In–Co–Zn/C–N catalysts derived from ZIFs for selective hydrogenation of CO₂ into methanol

Abstract

Recently, In₂O₃-based catalysts have shown great promise in methanol synthesis from CO₂ hydrogenation due to their high methanol selectivity. However, low CO₂ conversion restricts their application. Herein, we report In₂O₃ combined with zeolitic imidazolate framework (ZIF)-derived Co–Zn/C–N catalysts for the CO₂ hydrogenation reaction in a fixed-bed reactor. The catalysts were characterized by X-ray diffractometry, transmission electron microscopy, CO₂ and H₂ temperature desorption measurements, X-ray photoelectron spectroscopy, and in situ diffuse reflectance infrared Fourier transform spectroscopy. Compared to pure In₂O₃, In–Co–Zn/C–N-4 (PM) shows higher CO₂ conversion, with even higher methanol selectivity. A CO₂ conversion of 7.0% was achieved with a methanol selectivity over 77% and the space time yield (STY) of methanol was 3.3 mmol g_cat⁻¹ h⁻¹ under the conditions of H₂/CO₂ = 3 : 1, 2 MPa, 300 °C and GHSV = 6 L g_cat⁻¹ h⁻¹. The In–Co–Zn/C–N-4 (IMP) catalyst shows much lower methanol selectivity (46.6%) and stability due to the formation of Co₃InC_0.75. ZnO effectively weakens the interactions between Co and In, and the physically-mixed method further prevents the formation of Co₃InC_0.75. DRIFTS results were used to predict the possible methanol and CO production pathway through a formate intermediate for the In–Co–Zn/C–N catalyst.

---

1. Introduction

Methanol synthesis from CO₂ hydrogenation has recently attracted significant attention.¹ However, the preparation of highly active and stable catalysts for CO₂ hydrogenation into methanol is a key problem to solve. Catalyst research mainly focuses on two categories, i) metal-based catalysts, including modified Cu (ref. 2) and noble metals such as Au,³ Pd (ref. 4) and Pt,⁵ and ii) metal oxide catalysts with oxygen vacancies, such as In₂O₃ (ref. 6) and CeO₂.⁷ Co-Based catalysts have been widely investigated in the Fischer–Tropsch synthesis reaction. Although CO₂ conversion is quite high on Co-based CO₂ hydrogenation catalysts, the main product is methane due to its strong H₂ dissociation ability.⁸

CO₂ + 3H₂ → CH₃OH + H₂O, CO₂ + 4H₂ → CH₄ + 2H₂O

Bavykina et al. found that In/Co catalysts show high methanol selectivity. An indium oxide surface shows enriched oxygen vacancies upon the addition of Co, which are thought to be the active sites of In/Co catalysts in selective CO₂ conversion into methanol.^9,10 We also found that the appropriate proximity of In₂O₃ and Co/C–N is crucial for methanol synthesis. The impregnated 50% In₂O₃/Co/C–N catalyst exhibits 88.4% methanol selectivity at 300 °C and 2 MPa.¹¹ Aside from the combining of Co with In₂O₃, partial oxidation of Co/C–N into Co@Co₃O₄/C–N also increased methanol selectivity in our previous research. A decrease in surface metallic Co content weakens the H₂ dissociation ability of metallic Co. The oxygen vacancies on the Co₃O₄ surface also promote methanol production.¹² In conclusion, how to balance weak CO₂ adsorption and strong H₂ dissociative adsorption is a key challenge associated with Co-based CO₂ hydrogenation catalysts.

Metal–organic frameworks (MOFs), which are made up of metal ions and organic ligands, have been widely explored in various fields, especially electrocatalysis, due to their rich pore structures, high surface area and high thermal and chemical stability.^13–16 Among these MOF materials, zeolitic imidazolate frameworks (ZIFs) consisting of metal ions (Zn²⁺, Co²⁺) and imidazole are active catalysts for many reactions.¹⁷ ZIF materials can also act as precursors to prepare nitrogen-doped porous carbon supported metal or metal oxide catalysts under designed calcination conditions. For instance, Zhou et al. prepared Co/C–N catalysts via the pyrolysis of ZIF-67 for the synthesis of secondary amines.¹⁸ Li et al. prepared ZIF-67 derived Co-based porous carbon catalysts with cubic or rhombic dodecahedron morphology for the CO₂ methanation reaction.¹⁹ Zhou et al. exploited N-doped porous carbon incorporating Co nanoparticles and CoO species for the direct oxidation of alcohols to esters.²⁰

In light of these results, herein we explored ZIF-derived Co–Zn/C–N catalysts with or without In and their activity in methanol synthesis from the CO₂ hydrogenation reaction. The production selectivity varies according to different Zn/Co molar ratios. Moreover, the addition of In effectively prevents the production of CO and methane. The effect that the In addition method has on the catalytic performance was also thoroughly studied in detail. The optimized In–Co–Zn/C–N (PM) catalyst demonstrates high stability, where the space time yield (STY) of methanol is 3.3 mmol g_cat⁻¹ h⁻¹ (X_CO2 = 7.0%, S_MeOH = 77.7%) steadily within 50 h at 2 MPa and 300 °C.

2. Experimental section

2.1 Catalyst preparation

Preparation of core–shell ZIF-67@ZIF-8. In a typical process, Co(NO₃)₂·6H₂O (4.32 g) and 2-methylimidazole (9.75 g) were each separately dissolved in 150 mL of methanol. The homogeneous solution containing Co(NO₃)₂·6H₂O was then quickly dropped into the 2-methylimidazole solution and stirred for 10 min. The color of the solution changed from black to purple. Zn(NO₃)₂·6H₂O (2.23 g) and 2-methylimidazole (2.46 g) were then each dissolved separately in 150 mL of methanol. Then, the 2-methylimidazole solution was added into the above purple solution, and the resultant solution was stirred for 1 min. After that, the Zn(NO₃)₂·6H₂O in methanol solution was added to the abovementioned solution at the same speed, and the stirring of the resulting suspension was maintained for 30 min. The obtained purple suspension was centrifuged at room temperature for 24 h and then washed three times with methanol. ZIF-67@ZIF-8 was obtained by drying the purple solid at 60 °C for 24 h. Keeping the content of Co(NO₃)₂·6(H₂O) and 2-methylimidazole for ZIF-67 unchanged, the content of Zn(NO₃)₂·6H₂O and 2-methylimidazole were increased from 2.23 g and 2.46 g to 4.46 g and 4.93 g, 8.92 g and 9.86 g, and 17.84 g and 19.72 g. The obtained catalysts are named ZIF-67@ZIF-8-0.5, ZIF-67@ZIF-8-1, ZIF-67@ZIF-8-2, ZIF-67@ZIF-8-4, where the last number is the molar ratio of Zn to Co.

Preparation of the Co–Zn/C–N catalyst. The Co–Zn/C–N catalyst was prepared via the pyrolysis of the prepared ZIF-67@ZIF-8. The ZIF-67@ZIF-8 was heated from room temperature to 600 °C at a heating rate of 10 °C min⁻¹, and then kept at 600 °C for 5 h under a N₂ atmosphere. The as-prepared catalysts are named Co–Zn/C–N-0.5, Co–Zn/C–N-1, Co–Zn/C–N-2, and Co–Zn/C–N-4. For comparison, Co/C–N and Zn/C–N were prepared via the pyrolysis of ZIF-67 and ZIF-8 under the same conditions. C–N material was also synthesized via the calcination of ZIF-8 under a N₂ atmosphere. ZIF-8 was heated from room temperature to 600 °C at a heating rate of 10 °C min⁻¹, and then to 900 °C at a heating rate of 2 °C min⁻¹, before finally being held at 900 °C for 3 h.

Preparation of the In–Co–Zn/C–N catalyst. The In–Co–Zn/C–N-4 (IMP) catalyst was prepared via an incipient wetness impregnation method. 0.43 g of In(NO₃)₃·xH₂O was dissolved in 0.6 mL of deionized water, and was then added dropwise into 1.5 g of ZIF-67@ZIF-8-4. After impregnation for 12 h, the sample was dried at 100 °C for 10 h and calcined at 600 °C (10 °C min⁻¹) for 5 h under a N₂ atmosphere (20 mL min⁻¹). The weight percentage of In was 10%.

The In–Co–Zn/C–N-4 (PM) catalyst was prepared by physical mixing. In₂O₃ was prepared via the calcination of In(OH)₃ as before.¹⁹ 0.39 g of In₂O₃ powder was first dispersed in 1.0 mL of deionized water, then added dropwise into 1.5 g of ZIF-67@ZIF-8-4 with grinding. After this, the sample was dried at 100 °C for 10 h and calcined at 600 °C (10 °C min⁻¹) for 5 h under a N₂ atmosphere (20 mL min⁻¹). The weight percentage of In was the same as that of the In–Co–Zn/C–N-4 (IMP) catalyst.

2.2 Catalyst characterization

The powder X-ray diffraction (PXRD) patterns were recorded on an automated powder X-ray diffractometer (40 kV, 40 mA, Bruker-D8) using a Cu Kα radiation source (λ = 1.54056 Å) in the range of 10–80° with a step size of 2° min⁻¹.

Transmission electron microscopy (TEM) images of the samples were obtained using a FEI Tecnai G²20 instrument. The samples are prepared by directly suspending the catalyst in ethanol via ultrasonic treatment. A copper microscope grid covered with perforated carbon was dipped into the solution and then dried.

X-ray photoelectron spectroscopy (XPS) data were recorded on a Thermo VG scientific ESCA MultiLab-2000 spectrometer equipped with a monochromatized Al Kα source (1486.6 eV) at a constant analyzer pass energy of 25 eV. The binding energy was estimated to be accurate to within 0.2 eV. Co 2p, Zn 2p, N 1s and O 1s spectra were acquired, where the C 1s was set at 284.6 eV and taken as a reference for binding energy (BE) calibration. The software XPS PEAK was used to fit the XPS spectra using a mixture of Gaussian and Lorentzian functions in a 40/60 ratio and a Shirley background.

CO₂ temperature-programmed desorption (CO₂-TPD) measurements were performed using a BELCAT-II instrument from the MicrotracBEL Company. The catalyst was pretreated under a He atmosphere at 300 °C for 30 min and then cooled to 100 °C. Next, the samples were saturated with 10% CO₂/He at around 100 °C for 60 min and then flushed with He for 30 min to evacuate the surface adsorbed CO₂. Next, the catalyst was heated under a flow of He from 100 °C to 500 °C. Around 50 mg of catalyst was used at a heating rate of 10 °C min⁻¹ and a flow rate of 25 mL min⁻¹. The sample for the H₂-TPD experiments were pretreated under an Ar atmosphere, and then the samples were saturated with 10% H₂/Ar, where the other experiment conditions were the same as those used in the CO₂-TPD measurements.

In situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) was carried out on a Nicolet NEXUS 6700 spectrometer equipped with a liquid nitrogen-cooled mercury cadmium telluride (MCT) detector. The catalyst was pretreated using N₂ at 300 °C for 30 min, then cooled to 100 °C at a heating rate of 5 °C min⁻¹ and flow rate of 18 mL min⁻¹. Next, the flow rate was changed to 10 mL min⁻¹ and the background spectrum was scanned at 100 °C. Then, the catalyst was scanned in a flowing reactant gas mixture with a H₂/CO₂/N₂ ratio of 67.5/22.5/10.0 at 10 mL min⁻¹ at 100 °C for 1 h.

After that, the temperature was increased from 100 °C to 300 °C, with each temperature being maintained for 15 min.

2.3 Typical procedure of CO₂ hydrogenation

All catalytic tests were performed in a fixed-bed stainless steel mini-reactor (length of 53 cm and inside diameter of 8 mm). A mixed feed gas (22.5% CO₂, 67.5% H₂ and 10.0% N₂) was used without further purification. N₂ was added for interior label calculation. The pressure of the reactor was controlled using a backpressure regulator. For each experiment, 0.2 g of catalyst to be tested was mixed with 0.2 g of silica. The system was pressurized to 2 MPa using a mixed gas feed at 20 mL min⁻¹. The reaction for each catalyst was kept for around 50 h at the desired temperature. The gas products were analyzed online using an Agilent GC 3000 gas chromatograph equipped with thermal conductivity and flame ionization detectors. The liquid products were collected in a cold trap and analyzed offline using an Agilent GC 4890 gas chromatograph equipped with a flame ionization detector. The reaction parameters of CO₂ conversion (X_CO2), selectivity and STY are defined as below (where MeOH refers to methanol).

3. Results and discussion

The morphology of ZIF-67@ZIF-8 was characterized by TEM, as shown in Fig. S1,† wherein ZIF-67@ZIF-8 displays uniform rhombic dodecahedra that are 450–600 nm in size. The good morphology of the rhombic dodecahedra reveal their good crystallinity. After calcination at 600 °C for 5 h, the integrity of the rhombic dodecahedral shape is not destroyed (Fig. 1). The morphology of Co–Zn/C–N is much more distinct with an increase in the Zn/Co molar ratio. The metallic Co is well dispersed in the nitrogen doped carbon material.

Fig. 1 TEM images of Co–Zn/C–N.

The X-ray diffraction patterns of ZIF-67@ZIF-8 are shown in Fig. S2,† which confirm the ZiF-67 and ZIF-8 structures. The XRD patterns of Co–Zn/C–N (Fig. 2) show peaks with 2θ angles at 44.2°, 51.5°, 75.9°, which can be assigned to the (111), (200), and (220) crystalline planes of metallic Co (JCPDS Card No. 15-0806). The diffraction peaks of ZnO (JCPDS Card No. 36-1451) can also be observed for the Co–Zn/C–N-0.5 catalyst. However, the diffraction peaks of ZnO disappear, and the diffraction peaks of Co₃ZnC (JCPDS Card No. 29-0524) are observed with an increase in the Zn/Co molar ratio, which indicates a strong interaction between Co and Zn in the calcination process. The crystal size of metallic Co can be calculated as 9.7, 7.0, 7.1, and 9.3 nm for Co–Zn/C–N-0.5, Co–Zn/C–N-1, Co–Zn/C–N-2, and Co–Zn/C–N-4 respectively. PXRD patterns of the Co/C–N, Zn/C–N and C–N material are also shown in Fig. S3.† The diffraction peaks of metallic Co and ZnO can be observed in the patterns of the Co/C–N and Zn/C–N samples separately. There is only a broad peak at 24.3° ascribed to the (002) crystalline planes of graphite carbon in the pattern of the C–N sample. PXRD patterns of the spent Co–Zn/C–N catalysts are shown in Fig. S4.† The diffraction peaks of ZnO are observed after reaction for the Co–Zn/C–N-1 catalyst, which are not observed for the fresh Co–Zn/C–N-1 catalyst. The crystal size of the catalyst remains keeps constant after reaction, for instance, the crystal size of metallic Co is 9 nm both before and after reaction for the Co–Zn/C–N-4 catalyst.

Fig. 2 XRD patterns of Co–Zn/C–N.

Fig. 3 shows the XPS spectra of the Co–Zn/C–N catalysts. The binding energy distance between the peak Co 2p_3/2 and Co 2p_1/2 splitting is approximately 15.0 eV, which indicates the presence of a mixture of Co²⁺ and Co³⁺ (Fig. 3a).²¹ The weak satellite peaks at 786.8 eV and 802.6 eV can be attributed to Co²⁺ in the octahedral sites of CoO, which indicates the presence of abundant Co²⁺. The asymmetrical Co 2p_3/2 and 2p_1/2 signals of Co–Zn/C–N can be deconvoluted into three peaks corresponding to Co³⁺ and Co²⁺ and metallic Co. Two fitting peaks at 780.4 eV and 795.4 eV correspond to Co³⁺, the peaks at 782.2 eV and 797.3 eV can be ascribed to Co²⁺, while the peaks at 778.4 eV and 793.3 eV can be indexed to metallic Co.²² The presence of Co³⁺ and Co²⁺ is due to the surface of metallic cobalt nanoparticles in the catalyst being oxidized into cobalt oxides during their storage in air. The N 1s XPS spectrum (Fig. 3b) can be fitted into three peaks with BEs at 398.8, 400.3 and 401.2 eV, corresponding to pyridinic N, pyrrolic N and graphitic N, respectively.²³ The nitrogen content increases with an increase in the Zn/Co molar ratio. The surface molar ratio of Zn/Co is higher than the calculated number (Table 1), which probably indicates that Zn/C–N is mainly on the surface and that Co/C–N is mainly at the core of the catalyst.

Fig. 3 XPS spectra of Co–Zn/C–N: a) Co 2p and b) N 1s.

Table 1 XPS surface elemental analysis (molar percentage %)

Catalyst	Co	Zn	O	N	Zn/Co
Co–Zn/C–N-0.5	1.4	1.4	10.5	5.0	1.0
Co–Zn/C–N-1	1.1	1.7	10.3	5.1	1.5
Co–Zn/C–N-2	0.8	2.4	9.7	7.9	3.0
Co–Zn/C–N-4	1.0	5.3	8.4	14.8	5.3

---

CO₂-TPD profiles of Co–Zn/C–N (Fig. 4a) were measured to determine the basic properties of the catalysts. As can be seen from the data, the curves of the samples clearly drift upward to over 500 °C, which is probably due to the decomposition of a small amount of ZIF.²⁴ For the Zn/C–N catalyst, a broad peak appears in the data at around 170 °C, indicating weak basic sites. A similar desorption peak can be observed for other catalysts, but not for the C–N sample, which means that this CO₂ desorption peak can be ascribed to CO₂ adsorption on ZnO. Strong desorption can be observed at 470 °C for the C–N sample, which can be ascribed to the strong basic sites of N. In our catalytic activity experiments, the reaction temperature was set as 275 °C, so the strong basic sites mainly effect the catalyst activity and selectivity. Furthermore, compared with Zn/C–N and C–N, the strong basic sites move towards a lower temperature. The curves of Co–Zn/C–N-2 and Co–Zn/C–N-4 are similar to that of Zn/C–N, indicating that the catalyst surface is almost covered with Zn/C–N upon an increase in the ZIF-8 content. The strong basic sites of N element further move from 430 °C for Zn/C–N to 385 °C for the Co–Zn/C–N-1 catalysts. This is maybe due to the interactions between Co and ZnO weakening the strong basic sites.

Fig. 4 a) CO₂-TPD and b) H₂-TPD measurements of the Co–Zn/C–N catalysts.

The H₂-TPD data of the Co–Zn/C–N catalysts (Fig. 4b) show three major peaks: a low temperature peak (about 170 °C), a medium temperature peak (350–500 °C) and a high temperature peak (>500 °C), indicating the different adsorbed states of hydrogen. The low temperature peak is due to chemisorbed hydrogen on the surface of metallic Co, whereas the medium and high temperature peaks are maybe caused by the desorption of spillover hydrogen from the surface of metallic Co to the bulk.²⁰ Comparing Co–Zn/C–N-0.5 with Co/Zn/C–N-1, the H₂ desorption peak notably moves towards a low temperature. The peaks become less intense upon an increase in the content of Zn, which indicates that the surface of the metallic Co is partially covered with Zn.

Fig. 5 shows the effect of reaction temperature on the conversion of CO₂ over the Co/C–N, Zn/C–N and Co–Zn/C–N catalysts. The dotted lines indicate the equilibrium conversions of methanation, RWGS and the methanol synthesis reaction. From Fig. 5, it can be seen that the methanation and methanol synthesis reactions are exothermic reactions. The equilibrium conversion decreases with an increase in the reaction temperature, with the equilibrium conversion of methanation being much higher than that of the methanol synthesis reaction. On the contrary, the RWGS reaction is an endothermic reaction. CO₂ conversion increases with an increase in reaction temperature for all of the catalysts. This means that the reaction activity is mainly affected by dynamics. The CO₂ conversion of Co/C–N is nearly to the equilibrium conversion of methanation, and the main product of Co/C–N is methane at 275 °C. The CO₂ conversion decreases sharply with an increase in the Zn/Co ratio. For instance, the CO₂ conversions of Co–Zn/C–N-0.5, Co–Zn/C–N-1, Co–Zn/C–N-2 and Co–Zn/C–N-4 are 47.4%, 13.6%, 6.6% and 2.6%. Co–Zn/C–N-2 and Co–Zn/C–N-4 only show catalytic activity above 275 °C.

Fig. 5 Effect of reaction temperature on the conversion of CO₂ over the Co/C–N, Zn/C–N and Co–Zn/C–N catalysts.

The products selectivity and STY of methanol of Co–Zn/C–N at 275 °C are shown in Fig. 6. It is noted that the methanol selectivity increases with an increase in the Zn/Co molar ratio. For instance, the main product (S_CH4 = 91.6%) is methane for the Co–Zn/C–N-0.5 catalyst. Co–Zn/C–N-0.5 mainly shows metallic Co activity, which promotes the reaction of CO₂ hydrogenation into methane (CO₂ + 4H₂ → CH₄ + 2H₂O). With double the Zn content, the methane selectivity falls to 62.4% for the Co–Zn/C–N-1 catalyst, and the selectivity of methanol increases from 4.8% to 31.6%. CO is also observed for the Co–Zn/C–N-1 catalyst, and the selectivity of CO is 3.8%. The methanol selectivity increases with a further increase in the Zn/Co molar ratio. The methanol selectivity is 64.5% for the Co–Zn/C–N-4 catalyst. With an increase in the Zn/Co molar ratio, the function of metallic Co is weakened, and the interaction between Co and Zn increases and results in a new CO₂ hydrogenation process (RWGS, CO₂ + H₂ → CO + H₂O or CO₂ + 3H₂ → CH₃OH + H₂O). The STY values of methanol were calculated to be 1.4, 2.7, 1.9, and 1.0 mmol g_cat⁻¹ h⁻¹ for the Co–Zn/C–N-0.5, Co–Zn/C–N-1, Co–Zn/C–N-2, and Co–Zn/C–N-4 catalysts.

Fig. 6 CO₂ conversion and product selectivity of the Co–Zn/C–N catalysts. Reaction conditions: H₂ : CO₂ = 3, P = 2 MPa, T = 275 °C, and GHSV = 6 L g_cat⁻¹ h⁻¹.

In our earlier experiments, the 50% In₂O₃/Co/C–N catalyst showed stable 9.5% CO₂ conversion with 88.4% methanol selectivity at 300 °C. However, the methanol selectivity fell to 29.2% for the 10% In₂O₃/Co/C–N catalyst due to the formation of Co₃InC_0.75.¹⁹ The strong interaction between Co and In enhances the formation of mixed-metal carbide Co₃InC_0.75 phases. Zn maybe prevents the formation of Co₃InC_0.75 and improves the synthesis of methanol. Therefore, In–Co–Zn/C–N catalysts were prepared via incipient wetness impregnation and physical mixing methods to explore the function of In and Zn.

Unfortunately, Co₃InC_0.75 (JCPDS Card No. 89-7234) still forms in the In–Co–Zn/C–N-4 (IMP) catalyst because of the migration of Co on the catalyst surface (Fig. 7). However, no phase of Co₃InC_0.75 is formed in the In–Co–Zn/C–N-4 (PM) catalyst. As shown in Fig. 6, the PXRD patterns of In–Co–Zn/C–N-4 (PM) show peaks for cubic In₂O₃ (JCPDS Card No. 44-1087), metallic Co (JCPDS Card No. 15-0806) and ZnO (JCPDS Card No. 36-1451).Rhombohedral In₂O₃ (JCPDS Card No. 22-0336) is observed after reaction for the In–Co–Zn/C–N-4 (IMP) catalyst. Meanwhile, there are no crystal changes before and after reaction for the In–Co–Zn/C–N-4 (PM) catalyst (Fig. S6†).

Fig. 7 PXRD patterns of the In–Co–Zn/C–N-4 catalysts prepared via IMP and PM methods.

The CO₂-TPD profiles of the In–Co–Zn/C–N-4 catalysts are shown in Fig. 8a. For the In₂O₃ sample, there is only one desorption at 415 °C, which corresponds to the presence of chemisorbed CO₂ at the thermally induced oxygen vacancy sites of In₂O₃.²⁵ The In–Co–Zn/C–N-4 (IMP) catalyst shows a similar desorption peak temperature to that of the Co/C–N catalyst, which implies that the coating effect of ZnO was totally destroyed. This is also the reason for the formation of Co₃InC_0.75. However, In–Co–Zn/C–N-4 (PM) shows a similar desorption to that of In₂O₃. The CO₂-TPD results are consistent with the PXRD results. In–Co–Zn/C–N-4 (IMP) and In–Co–Zn/C–N-4 (PM) show similar H₂-TPD profiles, as shown in Fig. 8b, which means that the H₂ desorption is not the main effect factor.

Fig. 8 a) CO₂-TPD and b) H₂-TPD measurements of the In–Co–Zn/C–N-4 catalysts.

Fig. 9 shows the XPS spectra of the In–Co–Zn/C–N catalyst. Co mainly exists in the state of metallic Co in the In–Co–Zn/C–N (PM) catalyst. However, most of the Co is oxidized into Co₃O₄ in the In–Co–Zn/C–N (IMP) catalyst (Fig. 9a). The intensity of the In 3d of In–Co–Zn/C–N (PM) is much higher than that of the In–Co–Zn/C–N (IMP) catalyst (Fig. 9b). The peaks at 444.7 and 452.2 eV in the spectrum of the In–Co–Zn/C–N (PM) catalyst can be attributed to the 3d_5/2 and 3d_3/2 of In³⁺, respectively, which are similar to those of pure In₂O₃.²⁶ The shift of 0.3 eV of In 3d in the spectrum of In–Co–Zn/C–N (IMP) toward a higher energy compares to that of In–Co–Zn/C–N (PM). The XPS spectra of Co 2p and In 3d for In–Co–Zn/C–N (IMP) both affirm the incorporation of Co into the In₂O₃ crystal lattice, for instance, in the form of Co₃InC_0.75.

Fig. 9 XPS spectra of the In–Co–Zn/C–N-4 catalysts: a) Co 2p and b) In 3.

In order to address the function of In, the catalyst activity of In–Co–Zn/C–N-4 was evaluated, with the results shown in Fig. 10. Although the three catalysts exhibit similar CO₂ conversion, the addition of In effectively prevents the production of CO and methane. In particular, for the In–Co–Zn/C–N-4 (PM) catalyst, the selectivity of methane decreased to zero, and the selectivity of methanol increased to 77.7% with a 7.0% CO₂ conversion at 300 °C. The STY values of methanol increased almost linearly from 1.5 to 2.7 and 3.3 mmol g_cat⁻¹ h⁻¹ for the Co–Zn/C–N-4, In–Co–Zn/C–N-4 (IMP) and In–Co–Zn/C–N-4 (PM) catalysts.

Fig. 10 CO₂ conversion and product selectivity of the In–Co–Zn/C–N-4 catalyst. Reaction conditions: H₂ : CO₂ = 3, P = 2 MPa, T = 300 °C, and GHSV = 6 L g_cat⁻¹ h⁻¹.

Compared with the In–Co–Zn/C–N-4 (IMP) catalyst, In–Co–Zn/C–N-4 (PM) exhibits high stability, and no deactivation of its activity is observed within 50 h at 300 °C (Fig. 11). The STY of methanol reaches 3.3 mmol g_cat⁻¹ h⁻¹ steadily over 50 h. However, the In–Co–Zn/C–N-4 (IMP) catalyst exhibits a maximum methanol STY value of 2.7 mmol g_cat⁻¹ h⁻¹ over a reaction time of 20 h, and then falls slowly to 2.3 mmol g_cat⁻¹ h⁻¹ over 30 h. A comparison was made between In–Co–Zn/C–N-4 (PM) and some typical CO₂ hydrogenation catalysts. As shown in Table 2, in terms of CH₃OH STY, In–Co–Zn/C–N-4 (PM) exhibits a better performance than In₂O₃, and even 50% In₂O₃/Co/C–N with a high loading of In.¹⁹ Although Pt/In₂O₃,²⁷ Ni/In₂O₃ (ref. 28) and InNi₃C_0.5/m-ZrO₂ (ref. 29) present higher CH₃OH STY values, their activity tests were all performed at considerably higher GHSV (24 000 and 12 000 ml g_cat⁻¹ h⁻¹vs. 6000 ml g_cat⁻¹ h⁻¹) and higher reaction pressure (5 and 4 MPa vs. 2 MPa). The CH₃OH STY value of In–Co–Zn/C–N-4 (PM) is also higher than those of Cu-based catalysts.^30,31

Fig. 11 Catalyst stability experiments over the In–Co–Zn/C–N-4 catalysts.

Table 2 Performance comparison of CO₂ hydrogenation catalysts

Catalyst	GHSV (ml gcat^−1 h^−1)	P (MPa)	T (°C)	X CO2 (%)	S MeOH (%)	STYMeOH (mmol gcat^−1 h^−1)	Ref.
50% In2O3/Co/C–N	3000	2	300	9.5	88.4	2.5	19
In2O3	6000	2	300	4.4	64.5	1.7	19
In2O3	15 000	4	330	7.1	39.7	3.6	25
2.5% Pt/In2O3	24 000	2	300	8.3	41.0	8.7	27
9.7% Ni/In2O3	21 000	5	300	18.5	54.0	17.2	28
In–Co–Zn/C–N-4 (PM)	6000	2	300	7.0	77.7	3.3	This work
InNi3C0.5/m-ZrO2	12 000	4	300	11.2	85.4	19.3	29
Cu/Zn/ZrO2	3600	3	240	12.1	54.1	2.4	30
Cu/Zn/Al2O3	3600	3	230	18.7	43	2.2	31

---

In situ DRIFTS measurements were carried out by switching from N₂ to a CO₂ + H₂ mixture to identify the evolution of surface species over the In–Co–Zn/C–N-4 catalyst, and try to understand the mechanism of the CO₂ activation and hydrogenation reaction. As shown in Fig. 12a, for the In–Co–Zn/C–N-4 (PM) catalyst, the IR peaks at 1255 and 1545 cm⁻¹ can be assigned to monodentate HCOO* (m-HCOO*) species, which appear at 200 °C, and the intensity of these peaks increases with an increase in temperature. A substantial increase in bidentate HCOO* (b-HCOO*, 1421 cm⁻¹) is also observed from 100 °C to 300 °C. Meanwhile, the peak for bidentate H₃CO* (b-H₃CO*) at 1143 cm⁻¹ and terminal H₃CO* (t-H₃CO*) at 1096 cm⁻¹ start to appear at 200 °C, and then the signal of the methanol peak (2881 cm⁻¹) appears at the same temperature.^32,33 The presence of multiple peaks in the region between 3590 and 3730 cm⁻¹ can be attributed to CO₂ overtones (3590 and 3625 cm⁻¹) and OH vibrations (bridged hydroxyl groups, b-OH at 3697 cm⁻¹ and isolated hydroxyl groups, i-OH at 3730 cm⁻¹).³⁴ The intensity of both peaks decreases with an increase in the temperature, however, the CO₂ adsorption peaks are still present at 300 °C, which indicates the strong adsorption of CO₂ on the catalyst. Simultaneously, the signal for CH₃OH at 2881 cm⁻¹ is observed at around 225 °C.

Fig. 12 In situ DRIFTS spectra of a) In–Co–Zn/C–N-4 (PM) and b) In–Co–Zn/C–N-4 (IMP) at different temperatures.

The In–Co–Zn/C–N-4 (IMP) catalyst has a similar IR spectrum (Fig. 12b), however, a significant increase in the concentration of bridged adsorbed CO (1945 cm⁻¹) on metallic Co (ref. 35) is observed from 175 to 300 °C, which is not present in the spectrum of the In–Co–Zn/C–N-4 (PM) catalyst. The DRIFTS results are consistent with the catalyst reaction activity results, showing that the In–Co–Zn/C–N-4 (IMP) catalyst produces both CO and methanol, while the In–Co–Zn/C–N-4 (PM) catalyst mainly catalyzes the hydrogenation of CO₂ into methanol. From the DRIFTS results, it can be seen that the RWGS and methanol synthesis reaction seem to occur at the same time. A possible reaction pathway for the hydrogenation of CO₂ on the In–Co–Zn/C–N-4 catalyst was proposed based on the DRIFTS results (Fig. 13). In the first reaction period, CO₂ adsorbs on In₂O₃ that has reacted with active H* spillover from metallic Co into two types of formate (b-HCOO* and m-HCOO*) intermediates, the formate species further hydrogenate into H₃CO* and t-H₃CO* separately, and finally form methanol. Meanwhile, aside from the methanol pathway, some of the formate species hydrogenate into CO, especially for the In–Co–Zn/C–N-4 (IMP) catalyst.

Fig. 13 Possible reaction mechanism of the hydrogenation of CO₂ into methanol.

4. Conclusions

In conclusion, we successfully developed an In–Co–Zn/C–N catalytic system for the hydrogenation of CO₂ into methanol. In–Co–Zn/C–N-4 (PM) exhibits high stability, and the STY of methanol reaches 3.3 mmol g_cat⁻¹ h⁻¹ (X_CO2 = 7.0%, S_MeOH = 77.7%) steadily over 50 h at 2 MPa and 300 °C. In comparison, the In–Co–Zn/C–N-4 (IMP) catalyst exhibits a maximum methanol STY value of 2.7 mmol g_cat⁻¹ h⁻¹ over a reaction time of 20 h, and then falls slowly to 2.3 mmol g_cat⁻¹ h⁻¹ over 30 h under the same reaction conditions. The ZnO effectively weakens the interaction between Co and In, and the physical mixing method further prevents the formation of Co₃InC_0.75. The DRIFTS results were used to predict the possible methanol and CO production pathway via formate intermediates for the In–Co–Zn/C–N catalyst.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This project was supported by the Natural Science Foundation of Hubei Province (No. 2020CFB802).

References

1. K. Larmier, W. C. Liao, S. H. Tada, E. Lam, R. Verel, A. Bansode, A. Urakawa, A. Comas-Vives and C. Coperet, Angew. Chem., 2017, 129, 2358–2363.
2. S. Kattel, P. J. Ramírez, J. G. G. Chen, J. A. Rodriguez and P. Liu, Science, 2017, 355, 1296–1299.
3. Y. Hartadi, D. Widmann and R. J. Behm, J. Catal., 2016, 333, 238–250.
4. A. García-Trenco, E. R. White, A. Regoutz, D. J. Payne, M. S. P. Shaffer and C. K. Williams, ACS Catal., 2017, 7, 1186–1196.
5. B. Liu, B. Ouyang, Y. H. Zhang, K. L. Lv, Q. Li, Y. B. Ding and J. L. Li, J. Catal., 2018, 366, 91–97.
6. O. Martin, A. J. Martín, C. Mondelli, S. Mitchell, T. F. Segawa, H. Roland, C. Drouilly, D. Curulla-Ferré and J. Pérez-Ramírez, Angew. Chem., Int. Ed., 2016, 55, 626–6265.
7. B. Ouyang, W. L. Tan and B. Liu, Catal. Commun., 2017, 95, 36–39.
8. F. Q. Xie, S. Y. Xu, L. D. Deng, H. M. Xie and G. L. Zhou, Int. J. Hydrogen Energy, 2020, 45, 26938–26952.
9. A. Bavykina, I. Yarulina, A. J. A. Abdulghani, L. Gevers, M. N. Hedhili, X. H. Miao, A. R. Galilea, A. Pustovarenko, A. Dikhtiarenko, A. Cadiau, A. Aguilar-Tapia, J. L. Hazemann, M. K. Sergey, S. Oud-Chikh, L. Cavallo and J. Gascon, ACS Catal., 2019, 9, 6910–6918.
10. A. Pustovarenko, A. Dikhtiarenko, A. Bavykina, L. Gevers, A. Ramírez, A. Russkikh, T. Selvedin, A. Aguilar, J. L. Hazemann, S. Ould-Chikh and J. Gascon, ACS Catal., 2020, 10, 5064–5076.
11. T. F. Fang, B. Liu, Y. Lian and Z. H. Zhang, Ind. Eng. Chem. Res., 2020, 59, 19162–19167.
12. Y. Lian, T. F. Fang, Y. H. Zhang, B. Liu and J. L. Li, J. Catal., 2019, 379, 46–51.
13. H. Yang, X. W. He, F. Wang, Y. Kang and J. Zhang, J. Mater. Chem., 2012, 22, 21849–21851.
14. H. G. Zhang, S. H. Wang, M. Y. Wang, Z. X. Feng, S. Karakalos, L. L. Luo, Z. Qiao, X. H. Xie, C. M. Wang, D. Su, Y. Y. Shao and G. Wu, J. Am. Chem. Soc., 2017, 139, 14143–14149.
15. Y. Z. Chen, R. Zhang, L. Jiao and H. L. Jiang, Coord. Chem. Rev., 2018, 362, 1–23.
16. M. Y. Yang, L. Jiao, H. L. Dong, L. J. Zhou, C. Q. Teng, D. M. Yan, T. N. Ye, X. X. Chen, Y. Liu and H. L. Jiang, Sci. Bull., 2021, 66, 257–264.
17. J. Yang, F. J. Zhang, H. Y. Lu, X. Hong, H. L. Jiang, Y. Wu and Y. D. Li, Angew. Chem., Int. Ed., 2015, 54, 10889–10893.
18. P. Zhou, C. L. Yu, L. Jiang, K. L. Lv and Z. H. Zhang, J. Catal., 2017, 352, 264–273.
19. W. H. Li, A. F. Zhang, X. Jiang, C. Chen, Z. M. Liu, C. S. Song and X. W. Guo, ACS Sustainable Chem. Eng., 2017, 5, 7824–7831.
20. Y. X. Zhou, Y. Z. Chen, L. N. Cao, J. L. Lu and H. L. Jiang, Chem. Commun., 2015, 51, 8292–8295.
21. K. Chen, S. L. Bai, H. Y. Li, Y. J. Xue, X. Y. Zhang, M. C. Liu and J. B. Jia, Appl. Catal., A, 2020, 599, 117614–117621.
22. J. P. Hong, B. Wang, G. Q. Xiao, N. Wang, Y. H. Zhang, A. Y. Khodakov and J. L. Li, ACS Catal., 2010, 20, 5554–5566.
23. Z. L. Yuan, B. Liu, P. Zhou, Z. H. Zhang and Q. Chi, J. Catal., 2019, 370, 347–356.
24. Y. N. Miao, Y. Wang, D. H. Pan, X. H. Song, S. Q. Xu, L. J. Gao and G. M. Xiao, Catalysts, 2019, 9, 94–108.
25. K. H. Sun, Z. G. Fan, J. Y. Ye, J. M. Yan, Q. F. Ge, Y. N. Li, W. J. He, W. M. Yang and C. J. Liu, J. CO2 Util., 2015, 12, 1–6.
26. N. Akkharaphatthawon, N. Chanlek, C. K. Cheng, M. Chareonpanich, J. Limtrakul and T. Witoon, Appl. Surf. Sci., 2019, 489, 278–286.
27. Z. Han, C. Z. Tang, J. J. Wang, L. D. Li and C. Li, J. Catal., 2021, 394, 236–244.
28. X. Y. Jia, K. H. Sun, J. Wang, C. Y. Shen and C. J. Liu, J. Energy Chem., 2020, 50, 409–415.
29. C. Meng, G. F. Zhao, X. R. Shi, P. J. Chen, Y. Liu and Y. Lu, Sci. Adv., 2021, 7, eabi6012.
30. L. Li, D. S. Mao, J. Yu and X. M. Guo, J. Power Sources, 2015, 279, 394–404.
31. C. M. Li, X. D. Yuan and K. Fujimoto, Appl. Catal., A, 2014, 469, 306–311.
32. W. W. Wang, Z. P. Qu, L. X. Song and Q. Fu, J. Catal., 2020, 382, 129–140.
33. S. S. Dang, B. Qin, Y. Yang, H. Wang, J. Cai, Y. Han, S. G. Li, P. Gao and Y. H. Sun, Sci. Adv., 2020, 6, eaaz2060.
34. L. Proaño, E. Tello, A. Trevino, S. X. Wang, M. A. Arellano-Trevino and M. Cobo, Appl. Surf. Sci., 2019, 479, 25–30.
35. L. Nie, Z. Li, T. Kuang, S. Lu, S. X. Liu, Y. H. Zhang, B. Peng, J. L. Li and L. Wang, Chem. Commun., 2019, 55, 10559–10562.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/d1cy01663f

---