Photothermal CO₂ hydrogenation to hydrocarbons over trimetallic Co–Cu–Mn catalysts

Abstract

Photocatalytic CO₂ reduction is a highly vital process for converting CO₂ into valuable chemicals. However, the reaction always proceeds less efficaciously at low temperature. A combination of optical and thermal conditions is one of the feasible approaches to achieve the reaction with high efficiency and has gained much attention recently. In the present work, we prepared several Co–Cu–Mn trimetallic catalysts via a simple co-precipitation method, which were used in catalyzing photothermal CO₂ reduction to hydrocarbons. The metal composition and reduction temperature of the catalysts had important effects on their structural and photoelectrical characteristics and adsorption behaviors, further resulting in diverse catalytic performances. Among the prepared trimetallic catalysts, Co₇Cu₁Mn₁O_x(200), with a Co/Cu/Mn molar ratio of 7/1/1 and reduced at 200 °C in H₂ for 2 h, could produce CH₄ with an activity of 14.5 mmol g_cat⁻¹ h⁻¹ in 10% CO₂/30% H₂/60% N₂, and CH₄ and C₂₊ hydrocarbons with the activities of 15.9 and 7.5 mmol g_cat⁻¹ h⁻¹ in 25% CO₂/75% H₂, respectively. The present strategy for constructing trimetallic oxide catalysts for the photothermal reaction not only provides a highly active catalyst for CO₂ utilization, but also offers a potential possibility for reducing the high temperature of conventional thermal reactions.

---

Introduction

CO₂ is the primary greenhouse gas, whose concentration in the atmosphere continues to rise.¹ It is also an important C1 source and can be converted into many valuable chemicals such as CO,² hydrocarbons,^3–5 olefins,⁶ acids,^7–9 alcohols,^10–12etc. Although many efforts have been invested in this area, it is still very difficult to convert CO₂ into chemicals due to its thermodynamic stability and chemical inertness.¹ CO₂ hydrogenation to hydrocarbons represents an essential and green way for its utilization.¹³ However, the conventional thermal process is always conducted under harsh conditions, leading to the formation of carbon deposit, the increase of energy consumption, etc. The photocatalytic method is an ideal way for the reaction, which could be carried out under mild conditions and using solar energy. However, this route suffers from low activity especially at room temperature, impeding its further applications. Recently, the photothermal catalytic method has gained increasing attention, especially in CO₂ reduction.^14,15

The photothermal routes always contain four types, including thermal-assisted photocatalysis, photo-assisted thermocatalysis, photothermal co-catalysis, and photo-driven thermocatalysis.¹⁶ Generally, these routes combine the advantages of both photocatalysis and thermal catalysis including high activity, proceeding under mild conditions, efficient energy utilization, etc. To date, many catalysts such as Ru/silicon nanowire,¹⁷ RuO₂/SrTiO₃,¹⁸ RuO₂/3D silicon photonic crystals,¹⁹ Ru@FL-LDHs (Ru/Mg–Al LDHs matrix),²⁰ Ru/Al₂O₃,²¹ Ru/TiO₂,²² Pd@Nb₂O₅,²³ CoFeAl-LDH nanosheet,¹³ Ni/SiO₂·Al₂O₃,¹⁴ Cu₂O/Zn-MOF,²⁴ Cu₂O/graphene,²⁵etc. have been used for photothermal CO₂ hydrogenation, and their catalytic performances are given in Table S1.† Significant advances have been made in recent years; however, most of the above catalysts suffer from drawbacks such as using expensive noble metals, working at high temperature, and exhibiting low activity. Besides, the obtained products mainly focused on C1 chemicals such as CO and CH₄. The synthesis of C₂₊ products via a photothermal route is a more difficult but significant issue, which is rarely involved in the above developed catalytic systems.²⁶ Considering the above aspects, to develop a cost-efficient, highly active and reusable catalyst for photothermal CO₂ reduction into CH₄ and C₂₊ hydrocarbons is particularly important.

On the other hand, it has been well known that pure metals used in CO₂ hydrogenation always showed a low activity due to their low CO₂ binding capacity.²⁷ Oxides could not only anchor metal atoms but also provide vacancies for the reaction, and in this way, they could participate in the reaction and enhance the catalytic performances.²⁸ Trimetallic oxide catalysts show unique electronic and structural features different from their individual components. The complex composition of trimetallic catalysts could form multiple interfaces conducive to catalysis.²⁹ In this aspect, trimetallic catalysts have been widely used in many reactions such as CO and CO₂ hydrogenation,^30–32 ethanol steam reforming,³³ ammonia oxidation,³⁴etc. with good performances. In particular, complicated metal–trimetallic oxides can provide many vacancies and active sites for catalysis, and have great potential for application in CO₂ reduction. Herein, a series of Co–Cu–Mn trimetallic oxides were prepared and used for photothermal CO₂ reduction, and the screened Co₇Cu₁Mn₁O_x(200) catalyst, with a Co/Cu/Mn ratio of 7/1/1 and reduced at 200 °C in H₂, showed a high CH₄ formation activity (14.5 mmol g_cat⁻¹ h⁻¹) even at low CO₂ (10%) and H₂ (30%) concentrations, which lies in the highest activity range in Table S1.† Interestingly, after increasing the CO₂ and H₂ concentrations to 25% and 75%, 15.9 and 7.5 mmol g_cat⁻¹ h⁻¹ of CH₄ and C₂₊ hydrocarbons were produced with selectivities of 65.4% and 30.8%, respectively. Moreover, the catalyst has a good temporal stability under continuous-flow conditions. The metal/oxides were confirmed to be very important for the catalytic performance, in which the Co⁰ species were the active sites, which together with the oxides acted as supports and semiconductors for providing a local heating environment via the photothermal effects. The Cu species could promote the reduction of Co and Mn oxides, and also promote the coupling of C–C bonds to produce C₂₊ products especially at a high CO₂ concentration. The Mn species could on one hand enhance the adsorption of CO₂ and H₂, and also provide the photothermal effect to produce a local heating effect to promote the reaction. The present strategy for constructing trimetallic catalysts for the photothermal reaction does not only provide a simple but efficient way to convert CO₂ into valuable hydrocarbons, but also offer a general method for reducing the reaction temperature of conventional high-temperature reactions.

Results and discussion

Preparation and characterization of catalysts

Trimetallic Co–Cu–Mn catalysts were prepared by a simple co-precipitation method, and the details are given in the ESI.† The prepared catalysts were denoted as Co_aCu_bMn_cO_x(T), in which a/b/c and T represent the Co/Cu/Mn molar ratio and the reduction temperature, respectively. Co_aCu_bMn_cO_x means unreduced catalysts. The morphologies of the catalysts were examined by TEM tests. Initially, the Co₇Cu₁Mn₁O_x catalyst mainly showed the morphology of nanoparticles with the size range of 6–15 nm (Fig. 1a and b). The lattice spacings of 0.186, 0.246, 0.219, 0.152, 0.197, and 0.135 nm in the HRTEM image could be successively indexed to the (331) and (311) planes of Co₃O₄, the (411) and (611) planes of Mn₂O₃, and the (−112) and (−221) planes of CuO (Fig. 1c).

Fig. 1 TEM images (a, b, d, and e) and HRTEM images (c and f) for the Co₇Cu₁Mn₁O_x (a–c) and Co₇Cu₁Mn₁O_x(200) (d–f) catalysts, HAADF-STEM image for the Co₇Cu₁Mn₁O_x(200) catalyst (g) and the corresponding EDS elemental mapping of Co, Cu, Mn, and O (h–k). The insets in (a) and (d) are the size distribution histograms for the Co₇Cu₁Mn₁O_x and Co₇Cu₁Mn₁O_x(200) catalysts.

After reduction, the Co₇Cu₁Mn₁O_x(200) catalyst maintained the nanoparticle morphology in the size range of 6–12 nm (Fig. 1d and e). Co₃O₄ and Mn₂O₃ were also detected, as shown in Fig. 1f. The HAADF-STEM and EDX elemental mappings of the Co₇Cu₁Mn₁O_x(200) catalyst showed that Co, Cu, Mn, and O elements are homogeneously dispersed over the whole grain (Fig. 1g to k).

For comparison, the TEM images of catalysts with different Co/Cu/Mn ratios were also studied, and the results are shown in Fig. S1.† With the increasing Co content, the morphology changed from nanosheets to nanoparticles. For example, the Co₃Cu₁Mn₁O_x(200) and Co₅Cu₁Mn₁O_x(200) catalysts were nanosheets, while the Co₇Cu₁Mn₁O_x(200) and Co₉Cu₁Mn₁O_x(200) catalysts were mainly nanoparticles, which were beneficial for exposing more active sites.

The molar ratios of the bimetallic and trimetallic catalysts were tested by ICP-OES, and the results are given in Table S2.† The molar ratios of metals were consistent with the feed amounts.

XRD examinations showed that all the tested unreduced catalysts such as Co₇Cu₁Mn₁O_x, Co₇Mn₁O_x, and Co₇Cu₁O_x showed the XRD patterns of Co₃O₄ (JCPDS file no. 73-1701). No obvious CuO_x or MnO_x peaks were found, implying that they were easily incorporated into the Co₃O₄ crystals and were well dispersed (Fig. 2a). The Co₇Cu₁Mn₁O_x(200) catalyst showed the same peaks as that of the Co₇Cu₁Mn₁O_x catalyst, but with lower intensities, indicating the crystallinity decrease of the Co₃O₄ component after reduction.

Fig. 2 XRD patterns of (a) the unreduced Co₇Mn₁O_x, Co₇Cu₁O_x, and Co₇Cu₁Mn₁O_x catalysts; (b) Co₇Mn₁O_x(200), Co₇Cu₁O_x(200), and Co₇Cu₁Mn₁O_x(200); and (c) Co₇Cu₁Mn₁O_x catalysts reduced at different temperatures.

The XRD pattern of the Co₇Cu₁Mn₁O_x(200) catalyst was almost the same as that of the unreduced catalyst, and it is mainly attributed to the Co₃O₄ phase (Fig. 2b). However, two small peaks at 43.0° and 61.8° were carefully detected, which could be assigned to the CoO (200) and (220) planes (JCPDS file no. 70-2855). No Co⁰, Cu⁰, Cu₂O, or Mn₂O₃ species were detected due to their low amounts. The Co₇Mn₁O_x(200) catalyst showed three peaks at 36.8°, 42.7°, and 61.8°, which could be assigned to the CoO (111), (200), and (220) planes (JCPDS file no. 78-0431), respectively. Two phases of Co⁰ fcc and hcp were identified in the Co₇Cu₁O_x(200) catalyst, and the formation of two phases is commonly observed in the synthesis of cobalt nanostructures.³⁵ The peaks at 41.5°, 44.4°, 47.4°, and 75.9° could be indexed to Co⁰ hcp (100), (002), (101), and (110) facets (JCPDS file no. 01-071-4239), respectively. The peaks at 44.4° and 51.5° could be contributed to the Co⁰ fcc (111) and (200) planes (JCPDS file no. 01-071-4651).³⁵ The results indicated that Co₃O₄ could be reduced to Co⁰ by the promotion of Cu species, and this will be further verified by the following H₂-TPR tests. No MnO_x or CuO_x species were detected due to their good dispersion. It is worth noting that although Cu could promote the reduction of CoO_x species to Co⁰, no Co⁰ was detected in the bulk phase of the Co₇Cu₁Mn₁O_x(200) catalyst.

The XRD patterns of the catalysts with different Co/Cu/Mn ratios also showed the main peaks of Co₃O₄ and CoO (Fig. S2†); however, the Co₃Cu₁Mn₁O_x(200) and Co₅Cu₁Mn₁O_x(200) catalysts showed a wide peak at around 20°, indicating that they possess amorphous structures.

The results of the N₂ adsorption/desorption tests are shown in Fig. S3 and Table S3.† The reduction temperature affects the catalyst structure remarkably, and the BET surface areas of the tested trimetallic catalysts declined gradually with the increase of the reduction temperature, which is probably because the high reduction temperature leads to catalyst sintering. The bimetallic Co–Cu catalyst shows the lowest BET surface area, while the Co–Mn catalyst presents the highest. Indeed, the surface areas of the CoO_x and Co₇Cu₁Mn₁O_x(100, 200, and 300) catalysts did not differ greatly from one another, thus their diverse performances did not stem from the difference in their surface areas.

To distinguish the valences of the Co, Cu, and Mn species on the surface of the catalysts, the XPS of Co 2p, Cu 2p, Mn 2p and 3s, and AES of Cu L₃M_4,5M_4,5 were analyzed (Fig. 3).^5,36,37 In the XPS of the Co₇Cu₁Mn₁O_x catalyst, the Co element was composed of Co³⁺ and Co²⁺ species, which was confirmed by the peaks at 779.5 eV and 780.9 eV, which can be attributed to the Co³⁺ and Co²⁺ 2p_3/2 peaks (Fig. 3a).³⁸ The Cu²⁺ 2p XPS showed two peaks at 934.5 eV and 954.2 eV, which could be assigned to Cu²⁺ 2p_3/2 and 2p_1/2, respectively (Fig. 3b and c). The XPS peaks of Co species in the Co₇Cu₁Mn₁O_x(200) catalyst were the same as those in the unreduced catalyst; however, the Cu species was found to be Cu⁺. The Mn XPS indicated that the catalyst contains Mn³⁺ oxidation state (Fig. 3d and e). The XPS results indicated that the surface of the Co₇Cu₁Mn₁O_x(200) catalyst mainly contained Co²⁺, Co³⁺, Cu⁺, a small amount of Cu²⁺, and Mn³⁺, respectively, but no Co⁰ or Cu⁰ species were detected for air oxidation.

Fig. 3 XPS spectra of the unreduced Co₇Cu₁Mn₁O_x catalyst and the Co₇Cu₁Mn₁O_x(100, 200, 300, and 500) catalysts. (a) Co 2p XPS, (b) Cu 2p XPS, (c) Cu L₃M_4,5M_4,5 AES spectra, (d) Mn 2p XPS, and (e) Mn 3s XPS, respectively.

For comparison, the XPS spectra of the catalysts reduced at different temperatures were also tested. The Co element in these catalysts was mainly composed of Co²⁺ and Co³⁺ species but with different contents. The Co²⁺ component increased with the increase of the reduction temperature. No obvious Co⁰ species was found until the reduction temperature reached 500 °C. As for the Cu XPS, the area of the satellite peak around 942 eV decreased with increasing reduction temperature, indicating that Cu^0/+ 2p_3/2 increased. Besides, all the reduced catalysts possess the dominant Cu⁺ component. The Mn 3s XPS showed that the Co₇Cu₁Mn₁O_x(500) catalyst is composed of Mn²⁺ species, i.e., MnO, while the others showed Mn³⁺ species.

The XPS spectra of the Co–Cu–Mn trimetallic catalysts with different molar ratios were also investigated, and the results are given in Fig. S4.† All the elements had the same valences but contained different contents, such as Co²⁺, Co³⁺, Mn³⁺, Cu⁺, and a small amount of Cu²⁺, which are similar to the Co₇Cu₁Mn₁O_x(200) catalyst, indicating that they are probably active in the reaction. This was confirmed by the following catalytic activity studies (Table 1).

Table 1 Photothermal CO₂ hydrogenation over diverse catalysts^a

Entry	Catalyst	Conv. (%)	Activity^b (mmol gcat^−1 h^−1)			CH4^b Sele. (%)
			CH4	C2+	CO
a Reaction conditions: catalyst 50 mg, CO2/H2/N2 10%/30%/60%, full irradiation, 200 °C, 3 h. b Detected by GC and obtained from three successive runs. c N2 was replaced by Ar.
1	None	0	0	0	0	—
2	CoOx(200)	5.6	3.1 ± 0.1	0.4 ± 0.1	0	88.6
3	CuOx(200)	0	0	0	0	—
4	MnOx(200)	0	0	0	0	—
5	Cu1Mn1Ox(200)	0	0	0	0	—
6	Co7Mn1Ox(200)	0.8	0.2 ± 0.1	0	0.3 ± 0.1	40.0
7	Co7Cu1Ox(200)	14.0	5.9 ± 0.2	2.4 ± 0.1	0.4 ± 0.1	67.8
8	Co7Cu1Mn1Ox(200)	27.4	14.5 ± 0.2	1.4 ± 0.1	1.1 ± 0.1	85.3
9	Co3Cu1Mn1Ox(200)	12.4	7.2 ± 0.2	0.5 ± 0.2	0	93.5
10	Co5Cu1Mn1Ox(200)	16.8	9.4 ± 0.3	0.8 ± 0.2	0.2 ± 0.2	84.7
11	Co9Cu1Mn1Ox(200)	18.9	10.1 ± 0.2	0.9 ± 0.2	0.7 ± 0.1	90.4
12	Co1Cu3Mn1Ox(200)	3.1	1.6 ± 0.2	0.3 ± 0.1	0	84.2
13	Co1Cu1Mn3Ox(200)	2.1	1.1 ± 0.2	0.2 ± 0.1	0	84.6
14^c	Co7Cu1Mn1Ox(200)	28.4	15.1 ± 0.2	1.3 ± 0.1	1.2 ± 0.1	85.8
15	Co7Cu0.25Mn1.75Ox(200)	13.6	7.7 ± 0.1	0.4 ± 0.2	0.4 ± 0.1	90.6
16	Co7Cu0.75Mn1.25Ox(200)	22.9	12.7 ± 0.2	0.7 ± 0.1	0.9 ± 0.1	88.8
17	Co7Cu1.25Mn0.75Ox(200)	24.0	11.5 ± 0.2	2.6 ± 0.1	0.9 ± 0.1	76.7
18	Co7Cu1.75Mn0.25Ox(200)	14.1	5.0 ± 0.1	2.5 ± 0.1	1.3 ± 0.1	56.8

---

The Raman spectrum of the Co₇Cu₁Mn₁O_x(200) catalyst showed a peak at 680 cm⁻¹, which could be assigned to the A1 g mode of Co₃O₄ (Fig. S5†).³⁹ This profile is similar to those of the Co₇Cu₁Mn₁O_x and Co₇Cu₁Mn₁O_x (100, 300, and 400) catalysts, indicating that Co₃O₄ exists in these catalysts. Interestingly, the Co₇Cu₁Mn₁O_x(500) catalyst showed a wide peak at 540 cm⁻¹, assigned to the one-phonon LO mode of CoO,⁴⁰ which was formed by the partial oxidation of Co⁰ by local laser heating.⁴⁵ No obvious Cu or Mn species were detected, indicating that they have permeated into the Co₃O₄ crystal lattice.

The above characterization studies showed that the Co₇Cu₁Mn₁O_x(200) catalyst mainly possesses Co₃O₄ in the bulk phase, while the surface of the catalyst was composed of Co³⁺, Co²⁺, Mn³⁺, Cu⁺, and a small amount of Cu²⁺ species, which are in the forms of Co₃O₄, Mn₂O₃, Cu₂O, and CuO. The catalysts reduced at 100, 200, and 300 °C, and the catalysts with different Co/Cu/Mn molar ratios contained similar components but of different concentrations, indicating that they possess similar structures and properties.

Catalytic performances

Photothermal CO₂ hydrogenation was carried out in a tubular reactor (Fig. S6),† and the procedure details are given in the ESI.† The products of photothermal CO₂ reduction include CO, CH₄, and C₂₊ hydrocarbons (mainly C₂ to C₆) in the present work. Initially, the catalytic performances over different catalysts are given in Table 1. No products were detected in the absence of a catalyst, and the sole metal catalysts or the bimetallic catalysts showed a very low activity in the reaction. The Co₇Cu₁Mn₁O_x(200) catalyst showed the best catalytic performances with a CH₄ formation activity of 14.5 mmol g_cat⁻¹ h⁻¹ and a selectivity of 85.3% (entries 1–7 vs. 8). These activities were achieved under the irradiation of a 300 W Xe lamp (234 mW cm⁻², 300–1100 nm) and heat generated by an additional electric heater (Fig. S6†). Unlike the CuO_x(200) and MnO_x(200) catalysts, the CoO_x(200) catalyst could show a very low CH₄ formation activity (only 3.1 mmol g_cat⁻¹ h⁻¹) (entries 2–4), demonstrating that the Co species is probably the active species, while the CuO_x and MnO_x species are the promoters for the Co species. Indeed, the Cu and Mn species are always used as promoters or active sites for the Co-catalyzed CO₂ reduction.⁴¹ The activities of the bimetallic Cu₁Mn₁O_x or Co₇Mn₁O_x catalysts confirmed that Cu or Mn are not the active sites (entries 5–7). Interestingly, the Co₇Cu₁O_x(200) catalyst showed higher activities for the formation of CH₄ and C₂₊ hydrocarbons than those of the pure CoO_x(200) and CuO_x(200) catalysts, indicating that the Cu species could not only enhance the activity of the Co-based catalyst in CO₂ reduction to CH₄, but also promote the formation of C₂₊ products. For comparison, other trimetallic catalysts with different ratios were also assessed; however, all of them possessed lower activities than that of the Co₇Cu₁Mn₁O_x(200) catalyst (entries 8 vs. 9–13).

The activities achieved in CO₂/H₂/N₂ and CO₂/H₂/Ar were very similar (entries 8 vs. 14), indicating that the catalytic performances were not affected by the type of diluent gas. In addition, the C₂₊ activities of the Co₉Cu₁Mn₁O_x(200), Co₇Cu₁Mn₁O_x(200), Co₅Cu₁Mn₁O_x(200), and Co₃Cu₁Mn₁O_x(200) catalysts were only 0.9, 1.4, 0.8, and 0.5 mmol g_cat⁻¹ h⁻¹, respectively, and inconsistent with the change rule of the catalyst composition. Because the formation of C₂₊ products is closely related to the Cu species, the uneven distribution of Cu species on the catalyst surface is probably the reason for the discrepancy between the C₂₊ activity and the Co/Cu/Mn molar ratio of the catalyst.

The Co species are the active sites; however, the promoters of Cu and Mn species are also very important for the reaction. To confirm this, the catalysts with the same component of Co element were investigated, and the results indicated that the Cu and Mn concentrations affected the catalytic performances remarkably (entries 15–18). The Co₇Cu_0.75Mn_1.25O_x(200) and Co₇Cu_1.25Mn_0.75O_x(200) catalysts showed high CH₄ formation activities, while the Co₇Cu_0.25Mn_1.75O_x(200) and Co₇Cu_1.75Mn_0.25O_x(200) catalysts offered relatively low values, indicating that the low concentrations of Cu and Mn are not beneficial for catalysis.

The catalytic performance studies showed that the coexistence of Co, Cu, and Mn in the Co/Cu/Mn trimetal is very important for the reaction, and their molar ratios and the synergistic effects of these multicomponents are closely related to the catalytic performances, and a similar phenomenon could be found in the previous report.⁴²

Effect of the reduction temperature of the Co₇Cu₁Mn₁O_x catalysts

The above characterization results indicated that the reduction temperature affected the metal valence compositions of the catalysts, and it was proved to be able to affect their catalytic performances.^43,44 To verify this, the catalysts reduced at different temperatures were studied, and the results are shown in Fig. 4. Indeed, the Co₇Cu₁Mn₁O_x and Co₇Cu₁Mn₁O_x(100, 200, and 300) catalysts showed good catalytic performances during the reaction, indicating that they have similar catalyst compositions and properties. However, the Co₇Cu₁Mn₁O_x(400 and 500) catalysts showed decreased catalytic activities, which has a great relationship with the high concentration of Co⁰ in the bulk phase but less Co₃O₄ and Mn₂O₃.

Fig. 4 Effect of the reduction temperature of the Co₇Cu₁Mn₁O_x(T) catalysts on the catalytic performances of photothermal CO₂ reduction. Reaction conditions: catalyst 50 mg, CO₂/H₂/N₂ 10%/30%/60%, 200 °C, 3 h, and 300 W Xe lamp full irradiation.

The catalysts after being used for 9 h and 21 h were characterized by XPS and XRD in Fig. 5, and the results showed that Co⁰ and Cu⁰ species could be found during the reaction. Co⁰ was widely recognized as the active site for CO₂ hydrogenation. The XRD patterns of the catalysts used for 9 h showed similar peaks to those of Co₃O₄ and CoO, revealing that Co₃O₄ still exists in the bulk phase. Compared with the Co₇Cu₁Mn₁O_x(200) catalyst in Fig. 2, the peak at 36.6° decreased remarkably. The catalyst used for 21 h showed similar results but with more Co⁰ and Cu⁰ species. The XPS and XRD characterization studies showed that the catalyst compositions were stable during the reaction.

Fig. 5 XPS tests (a, Co 2p; b, Cu 2p; c, Cu 3s LMM AES) and XRD pattern (d) for the catalysts after 9 h and 21 h of uses, respectively.

The highest CH₄ activity was achieved over the Co₇Cu₁Mn₁O_x(200) catalyst. Besides, the Co₇Cu₁Mn₁O_x(100 and 300) catalysts also showed acceptable CH₄ formation activities. Generally, in thermal catalysis, Co⁰ species are recognized as the active components. However, the Co₇Cu₁Mn₁O_x(500) catalyst showed the lowest activity in the reaction, which mainly contained the Co⁰ species but no Co₃O₄ or Mn₂O₃. It should be noted that in the present photothermal catalysis, the oxides including Co₃O₄, Mn₂O₃, etc. could promote the reaction via the interaction effect between the Co species and other oxides.

The H₂-TPR tests were carried out to study the interaction effect of different metal oxides, and the results are shown in Fig. 6. Pure Co₃O₄ showed two peaks at 273 °C and 336 °C, which could be assigned to the reduction peaks of Co₃O₄ to CoO and CoO to Co⁰, respectively.⁴⁶ Mn₂O₃ showed two peaks located at 292 °C and 388 °C, which could be assigned to the reduction of Mn₂O₃ to Mn₃O₄ and Mn₃O₄ to MnO, respectively.^47,48 The Co₇Cu₁O_x catalyst had two peaks at lower temperatures of 156 °C and 210 °C indicating that Cu could obviously promote the reduction of Co₃O₄ and CoO, which is consistent with previous reports.⁴⁵ However, the Co₇Mn₁O_x catalyst showed two main peaks at 280 °C and 428 °C, which could be assigned to the reduction of (Co,Mn)₃O₄ to (Co,Mn)O, and (Co,Mn)O to Co⁰ and MnO,^5,49,50 respectively, revealing that Mn species impedes the reduction of CoO_x. The Co₇Cu₁Mn₁O_x catalyst had three peaks at 128, 183, and 284 °C, which could be attributed to the reduction peaks of CuO to Cu⁰, Co₃O₄ to CoO, and CoO to Co⁰ and MnO, respectively.⁵⁰ The H₂-TPR results indicated that the Co, Cu, and Mn species show strong interaction effects, and the previous report showed that the presence of Cu could alter the electronic interactions with Co and Ni in the Co–Cu–Ni trimetallic catalysts.³⁰

Fig. 6 H₂-TPR(a), CO₂-TPD (b), and H₂-TPD (c) tests for the different catalysts.

The surface adsorption capacities could significantly affect the catalytic activity. Thus, we conducted the CO₂-TPD and H₂-TPD tests for the diverse trimetallic catalysts and the results indicated that with the increase of the reduction temperature, the adsorption amounts of CO₂ and H₂ on the catalysts decreased. The Co₇Cu₁Mn₁O_x(100 and 200) catalysts contained higher CO₂ and H₂ adsorption amounts than the others, which is beneficial for them to catalyze the CO₂ photothermal reduction.

To investigate the effects of the Mn and Cu species on the Co-based catalysts, the CO₂-TPD tests for Co–Mn and Co–Cu bimetallic catalysts were performed (Fig. S7†). The Mn species was found to be able to improve CO₂ and H₂ adsorption for the Co-based catalysts, while the Cu species reduced their adsorption since the Co–Cu bimetallic catalyst could be deeply reduced at 200 °C. Even so, the Co₇Cu₁O_x(200) catalyst exhibited a strong CO₂ adsorption peak at around 350 °C, showing that the Cu species could promote strong adsorption of CO₂ on the catalyst.

We investigated the catalytic performances of the Co–Cu and Co–Mn bimetallic catalysts reduced at different temperatures, and the results are given in Table S4.† With the increase of the reduction temperature, the activities of the Co–Cu and Co–Mn catalysts increased initially but were then reduced at high reduction temperatures of 400 and/or 500 °C. The highest activities were lower than that of the Co₇Cu₁Mn₁O_x(200) catalyst (Table 1, entry 8), indicating that the Cu and Mn species were both needed for the photothermal CO₂ reduction.

Effect of the reaction conditions

The above investigation showed the composition and pretreatment of the catalyst affected the activity remarkably. In addition, the reaction conditions also had a great influence on the catalytic performance. The effects of the reaction conditions on the catalytic performance were also investigated, and the results are shown in Fig. 7. Almost no product was detected in the absence of CO₂ or H₂, indicating that the products are generated from CO₂. This was further confirmed by a control experiment of adding ¹³CO₂ in the reaction, in which ¹³CH₄ was detected with a mass spectrometer (Fig. S8†).

Fig. 7 Effects of the reaction conditions on the catalytic performances of photothermal CO₂ reduction over the Co₇Cu₁Mn₁O_x(200) catalyst; (a) effect of the CO₂/H₂ ratio, (b) effect of the total concentration of CO₂ and H₂, (c) effect of the reaction temperature, and (d) effect of the reaction time. Reaction conditions: catalyst 50 mg, 300 W Xe lamp full irradiation and (a) (CO₂ + H₂)/N₂ 40%/60%, 200 °C, 3 h; (b) CO₂/H₂ 1/3, 200 °C, 3 h; (c) CO₂/H₂/N₂ 10%/30%/60%, 3 h; and (d) CO₂/H₂/N₂ 10%/30%/60%, 200 °C, spades (/) photothermal conditions, clubs (♣) thermal conditions, and diamonds (◆) photocatalytic conditions respectively.

The best performances for CH₄ formation were achieved in CO₂/H₂ with a molar ratio of 1/3 (Fig. 6a). Surprisingly, changing the CO₂ and H₂ concentrations slightly affected the activity of CH₄ formation (around 15 mmol g_cat⁻¹ h⁻¹), but had an influence on C₂₊ hydrocarbon formation. In the absence of a dilute gas, the highest C₂₊ activity reached 7.5 mmol g_cat⁻¹ h⁻¹ with a 30.8% selectivity. Generally, C₂₊ hydrocarbon formation involves CO₂ reduction and C–C coupling steps, and is much more difficult to achieve under photothermal conditions.⁵¹ Thus, the present Cu₇Cu₁Mn₁O_x(200) catalyst could catalyze the C₂₊ hydrocarbon formation, but only the CO₂ and H₂ concentrations need to be enhanced. The C₂₊ formation is strongly linked to the amount of surface adsorbed carbon species such as CO, –CH₂–, etc. A high amount of carbon species can be coupled to generate C₂₊ hydrocarbons over a suitable catalyst, thus an increase of the CO₂ concentration could enhance the C₂₊ selectivity. However, generally, to obtain C₂₊ products, high pressure and/or CO₂/H₂ molar ratio are always needed.⁴² The present Co₇Cu₁Mn₁O_x(200) catalyst has a good CO₂ adsorption capacity, which was confirmed by the CO₂-TPD tests. The high C₂₊ selectivity indicates that the Co₇Cu₁Mn₁O_x(200) catalyst possesses good CO₂ reduction and C–C coupling abilities even under ambient pressure and a low CO₂/H₂ molar ratio (1/3). These multi-functional applications of synthesis of CH₄ and C₂₊ hydrocarbons stemmed from the multiple interfaces created by the complicated compositions of Co⁰, Co₃O₄, Mn₂O₃, CuO and Cu₂O species, etc.

The reaction temperature has a significant influence on the catalytic performance, which was monitored using a thermocouple positioned above the catalyst level (Fig. S6†). As shown in Fig. 7c, high temperature promoted the reaction remarkably. With the increase of the reaction temperature, the activity of CH₄ formation increased. The C₂₊ selectivity increased initially and then slightly decreased at a temperature higher than 200 °C. This is probably because at high temperature, the catalyst was sintered and reduced, which affected the product distribution especially with the CH₄ selectivity increased and C₂₊ products decreased.

The evolution of activity on reaction time was also studied and the results are shown in Fig. 7d. In the first 30 min of reaction, CH₄ was formed with a very low activity, indicating that the catalyst had an induction period. During this period, the surface Co species, especially the surface CoO was reduced to the active Co⁰ species. With prolonged time, the CH₄ formation activity increased remarkably. During our experiments, we found that the pressure of the reaction gas decreased remarkably after 3 h, but further prolonging the time did not show any remarkable decrease, indicating that the reaction has reached equilibrium. Thus, we chose 3 h as the optimum reaction time. In addition, control experiments were conducted to investigate the effect of optical and thermal conditions. The reaction could hardly occur under light irradiation but without extra heating, and the sole thermal conditions offered a very low CH₄ formation activity (only one-sixth that of the photothermal reaction). However, their combination is beneficial for the CO₂ reduction, indicating that the excellent catalytic performances stemmed from their synergistic promotion effect.

As mentioned above, the present catalytic system could achieve the photothermal CO₂ reduction to C₂₊ hydrocarbons, but only needs to enhance the concentrations of CO₂ and H₂ to 25% and 75%. Fig. S9† shows the catalytic performances of the reaction. Under the conditions of 200 °C and 3 h, the highest C₂₊ selectivity and activity reached 30.8% and 7.5 mmol g_cat⁻¹ h⁻¹ obtained after 3 h of reaction.

The results indicated that the multifunctional applications of the trimetallic Co–Cu–Mn catalysts derive from not only the multicomponents of the catalysts, but also the interaction effects between the components.

Photoelectric properties of the Co–Cu–Mn catalysts

The study of the catalytic performances confirmed that Co₃O₄ and Mn₂O₃ oxides are very important for the reaction, and this differs from the fully reduced Co-based catalysts in H₂ such as Co/CoO_x²⁹ and Co₆/MnO_x⁵ catalysts in CO₂ thermal hydrogenation. This diversification is probably because photothermal catalysis needs more light-sensitive components such as Co₃O₄ and Mn₂O₃ due to their semiconductor nature,^52,53 while the thermal catalyst favours low-valent metals such as Co⁰.

To further investigate the light-sensitive properties of the Co–Cu–Mn trimetallic catalysts, the UV-vis DRS spectra were investigated, and the results are shown in Fig. S10a.† The unreduced catalysts including Co₇Mn₁O_x, Co₇Cu₁O_x, and Co₇Cu₁Mn₁O_x showed higher absorption than the corresponding reduced catalysts throughout the UV-vis region. This was because the former mainly contain Co₃O₄, which has strong absorption of d–d transitions.⁵⁴ Upon reduction, the Co₃O₄ amount decreased, leading to the decrease of absorption accordingly. As for the Co₇Cu₁Mn₁O_x(200, 300, 400, and 500) catalysts, with the increase of the reduction temperature, the absorption strength decreased gradually for the deep reduction of Co₃O₄ to CoO and Co⁰ species. Both the Co₇Cu₁O_x(200) and Co₇Cu₁Mn₁O_x(500) catalysts showed a lower absorbance than the other samples because they did not contain the Co₃O₄ species, which was confirmed by the XRD tests.

The band gap energies (E_g) for the tested catalysts were calculated using the Kubelka–Munk (K–M) model.⁵⁵ The Co₇Cu₁Mn₁O_x(200) catalyst had a lower E_g value (1.86 eV) than those of the catalysts reduced at higher temperatures (Fig. S10b†). The Co₇Cu₁Mn₁O_x(500) catalyst, however, had the highest E_g (3.26 eV) value and showed a very low activity in the reaction. The low E_g value of the Co₇Cu₁Mn₁O_x(200) catalyst led to the feature that the electrons in the semiconductor could be easily excited under light irradiation. The lower band gap is conducive to visible light excitation, which was confirmed by the control experiments using different types of light including full irradiation (300–1100 nm), UV (300–420 nm), and visible light (420–800 nm). The catalytic performances obtained under visible light illumination are almost similar to those under UV-vis and full irradiation, as shown in Table S5.†

The photocurrent–time curves of the Co₇Cu₁Mn₁O_x(100, 200, 300, and 400) catalysts are shown in Fig. S10c.† All of these catalysts showed obvious photocurrent response behavior, indicating that they could be excited to generate electron–hole pairs.

Co₃O₄, Mn₂O₃, Cu₂O, and CoO are p-type semiconductors; however, after reduction, oxygen vacancies were created and regulated the p-type to a n-type semiconductor, which is confirmed by the Mott–Schottky curves shown in Fig. S10d.† All the catalysts showed dipolar (p-type and n-type) semiconductor properties, resulting from the multiple interfaces of the complex trimetallic catalyst.

The EIS Nyquist plots of the tested samples including Co₇Cu₁Mn₁O_x(100, 200, 300, and 500) showed similar arc radii, revealing that they have similar impedance values and electron transfer rates (Fig. S10e†).⁵⁶

The photothermal effect was investigated via testing the stable temperature of the samples under vacuum (Fig. S11†). Under similar conditions, the catalyst temperatures were monitored using an infrared camera. Among the tested catalysts, the MnO_x catalyst showed the highest temperature (104.6 °C) under the full irradiation, while the CuO_x catalyst offered the lowest (88.3 °C). The Co₇Cu₁Mn₁O_x(200) catalyst showed a higher temperature (102.2 °C) than that of the Co₇Cu₁Mn₁O_x(500) (95.8 °C), indicating that the Co₇Cu₁Mn₁O_x(200) catalyst could provide a higher local heating environment. The results revealed that the presence of Mn could provide a higher local temperature for the catalyst.

In summary, the results obtained due to the photoelectrical characteristics indicated that the Co₇Cu₁Mn₁O_x(200) catalyst is a p-type and n-type semiconductor, showing a low band gap. All of these properties are closely related to the excellent catalytic activity in photothermal CO₂ reduction. However, it should be noted that the unique catalytic performances of the Co₇Cu₁Mn₁O_x(200) catalyst were derived from the synergistic effect of the adsorption effect, light-response properties, and photothermal effect.

Light effect and the proposed mechanism for photothermal CO₂ reduction

In the present catalytic system, the synergistic effect between light and heat had a great influence on the catalytic activity. Under light irradiation, the electrons of the Co₇Cu₁Mn₁O_x(200) catalyst could be excited to generate the electron–hole pairs. Generally, in photocatalysis, the excited electrons and holes can participate in reactions.⁵⁷ However, under photothermal conditions, the E_g value of the semiconductor is normally found to be lower, and the excited electrons recombine more easily with the photogenerated holes than under the traditional photocatalysis without heating.

To investigate the effects of the generated electron–hole pairs, we carried out two control experiments by adding sacrificial electron donors including p-xylene (∼2.18 V vs. Ag/AgCl_sat) and anisole (1.92 V vs. Ag/AgCl_sat) in the absence of H₂.^18,25 As shown in Fig. S12,† both the reactions gave a very low CH₄ yield despite the fact that p-xylene and anisole could quench the photogenerated holes more efficiently than H₂. These results indicated that photocatalysis plays a minor role in the present photothermal CO₂ reduction. In this aspect, similar to the Ru-,¹⁸ Rh-,⁵⁸ Au-,⁵⁸ and Ag-based⁵⁹ catalysts, the charge separation and recombination under light irradiation resulted in a local thermal effect at the surface of the catalyst, offering more energy to promote CO₂ reduction.

The mechanism for the photothermal CO₂ reduction to hydrocarbons is still unclear, and two reaction routes are generally accepted.⁶⁰ The first involves a CO intermediate, in which CO₂ is initially converted to CO, and CO could be further hydrogenated to hydrocarbons.⁶¹ The other route is the direct CO₂ conversion. In this route, CO₂ is hydrogenated to carbonate, formate, and methoxy intermediates, and then hydrogenated to hydrocarbons. In this work, the CO formation activity was very low. To study the reaction pathway, we conducted the control experiments of replacing CO₂ by CO under the conditions given in Table 1, entry 8. Interestingly, the activities of CH₄ and C₂₊ were 4.9 and 7.3 mmol g_cat⁻¹ h⁻¹, and their selectivities were 40.2% and 59.8%, respectively (Table 1, entry 8, Fig. S13, S14 vs. Table S6, entry 2, Fig. S15†). To exclude that the high CO concentration may occupy the active sites on the catalyst surface, a mixed gas of 2%CO/8%CO₂ (total 10%) was also reacted (Table S6, entry 3 and Fig. S16†). However, the activities of CH₄ and C₂₊ were 7.2 and 2.8 mmol g_cat⁻¹ h⁻¹ with the selectivities of 72.0% and 28.0%, respectively. The results implied that the presence of CO could not improve but reduce the CH₄ activity, meaning that in the photothermal CO₂ reduction, CH₄ is probably not formed via the CO intermediate route. However, CO was found to be beneficial for the C₂₊ synthesis, revealing that it is probably an intermediate for the formation of the C₂₊ hydrocarbon product.

The characterization results confirmed that the trimetallic Co₇Cu₁Mn₁O_x(200) catalyst contains CoO, Co₃O₄, Cu₂O, Mn₂O₃, etc. During the reaction, the CoO species could be reduced to Co⁰, which was recognized as the active site. The CoO and Cu₂O species could be reduced by H₂ to Co⁰ and Cu⁰ during the reaction, which was confirmed by the XPS results (Fig. 5). In the present photothermal CO₂ hydrogenation, metal oxides such as Co₃O₄ and Mn₂O₃ were very important for the reaction because they are semiconductors, which could generate electrons and holes under light illumination. A quick recombination of electrons and holes provides a local heating environment for the catalysts, and further promotes the reaction.⁵ In addition, Mn₂O₃ could offer Lewis acidic sites to enhance CO₂ and H₂ adsorption capacity of the catalyst.

Cu species could promote the partial reduction of cobalt oxides to CoO and Co⁰ and maintain the presence of CoO and Co₃O₄ prior to and during the reaction, and reduce the temperature of catalyst reduction. Besides, the CO₂-TPD tests and the control experiments confirmed that the Cu species or the Cu–Co interfaces could provide strong CO₂ adsorption sites to convert CO₂ to CO, and further form C₂₊ hydrocarbons.

It should be mentioned that, although CH₄ was probably not formed via the CO route, CO could be produced with an activity of 1.1 mmol g_cat⁻¹ h⁻¹ (entry 8, Table 1). Importantly, the higher concentration of CO could make it easily coupled into the C₂₊ products over the same Co₇Cu₁Mn₁O_x(200) catalyst. Thus, C₂₊ hydrocarbons could be formed by increasing the CO₂ concentration.

In addition, the present protocol could be extended to photothermal CO hydrogenation. We conducted the reaction over the Co₇Cu₁Mn₁O_x(200) catalyst under the similar conditions given in Table 1, entry 8, and the results are shown in Fig. S17.† The photothermal activity is about 7 times higher than that of the thermal activity, and the selectivity of the C₂₊ products is higher than that of CH₄, which is similar to CO₂ hydrogenation. The results indicated that the present trimetallic Co–Cu–Mn catalyst has multifunctional applications as shown in photothermal CO hydrogenation.

The reusability and stability studies of the Co₇Cu₁Mn₁O_x(200) catalyst

The reusability of the Co₇Cu₁Mn₁O_x(200) catalyst was investigated, and the results are shown in Fig. 8a. For comparison, the CO₂ reduction reactions were alternatively conducted under photothermal and thermal conditions. The results revealed that the present catalyst possesses good reusability in the reaction, and after using for 18 times, it still offered an activity of higher than 10 mmol_CH4 g_cat⁻¹ h⁻¹. Besides, the results also confirmed that the activities under photothermal conditions were much higher than those under thermal conditions. The used catalysts were characterized by XRD and XPS, and the results are shown in Fig. 5, respectively.

Fig. 8 Reusability (a) and temporal stability (b) studies of Co₇Cu₁Mn₁O_x(200) in photothermal CO₂ reduction. Reaction conditions: (a) catalyst 50 mg, 200 °C, CO₂/H₂/N₂ 10%/30%/60%, full irradiation for photothermal conditions; (b) catalyst 50 mg, 200 °C, CO₂/H₂/N₂ 10%/30%/60% 20 mL min⁻¹, full irradiation.

The activity loss is probably caused by catalyst agglomeration, which was confirmed by the TEM images of the catalyst after being used 18 times (Fig. S18).† The size of the used catalyst increased in the range 40–90 nm.

Fig. 8b shows the temporal stability of the Co₇Cu₁Mn₁O_x(200) catalyst in a continuous flow operation (details are given in the ESI†). The catalyst offered approximately 20 mmol g_cat⁻¹ h⁻¹ of CH₄ after a short induction period of 3 h, and maintained higher than 18 mmol g_cat⁻¹ h⁻¹ even after 25 hours of reaction. The result showed that the present Co₇Cu₁Mn₁O_x(200) catalyst has a long-term temporal stability, confirming that it has a potential industrial application.

Conclusions

In summary, a trimetallic Co–Cu–Mn catalytic system for photothermal CO₂ reduction was constructed by a simple co-precipitation method. The metal compositions and reduction temperatures of the catalysts affected their activities remarkably, and the screened Co₇Cu₁Mn₁O_x(200) catalyst could offer 14.5 mmol g_cat⁻¹ h⁻¹ of CH₄ at low CO₂ (10%) and H₂ (30%) concentrations. Importantly, the catalyst system could produce C₂₊ hydrocarbons with an activity of 7.5 mmol g_cat⁻¹ h⁻¹ and a selectivity of 30.8% by only improving the CO₂ and H₂ concentrations to 25% and 75%, respectively. High activities stemmed from the multicomponents of the catalyst, especially the metallic Co⁰, Cu⁰, and CoO, Co₃O₄, Cu₂O, and Mn₂O₃ semiconductors. The Cu species could tune the reduction characteristics, and further tune the balance of Co²⁺ and Co³⁺. The Mn species could enhance the CO₂ and H₂ adsorption, and also offer a high local temperature for the nanocatalyst. The present strategy of partial reduction of the Co–Cu–Mn trimetallic catalyst not only highlights the route for using solar energy to produce valuable chemicals under mild conditions especially low CO₂ concentration, but also constructs the highly efficient trimetallic catalysts with multifunctional applications in photothermal catalysis.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors gratefully acknowledge the financial supports from the National Natural Science Foundation of China (21978160, 21776170, 22078182, 21706152, and 21908139), the Natural Science Basic Research Plan in Shaanxi Province of China (Program No. 2019JLM-16, 2019JQ-772, and 2019JQ-782) and the Key Industrial Innovation Project of Shaanxi Provincial Science and Technology Department (2019ZDLGY06-04).

Notes and references

1. M. He, Y. Sun and B. Han, Angew. Chem., Int. Ed., 2013, 52, 9620–9633.
2. L. Wang, Y. Dong, T. Yan, Z. Hu, A. A. Jelle, D. M. Meira, P. N. Duchesne, J. Y. Y. Qiu, E. E. Storey, Y. Xu, W. Sun, M. Ghoussoub, N. P. Kherani, A. S. Helmy and G. A. Ozin, Nat. Commun., 2020, 11, 2432.
3. B. Zhou, P. Ou, N. Pant, S. Cheng, S. Vanka, S. Chu, P. T. Rashid, G. Botton, J. Song and Z. Mi, Proc. Natl. Acad. Sci. U. S. A., 2020, 117, 1330–1338.
4. M. Cui, Q. Qian, J. Zhang, Y. Wang, B. B. A. Bediako, H. Liu and B. Han, Chem, 2021, 7, 726–737.
5. Z. He, M. Cui, Q. Qian, J. Zhang, H. Liu and B. Han, Proc. Natl. Acad. Sci. U. S. A., 2019, 116, 1254–12659.
6. Z. Ma and M. D. Porosoff, ACS Catal., 2019, 9, 2639–2656.
7. R. De, S. Gonglach, S. Paul, M. Haas, S. S. Sreejith, P. Gerschel, U.-P. Apfel, T. H. Vuong, J. Rabeah, S. Roy and W. Schöfberger, Angew. Chem., Int. Ed., 2020, 59, 10527–10534.
8. Y. Wang, X. Shang, J. Shen, Z. Zhang, D. Wang, J. Lin, J. C. S. Wu, X. Fu, X. Wang and C. Li, Nat. Commun., 2020, 11, 3043.
9. A. Weihard, S. P. Argent and V. Sans, Nat. Commun., 2021, 12, 231.
10. Z. He, Q. Qian, J. Ma, Q. Meng, H. Zhou, J. Song, Z. Liu and B. Han, Angew. Chem., Int. Ed., 2016, 55, 737–741.
11. L. Wang, E. Guan, Y. Wang, L. Wang, Z. Gong, Y. Cui, X. Meng, B. C. Gates and F.-S. Xiao, Nat. Commun., 2020, 11, 1033.
12. M. Rahaman, K. Kiran, I. Z. Montiel, V. Grozovski, A. Dutta and P. Broekmann, Green Chem., 2020, 22, 6497–6509.
13. G. Chen, R. Gao, Y. Zhao, Z. Li, G. I. N. Waterhouse, R. Shi, J. Zhao, M. Zhang, L. Shang, G. Sheng, X. Zhang, X. Wen, L.-Z. Wu, C.-H. Tung and T. Zhang, Adv. Mater., 2017, 1734663.
14. F. Sastre, A. V. Puga, L. Liu, A. Corma and H. García, J. Am. Chem. Soc., 2014, 136, 6798–6801.
15. J. Albero, E. Dominguez, A. Corma and H. Garcia, Sustain. Energy Fuels, 2017, 7, 1303–1307.
16. F. Zhang, Y.-H. Li, M.-Y. Qi, Y. M. A. Yamada, M. Anpo, Z.-R. Tang and Y.-J. Xu, Chem. Catal., 2021, 1, 1–26.
17. P. G. O'Brien, A. Sandhel, T. E. Wood, A. A. Jelle, L. B. Hoch, D. D. Perovic, C. A. Mims and G. A. Ozin, Adv. Sci., 2014, 1, 1400001.
18. D. Mateo, J. Albero and H. García, Joule, 2019, 3, 1–14.
19. A. A. Jelle, K. K. Ghuman, P. G. O'Brien, M. Hmadeh, A. Sandhel, D. D. Perovic, C. V. Singh, C. A. Mims and G. A. Ozin, Adv. Energy Mater., 2018, 8, 1702277.
20. J. Ren, S. Ouyang, H. Xu, X. Meng, T. Wang, D. Wang and J. Ye, Adv. Energy Mater., 2016, 1601657.
21. X. Meng, T. Wang, L. Liu, S. Ouyang, P. Li, H. Hu, T. Kako, H. Iwai, A. Tanaka and J. Ye, Angew. Chem., Int. Ed., 2014, 53, 11478–11482.
22. C. Wang, S. Fang, S. Xie, Y. Zheng and Y. H. Hu, J. Mater. Chem. A, 2020, 8, 7990–7394.
23. J. Jia, H. Wang, Z. Lu, P. G. O'Brien, M. Ghoussoub, P. Duchesne, Z. Zheng, P. Li, Q. Qiao, L. Wang, A. Gu, A. A. Jelle, Y. Dong, Q. Wang, K. K. Ghuman, T. Wood, C. Qian, Y. Shao, C. Qiu, M. Ye, Y. Zhu, Z.-H. Lu, P. Zhang, A. S. Helmy, C. V. Singh, N. P. Kherani, D. D. Perovic and G. A. Ozin, Adv. Sci., 2017, 1700252.
24. M. Cabrero-Antonino, S. Remiro-Buenamañana, M. Souto, A. A. García-Valdivia, D. Choquesillo-Lazarte, S. Navalón, A. Rodríguez-Diéguez, G. M. Espallargas and H. García, Chem. Commun., 2019, 55, 10932–10935.
25. D. Mateo, J. Albero and H. García, Energy Enviorn. Sci., 2017, 10, 2392–2400.
26. M.-P. Jiang, K.-K. Huang, J.-H. Liu, D. Wang, Y. Wang, X. Wang, Z.-D. Li, X.-Y. Wang, Z.-B. Geng, X.-Y. Hou and S.-H. Feng, Chem, 2020, 6, 2335–2346.
27. J. A. Rodriguez, P. Liu, D. J. Stacchiola, S. D. Senanayake, M. G. White and J. G. Chen, ACS Catal., 2015, 5, 6696–6706.
28. R. Lang, X. Du, Y. Huang, X. Jiang, Q. Zhang, Y. Guo, K. Liu, B. Qiao, A. Wang and T. Zhang, Chem. Rev., 2020, 120, 11986–12043.
29. K. Zhao, M. Calizzi, E. Moioli, M. Li, A. Borsay, L. Lombardo, R. Mutschler, W. Luo and A. Zuttel, J. Energy Chem., 2021, 53, 241–250.
30. M. Ao, G. H. Pham, V. Sage, V. Pareek and S. Liu, Fuel Process. Technol., 2019, 193, 141–148.
31. Y. Xiang, R. Barbosa, N. Kruse and X. Li, ACS Catal., 2015, 5, 2929–2934.
32. Y. Wang, S. Kattel, W. Gao, K. Li, P. Liu, J. G. Chen and H. Chen, Nat. Commun., 2019, 10, 1166.
33. A. A. S. Goncalves, P. B. Faustino, J. M. Assaf and M. Jaroniec, ACS Appl. Mater. Interfaces, 2017, 9, 6079–6092.
34. Y. Li, H. S. Pillai, J. Lattimer, N. M. Adli, S. Karakalos, M. Chen, L. Guo, H. Xu, J. Yang, D. Su, H. Xin and G. Wu, ACS Catal., 2020, 10, 3945–3957.
35. M. Sharma, J.-H. Jang, D. Y. Shin, J. A. Kwon, D.-H. Lim, D. Choi, H. Sung, J. Jang, S.-Y. Lee, K. Y. Lee, H.-Y. Park, N. Jung and S. J. Yoo, Energy Environ. Sci., 2019, 12, 2200–2211.
36. L. Martin, H. Martinezz, D. Poinot, B. Pecquenard and F. L. Cras, J. Phys. Chem. C, 2013, 117, 4421–4430.
37. Z. Geng, Y. Wang, J. Liu, G. Li, L. Li, K. Huang, L. Yuan and S. Feng, ACS Appl. Mater. Interfaces, 2016, 8, 27825–27831.
38. L. Lukaskuk, N. Yigit, R. Rameshan, E. Kolar, D. Teschner, M. Havecker, A. Knop-Gericke, R. Schlogl, K. Fottinger and G. Rupprechter, ACS Catal., 2018, 8, 8630–8641.
39. J. Su, Z. Zhang, D. Fu, X. Xu, B. Shi, X. Wang, R. Si, Z. Jiang, J. Xu and Y. Han, J. Catal., 2016, 336, 94–106.
40. B. Rivas-Murias and V. Salgueiriño, J. Raman Spectrosc., 2017, 48, 837–841.
41. X. Jiang, X. Nie, X. Guo, C. Song and J. G. Chen, Chem. Rev., 2020, 120, 7984–8034.
42. Y. Xiang, V. Chitry, P. Liddicoat, P. Felfer, J. Cairney, S. Ringer and N. Kruse, J. Am. Chem. Soc., 2013, 135, 7114–7117.
43. Z.-H. He, C.-S. Jiang, K. Wang, Z.-Y. Wang, N. Li, W.-T. Wang and Z.-T. Liu, Catal. Today, 2020, 356, 579–588.
44. M. Tahir, B. Tahir, N. A. S. Amin and A. Muhammad, Energy Convers. Manag., 2016, 119, 368–378.
45. Z.-H. He, N. Li, K. Wang, W.-T. Wang and Z.-T. Liu, Mol. Catal., 2019, 470, 120–126.
46. O. A. Bulavchenko, E. Y. Gerasimov and T. N. Afonasenko, Dalton Trans., 2018, 47, 17153–17159.
47. W. Yang, Y. Peng, Y. Wang, Y. Wang, H. Liu, Z. Su, W. Yang, J. Chen, W. Si and J. Li, Appl. Catal., B, 2020, 278, 119279.
48. W. Yang, S. Wang, K. Li, S. Liu, L. Gan, Y. Peng and J. Li, Chem. Eng. J., 2019, 364, 448–451.
49. C.-I. Ahn, D.-W. Jeong, J. M. Cho, H.-S. Na, W.-J. Jang, H.-S. Roh, J.-H. Choi, S.-H. Um and J.-W. Bae, Microporous Mesoporous Mater., 2016, 221, 204–211.
50. O. A. Bulavchenko, E. Y. Gerasimov and T. N. Afonasenko, Dalton Trans., 2018, 47, 17153–17159.
51. G. Chen, R. Gao, Y. Zhao, Z. Li, G. I. N. Waterhouse, R. Shi, J. Zhao, M. Zhang, L. Shang, G. Sheng, X. Zhang, X. Wen, L. Wu, T. Tung and T. Zhang, Adv. Mater., 2018, 30, 1704663.
52. J. Zhao, Z. Zhao, N. Li, J. Nan, R. Yu and J. Du, Chem. Eng. J., 2018, 353, 805–813.
53. Y. Zeng, H. Li, Y. Xia, L. Wang, K. Yin, Y. Wei, X. Liu and S. Luo, ACS Appl. Mater. Interfaces, 2020, 12, 44608–44616.
54. Z. Li, J. Liu, Y. Zhao, R. Shi, G. I. N. Waterhouse, Y. Wang, L.-Z. Wu, C.-H. Tung and T. Zhang, Nano Energy, 2019, 60, 467–475.
55. A. Umasadharshini, M. Bououdina, M. Venkateshwarlu, C. Manoharan and P. Dhamodharan, Surf. Interfaces, 2020, 19, 100535.
56. Q. Wan, J. Zhang, B. Zhang, D. Tan, L. Yao, L. Zheng, F. Zhang, L. Lin, X. Cheng and B. Han, Green Chem., 2020, 22, 2750–2754.
57. N. Li, M. Liu, B. Yang, W. Shu, Q. Shen, M. Liu and J. Zhou, J. Phys. Chem. C, 2017, 121, 2923–2932.
58. X. Zhang, X. Li, D. Zhuang, N. Q. Su, W. Yang, H. O. Everitt and J. Liu, Nat. Commun., 2017, 8, 14542.
59. P. Christopher, H. Xin and S. Linic, Nat. Chem., 2011, 3, 467–472.
60. B. Miao, S. S. K. Ma, X. Wang, H. Su and S. H. Chan, Catal. Sci. Technol., 2016, 6, 4048–4058.
61. B. Tahir, M. Tahir and N. S. Amin, Energy Convers. Manage., 2015, 90, 272–281.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/d1gc01152a

---