdNP species through pathway A shown in Fig. 4.
Fig. 4 Possible pathways of photocatalytic selective reduction of functionalized nitroarenes to anilines. |
Fig. 5 Proposed mechanism of photocatalytic water splitting and glucose reforming for selective reduction of nitrobenzenes to anilines catalyzed by Pd/TiO2. |
Inspired by the results obtained from the selective photocatalytic reduction of nitrobenzenes, we explored the hydrogenation of other kinds of unsaturated organic compounds using the same method, and the results are illustrated in Table 2. It was found that the hydrogen from water splitting and glucose reforming could be used for efficiently hydrogenating benzaldehdye and hydroazobenzene to the corresponding saturated products with high yields. The products from glucose were also mainly arabinose, erythrose, formic acid, and hydroxyacetic acid. Control reactions conducted in D2O showed that the content of deuterohydrogen in the products was 39% and 30%, respectively. This result further proved that the hydrogen for the reductions originated from both water and glucose.
Entry | Reactant | Product | t (h) | Con.b (%) | Yield (%) | Hydrogen from water |
---|---|---|---|---|---|---|
a Reaction conditions: substrates, 0.1 mmol; glucose, 0.5 mmol; water, 1 mL; temperature, 25 °C; 0.025 g Pd/TiO2 with 2 wt% Pd; UV-light (350 nm, 4 mW cm−2); illuminated area, 2 cm2. b Con. = conversion. | ||||||
1 | 24 | 96 | 89 | 39% | ||
2 | 10 | 100 | 90 | 30% |
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5sc03178h |
This journal is © The Royal Society of Chemistry 2016 |