Integrated CO₂ capture and reverse water–gas shift reaction over CeO₂-CaO dual functional materials

Abstract

Achieving carbon neutrality is one of the most important tasks to meet the environmental challenges due to excessive CO₂ emissions. Integrated CO₂ capture and utilization (ICCU) represents an effective process for direct utilization of CO₂-contained exhaust gas (e.g. flue gas), in which converting the captured CO₂ into CO via reverse water–gas shift (RWGS) reaction is a promising route. The dual functional materials (DFMs), containing CO₂ adsorbents and catalysts, are widely applied to achieve ICCU. The conventional active metals (Ni, Fe, etc.)-based DFMs and non-transition metal DFMs (e.g. CaO) are restricted by low CO selectivity, catalytic efficiency or CO generation in the CO₂ capture step. To address the above obstructs in the application of DFMs, the metal oxides-based DFMs, MO_x-CaO (M = Al, Ce, Ti or Zr), are synthesized and evaluated. The CeO₂-CaO outperformed the other metal oxides-based DFMs and possessed significantly improved catalytic performance. It is found that 33% CeO₂-CaO DFM displayed approximately 49% CO₂ conversion and approximately 100% CO selectivity in integrated CO₂ capture and reverse water–gas shift reaction (ICCU-RWGS) at 650°C, while CaO-alone only achieved approximately 20% CO₂ conversion at the same condition. The surface basicity of CeO₂ is revealed to contribute to the improved catalytic performance by enhancing CO₂ chemisorption and activation in the hydrogenation step. Furthermore, CeO₂-CaO material possessed excellent cycle stability in 20 cycles ICCU-RWGS, achieving a sustainable and high-efficient performance in CO₂ conversion and CO selectivity.

1. Introduction

Numerous countries have pledged to achieve carbon neutrality around the mid-twenty-first century to eliminate the severe greenhouse effect and accompanying environmental issues [1]. However, in the foreseeable future, it is still inevitable to use fossil fuels to meet the energy demands [2], which would emit a huge amount of CO₂. Carbon dioxide capture and utilization or storage (CCUS) processes are believed to be effective and essential solutions to meet the great challenges for carbon neutrality [3–6]. However, the high capital costs of CCUS processes [7], including CO₂ enrichments, transportation and heat management, obstruct the industrial deployments.

To directly use the diluted CO₂ in the exhaust gas, integrating CO₂ capture with utilization (ICCU) [8–10] into one process exhibits an impressive performance and attractive application potential [11]. Specifically, the ICCU process can be achieved by swinging the inlet gas between exhaust gas (e.g. flue gas) and reducing agent (e.g. H₂) isothermally over the dual functional materials (DFMs). The majority of researchers paid attention to converting captured CO₂ into CH₄ [8,12–16], which is mainly re-used as fuel with equivalent carbon emissions. Recently, researchers have achieved CO generation via ICCU by reverse water–gas shift reaction (ICCU-RWGS) [17–20]. The unconsumed H₂ mixed with CO (syn-gas) can be further introduced into the Fischer–Tropsch synthesis process to produce high-end chemicals (e.g. olefins), which possess a longer life cycle compared with fuels [21]. Various metal-functionalized CaO DFMs were demonstrated to realize ICCU-RWGS, such as Ni-CaO [18,19,22,23], Fe_xCo_yMg₁₀CaO [17], FeCrCu/K/hydrotalcite [24] and Fe-CaO [19,25]. Although the introduction of metals (e.g. Ni, Fe or Co) contributes to the effective catalytic RWGS in ICCU, the CO selectivity and undesirable CO generation during CO₂ capture restrict the further deployment [19]. Specifically, the CO generation in the CO₂ capture process is attributed to the reaction between CO₂ and reduced metallic metal (e.g. Fe) [25]. In recent study, removing transition metals from DFMs shows reduced CO generation during CO₂ capture and improved CO selectivity during the conversion of adsorbed CO₂ [26,27]. However, the absence of active metals significantly reduced the hydrogenation efficiency of DFMs and further restricted the cycle efficiency of ICCU. In short, there is a trade-off between impurity (i.e. CO) generation in the CO₂ capture process and the catalytic efficiency and CO selectivity in the RWGS process. It is necessary to develop novel DFMs to avoid CO generation in the CO₂ capture process while achieving enhanced RWGS efficiency with excellent CO selectivity.

Metal oxides (such as CeO₂ and TiO₂) are believed to be catalytically active in many reaction processes [28], such as reforming processes, photocatalysis and water–gas shift reaction [29]. Those metal oxides so far are mainly applied in ICCU by acting as the catalyst support for the active metal species [18,30,31]. However, there is still a knowledge gap in understanding the catalytic roles and other promotion effects of the metal oxides in DFMs during the ICCU process. Herein, we investigated ICCU performance over the DFMs composing various metal oxides (CeO₂, TiO₂, ZrO₂ and Al₂O₃)-CaO DFMs, in which CeO₂ is identified as the active metal oxide, while other metal oxides are benchmarks. The DFMs were produced by physically mixing the metal oxides and CaO. As illustrated in figure 1, the DFMs firstly act as the adsorbents to reduce the CO₂ emissions via carbonation, subsequently, the carbonated DFMs are converted in the H₂ atmosphere with the formation of CO. The catalytic performances of various DFMs and cycle stability were real-time studied using an online gas analyser and discussed with characterizations, in order to reveal the effect of the non-active metal containing CeO₂-CaO DFM on promoting the ICCU-RWGS.

Figure 1.

Schematic diagram of ICCU-RWGS.

2. Experimental section

2.1. Preparation of MO_x-CaO (M = Al, Ce, Ti or Zr) dual functional materials

The CeO₂ was synthesized using a hydrothermal method as previously reported [15,16]. Specifically, 5.21 g Ce(NO₃)₃·6H₂O (Sigma-Aldrich, 99%) was dissolved in deionized water (30 ml) to prepare a Ce source solution, followed by the dissolution of 57.6 g NaOH (Sigma-Aldrich, 99%) in deionized water (210 ml) to prepare the precipitant. The Ce source was mixed with the precipitant dropwise for 30 min at room temperature to obtain a slurry. The slurry was transferred into a stainless-steel autoclave and kept at 100°C for 24 h. The precipitate was washed and separated by vacuum filtration using distilled water and ethanol to neutrality and dried at 120°C overnight, to produce a yellow powder, labelled as CeO₂. The ZrO₂ (Sigma-Aldrich, 99%), TiO₂ (Sigma-Aldrich, 99.5%), Al₂O₃ (Sigma-Aldrich, 99.5%) and CeO₂ were calcined at 800°C for 2 h with a heating rate of 5°C min⁻¹ before mixing with CaO.

The CaO was derived by a sol–gel method as reported in previous literatures [19,25]. Briefly, 23.6 g Ca(NO₃)₂·4H₂O (Sigma-Aldrich, 99%) and 19.2 g citric acid monohydrate (Sigma-Aldrich, 99.5%) were dissolved into 72 ml distilled water, stirred at room temperature at 80°C and dried at 120°C overnight. The sample was ground and calcined at 850°C for 5 h at a heating rate of 5°C min⁻¹ to obtain CaO.

The MO_x (M = Al, Ce, Ti or Zr) and CaO are physically mixed by grinding (mass ratio: MO_X : CaO = 1 : 2) to prepare the MO_x-CaO DFMs.

2.2. Characterizations

X-ray diffraction (XRD) patterns of MO_x (M = Al, Ce, Ti or Zr) were measured using a PANalytical Empyrean Series 2 diffractometer with a Cu Ka X-ray source. The CO₂ temperature-programmed desorption (CO₂-TPD) patterns of MO_x (M = Al, Ce, Ti or Zr) were measured by a Micromeritics Autochem II 2920 analyser equipped with a TCD detector. Briefly, the MO_x were in situ reduced at 550°C in H₂ for 1 h and then cooled down to 30°C under He. After adsorbing CO₂ in a 10% CO₂/He gas mixture at 30°C, the temperature was increased to 800°C in He at a heating rate of 10°C min⁻¹. Scanning electron microscopy coupled with an energy-dispersive X-ray spectrometer (SEM-EDX, FEI Quanta FEG) was used to characterize the morphology and element dispersion.

2.3. Integrated CO₂ capture and reverse water–gas shift reaction evaluation

The ICCU-RWGS performances of MO_x-CaO DFMs were evaluated on a tubular fixed-bed reactor (stainless-steel tube; 500.0 mm in length and 10.2 mm in inner diameter). The reactor was placed in the middle of the furnace (Elite TSH-2416CG), filled with 0.3 g DFMs catalyst with quartz wool on both end of the catalysts. The thermocouple was placed in the middle of DFMs to control the temperature. The flow rates of the inlet gases were controlled by mass flow meters (OMEGA FMA2300), and the outlet gas (CO₂, CO and CH₄) was monitored by an online gas analyser (Kane Autoplus 5).

The typical ICCU-RWGS reaction procedure includes mainly two steps, i.e. carbonation and hydrogenation. In this work, all the DFMs catalysts were pretreated in 100 ml min^–1 5% H₂/N₂ at 550°C for 1 h to clean the surface of the catalysts and then equilibrated to the defined evaluation temperature in the range of 600–750°C in 100 ml min^–1 N₂. In the carbonation step, 100 ml min^–1 17% CO₂/N₂ (no added steam and O₂) was introduced for 1700 s to ideally simulate flue gas CO₂ capture. Subsequently, the hydrogenation step is to switch the gas to 5% H₂/N₂ at 100 ml min^–1 to convert CO₂ and regenerate the adsorbent. The cycle evaluations were carried out with extra 5 min N₂ purge among each ICCU-RWGS procedure.

The real-time CO₂ conversion, CO generation rate and CO selectivity in RWGS were calculated as equations (2.1)–(2.3).

$$C_{\mathrm{CO}2}={\displaystyle \frac{\mathrm{CO}\hspace{0.17em}+{\mathrm{CH}}_4}{\mathrm{CO}\hspace{0.17em}+{\mathrm{CH}}_4+{\mathrm{CO}}_2}\text{\%},}$$ 2.1

$$Y_{\mathrm{CO}}={\displaystyle \frac{\mathrm{CO}(\text{\%})\times 1.667\hspace{0.17em}\mathrm{ml}\hspace{0.17em}{\mathrm{s}}^{-1}}{0.0224\hspace{0.17em}\mathrm{ml}\hspace{0.17em}\mu {\mathrm{mol}}^{-1}\times 0.30\hspace{0.17em}\mathrm{g}}}$$ 2.2

$$\mathrm{and}{S}_{\mathrm{CO}}={\displaystyle \frac{\mathrm{CO}}{\mathrm{CO}\hspace{0.17em}+{\mathrm{CH}}_4}\text{\%}\hspace{0.17em},}$$ 2.3

where C_CO2, Y_CO and S_CO represent CO₂ conversion (%), CO generation rate (μmol g_DFM s⁻¹) and CO selectivity (%). The catalytic performance (CO₂ conversion, CO yield and CO selectivity) throughout the hydrogenation step was evaluated by the integration of real-time CO₂ conversion, CO generation rate and CO selectivity.

3. Results and discussions

3.1. Characterizations of MO_x (M = Al, Ce, Ti or Zr)

The XRD patterns of the produced metal oxides (MO_x) are shown in figure 2a. The CeO₂ and the benchmarks Al₂O₃, TiO₂ and ZrO₂ all possess pure crystal phase after elevated temperature pretreatment and are consistent with PDF75-1864, PDF78-0694, PDF75-1753 (rutile, rTiO₂), PDF83-2243 (anatase, aTiO₂) and PDF86-1451, respectively. Al₂O₃, ZrO₂ and CeO₂ performed no interaction with CaO during ICCU evaluation, while the TiO₂ formed a small amount of CaTiO₃ with CaO (figure 2b). The CO₂-TPD profiles are shown in figure 2c to evaluate the basicity of MO_x. Notably, only CeO₂ possessed distinct CO₂ desorption peaks, while the other three benchmark MO_x exhibited negligible basic property. For the CO₂ desorption on CeO₂, two major CO₂ desorption peaks appear at 110 and 234°C, representing the weak and medium basic sites, respectively [32,33]. Furthermore, the CeO₂ material exhibits weak high-temperature CO₂ desorption signal (400–600°C), which might be attributed to the strong interaction of CO₂ and CeO₂ [33]. The basicity of the catalyst is believed to benefit the adsorption and catalytic activation of CO₂ in CO₂ reduction process. The porosity of MO_x is highly related to the diffusion of reactants and exposure of active sites. As summarized in table 1, the CeO₂ exhibited the most abundant pores, which might contribute to the CO₂ diffusion, chemisorption and then activation in ICCU.

Figure 2.

XRD patterns of (a) M

O_x and (b) spent M

O_x-CaO; (c) CO₂-TPD profiles of MO_x (M = Al, Ce, Ti or Zr).

Table 1.

Pore information of M

O_x materials.

materials	BET surface area m2 g−1	pore volume cm3 g−1
ZrO2	5.46	0.02
TiO2	11.85	0.03
CeO2	65.76	0.25
Al2O3	0.65	<0.01

3.2. Integrated CO₂ capture and reverse water–gas shift reaction performance over MO_x-CaO dual functional materials

The real-time catalytic performances of ICCU-RWGS using MO_x-CaO (M = Al, Ce, Ti or Zr) DFMs are presented in figure 3. The Al₂O₃ TiO₂ and ZrO₂ are widely recognized as inert materials in thermal catalytic processes, which are applied as the benchmark in this work. To strictly exclude the potential effects of inert metal oxides on ICCU-RWGS, 0.2 g CaO without any MO_x (M = Al, Ce, Ti or Zr) was also evaluated for comparison. In the previous work [19,26], 650°C was suggested as the optimal temperature for CaO in ICCU based on the carbonation-decarbonation kinetics. Herein, the ICCU-RWGS evaluations using physically mixed MO_x and CaO formed DFMs (MO_x-CaO) were firstly carried out at 650°C (figure 3).

Figure 3.

ICCU-RWGS performance of CeO₂-CaO at 650°C (a); real-time CO₂ conversion and CO selectivity (b) and CO generation rate (c) of ICCU-RWGS over MO_x-CaO (M = Al, Ce, Ti or Zr) DFMs at 650°C.

As shown in figure 3a, the CaO in DFMs capture CO₂ via carbonation from a 100 ml min^–1 17% CO₂/N₂ gas mixture (simulating flue gas) for 1700 s. After that, the carbonated MO_x-CaO DFMs were reduced in a 5% H₂/N₂ at 100 ml min^–1 for reverse water–gas shift reaction (RWGS), as shown in figure 3a. The real-time carbonated CO₂ conversion possessed a gradually increasing trend as a function of time in the hydrogenation step. In the initial stage, the CO₂ conversion was hindered by the fast CO₂ release rate ascribed to the rapid decomposition of surface carbonates [19], which decreases the H₂ partial pressure (i.e. concentration) around DFMs and hinders the CO₂ conversion with H₂. With the consumption of the surface carbonates, the subsurface carbonate species exhibit a slower CO₂ release rate, then the increased H₂ partial pressure over the surface of the DFMs enhanced CO₂ hydrogenation performance with increased CO₂ conversion as shown in figure 3b. The inhibition of CO₂ release on the enhancement of the catalytic performance can be further evidenced by the real-time CO generation rate (figure 3b). In the initial approximately 250 s hydrogenation, the CO generation rate was even lower than 0.5 μmol g_DFM s⁻¹ on all tested DFMs. Notably, all the tested DFMs exhibited excellent CO selectivity (greater than 99%), which outperformed the commonly applied Ni-based DFMs [19].

The Al₂O₃-CaO and ZrO₂-CaO DFMs, assigned as inert benchmarks, possessed similar ICCU-RWGS performance compared with CaO-alone. In the previous research [26], the CaO-alone was proven active for ICCU-RWGS via direct hydrogenation of carbonates. It can also be concluded that TiO₂ is inactive in ICCU-RWGS according to the poor performance of TiO₂-CaO DFM. And the interaction between TiO₂ and CaO (CaTiO₃) performs no promotion effect on CO₂ hydrogenation in ICCU. The CeO₂-CaO outperformed all the other tested DFMs and possessed superior catalytic CO₂ conversion (approx. 60%) and CO generation rate (approx. 1.5 μmol g_DFM s⁻¹) at 650°C (figure 3c).

3.3 . The effect of CeO₂ mass fraction on integrated CO₂ capture and reverse water–gas shift reaction performance

To identify the optimal fraction of CeO₂ in the CeO₂-CaO DFM, a set of DFMs (CeO₂ fraction: 10–67 wt%) were evaluated for ICCU-RWGS performance at a temperature range of 600°C to 750°C (figure 4). The 33 wt% CeO₂ in DFM exhibited the optimal CO₂ conversion at the tested temperatures. The low CeO₂ fraction can restrict the catalytic performance due to the insufficient catalytic sites, while a CeO₂ fraction over 50% can lead to a restriction of the adsorbent amount and hinder the overall CO₂ capture performance of DFM. As can be seen from the CO yield results in figure 4b, a lower CeO₂ fraction, representing a higher CaO fraction, could provide more carbonates and thus achieve a higher CO yield. For example, 10% and 67% CeO₂-CaO DFMs achieved 3.5 and 1.0 mmol g⁻¹_DFM CO yield at 700°C, respectively.

Figure 4.

The effect of CeO₂ fraction in CeO₂-CaO DFMs over ICCU-RWGS at various temperature: (a) CO₂ conversion and (b) CO yield.

The reaction temperature is another key parameter in ICCU-RWGS using CeO₂-CaO DFMs. The RWGS is an endothermic reaction [34], which prefers a higher temperature. However, the simultaneous decomposition of CaCO₃ is more intense to release excessive CO₂ at higher reaction temperature [35], especially in the tested temperature range. The trade-off between RWGS performance and carbonates decomposition can reflect on CO₂ conversion and CO yield (figure 4). As shown in figure 4a, the CeO₂-CaO DFMs possessed relatively higher CO₂ conversion and reasonable CO yield at 650°C in approximately 1700 s hydrogenation, which is proven a suitable temperature for ICCU-RWGS.

3.4. Cycle performance of integrated CO₂ capture and reverse water–gas shift reaction using 33% CeO₂-CaO dual functional material

The cycle stability of the optimal 33% CeO₂-CaO DFM was further evaluated on ICCU-RWGS at the determined suitable reaction temperature of 650°C. As shown in figure 5, 33% CeO₂-CaO possessed impressive stable catalytic performance in 20 reaction cycles. Specifically, the CO₂ conversion and CO selectivity using 33% CeO₂-CaO were sustained at approximately 49% and greater than 99%, respectively. As shown in figure 6, the morphology and chemical composition of CeO₂ and CaO in the catalyst were stable after cyclic evaluation, indicating the outstanding stability of CeO₂-CaO DFM in ICCU-RWGS. Furthermore, the 33% CeO₂-CaO DFM possessed superior catalytic activity (approx. 49%) compared with CaO-alone, which achieved only approximately 20% CO₂ conversion during ICCU-RWGS cycles.

Figure 5.

Cycle performance of ICCU-RWGS using 33% CeO₂-CaO DFM at 650°C.

Figure 6.

SEM-mapping images of fresh and spent (after 20 cycles) CeO₂-CaO DFM.

In the previous research, the CeO₂ possessed excellent synergistic catalytic performance due to the oxygen vacancy and abundant coordination defect [18,23,36]. In ICCU-RWGS, the CeO₂ can provide a basic surface (figure 2c) to promote CO₂ chemisorption and activation [23]. It is speculated that the CO₂ is decomposed from carbonates, adsorbed on the surface of CeO₂ and activated by its oxygen vacancies [37]. It is also believed that H₂ could interact with ceria to promote CO₂ hydrogenation [38,39]. Notably, the CeO₂-CaO DFM can effectively and selectively (greater than 99%) convert CO₂ into CO in the absence of active metals (e.g. Ni and Ru). It is known that CH₄ formation is highly related to H₂ dissociation [40,41], which hardly occurred in the absence of H₂-sensitive active metals. Although some metals can achieve excellent CO selectivity in ICCU-RWGS via the redox pathway [19], such as Fe, a new drawback arises. Specifically, the Fe will be in the form of a metallic state after hydrogenation, which leads to undesirable CO formation in the following CO₂ capture (equation (3.1)). Compared with that, the CeO₂-CaO DFM demonstrated in this study is evidenced to address the above obstructions on undesirable CO formation but with superior and sustainable catalytic activity to realize a more promising ICCU-RWGS process.

$$M+XC{O}_2\to XCO+M{O}_X\hspace{0.17em}(M=\mathrm{Fe},\hspace{0.17em}\mathrm{Co}\hspace{0.17em}\mathrm{etc}.).$$ 3.1

4. Conclusion

ICCU is an emerging process, which provides a more direct path between CO₂-contained exhaust gas and catalytic conversion. The existing DFMs for ICCU-RWGS reaction meet obstructions on low CO selectivity, catalytic efficiency or undesirable CO generation in CO₂ capture. Herein, a non-active metal containing CeO₂-CaO DFM was synthesized to overcome the obstructions using simple and easy-access physically mixing method. The investigation of the effect of different MO_x-CaO (M = Al, Ce, Ti or Zr) DFMs with different mass fraction of MO_x show that CeO₂-CaO DFM with 33 wt% of CeO₂ possessed significantly enhanced and sustainable (20 reaction cycles) ICCU-RWGS catalytic performance. Specifically, approximately 49% CO₂ conversion and approximately 100% CO selectivity were achieved over the 33% CeO₂-CaO DFM. As a comparison, CaO-alone could only realize approximately 20% CO₂ conversion. The superior surface basicity on CeO₂ is believed to contribute to CO₂ chemisorption and activation among all tested MO_x (M = Al, Ce, Ti or Zr). The CeO₂-CaO DFMs exhibit a cost-effective, noble metal-free, stable and highly efficient materials candidates for promising industrial application of ICCU.

Data accessibility

This article has no additional data.

Authors' contributions

S.S.: conceptualization, data curation, investigation, methodology, validation, writing—original draft and writing—review and editing; C.Z.: data curation and investigation; S.C.: data curation and methodology; X.Z.: data curation and investigation; Y.W.: data curation and investigation; S.X.: conceptualization, data curation, funding acquisition, methodology, project administration, resources, supervision, writing—original draft and writing—review and editing; C.W.: conceptualization, data curation, investigation, methodology, project administration, resources, supervision, validation, visualization, writing—original draft and writing—review and editing.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We declare we have no competing interests.

Funding

The authors gratefully acknowledge financial support from the China Scholarship Council (reference number: 201906450023). This project has received funding from the European Union's Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement no. 823745. The UK Catalysis Hub is kindly thanked for the resources and support provided via our membership of the UK Catalysis Hub Consortium and funded by EPSRC grant: EP/R026939/1, EP/R026815/1, EP/R026645/1, EP/R027129/1 or EP/M013219/1(biocatalysis). The University of Manchester is kindly thanked for funding this research.