Porous Copper/Zinc Bimetallic Oxides Derived from MOFs for Efficient Photocatalytic Reduction of CO₂ to Methanol

Abstract

A novel metal organic framework (MOF)-derived porous copper/zinc bimetallic oxide catalyst was developed for the photoreduction of CO₂ to methanol at a very fast rate of 3.71 mmol g_cat⁻¹ h⁻¹. This kind of photocatalyst with high activity, selectivity and a simple preparation catalyst provides promising photocatalyst candidates for reducing CO₂ to methanol.

The rapid development of industrial society has been accompanied by the consumption of a large number of carbon-containing high-energy compounds, resulting in a large consumption of fossil fuels and a great burden on the natural carbon cycle [1,2]. In recent years, it has become a trend to convert CO₂ into high-energy carbon compounds using solar energy or electricity from other renewable sources [3,4,5]. Compared with other forms of clean energy, solar energy is favored because it is more convenient to use and requires no further energy consumption. At present, most photocatalytic CO₂ reduction products are carbon monoxide, formic acid, methane and other products of low industrial value [6,7,8]. As an important fine chemical, methanol is widely used in the preparation of fine chemicals such as formaldehyde, acetic acid and ethylene glycol, and can be used as extraction agent and solvent [9,10,11]. Therefore, using solar energy to convert CO₂ into methanol has high industrial value and scientific significance.

Metal organic frameworks (MOFs) are new porous materials whose composition and morphology can be precisely regulated by adjusting the types of metal and ligand [12,13,14,15,16]. This gives them wide application in the field of photocatalysis. Therefore, the derived metal compound materials also have adjustable structure, composition and application. Cuprous oxide, derived from copper-based MOFs, for example, degrades organic dyes by more than 99% under sunlight [17,18,19]. Cadmium sulfide derived from cadmium-base MOFs can effectively promote photocatalytic oxidation of phenyl methyl sulfide [20,21,22]. Bismuth-based MOFs can convert CO₂ into formic acid under photocatalysis [23,24,25]. Zinc-based MOFs can use sunlight to reduce CO₂ to methanol [26,27,28,29,30,31]. Even so, in many reports of photocatalytic reduction of CO₂ into methanol, the rate of methanol generation is relatively low. Both sulfides and selenides are known to be highly toxic, so it is of great importance to develop a green and efficient catalyst for the photoreduction of carbon dioxide to methanol.

Herein, we design a novel porous bimetallic oxide material derived from copper and zinc bimetal MOFs for efficient photocatalysis of CO₂ to methanol. With the help of simulated sunlight, up to 3 mmol of methanol was generated. This copper/zinc oxide material has a rich mesoporous structure that facilitates the exposure of more active sites, the absorption and mass transfer of CO₂, and the separation of products.

The catalyst was prepared by a simple method (Supporting Information). Appropriate amounts of copper acetate monohydrate, zinc acetate dihydrate, and terephthalic acid were dissolved in dimethylformamide and reacted at room temperature for three days. After centrifugation, washing and drying, the product was calcined in a muff furnace at 350 °C for 1 h to obtain the desired Cu/Zn bimetallic oxide catalyst. Figure 1 shows the characterizations of the catalyst synthesized with Zn content of 48.2 wt%, which was determined by inductively coupled plasma-atomic emission spectrometry (ICP-AES). It presents a fish-scale structure with many holes in each sheet, which is about 250 nm in diameter, as revealed by scanning electron microscopy (SEM) and transmission electron microscopy (TEM) (Figure 1a,b). A high-resolution TEM (HRTEM) image reveals two different lattices with two interplanar distances of 0.198 nm and 0.202 nm (Figure 1c), corresponding to the (11-1) plane of face-centered tenorite and the (101) plane of zincite, respectively. The X-ray diffraction (XRD) pattern (Figure 1d) presents several sharp diffractions at 31.8, 34.4, 35.4, 36.3 and 38.7°. The diffractions 31.8°, 34.4° and 36.3° correspond to (100), (002) and (101) planes of zincite, respectively (Simulated ZnO, PDF #36-1451). At the same time, the diffractions at 35.4° and 38.7° correspond to the (002) and (11-1) planes of tenorite, respectively (Simulated CuO, PDF #48-1548). The above two experimental results show that the catalyst is composed of ZnO and CuO with a ratio of 3:2.

X-ray photoelectron spectroscopy (XPS) analysis was performed to inspect the surface composition and chemical states of the double metal oxide sample. As shown in Figure S1 the high-resolution of the Zn 2p spectrum was fitted to two binding energy peaks at 1022.28 eV and 1045.35 eV corresponding to Zn 2p_3/2 and Zn 2p_1/2, respectively. The spin-orbit coupling value of 23.07 eV indicates that the Zn element in the hybrid structure existed mainly in the form of Zn²⁺ of the wurtzite ZnO structure. In addition, The Cu 2p spectrum in Figure S1 shows the two peaks at 933.65 eV and 953.50 eV corresponding to Cu 2p_3/2 and Cu 2p_1/2 of Cu²⁺, respectively.

The porosity properties of the catalyst was determined by N₂ adsorption-desorption measurement. The Brunauer-Emmett-Teller (BET) surface area and pore volume of catalyst are 94.3 m² g⁻¹ and 0.18 cm³ g⁻¹, respectively, with a hysteresis loop at a relative pressure (P/P₀) of 0.42–0.97 (Figure 2a), which is indicative of a typical IV profile with a classical H3-type hysteresis loop, evidencing the presence of mesopores resulting from the highly cross-linked linear particles. The mesopore size distribution of the catalyst, calculated from Barrett-Joyner-Halenda (BJH) analysis, shows a narrow pore size distribution centered at ~4.0 nm (Figure 2b). These results reveal that a large number cross-linked mesopores are formed and distribute rather evenly throughout the sample, which resulted from the removal of CO₂ at high temperatures. In view of its unique pore structure, CO₂ adsorption was characterized. For comparison, we compared copper oxide and zinc oxide prepared by copper and zinc MOFs with the same ligand, and named them, respectively, C350 and Z350, also named as double metal oxide CZ350. The absorption threshold of CZ350 and C350 is higher than that of Z350, indicating that CZ350 and C350 have stronger adsorption capacity for CO₂. When the partial pressure reaches 0.3, CZ350 begins to exceed C350, its adsorption ability is 3.5 times of Z350 and 1.5 times of C350, respectively, which indicates that the absorption of CO₂ by copper oxide and zinc oxide is not independent, but has a certain synergism. The results show that CZ350 is more conducive to the adsorption of CO₂ and has stronger adsorption ability.

The photocatalytic performance of CZ350 for CO2 reduction reaction (referred to CO₂RR) was detected. Firstly, to gain further insight into the interfacial charge transport behavior of the photocatalysts, an electrochemical impedance spectroscopy (EIS) experiment was performed. A single fluid cell with 0.05 M NaSO₄ electrolyte was used and the frequencies are in the range of 10⁻¹–10⁵ HZ. As shown in Figure 3a, CZ350 exhibits earlier onset potential in 0.5 M NaSO₄ solutions than C350 and Z350, implying its higher photocatalytic CO₂RR performance. The photocurrent density of CZ350 is ~2 times higher than that of Z350 (Figure 3b), which can be ascribed to the longer lifetime and faster transfer of photogenerated electron-hole pairs. The above results confirm that CZ350 has more efficient electron-hole pair separation than C350 and Z350. The activities of the three catalysts for photocatalytic CO₂RR were evaluated in a round-bottom flask. Before the test, a half hour’s supply of nitrogen gas was pumped into the flask to expel the air, followed by a half hour’s supply of high-purity CO₂ gas to ensure that the entire test was conducted in a CO₂ atmosphere. The gaseous and liquid products were analyzed by gas chromatography (GC) and 1H nuclear magnetic resonance (1H NMR) spectroscopy, respectively. There were only liquid products (CH₃OH) and no gas product was detected (Figure 3c), from which a clear peak of the -CH₃ group for CH₃OH can be seen. With the extension of reaction time, CZ350 still showed a strong catalytic capacity without obvious attenuation (Figure 3d). This shows that the catalyst has high stability in the catalytic process. To further measure the stability of the catalyst, we repeatedly used the catalyst. As can be seen from Figure 3e, after five cycles, the catalyst had only a little attenuation and still maintained a high catalytic activity. The XRD patterns before and after the cyclic reaction indicated that the chemical composition of the catalyst remained stable with the increase of the cycle times, and no obvious change occurred (Figure 3f). To the best of our knowledge, this catalyst is among the most effective Zn-based catalyst for photocatalytic CO₂RR to methanol (for more details see Table S1).

In order to explore the influence of different metal ratios on catalytic properties, we synthesized copper-oxide bimetallic MOFs composed of different metals and prepared bimetallic oxides with different proportions. According to the molar ratio and calcination temperature of the added copper and zinc metal salts, the samples were named as CZ350-64, CZ350-55 and CZ350-28, respectively. As shown in Figure 4a–c, with the increase of zinc salt content, the prepared samples gradually change from solid block structure to thin flake porous structure. This may be attributed to the fact that metallic zinc is more active than metallic copper, and the production rate of zinc oxide is higher than that of copper oxide, which is more conducive to the rapid dissociation of MOF and the formation of a porous structure. XRD patterns of the three samples show that with the increase of the amount of zinc salt added, the peak attributed to zinc oxide in the sample increased, indicating that the content of zinc oxide in the sample increased (Figure 4d). The Zn contents in those samples were determined by ICP-AES to be 40.2 wt%, 51.3 wt% and 58.7 wt%. This result is just consistent with the XRD test result. The photocatalytic performance of the three samples to CO₂RR was detected. All these samples can photocatalyze CO₂RR to produce methanol (Figure 4e). As shown in Figure 4f, the production rate of methanol increases with the increase of zinc content, which may be due to the fact that with the increase of zinc oxide content, the catalyst’s absorption efficiency of sunlight improved, leading to higher photoelectron production efficiency. It is worth noting that the formation rate of methanol reaches its maximum at the zinc content of 48.2% and then decreases, which was probably because an appropriate amount of copper oxide can effectively convert photoelectrons absorbed by zinc oxide into intermediates in the methanol formation process during the catalytic process. If the content of copper oxide is too low, the photoelectrons absorbed by zinc oxide cannot be effectively transferred to the reaction intermediate. By contrast, the production of the intermediates slowed down due to the decrease of the content of copper oxide, leading to the decrease of the formation rate of methanol. The results show that the absorption and transfer of photoelectrons in the process of photocatalytic CO₂RR to methanol does not depend solely on ZnO. The addition of proper proportion of copper oxide into the catalyst can transfer photoelectrons to the reaction intermediate more effectively and promote the formation of methanol. These results indicate that the catalytic process is determined by the cooperating effect of the copper oxide and zinc oxide, which is also consistent with the result of CO₂ adsorption.

Based on the above results, the plausible reaction mechanism was proposed that CuO possesses typical p-type character with an energy band gap of 1.4 eV. Based on the calculated conduction band gap between ZnO and CuO is 0.43 eV. Electrons and holes are generated and separated in the conduction band (CB) and valence band (VB) of CuO when the photocatalyst was photoexcited. Then, the excited electrons of CuO move directly to the CB of ZnO due to the lower CB of ZnO than that of CuO. The consequent increased excess of electrons of ZnO induces the Fermi level of ZnO to shift toward the negative direction, which provides the required overvoltage for efficient CO₂ reduction reaction. The possible reason for the high activity of the as-synthesized CZ350 can be proposed. First, the bimetallic oxides derived from Cu/Zn bimetallic MOFs have a large specific surface area and unique mesoporous structure, which favors the exposure of accessible active sites, CO₂ adsorption and diffusion of CO₂ and the photocatalytic product. In addition, in the process of photocatalytic reduction of CO₂, copper oxide absorbs CO₂ more easily than zinc oxide, which is more conducive to the activation of CO₂. Meanwhile, because zinc oxide is a semiconductor while copper oxide is a conductor, CuO is more efficient in electron transfer. In the catalytic process, CuO can transfer photoelectrons absorbed by ZnO to the reaction intermediate more efficiently, avoiding photoelectron annihilation due to slow transmission. The synergizing effect of CuO and ZnO avoids the slow photoelectron absorption and the slow catalytic efficiency of part of CO₂ due to the lack of photoelectrons and avoids the waste of light energy caused by the slow photoelectron transfer. Owing to these unique features, the catalyst can efficiently reduce CO₂ photocatalysis to methanol under simulated sunlight.

Conclusions

In conclusion, we have developed a novel MOF-derived bimetallic oxide for photocatalytic CO₂ reduction. The catalyst has a large specific surface area and a unique mesoporous structure, which is conducive to the adsorption and activation of CO₂ and the release of the product. Based on the above characteristics, the catalyst can effectively convert CO₂ into methanol with a maximum generation rate of 3.71 mmol⁻¹ g_cat⁻¹ h⁻¹. This study provides a new strategy for producing methanol from photocatalytic CO₂RR by combining it with nano-engineering.