Introduction

Nanomaterials, with the merits of high catalytic activity and high atomic utilization, play a crucial role in numerous fields such as materials, energy, and chemistry1,2,3,4,5. However, due to the high surface activity, nanomaterials tend to sinter into larger particles at elevated temperatures6,7,8, resulting in catalytic deactivation9,10. Representatively, Cu-based nanomaterials are the benchmark catalysts of diverse industrial processes, such as methanol steam reforming11, methanol synthesis12,13, water gas shift reaction14, and emerging photothermal catalysis9,15. But, the Taman temperature (~400 °C) of Cu-based nanocatalysts is always lower than the operating temperature of industrial processes and photothermal catalysis (450 °C)16,17, which shortens the service life of industrial catalytic systems and reduces the solar to chemical energy conversion efficiency. To date, strong metal-support interaction (SMSI) is the predominant approach for enhancing the sintering resistance of nanocatalysts18,19. Typically, Sun et al. have reported that the 2D silica supported Cu nanoparticles exhibit a stable CO2 hydrogenation rate of ~60 mmol g−1 h−1 at 500 °C via SMSI20. However, SMSI involves partial or complete encapsulation of Cu-based nanoparticles by heterogeneous materials overlayers21,22, which blocks the active Cu sites, impedes the transport of reactants and loses catalytic activity23,24. Therefore, regulating the structure of Cu-based nanocatalysts to obtain high catalytic stability and activity at high temperatures is important for the catalytic science25,26,27.

The structural rigidity of materials is proportional to the structural free energy (ΔG = ΔH-TΔS), where ΔH, ΔS represent the enthalpy change and entropy change, respectively28,29. In physical essence, previously reported stabilization strategies primarily focus on enhancing the structural enthalpy (ΔH)30,31. Herein, we proposed a high-entropy concept to strengthen the structural rigidity of Cu based nanocatalysts32, and a PVP templated method could generally and large-scale synthesize high-entropy two-dimensional (2D) materials. Due to the high structural entropy, 2D Cu2Zn1Al0.5Ce5Zr0.5Ox exhibited superior activity and stability for the reverse water gas shift reaction (RWGS, CO2 + H2 → CO + H2O) under high temperature and H2/air corrosion33,34. Consequently, the 2D Cu2Zn1Al0.5Ce5Zr0.5Ox could be extended to photothermal RWGS under harsh conditions, demonstrating unexpected CO2 conversion rate and solar to chemical energy conversion efficiency. In an outdoor photothermal catalysis test, solar-driven RWGS for 7 continuous days was realized by using 2D Cu2Zn1Al0.5Ce5Zr0.5Ox. This work offers a new pathway for low-temperature synthesizing high-entropy metal oxide nanocatalysts to realize the synergism of catalytic stability and activity of Cu based nanocatalysts.

Results

Low temperature synthesis of high-entropy two-dimensional materials

Several methods have been applied to synthesize high entropy materials35, such as the carbothermal shock technique and the fast-moving bed pyrolysis technique14,36. However, the synthetic accessibility of these methods is limited by high temperature (usually >1000 °C), specialized equipment and tedious procedures37,38. Therefore, it is urgent to develop a low temperature and simple method for preparing high entropy nanocatalysts. As illustrated in Fig. 1a, a polyvinylpyrrolidone (PVP) templated method was employed to synthesize high-entropy two-dimensional (2D) materials. High entropy materials generally contain more than 5 kinds of metal elements, and all elements are generally in equal proportion. To verify the universality of this method, 10 kinds of metal ions (Ce3+, Cu2+, Mn2+, Mg+, Al3+, Co2+, La3+, Zr4+, Ca2+, Y3+) with equal atomic proportion were added into this solution. During aging, the PVP was self-assembled into a 2D micelle (Supplementary Fig. 1). The freeze-drying process was applied to obtain solids of 2D PVP micelles loaded with various metal ions. After annealing the precursors in air at 450 °C, the mixed metal ions formed 2D materials. Figure 1b shows that 2D Ce1Cu1Mn1Mg1Al1Co1La1Zr1Ca1Y1Ox was grown in 2D morphology, and the eleven elements of Ce, Cu, Mn, Mg, Al, Co, La, Zr, Ca, Y, and O were all evenly distributed on the surface of 2D materials, which is the fingerprint feature of high-entropy materials36,39. The powder X-ray diffraction (XRD) pattern of 2D Ce1Cu1Mn1Mg1Al1Co1La1Zr1Ca1Y1Ox showed a single cubic fluorite phase (Supplementary Fig. 2)40,41, that belongs to a characteristic crystal structure of high-entropy metal oxide37,42. This evidence demonstrates that this method successfully synthesized the 2D high-entropy metal oxides. The preparation temperature of this PVP templated method is only 450 °C, significantly lower than the traditional high entropy material preparation methods (usually >1000 °C)37,38. Meanwhile, the instruments, chemicals, and steps required for this PVP templated method are simple and inexpensive. Using this method, we also prepared Cu based 2D high-entropy metal oxides. To optimize catalytic performance, the proportion of Cu2+, Zn2+, Al3+, Ce3+, Zr4+ was 2:1:0.5:5:0.5. Figure 1c presents the typical transmission electron microscopy (TEM) image of the as-prepared sample. It was clearly observed that the sample had a 2D morphology and no heterogeneous nanoparticles were grown on its surface. The corresponding XRD pattern exhibited four peaks at around 29.4°, 33.7°, 48.2°, and 57.2°, which are indexed to the (111), (200), (220), and (311) crystal planes of the single cubic fluorite phase (Fig. 1d)40,41. Atomically level high-angle annular dark-field scanning TEM (HAADF-STEM) revealed an inter-plane spacing measured to be 3.12 Å, representing the (111) planes of face centered cubic (FCC) phase (Fig. 1e)43,44. Furthermore, the elemental mapping images demonstrated the homogeneous distribution of Cu, Zn, Al, Ce, Zr, and O over the whole nanosheet (Fig. 1f). This sample was named 2D Cu2Zn1Al0.5Ce5Zr0.5Ox. Atomic force microscopy (AFM) confirmed that the thickness of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox was 4 nm, revealing its ultrathin nature (Supplementary Fig. 3). X-ray photoelectron spectroscopy (XPS) was employed to analyze the elemental chemical states of the prepared sample. The XPS analysis showed that all constituting metal elements of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox were in oxidation states (Supplementary Fig. 4).

Fig. 1: The preparation and characterizations of high-entropy 2D materials.
figure 1

a The preparation diagram of 2D high-entropy materials. b The TEM image and Ce, Cu, Mn, Mg, Al, Co, La, Zr, Ca, Y, O elemental mapping images of 2D Ce1Cu1Mn1Mg1Al1Co1La1Zr1Ca1Y1Ox. c TEM image, d XRD pattern, e HAADF-STEM image of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox. f Cu, Zn, Al, Ce, Zr, O elemental mapping images of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox. The scale bars in b, c, e, f are 300 nm, 2 μm, 2 nm, 50 nm, respectively.

Full size image

The CO2 hydrogenation activity

Cu-based nanocatalysts are active for reverse water gas shift reaction (RWGS, CO2 + H2 → CO + H2O), which is a fundamental reaction for the synthesis of methanol (CAMERR process) and alkanes (Fischer-Tropsch processes)45,46. 2D Cu2Zn1Al0.5Ce5Zr0.5Ox was applied for RWGS, commercial CuZnAlOx catalyst (Supplementary Fig. 5, named as Cu6Zn3Al1) and Cu doped CeO2 nanosheets (named as 2D Cu2Ce7Ox, Supplementary Figs. 6–8) were selected as reference samples. The 2D Cu2Ce7Ox had the SMSI effect of stabilizing the high dispersion of Cu species47,48. Figure 2a shows the RWGS CO production rates of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox, 2D Cu2Ce7Ox and Cu6Zn3Al1 at different temperatures. 2D Cu2Ce7Ox and Cu6Zn3Al1 showed peak CO generation rates of 50.1 mmol g−1 h−1 and 35.8 mmol g−1 h−1 at 450 °C and 400 °C, respectively. Then, the CO generation rates of 2D Cu2Ce7Ox and Cu6Zn3Al1 slowly dropped along with the further increase operation temperature, indicating their thermal instability. Moreover, the CO production rate of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox was monotonically increased to 417.2 mmol g−1 h−1 at 500 °C, which was higher than the reported advanced catalysts for RWGS at 500 °C as far as we know (Fig. 2b). For example, Cu/CeO2-δ (106.2 mmol g-1 h−1)49, FeCu/CeAl (102.9 mmol g−1 h−1)50, Pd/TiO2 (80 mmol g−1 h−1)51, Cu/2D silica (60 mmol g−1 h−1)20, Pt/CeO2 (45 mmol g−1 h−1)52, Co-Fe/Al2O3 (18 mmol g−1 h−1)53. Figure 2c displays the thermal RWGS stability of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox at 450 °C for 72 h. The CO production rate of 2D Cu2Ce7Ox was reduced from 50 mmol g−1 h−1 to ~25 mmol g−1 h−1 after 72 h test, corresponding to 50% inactivation, and the CO production rate of Cu6Zn3Al1 was reduced from 36 mmol g−1 h−1 to 7.2 mmol g−1 h−1 after 72 h test, corresponding to 80% inactivation. Meanwhile, the CO production rate of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox was maintained at ~355 mmol g−1 h−1 for 72 h, confirming its thermal stability. Additionally, the 2D Cu2Zn1Al0.5Ce5Zr0.5Ox showed 100% selectivity for CO2 converted as CO (Fig. 2d).

Fig. 2: Thermal RWGS performance of catalysts.
figure 2

a Thermal RWGS performance of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox, 2D Cu2Ce7Ox, commercial CuZnAlOx (Cu6Zn3Al1). b Visual contrast diagram of the RWGS CO production rates of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox (This work) and other advanced catalysts at 500 °C. c The RWGS stability of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox, 2D Cu2Ce7Ox and Cu6Zn3Al1 under 450 °C. d The CO selectivity of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox for thermal RWGS at different temperature. The errors of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox show standard deviation.

Full size image

In-situ characterizations

To directly observe the structure evolution of catalysts during RWGS, in-situ characterization was carried out by using an environmental transmission electron microscope (ETEM) setup, in which CO2 + H2 acted as feeding gas and the catalysts were heated by a chip to simulate RWGS. The RWGS was carried out on the pristine 2D Cu2Zn1Al0.5Ce5Zr0.5Ox and resulted in no sintering phenomenon during the heating ramp up from 400 °C to 800 °C (Fig. 3a). XRD pattern, HAADF-STEM images and electron diffraction pattern revealed that the crystal structure of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox was also robust after experiencing 800 °C of RWGS (Supplementary Figs. 9–11). In comparison, it was observed that several nanoparticles were sintered on the surface of 2D Cu2Ce7Ox during the heating process of RWGS from 400 °C to 800 °C (Fig. 3b). The high-resolution (HR)TEM image confirmed that these nanoparticles were metallic Cu (Supplementary Fig. 12). The H2 temperature-programmed reduction (H2-TPR) was applied to detect their evolution under reduction atmosphere (Fig. 3c). The H2-TPR curve of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox had no obvious fluctuations throughout the whole 100–500 °C temperature range (Fig. 3c), revealing the chemical stability of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox under H2 corrosion. While the H2-TPR curve of 2D Cu2Ce7Ox showed a reduction peak in the range of 220–300 °C (Fig. 3c), indicating that the Ce and O in 2D Cu2Ce7Ox had no obvious valence change (Supplementary Fig. 13), this H2-TPR peak indicated that Cu species undergo a Cu2+-Cu0 conversion during the RWGS process54. Then, the H2 reduced samples experienced an oxidation process by annealing in air at 300 °C. XPS was used to characterize the chemical states of Cu element before and after O2 corrosion. The XPS spectra of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox shown in Fig. 3d illustrated a similar oxidation state of Cu before and after O2 corrosion. Further, the XPS spectra shown in Fig. 3e confirmed that the Cu species in 2D Cu2Ce7Ox was changed from Cu0 to Cu2+55. It proved that the high-entropy 2D Cu2Zn1Al0.5Ce5Zr0.5Ox had an ultra-stable chemical state under the corrosion of H2 and air. In addition, 2D Cu2Zn1Al0.5Ce5Zr0.5Ox also showed the activity of CO oxidation (Supplementary Fig. 14)56, indicating the potential for catalytic versatility.

Fig. 3: In-situ characterizations of catalysts.
figure 3

a, b In-situ TEM observations of the 2D Cu2Zn1Al0.5Ce5Zr0.5Ox, 2D Cu2Ce7Ox at different temperatures of RWGS. c H2-TPR curves of the 2D Cu2Zn1Al0.5Ce5Zr0.5Ox and 2D Cu2Ce7Ox. d, e The Cu 2p XPS spectra of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox and 2D Cu2Ce7Ox before and after the oxidation process. The scale bars in a, b are 1 μm.

Full size image

Theoretical calculations

Density functional theory (DFT) was applied to investigate the mechanism of sintering resistance and CO2 hydrogenation activity of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox. Figure 4a illustrates the atomic structures and metallic Cu precipitation energy barriers of 2D Cu2Ce7Ox and 2D Cu2Zn1Al0.5Ce5Zr0.5Ox. The metallic Cu precipitation energy barrier of 2D Cu2Ce7Ox was 6.61 eV, which was significantly greater than the metallic Cu precipitation energy of pure CuO (1.69 eV, Supplementary Fig. 15). It was confirmed that using metal oxides such as CeO2 as support to introduce SMSI can weaken the sintering of Cu species. Meanwhile, the metallic Cu precipitation energy barrier in 2D Cu2Zn1Al0.5Ce5Zr0.5Ox was as high as 8.85 eV (Fig. 4a), clearly higher than that of 2D Cu2Ce7Ox (6.61 eV) and CuO (1.69 eV). Therefore, the sintering resistance of Cu species in 2D Cu2Zn1Al0.5Ce5Zr0.5Ox could be mainly attributed to the high-entropy change. Then we simulated the CO2 hydrogenation (CO2RR) of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox. Since the in-situ characterization revealed that the Cu species sintered in 2D Cu2Ce7Ox during RWGS, the model of 2D Cu2Ce7Ox was changed to the atomic structure shown in Fig. 4b, in which the metallic Cu nanoparticle was supported on CeO2 (Cu NP/CeO2). Figure 4c demonstrates the free-energy diagrams and the intermediate pathways of CO2RR on the Cu NP/CeO2 and 2D Cu2Zn1Al0.5Ce5Zr0.5Ox. For the case of Cu NP/CeO2, the release of Cu-CO intermediate (CO* + H2O(g)→CO(g)+H2O(g)) exhibited a free-energy change of 1.46 eV, marking it as the rate-limiting step. In comparison, the free-energy change of the rate-limiting step of CO2RR through 2D Cu2Zn1Al0.5Ce5Zr0.5Ox was calculated as 0.74 eV (the formation of Cu-COOH intermediate), which is 0.72 eV lower than that of Cu NP/CeO2 (1.46 eV). In terms of valence electron cloud distribution, the Bader charge of Cu in Cu NP/CeO2, Cu in 2D Cu2Zn1Al0.5Ce5Zr0.5Ox, and C in CO* was calculated as +0.15, +1.38, and −0.22 |e | , respectively. The electronegative value difference revealed that the coordination of Cu-CO in Cu NP/CeO2 and 2D Cu2Zn1Al0.5Ce5Zr0.5Ox were covalent and ionic, respectively57. Since the bond energy of ionic Cu-CO was lower than that of covalent Cu-CO, the dissociation of Cu-CO in 2D Cu2Zn1Al0.5Ce5Zr0.5Ox was easier than that of Cu-CO in Cu NP/CeO2. Therefore, the above results indicated that Cu-CO preferentially dissociation on 2D Cu2Zn1Al0.5Ce5Zr0.5Ox compared to Cu NP/CeO2 due to the transformation of Cu-CO bonding from covalent to ionic.

Fig. 4: Theoretical calculation of Cu precipitation and CO2RR.
figure 4

a Atomic structures of 2D Cu2Ce7Ox and 2D Cu2Zn1Al0.5Ce5Zr0.5Ox before and after metallic Cu precipitation with corresponding free energy changes. b Atomic structure of Cu NP/CeO2. c Energy profiles for CO2RR on Cu NP/CeO2 and 2D Cu2Zn1Al0.5Ce5Zr0.5Ox. The X-axis illustrates the intermediates; the Y-axis illustrates the energy levels of each stage.

Full size image

Photothermal RWGS

Photothermal catalysis is a new mode of photochemical energy conversion that can effectively convert solar energy to chemical energy via a pathway involving sunlight-thermal energy-chemical energy, especially under intense sunlight iirradiation58,59. This intense irradiation can cause high temperature in photothermal catalysis, thus deactivating nanocatalysts. Due to the excellent sintering resistance, chemical stability and RWGS activity, 2D Cu2Zn1Al0.5Ce5Zr0.5Ox was applied to the photothermal RWGS. The 2D Cu2Zn1Al0.5Ce5Zr0.5Ox was loaded into a homemade TiC-based photothermal device that was designed to fully absorb solar spectrum and convert it to thermal energy, which was used to heat the catalyst (detailed device synthesis can be found in the Supplementary methods and Supplementary Fig. 16). Under 1 sun (1 kW m−2 intensity) and 2 suns irradiation, the temperature of the 2D Cu2Zn1Al0.5Ce5Zr0.5Ox catalyst reached 350 °C and 459 °C, respectively (Fig. 5a). The photothermal RWGS measurement revealed that the CO production started at 0.6 sun irradiation, and the CO production rate was 37.4 mmol g−1 h−1 under 1 sun irradiation.(Fig. 5a). To make a comparison, we listed the state-of-the-art solar driven RWGS in Table 1. Most catalysts irradiated by 1 sun had little RWGS activity and the advanced reported CO generation rates were 0.0013 mmol g−1 h−1 for Bi2In2-zO3-x(OH)y60, 0.0012 mmol g−1 h−1 for In2O3-x(OH)y61. Therefore, the 1 sun driven photothermal CO generation rate of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox (37.4 mmol g−1 h−1) was far higher than the previously reported highest value. When the light intensity increased to 2 suns, the CO generation rate of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox was increased to 248.5 mmol g−1 h−1, at least 31 times higher than the record photothermal RWGS reported under concentrated sunlight (>20 suns) irradiation, e.g., BixIn2-xO3 (8 mmol g−1 h−1)62, Pd@HyWO3-x (3 mmol g−1 h−1)63, Pd/Nb2O5 (1.8 mmol g−1 h−1)64, In2O3-x(OH)y/Nb2O5 (1.4 mmol g−1 h−1)65, Pt/NaTaO3 (0.139 mmol g−1 h−1)66, Pd@SiNS (0.01 mmol g−1 h−1)67. The air corrosion photothermal RWGS through 2D Cu2Zn1Al0.5Ce5Zr0.5Ox is shown in Fig. 5b. During the initial 2 suns driven photothermal RWGS, the CO generation rate of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox was remained at ~250 mmol g-1 h-1 for ~10 h. Then the light and the feeding gas of CO2 + H2 were cut off for ~5 h. After the light was turned back on, the restarted photothermal RWGS still maintained ~250 mmol g−1 h−1.

Fig. 5: The photothermal RWGS performance of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox.
figure 5

a The temperature of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox and CO production rate of photothermal RWGS through 2D Cu2Zn1Al0.5Ce5Zr0.5Ox under different sunlight irradiation. b The CO generation rate of photothermal RWGS through 2D Cu2Zn1Al0.5Ce5Zr0.5Ox under 2 suns irradiation and light off working conditions. c The CO generation rate of photothermal RWGS demonstration with 100 g of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox under different sunlight irradiation. d The STC efficiency of photothermal RWGS demonstration with 100 g of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox under different sunlight irradiation. e The photograph of photothermal RWGS demonstration equipped with reflector in Hebei University. f The CO yield of photothermal RWGS demonstration equipped with reflector under outdoor sunlight irradiation, on December 12, 13, 14, 17, 18, 20, 21, 2021, in Baoding City, China. The errors show standard deviation.

Full size image
Table 1 The advanced solar driven RWGS through different catalysts
Full size table

In view of the mass production of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox and the photothermal device, we filled the photothermal device with 100 g 2D Cu2Zn1Al0.5Ce5Zr0.5Ox to make a demonstration to test its application potential (Supplementary Fig. 17). Under 1 sun, 2 suns irradiation, the CO generation rate of this demonstration was 12.3 L h−1, 61.5 L h−1, respectively (Fig. 5c). According to the experimental data, Fig. 5d showed that the solar to chemical energy conversion efficiency of the demonstration were calculated to be 14.4% and 36.2% under 1- and 2 suns irradiation, respectively (Details seen in Methods). As far as we know, the reported highest solar to chemical efficiency was ~31%9,68. This work reveals that the high entropy 2D materials made photothermal catalysis the highest photochemical energy conversion mode in the world. The demonstration was used for industrial outdoor photothermal RWGS. Figure 5e depicts the photograph of an outdoor photothermal RWGS demonstration, in which a TiC/Cu-based device was equipped with a parabolic reflector to concentrate the sparse outdoor sunlight to maintain a high solar driven temperature all day. The photothermal RWGS was tested on 7 successive sunny days in December 2021 in Baoding City of Hebei Province, China. In this continuous outdoor photothermal RWGS, the working time was from 9:00 AM to 16:00 PM, and the rest was the air corrosion time without the supply of feeding gas (CO2 + H2). As shown in Fig. 5f, the CO yield was 77.6, 62.6, 46.8, 98.2, 88.1, 118.8, 78.7 L on December 12, December 13, December 14, December 17, December 18, December 20, December 21, respectively. And the CO yield difference of each day was originated from the change of sunshine and solar irradiated temperature of the catalyst (Supplementary Figs. 18 and 19). It revealed that the 2D Cu2Zn1Al0.5Ce5Zr0.5Ox could realize the continuous operation of outdoor photothermal RWGS.

Discussion

In this work, a PVP assisted templated method was developed to synthesize high-entropy two-dimensional (2D) materials of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox, 2D Ce1Cu1Mn1Mg1Al1Co1La1Zr1Ca1Y1Ox, which showed a single cubic fluorite phase, ~4 nm thickness and uniform elemental distribution. The 2D Cu2Zn1Al0.5Ce5Zr0.5Ox for RWGS showed a stable 417.2 mmol g−1 h−1 of CO production rate at 500 °C and 100% CO selectivity. The in-situ characterizations revealed that the morphology and crystal structure of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox were robust under 800 °C of RWGS, and the chemical state of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox was rigid under H2 and air corrosion. DFT calculations revealed that the Cu precipitation energy barrier and RWGS reaction energy barrier over 2D Cu2Zn1Al0.5Ce5Zr0.5Ox was 8.85 eV and 0.74 eV, respectively, due to the high-entropy structure. Under 2 suns irradiation, the 2D Cu2Zn1Al0.5Ce5Zr0.5Ox loaded in a TiC-based device showed a 459 °C temperature of the catalyst, a RWGS CO generation rate of 248.5 mmol g−1 h−1 and 36.2% solar to chemical energy conversion efficiency. Furthermore, this demonstration was used for outdoor photothermal RWGS for continuous 7 days, exhibiting a CO yield was 77.6, 62.6, 46.8, 98.2, 88.1, 118.8, 78.7 L on December 12, December 13, December 14, December 17, December 18, December 20, December 21, 2021, respectively, under severe changes of natural sunlight. This study indicated that the high-entropy strategy is a new route for designing nanocatalysts with high activity and stability simultaneously, and promote the application of nanocatalysts. In view of the drastic changes in temperature and atmosphere of natural photothermal catalysis, the high-entropy 2D materials may also provide a cornerstone for the industrialization of natural photothermal catalysis.

Methods

Chemicals

Cu(NO3)2, Al(NO3)3·9H2O, Zr(NO3)4·5H2O, In(NO3)3, Mn(NO3)2·4H2O, Ca(NO)3·4H2O, Y(NO)3·6H2O, Mg(NO)2·6H2O were purchased from Macklin Co., Ltd. Ce(NO3)3·6H2O, La(NO3)3·6H2O, Co(NO3)2·6H2O were purchased from Kermel Co., Ltd. Zn(NO3)2·6H2O and PVP K30 were purchased from Fuchen Chemical Co., Ltd. The Cu6Zn3Al1 catalyst was purchased from Sichuan Shutai Chemical Technology Co., Ltd.

The synthesis of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox

Firstly, 4 g of PVP was dissolved in 20 ml of H2O. Then the solution was stirred by a magnetic agitator with the addition of 0.462 g of Cu(NO3)2, 0.366 g of Zn(NO3)2·6H2O, 0.231 g of Al(NO3)3·9H2O, 2.675 g of Ce(NO3)3·6H2O and 0.264 g of Zr(NO3)4·5H2O, in which the PVP/metal salts weight ratio was 1. After 0.5 h of stirring, the uniform solution was dripped into liquid nitrogen to make it freeze into ice quickly and it was freeze-dried for 48 h to remove H2O. The dried product was calcined in a muffle furnace at 450 °C (heating rate 1 °C min−1) for 6 h, and the obtained was named 2D Cu2Zn1Al0.5Ce5Zr0.5Ox.

The synthesis of 2D Ce1Cu1Mn1Mg1Al1Co1La1Zr1Ca1Y1Ox

The synthesis of 2D Ce1Cu1Mn1Mg1Al1Co1La1Zr1Ca1Y1Ox was similar to the preparation of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox, and the only difference was that the metal salts was the mixture of Ce(NO3)3·6H2O, Cu(NO3)2, Mn(NO3)2·4H2O, Mg(NO)2·6H2O, Al(NO3)3·9H2O, Co(NO3)2·6H2O, La(NO3)3·6H2O, Zr(NO3)4·5H2O, Ca(NO)3·4H2O, Y(NO)3·6H2O with 1:1:1:1:1:1:1:1:1:1 mole ratio.

The synthesis of 2D Cu2Ce7Ox

The synthesis of 2D Cu2Ce7Ox was similar to the preparation of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox, and the only difference was that the metal salts was the mixture of Cu(NO3)2, Ce(NO3)3·6H2O with 2:7 mole ratio.

Thermocatalytic RWGS

The thermocatalytic activity of catalysts for RWGS was tested by the fixed-bed reactor (XM190708-007, DALIAN ZHONGJIARUILIN LIQUID TECHNOLOGY CO., LTD) in continuous flow form. Typically, 20 mg of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox or 200 mg of 2D Cu2Ce7Ox or 200 mg of Cu6Zn3Al1 catalyst was placed in a quartz flow reactor and the feeding gas of CO2/H2 = 1/1 with 40 sccm of flow rate was regulated by the mass flow controller. The reaction products were tested by gas chromatography (GC) 7890 A equipped with FID and TCD detectors. Before thermal RWGS, the 2D Cu2Ce7Ox and Cu6Zn3Al1 with 200 mg weight were reduced by 10% H2/Ar mixture at 300 °C for 4 h with a flow rate of 100 sccm.

H2-TPR

Hydrogen temperature-programmed oxidation (H2-TPR) was carried out using an online gas chromatograph (GC-7090A) equipped with a TCD detector. In a typical process, 50 mg of catalyst was placed in a quartz tube (6 mm ID). Subsequently, TPR was performed by heating the samples from room temperature to 500 °C at the heating rate of 5 °C min−1, in the presence of a 10% H2/He mixture (50 sccm) flowing.

Photothermal RWGS

The photothermal RWGS of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox was similar to Thermocatalytic RWGS over 2D Cu2Zn1Al0.5Ce5Zr0.5Ox and the difference was that the 2D Cu2Zn1Al0.5Ce5Zr0.5Ox was loaded into a TiC/Cu-based device irradiated by a simulate solar light source (DL3000).

The photothermal RWGS demonstration

The photothermal RWGS demonstration was 100 g of 2D Cu2Zn1Al0.5Ce5Zr0.5Ox loaded into a TiC/Cu-based device with 4.4 cm of diameter and 45 cm of length (Supplementary Fig. 16). 120 L h−1 of CO2 and 120 L h−1 of H2 were simultaneously put into the photothermal RWGS demonstration, which was controlled by a mass flow controller. The photothermal RWGS demonstration was irradiated by a solar light source (DL3000). As shown in Supplementary Fig. 16, the demonstration is irradiated up and down by the light source, so its irradiation area is calculated as 4.4 cm*45 cm*2 = 396 cm2. The composition of produced gas was tested by GC 7890 A equipped with FID and TCD detectors.

Enthalpy change energy of chemicals

The enthalpy change energy of CO2 (g), CO (g), H2 (g), H2O (g) was −393.505, −110.541, 0, −241.818 kJ mol−1, respectively. And they are all in gas state.

Solar to chemical energy conversion efficiency (STC) calculation of photothermal RWGS demonstration

The STC of photothermal RWGS demonstration was calculated as follows:

$${{{{{\rm{STC}}}}}}=(\Delta {{{{{{\rm{H}}}}}}}\, *\, \varepsilon /24.5)/({I}\, *\, {S}\,*\,3600)$$
(1)

ΔH was the reaction Enthalpy change energy (CO2 (g) + H2 (g) → CO (g) + H2O (g), RWGS, ΔH = 41.15 kJ/mol), ε (L h−1) was the CO generation amount per hour detected by a flowmeter, I was the light intensity (kW m-2), S was the irradiated area of demonstration (0.0396 m2). The ε irradiated by 1 sun, 2 suns was 12.3 L h−1, 61.5 L h−1, respectively, corresponding to 14.4%, 36.2% of STC, severally.

Outdoor photothermal RWGS

The photothermal RWGS demonstration was first equipped with a reflector (30 cm width and 55 cm length). 60 L h−1 of CO2 and 60 L h−1 of H2 were simultaneously put into the demonstration for photothermal RWGS. The composition of produced gas was tested by GC 7890 A equipped with FID and TCD detectors. From 17:00 PM to night to 8:00 AM every day, the supply of CO2 and H2 was stopped from the photothermal RWGS demonstration.