One-Pot Method Synthesis of Bimetallic MgCu-MOF-74 and Its CO2 Adsorption under Visible Light.

A magnesium-based metal-organic framework (Mg-MOF-74) exhibits excellent CO2 adsorption under ambient conditions. However, the photostability of Mg-MOF-74 for CO2 adsorption is poor. In this study, Mg x Cu1-x -MOF-74 was synthesized by using a facile "one-pot" method. Furthermore, the effects of synthesis conditions on the CO2 adsorption capacity were investigated comprehensively. X-ray diffraction, Fourier transform infrared, scanning electron microscopy, thermo gravimetric analysis, inductively coupled plasma atomic emission spectroscopy, ultraviolet-visible spectroscopy and photoluminescence spectroscopy, and CO2 static adsorption-desorption techniques were used to characterize the structures, morphology, and physicochemical properties of Mg x Cu1-x -MOF-74. CO2 uptake of Mg x Cu1-x -MOF-74 under visible light illumination was measured by the CO2 static adsorption test combined with the Xe lamp. The results revealed that Mg x Cu1-x -MOF-74 exhibited excellent photocatalytic activity. Furthermore, the CO2 adsorption capacity of Mg x Cu1-x -MOF-74 was excellent at a synthesis temperature and time of 398 K and 24 h in dimethylformamide (DMF)-EtOH-MeOH mixing solvents, respectively. Mg x Cu1-x -MOF-74 retained a crystal structure similar to that of the corresponding monometallic MOF-74, and its CO2 uptake under visible light was superior to that of the corresponding monometallic MOF-74. Particularly, the CO2 uptake of Mg0.4Cu0.6-MOF-74 under Xe lamp illumination for 24 h was the highest, up to 3.52 mmol·g-1, which was 1.18 and 2.09 times higher than that of Mg- and Cu-MOF-74, respectively. The yield of the photocatalytic reduction of CO2 to CO was 49.44 μmol·gcat -1 over Mg0.4Cu0.6-MOF-74 under visible light for 8 h. Mg2+ and Cu2+ functioned as open alkali metal that could adsorb and activate CO2. The synergistic effect between Mg and Cu metal strengthened Mg x Cu1-x -MOF-74 photostability for CO2 adsorption and broadened the scope of its photocatalytic application. The "bimetallic" strategy exhibits considerable potential for use in MOF-based semiconductor composites and provides a feasible method for catalyst design with remarkable CO2 adsorption capacity and photocatalytic activity.

Introduction

Metal–organic framework (MOF) materials, featuring highly effective structures (i.e., ultrahigh surface areas, extraordinary porosity, a highly ordered porous structure, and homogeneous active sites), exhibit considerable potential for use in carbon capture[1−3] and sequestration[4−6] (CCS) and heterogeneous catalysis,[7−9] especially in photocatalysis.[10−12] Among the numerous MOFs, Mg-MOF-74 (or CPO-27) exhibits the best CO2 uptake in low pressures because of its open metal sites.[13,14] Yazaydin et al. analyzed the CO2 adsorption capacities of M-MOF-74 (M = Ni, Zn, Co, and Mg) with simulation and experimental methods at 1 bar and 298 K and proved that the CO2 uptake of Mg-MOF-74 was highest, up to 8 mmol·g–1.[15] Bao et al. proved that the CO2 uptake of Mg-MOF-74 was considerably higher than that of zeolite 13X, up to 8.61 mmol·g–1. Mg-MOF-74 exhibits open metal sites. Compared with Na+ in zeolite 13X, unsaturated Mg2+ has a smaller ionic radius and a larger ionic valence, which results in stronger adsorbate–metal interactions.[16] MOFs with excellent CO2 adsorption capacity have been used as co-catalysts in semiconductor composites for CO2 photocatalytic reduction.[17,18] Zhao et al. reported a novel strategy for incorporating Mg-MOF-74 as a co-catalyst to prepare Zn2GeO4/Mg-MOF-74 and revealed that the CO2 photocatalytic reduction of the semiconductor composites was considerably higher than that of single semiconductor materials. The productive rate of CO (12.94 μ mol·gcat–1) was 7 times higher than that of Zn2GeO4 (1.8 μmol·gcat–1).[19] The high concentration of open alkaline metal (Mg2+) of Mg-MOF-74 was beneficial to the CO2 adsorption and activation, which improved electron transfer from Zn2GeO4 to Mg-MOF-74 and effectively inhibited the excited electron–hole recombination of composites. Wang et al. revealed that Mg-MOF-74 (CPO-27)/TiO2 exhibited enhanced photocatalytic reduction of CO2 to CO and CH4. Furthermore, Mg-MOF-74 exhibited excellent CO2 adsorption capacity and open alkaline metal sites, which enhanced the photocatalytic performance of the composites.[20] Furthermore, although the CO2 uptake of the Cu-based MOF was lower than that of Mg-MOF-74, the semiconductor composite of the Cu-based MOF exhibited excellent CO2 photocatalytic reduction and physicochemical stability. Li et al. prepared a Cu3(BTC)2@TiO2 semiconductor composite, and the yield of photocatalytic reduction of CO2 to CH4 was up to 2.64 μmol·gcat–1·h–1. Cu3(BTC)2 functioned as a co-catalyst for Cu3(BTC)2@TiO2 that exhibited excellent stability of the morphology and compositions after CO2 photocatalytic reduction.[21] A strategy using Mg-MOFs as a co-catalyst and with excellent CO2 adsorption capacity is required to enhance the photocatalytic performance of composite semiconductors.[22] However, the CO2 uptake of Mg-MOF-74 is poor under ambient conditions. The existing studies of Mg-MOF-74 showed improved CO2 uptake under humid conditions (i.e., water stability of CO2).[23,24] The CO2 uptake of Mg-MOF-74 under long-time illumination (i.e., photostability of CO2) is yet to be discussed.

In bimetallic MOFs, two inorganic metal nodes are used to integrate two monometallic MOFs. Bimetallic MOFs considerably outperform their corresponding monometallic MOFs.[25−27] Although bimetallic Mg-MOF-74 is widely used in thermocatalysis[28,29] and adsorption,[30,31] its use in CO2 photocatalytic reduction is limited.[32] Guo et al. prepared bimetallic NiMg-MOF-74 catalysts by using a solvothermal method and revealed that CO2 can be photocatalytically reduced into formate over NiMg-MOF-74 under extremely harsh flue gas conditions. Furthermore, the excellent stability of NiMg-MOF-74 in the photocatalytic reaction was demonstrated through five cyclic photocatalytic experiments and X-ray diffraction (XRD) patterns after the photocatalytic reaction.[32] Consequently, based on the good CO2 uptake of Mg-MOF-74 and the photostability of the Cu-based MOF, a bimetallic strategy was proposed to strengthen Mg-MOF-74 photostability for CO2 adsorption.

We used the one-pot method to prepare a series of bimetallic MgCu1–-MOF-74 (x = 0.4, 0.2, and 0.17). The effects of the preparation conditions (i.e., temperatures, time, and solvents) and Mg/Cu molar ratio on CO2 adsorption were discussed. The relationship between the structure and CO2 adsorption under visible light illumination was investigated through inductively coupled plasma atomic emission spectroscopy (ICP-AES), XRD, Fourier transform infrared (FT-IR), scanning electron microscopy (SEM), thermo gravimetric analysis (TGA), and density functional theory (DFT). The photocatalytic performance of the MgCu-MOF was verified. In this study, an executable and competitive method was proposed for design of MOF-based semiconductor composites for excellent CO2 adsorption capacity and considerable photocatalytic activity.

Materials and Experiment

Materials

All organic and inorganic chemicals in this study were of commercially available analytical grade and were used without further purification. Magnesium nitrate hexahydrate [Mg(NO3)2·6H2O, 99.0%] and copper nitrate hydrate [Cu(NO3)2·3H2O, 99.0%] were purchased from Kermel. 2,5-Dihydroxyterephthalic acid (H4dhtp, 98.0%) and N,N-dimethylformamide (DMF, 99.8%) were purchased from Macklin. Anhydrous ethanol (EtOH, 99.7%), anhydrous methanol (MeOH, 99.8%), and 2-propanol (IPA, 99.7%) were purchased from Fuyu Fine Chemical Co., Ltd. (Tianjin, China).

Characterization

The crystal structure was identified using an X-ray diffractometer (7000S/L, Shimadzu) with Cu Kα radiation. The morphology was observed using a scanning electron microscope (S-4800, Hitachi) with a magnification of 2000–20,000 times and an acceleration voltage of 5 kV. The chemical structure and functional groups were determined using a FT-IR spectroscope (Tensor 27, Bruker); each spectrum was obtained from the acquisition of 32 scans from 4000 to 400 cm–1 with 2 cm–1. Thermogravimetry and derivative thermogravimetry (Mettler Toledo) were performed using a TGA analyzer. Elemental quantitative analysis was performed using an inductively coupled high-frequency plasma emission spectrometer (715-ES, Agilent). Before the test, the sample was dissolved completely with concentrated nitric acid and subsequently diluted to 250 mL with ultrapure water. The band gap width was measured using an ultraviolet–visible spectrophotometer (UV-2600, Shimadzu). The photoluminescence (PL) spectra were recorded on a fluorescence spectrophotometer (PL, F-4600, Hitachi). The excitation wavelength was set to 370–580 nm. The pore structures were determined by N2 adsorption–desorption isotherms at 77 K using the Micromeritics ASAP 2460 adsorption apparatus. All samples were as-treated by heating at 473 K for 10 h in a dynamic vacuum before measurement. The surface area and micropore diameter were calculated through DFT. Micropore volume and average micropore diameter were calculated using the Horvath–Kawazoe (H–K) method.

Synthesis of MgCu1–-MOF-74

To obtain MgCu1–-MOF-74, Mg(NO3)2·6H2O and Cu(NO3)2·3H2O were mixed in the molar ratio of 2:1 (Mg0.67Cu0.33), 1.5:1.5 (Mg0.5Cu0.5), and 1:2 (Mg0.33Cu0.67). Next, the synthesis method of MgCu1–-MOF-74 (x = 0.67, 0.5, and 0.33) was consistent with the synthesis of Mg-MOF-74 and Cu-MOF-74. Furthermore, Mg0.5Cu0.5-MOF-74 was synthesized at the three temperatures (398, 408, and 418 K), at three times (12, 24, and 36 h), and in three solvents (A, B, and C). A is DMF-MeOH-EtOH, B is DMF-EtOH-IPA, and C is DMF-MeOH-IPA.

CO2 Adsorption Measurement

The CO2 static adsorption–desorption isotherms of activated samples were obtained using the Micromeritics ASAP 2460 adsorption apparatus, which were measured at 298 K and gas pressure up to 760 mm Hg. The temperature control system was achieved using a Dewar bottle with a circulating sleeve connected to a thermostatic bath.

Photostability of MgCu1–-MOF-74 for CO2 Adsorption Measurement

The activated samples were treated using a Xe lamp (CEL-HXL300, China Education Au-Light Co., Ltd., Beijing) at three times (12, 24, and 36 h), and the CO2 uptake was evaluated. The test method was the same.

Photocatalytic Activity of MgCu1–-MOF-74 for CO2 Measurement

A 300 W Xe lamp was used as the visible light. The samples (20 mg) and deionized water (2 mL) were both placed in a closed reactor (CEL-HPR100T) with a sapphire window and temperature control system. Before the reaction, the reactor was blown with pure CO2 with a flow rate of 20 mL·min–1 for 15 min to ensure elimination of all air. The pressure and temperature of the sealed reaction system were increased to 1 MPa and 150 °C, respectively, and stirred in the dark for 30 min to ensure that the catalysts reached the adsorption–desorption equilibrium. After a certain time, the gases produced were analyzed and quantified through BFRL SP-3510 gas chromatography (GC) and quantified with a TCD and FID.

Results and Discussion

Metal Composition Measurement of As-Prepared Samples

To confirm the metal composition of samples, ICP-AES analyses were conducted (Table S1). According to the ICP-AES results, Mg and Cu coexist in MgCu1–-MOF-74. With the increase in the Cu content, the Mg/Cu molar ratio decreased. The results are consistent with the experimental system and proved that MgCu1–-MOF-74 with various Mg/Cu metal ratios can be synthesized. However, the Cu content in the MgCu1–-MOF-74 structure was considerably higher than the Mg content. This phenomenon can be attributed to the higher stability of Cu-MOF-74 than that of Mg-MOF-74. Cu is more easily coordinated with the H4dhtp to form MOF-74 than Mg, and the synthesis temperature of Cu-MOF-74 (353 K) is lower than that of Mg-MOF-74 (398 K).[33,34] Therefore, Cu-MOF-74 is preferentially formed in the synthesis process of MgCu1–-MOF-74, and Mg2+ replaces part of Cu2+ into the lattice. Because the Cu–O bond length is longer than that of the Mg–O bond, the coordination of the MgCu1–-MOF is distorted and deformed because of the Jahn–Teller effect.[35] The coordination distortion affects the synthesis of MgCu1–-MOF-74, which results in a lower Mg content than Cu. The result is consistent with TG analysis mentioned previously.

According to ICP-AES analysis, the Mg/Cu metal ratio was adjusted to Mg0.4Cu0.6, Mg0.2Cu0.8, and Mg0.17Cu0.83. Furthermore, all structures and properties of MgCu1–-MOF-74 in the study are discussed with reference to the adjusted Mg/Cu molar ratio.

Influence of Synthesis Conditions and the Mg/Cu Molar Ratio on the Crystal Structure and Morphology

The crystal structure of samples under four conditions of temperatures, time, solvents, and the metal molar ratio was confirmed through XRD (Figure S1). The XRD pattern of Mg0.2Cu0.8-MOF-74 under three conditions (temperatures, time, and solvents) exhibits two diffraction peaks at 2θ = 6.8° and 11.8°, which correspond to the (210) and (300) crystal faces of Mg-MOF-74, respectively. Furthermore, the pattern is same as that of Mg-MOF-74.[27,36,37] A novel diffraction peak at 2θ = 10.4° (Figure S1b) did not belong to MgO or CuO, which could be attributed to the impurity resulting from the coordination reaction of Mg, Cu, and DMF.[38] The XRD patterns of MgCu1–-MOF-74 were well consistent with those of Mg- or Cu-MOF-74 (Figure a). MgCu1–-MOF-74 revealed major diffraction peaks at 2θ = 6.8°, 11.8°, 17.3°, 21.9°, 24.8°, 25.6°, 27.4°, 31.4°, and 42°, which are consistent with those of reported MOF-74.[39,40] Replacing Mg with Cu does not affect the crystal structure of Mg-MOF-74, and MgCu1–-MOF-74 is successfully synthesized.

Figure 1

(a) Powder XRD patterns and (b) FT-IR spectra of Mg-, Cu-MOF-74, and MgCu1–-MOF-74.

FT-IR was performed to detect the surface functional groups of samples under four synthesis conditions of various temperatures, time, solvents, and metal molar ratios (Figure S1d–f). According to the coordination characteristics of the MOF, all vibrations can be distinguished into two distinctive regions.[41,42] The characteristic vibrations above the 700 cm–1 region are organic ligands and those below 700 cm–1 regions mainly belong to metal centers of Mg and Cu.[42] Several sharp and clear absorption peaks were observed at 1522, 1420, 890, and 823 cm–1, which were ascribed to ν(−C=O), ν(−COO−), and ν(C–H) of benzene rings in H4dhtp (Figure S1d).[28,42] The sharp and weak peak at 1240 cm–1 is attributed to the stretching vibration of the C–O band of the phenolate group. Mg0.2Cu0.8-MOF-74 obtained under various preparation conditions exhibits the aforementioned characteristic peaks (from Figure S1d–f),[38] which proves that Mg0.2Cu0.8-MOF-74 was successfully synthesized.

Notably, the functional groups of Mg0.2Cu0.8-MOF-74 were primarily affected by the synthesis temperature and less affected by synthesis time and solvents. The characteristic peaks at 889 and 823 cm–1 of Mg0.2Cu0.8-MOF-74 nearly disappeared at a synthesis temperature of 145 °C, which indicated that the synthesis temperature affected the coordination reaction between metal ions and organic ligands (Figure S1d).[38] The Cu metal ratio of MgCu1–-MOF-74 exceeded 50%, and the characteristic peaks of 1589 and 1191 cm–1 in the spectrum of Mg-MOF-74 transferred to 1552 and 1191 cm–1, respectively (Figure b). This phenomenon is attributed to the conjugated effect of MgCu1–-MOF-74. The C=O group conjugates with the C=C bond, and delocalization of the π-electrons occurred between two unsaturated bands. The double bond characteristics of C=O were reduced, which resulted in the shift of the absorption frequency toward a lower wave number.[36] Therefore, strongly coordinated olefin molecules existed in the metal centers of MgCu1–-MOF-74.

The morphology and structures of samples at three synthesis temperatures (398, 408, and 418 K) were observed through SEM (Figure S2a). The morphology structure of MgCu1–-MOF-74 was affected by synthesis temperature. Furthermore, Mg0.2Cu0.8-MOF-74 at 398 K exhibited a spherical crystal structure formed by the reaggregation of needle-like crystals in diameters less than approximately 1 μm (Figure S2a,b).[36] The morphology of Mg0.2Cu0.8-MOF-74 at 418 K exhibited flake crystal branches (Figure S2d). Mg0.2Cu0.8-MOF-74 at 408 K formed flake crystal branches (Figure S2c), which are slightly larger than the needle crystal branches of Mg0.2Cu0.8-MOF-74 at 398 K. The temperature can promote crystal growth; however, excessively high temperature leads to grain agglomeration.[43] With the increase in temperature, the dispersion degree of the attachment energy of each modified crystal surface increased, which resulted in various growth rates between the crystal surfaces.[44,45] The morphology of Mg0.2Cu0.8-MOF-74 at 418 °C was nonspherical. The morphology structure of MgCu1–-MOF-74 was also affected by the Mg/Cu molar ratio (Figure ). Mg-MOF-74 exhibited a chrysanthemum-like morphology formed by polyhedral prism crystal branches (Figure a,b). The spherical structures of Cu-MOF-74 consisted of needle-like crystal branches, and the average crystal is less than 5 μm (from Figure c–f). With the increase in the Cu content, the petal structure of MgCu1–-MOF-74 changed from prismatic petals to needle-like petals. Therefore, the average particle size of Mg0.17Cu0.83-MOF-74 with the lowest Mg/Cu molar ratio (nMg/nCu = 0.20) was close to 5 μm. MgCu1–-MOF-74 with a hydrangea-like morphology exhibited a structure denser than that of the corresponding monometallic MOF-74, which results in its specific surface area becoming larger than that of Cu-MOF-74 and its thermal stability being superior to that of Mg-MOF-74.

Figure 2

(a) SEM images of synthesized (b) Mg-, (g) Cu-, and (c,d) Mg0.4Cu0.6-; (e) Mg0.2Cu0.8-; and (f) Mg0.17Cu0.83-MOF-74.

Influence of the Mg/Cu Molar Ratio on Thermal Stability

Two main mass loss stages exist in the TG curves of MgCu1–-MOF-74 in a N2 atmosphere, which revealed the weight loss temperature of each sample (Figure a). The first mass loss stage started with heating and lasts until the start of the second stage at various temperatures. For Cu-MOF-74, the starting and finishing temperatures reduced to 308 and 475 °C, respectively. For Mg-MOF-74, the starting and finishing temperatures reduced to 267 and 609 °C, respectively. The mass loss temperatures of MgCu1–-MOF-74 in the second stage were between those of Mg- and Cu-MOF-74. In the first stage, the mass loss was attributed to the removal of adsorbed water, gas molecules, and residual solvents, such as methanol, ethanol, or DMF.[27,46] The weightlessness in the second stage corresponds to the collapse of the framework structure, which disintegrated organic ligands and formed metal oxides.[35,47] The TG curve of Mg-MOF-74 reveals that the thermal stability of Mg-MOF-74 was not satisfied. However, the thermal stability of MgCu1–-MOF-74 was enhanced considerably. The considerable difference in the thermostability of MgCu1–-MOF-74 could be attributed to the synergy between Mg and Cu.[48] To avoid the structural collapse of MgCu1–-MOF-74, its thermal treatments were strictly performed at less than 200 °C.

Figure 3

(a) TG (N2 atmosphere), (b) N2 adsorption–desorption isotherms, (c) pore size distribution curves, (d) UV–vis DRS, (e) Eg, and (f) PL spectra of MgCu1–-MOF-74.

Influence of the Mg/Cu Molar Ratio on Pore Structures

To investigate the influence of the Mg/Cu molar ratio on the pore structures of MgCu1–-MOF-74, N2 adsorption–desorption isotherms of Mg-, Cu-, and MgCu1–-MOF-74 at 77 K were measured (Figure b,c). MgCu1–-MOF-74 exhibited a typical type I isotherm and type H3 hysteresis loops, which indicated that both microporous and mesoporous structures were the same. The micropores of MgCu1–-MOF-74 were predominantly between 0.84 and 1.70 nm, which is consistent with the previous report (Figure c).[24,36] The surface area of Mg0.4Cu0.6-MOF-74 calculated according to the DFT method was 1561.96 m2·g–1, which was approximately 3.18% higher than that of Mg-MOF-74 and 50.96% higher than that of Cu-MOF-74 (Table ). The micropore volume of Mg0.4Cu0.6-MOF-74 calculated according to the H–K method is the highest among MgCu1–-MOF-74 samples, at approximately 0.39 cm3·g–1. The surface area and micropore volume of Mg0.17Cu0.83-MOF-74 were the lowest. The surface area and micropore size of MgCu1–-MOF-74 decreased with the increase in the Cu content, which is consistent with the results of the ICP-AES analysis.

Table 1

Pore Structure Analyses of MgCu1–-MOF-74

sample	total Area in porea, m2/g	Langmuir specific surface area, m2/g	micropore volumeb, cm3/g	average pore sizeb, nm
Mg-MOF-74	1513.75	1448	0.45	0.75
Mg0.4Cu0.6-MOF-74	1561.96	1278	0.39	0.77
Mg0.2Cu0.8-MOF-74	1235.0	1061	0.32	0.76
Mg0.17Cu0.83-MOF-74	1012.50	978	0.27	0.77
Cu-MOF-74	1034.58	989	0.28	0.77

Total area in the pore calculated by the DFT method.

Micropore volume and average pore size calculated by the H–K method.

Influence of the Mg/Cu Molar Ratio on Optical Properties

UV–vis spectra were analyzed to confirm the optical performance of MgCu1–-MOF-74 (Figure d). The absorption spectrum of Mg-MOF-74 exhibited a strong absorption in the wavelength range of 200–413 nm, and the light absorption decreased considerably in the visible region (above 420 nm). Light absorption capacities of Cu-MOF-74 and MgCu1–-MOF-74 are better than those of Mg-MOF-74. Their absorption spectra revealed a stronger absorption in the wavelength range 200–530 nm but decreased slightly above 530 nm (Figure d). With the increase in the Cu content in MgCu1–-MOF-74, its absorption band edges exhibited a red shift with enhanced visible light absorption. The light absorption capacity of MgCu1–-MOF-74 was stronger than that of Cu-MOF-74 and Mg-MOF-74 in the UV region. However, the light absorption capacity of Mg0.4Cu0.6-MOF-74 and Mg0.2Cu0.8-MOF-74 was lower than that of Cu-MOF-74 in the visible region. The visible light absorption capacity of Mg0.17Cu0.83-MOF-74 above 480 nm was highest because the absorption spectrum of Cu-MOF-74 exhibited strong absorption in both UV and visible regions. Replacing the Mg of Mg-MOF-74 with Cu can enhance the light absorption capacity of MgCu1–-MOF-74 in the visible region. Moreover, Cu and Mg exhibit distinct coordination structures with H4dhtp in the process of synthesizing MgCu1–-MOF-74, which affects their visible light absorption ability. The band gap energy (Eg) was calculated from the UV absorption spectrum of MgCu1–-MOF-74. The Kubelka–Munk transformation spectra of MgCu1–-MOF-74 were estimated, and the tangent line was drawn and calculated by using software. The results revealed that the Eg values of Mg-, Mg0.4Cu0.6-, Mg0.2Cu0.8-, Mg0.17Cu0.83-, and Cu-MOF-74 were 2.628, 1.810, 1.718, 1.571, and 1.671 eV, respectively (Figure e). The absorption band edge of MgCu1–-MOF-74 exhibited a red shift with the decrease in Eg, and the visible light absorption was enhanced. Here, the Eg of Mg0.17Cu0.83-MOF-74 was the lowest, which was attributed to its highest response properties in visible light above 480 nm. The result revealed that replacing the Mg of Mg-MOF-74 with Cu can reduce the band gap and enhance the visible light response, which improves the photocatalytic activity.

PL emission spectroscopy was studied to verify the photocatalytic activity of MgCu1–-MOF-74 (Figure f). The emission peaks of the PL spectra were attributed to the recombination of excited free carriers. The emission intensities of Mg0.4Cu0.6-, Mg0.2Cu0.8-, and Mg0.17Cu0.83-MOF-74 were all lower than those of Mg- and Cu-MOF-74. With the increase in Cu, the PL peak intensities of MgCu1–-MOF-74 decreased considerably. Replacing the Mg of Mg-MOF-74 with Cu can inhibit the recombination of electron–holes and enhance photocatalytic activity because the lower intensity of PL indicates higher separation efficiency of electron–hole pairs, that is, the superior photocatalytic activity.[12,49,50]

Influence of Synthesis Conditions on CO2 Adsorption Capacity

To confirm the effect of synthesis conditions on the CO2 uptake of MgCu1–-MOF-74, the CO2 uptake of S at 298 K and 1 bar under three synthesis conditions (time, temperatures, and solvents) was tested (Table S2). The CO2 uptake of S-398-24-A (i.e., 398 is temperature, 24 is time, and A is the solvent) was the highest, up to 3.24 mmol·g–1. Thus, the best synthesis condition was 398 K, 24 h, and in A.

To verify that MgCu1–-MOF-74 was successfully synthesized, we obtained the mechanical combination of two corresponding monometallic MOF-74 in a molar ratio of 2:1 and tested CO2 uptake (Figure a). The CO2 uptake of Mg0.4Cu0.6-MOF-74 outperformed that of the mechanically mixed MOF-74(M), which proved that MgCu1–-MOF-74 was synthesized successfully. The synergistic effects originating from two adjacent Mg and Cu metals considerably improved the CO2 adsorption capacity.[28]

Figure 4

(a) CO2 adsorption–desorption isotherms of Mg0.4Cu0.6- and MOF-74(M), (b) CO2 adsorption–desorption isotherms of MgCu1–- at 298 K and 1 bar, (c) CO2 uptake of MgCu1–- treated using the X-lamp for 12, 24, 36, and 48 h, and (d) CO2 decrease rate of MgCu1–-MOF-74.

Evaluation of the Photostability of MgCu1–-MOF-74 for CO2 Adsorption

To investigate the effect of the Mg/Cu molar ratio on the photostability of MgCu1–-MOF-74 for CO2 adsorption, the adsorption–desorption isotherms at 298 K and 1 bar were compared to those obtained by the static volumetric CO2 adsorption method (Figure b). The CO2 uptake of MgCu1–-MOF-74 was higher than that of Cu-MOF-74 but lower than that of Mg-MOF-74. Specifically, the CO2 uptake of Mg0.4Cu0.6-MOF-74 was the highest, up to 4.58 mmol·g–1, and that of Mg0.17Cu0.83-MOF-74 was the lowest, up to 1.52 mmol·g–1. This phenomenon could be attributed to the surface area and the pore volume of Mg0.4Cu0.6-MOF-74, which is consistent with the pore structure analysis. Moreover, the CO2 adsorption capacity of MgCu1–-MOF-74 was influenced by the Mg content. With the decrease in the Mg content, the CO2 uptake of MgCu1–-MOF decreased, which was attributed to the CO2 adsorption heat of Mg ions being larger than that of Cu ions in a low-pressure region.[35]

To investigate the effect of the Mg/Cu molar ratio on the chemical stability and photostability of MgCu1–-MOF-74, the CO2 uptake of the MgCu1–-MOF-74 treated using a Xe lamp was measured by using the static volumetric CO2 adsorption method (Figure c). The CO2 uptake of MgCu1–-MOF-74 decreased with the increase in the illumination time, and the CO2 uptake of those treated using an X-lamp for more than 24 h was the lowest and stable. Furthermore, the CO2 uptake values of Mg-, Mg0.4Cu0.6-, Mg0.2Cu0.8- Mg0.17Cu0.83-, and Cu-MOF-74 treated by using the X-lamp for 24 h were 2.98, 3.52, 2, 1.69, and 1.68 mmol·g–1, respectively. To analyze the effect of the Mg/Cu molar ratio on the photostability of MgCu1–-MOF-74 for CO2 adsorption, the decrease rate in the CO2 uptake of those treated by the X-lamp for 12, 24, 36, and 48 h was quantitatively discussed (Figure c). The CO2 uptake decrease rate (100%) of MgCu1–-MOF-74 under long-term illumination is defined as followsHere, Y indicates the CO2 decrease rate; Cn indicates the CO2 uptake of MgCu1–-MOF-74 treated using the X-lamp for 12, 24, 36, and 48 h; and C0 indicates the CO2 uptake of MgCu1–-MOF-74.

Mg-MOF-74 treated using the X-lamp for 48 h exhibited the most severe decrease in CO2 uptake, and the CO2 uptake decrease rate was as high as 65.59%, which is approximately 3 times that of Cu-MOF-74 (22.90%). The CO2 uptake decrease rate of MgCu1–-MOF-74 treated using the X-lamp for 36 h reached photostability and that of Mg0.4Cu0.6-, Mg0.2Cu0.8-, and Mg0.17Cu0.83-MOF-74 was 29.34, 45.42, and 35.51%, respectively (Figure d). The CO2 uptake of Mg-MOF-74 was the highest and reached 7.18 mmol·g–1. However, its photostability was the weakest (Figure c,d). The photostability of Cu-MOF-74 for CO2 uptake was the best. The CO2 uptake of Cu-MOF-74 was the lowest, only up to 2.09 mmol·g–1. MgCu1–-MOF-74 photostability of CO2 adsorption was superior to that of two corresponding monometallic MOF-74 (Figure c). Specifically, the photostability of Mg0.4Cu0.6-MOF-74 treated using the X-lamp for 24 h was the best, and its CO2 uptake and decrease rate were 3.52 mmol·g–1 and 23.13%, respectively. A strong relationship exists between photostability and metal composition of MgCu1–-MOF-74. Because of the synergistic effect between Cu and Mg, the photostability of MgCu1–-MOF-74 was enhanced considerably.[32] Furthermore, the stability of Mg-MOF-74 was highly affected by the M–O bond on the top of the metal–organic frame (M represents the metal species) and the metal center atoms (Mg).[31] The distortion and deformation of the coordination environment of Cu2+ (i.e., Jahn–Teller effect) lead to the contraction of M–O bonds and improved photostability of MgCu1–-MOF-74.[35]

Evaluation of the CO2 Photocatalytic Activity of Mg0.4Cu0.6-MOF-74

To confirm the potential application of Mg0.4Cu0.6-MOF-74 as a co-catalyst for MOF-based semiconductor composites, its CO2 photoreduction activity was investigated without any sacrificial agent (Table ). The main gaseous product of CO2 photoreduction in an aqueous system was CO. The yield of CO was 49.44 μmol·gcat–1 over Mg0.4Cu0.6-MOF-74 under visible light for 8 h. The yield of CO was only 6.26 μmol·gcat–1 over Mg0.4Cu0.6-MOF-74 in the dark, which is considerably lower than that under visible light. The result revealed that Mg0.4Cu0.6-MOF-74 exhibits an excellent CO2 photoreduction activity. Replacing the Mg of Mg-MOF-74 with Cu can considerably improve the CO2 photocatalytic activity, which is consistent with the UV and PL analyses.

Table 2

CO2 Photoreduction of Mg0.4Cu0.6-MOF-74 and Mg-MOF-74

samples	product	CO yield/(μmol·gcat–1)	experimental conditions
Mg0.4Cu0.6-MOF-74	CO	49.44	X-lamp (8 h)
		6.26	without X-lamp
Mg-MOF	CO	0	X-lamp (8 h)
		0	without X-lamp

Conclusions

MgCu1–-MOF-74 was successfully constructed at 398 K, 24 h, and in DMF-EtOH-MeOH solvents. The CO2 adsorption capacity of MgCu1–-MOF-74 was considerably affected by the temperatures, time, solvents, and Mg/Cu molar ratio. The photostability of MgCu1–-MOF-74 for CO2 adsorption was affected by the Mg/Cu molar ratio. The synergistic effect of adjacent Mg and Cu enhanced the CO2 adsorption capacity and photocatalytic activity. The CO2 uptake of Mg0.4Cu0.8-MOF-74 under the Xe lamp for 24 h was the best, up to 3.52 mmol·g–1. The yield of the photocatalytic reduction of CO2 to CO was 49.44 μmol·gcat–1 over Mg0.4Cu0.8-MOF-74 under visible light for 8 h. The stability of Mg-MOF-74 was primarily affected by the Mg–O bond on the top of the metal–organic frame (M represents the metal species) and the metal center atoms (Mg). Replacing Mg2+ with Cu2+ can result in asymmetric defects in the skeleton structure. The distortion and deformation of the coordination environment of Cu2+ (i.e., Jahn–Teller effect) lead to the contraction of Mg–O bonds, which enhanced Mg–O bond photostability. The bimetallic strategy is a feasible method for photocatalyst design to achieve remarkable CO2 adsorption and photocatalytic properties in the future.