Application of Li-, Mg-, Ba-, Sr-, Ca-, and Sn-doped ceria for solar-driven thermochemical conversion of carbon dioxide

The redox reactivity of the Li-, Mg-, Ca-, Sr-, Ba-, and Sn-doped ceria (Ce0.9A0.1O2−δ) toward thermochemical CO2 splitting is investigated. Proposed Ce0.9A0.1O2−δ materials are prepared via co-precipitation of the hydroxide technique. The composition, morphology, and the average particle size of the Ce0.9A0.1O2−δ materials are determined by using suitable characterization methods. By utilizing a thermogravimetric analyzer setup, the long-term redox performance of each Ce0.9A0.1O2−δ material is estimated. The results obtained indicate that all the Ce0.9A0.1O2−δ materials are able to produce steady amounts of O2 and CO from cycle 4 to cycle 10. Based on the average nO2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ n_{{{\text{O}}_{2} }} $$\end{document} released and nCO\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ n_{\text{CO}} $$\end{document} produced, the Ce0.899Sn0.102O2.002 and Ce0.895Ca0.099O1.889 are observed to be the top and bottom-most choices. When compared with the CeO2 material, all Ce0.9A0.1O2−δ materials showed elevated levels of O2 release and CO production.


Introduction
The rise in the world's population is one of the primary reasons for the increase in the consumption of energy. As per the sources, the energy requirement of the world will increase up to 30 TW by 2050 [1]. More than 80% of the current energy necessity has been fulfilled by utilizing fossil fuels such as natural gas, oil, and coal. This excessive use of fossil fuels increased the atmospheric concentration of carbon dioxide. Plenty of research is currently focused on capturing the CO 2 from the industrial off-gases and sequestering it in the possible storage locations [2][3][4][5]. The utilization of CO 2 for the manufacturing of valuable products is a possible alternative to deal with CO 2 -related issues. One of the options is to convert CO 2 into syngas, which can be further utilized in the Fischer Tropsch process [6]. Two-step metal oxide (MO)-based solar thermochemical cycle (STC) is one of the possible ways for the production of syngas via H 2 O (WS) and CO 2 splitting (CS) [7][8][9]. The requirement of lower operating temperatures compared to direct thermolysis and utilization of the same MO in multiple cycles are some of the major advantages associated with the STCs. Besides, the H 2 /CO and O 2 are produced in two different steps; hence, the need for a cost-intensive separation method is avoided. Previously, zinc oxide [10,11], tin oxide [12,13], iron oxide [14,15], ceria and doped ceria [16,17], ferrites [18][19][20][21][22], perovskites [23][24][25][26][27], and others [28][29][30][31] have been considered for the STCs. Among these MOs, ceria-based oxides showed long-term stability and faster reaction kinetics for multiple STCs. In general, the ceria-based oxide is thermally reduced by releasing lattice O 2 at higher temperatures. This reduced ceria-based oxide is then re-oxidized either by WS or CS reaction resulting in the formation of H 2 or CO.
Ceria was first investigated for the solar thermochemical conversion of H 2 O into H 2 by Abanades and Flamant in 2006 [32]. Rhodes et al. [33] examined ceria toward the CS reaction in more than 2000 STCs. Chueh et al. [34] analyzed the redox reactivity (RR) of ceria in 500 cycles by using a cavity-receiver reactor driven by concentrated solar power. Venstrom et al. [35] studied the isothermal operation of ceria-based thermochemical CS by performing 100 cycles. Scheffe and Steinfeld [36] thermodynamically scrutinized the application of Gd-, Y-, Sm-, Ca-, and Sr-doped ceria toward the splitting of H 2 O and CO 2 at different temperatures and oxygen partial pressures. The H 2 production capacity of the combustion synthesized Cu-, Ni-, Mn-, and Fe-doped ceria via thermochemical WS reaction was estimated by Kaneko et al. [37]. Likewise, the Zr-, Sr-, Sc-, Y-, Dy-, Mg-, Hf-and Cadoped ceria were prepared and tested toward WS reaction by Meng et al. [38]. Bhosale et al. [39] developed and tested Hf ?4 -and Zr ?4 -doped ceria toward the thermochemical CS reaction. Recently, Takalkar et al. [40,41] examined multiple transition metals and lanthanides as the potential dopants for the ceria.
The results reported in the previous studies mainly show that the doped ceria possesses better fuel production capacity as compared to the undoped ceria. To explore further, this investigation reports the utilization of metal cations from the alkali (Li), alkaline earth (Mg, Ca, Sr, Ba), and post-transition (Sn) section of the periodic table as the possible dopants for the ceria material. In previous studies [38,42,43], the Mg, Ca, Sr, and Li-doped ceria materials were explored for the WS application; however, their utilization toward CS is not investigated. Furthermore, the Sndoped ceria was considered for the CS application [44]; nevertheless, the molar concentration of the dopant was 20%. This study reports the utilization of Li-, Mg-, Ca-, Sr-, Ba-, and Sn-doped ceria (Ce 0.9 A 0.1 O 2-d , where A = dopant) for the thermochemical splitting of CO 2 .

Material preparation and characterization
The synthesis of Ce 0.9 A 0.1 O 2-d materials was carried out by using a co-precipitation of the hydroxide method. The metal nitrates and an aqueous solution of ammonium hydroxide (28% NH 4 OH) were acquired from Sigma-Aldrich and utilized during the synthesis step without any pre-treatment. Deionized water was used for the dissolution of metal nitrates. Ultrapure Inert Ar (purity = 99.999%) needed as a carrier gas, and a gas mixture containing 50%CO 2 and 50%Ar (utilized as the reactive gas) were procured from Buzwair Scientific and Technical Gases, Doha, Qatar.
The dopant amounts required for the synthesis of Ce 0.9 A 0.1 O 2-d materials (basis 1 g) were controlled by estimating them based on the mole balance. The calculated amounts of metal nitrates were dissolved in 300 ml of deionized water at room temperature. Once the salts were dissolved entirely in deionized water, aqueous ammonium hydroxide was added to the solution in a dropwise manner (to attain a pH of the solution * 10). The mixture was agitated for 24 h with a sustained pH * 10. During the continuous stirring, due to precipitation, the color of the solution changes from colorless (at time = 0 h) to pale yellow (at time = 24 h). The next day, the stirring was stopped, and the solution obtained was kept undisturbed for 24 h allowing the precipitate to settle at the bottom of the beaker. The precipitated solids were recovered by decanting the supernatant liquid. The solids obtained were washed with water (using a vacuum filtration unit) to eliminate the unreacted chemicals. The solids obtained after vacuum filtration was dried at 120°C for 5 h. The powder obtained after drying was crushed and annealed up to 1000°C in the air for 4 h. Before conducting the TGA experiments, the calcined powder of Ce 0.9 A 0.1 O 2-d materials was analyzed by using the following techniques: a. Panalytical XPert powder X-ray diffractometer with CuKa radiation (voltage = 45 kV, current = 20 mA, k = 0.15418 nm. b. Scanning electron microscope (SEM Nova Nano 450, FEI) equipped with energy-dispersive spectroscopy (EDS) Thermogravimetric CS experiments Figure 1 shows the SETSYS Evolution thermogravimetric analyzer (TGA) setup used in this investigation. The details associated with the various parts of this setup are already reported in our previous investigations [40,41]. During the TGA experiments, the graphite heater is protected from the oxidizing gases by utilizing an inert Ar. It was also applied as a carrier gas to avoid probable vapor oxidation of the reactive sample due to gases evolved during the thermochemical reactions. The temperature of the gas streams exiting the TGA setup was controlled and maintained in a safe range by using a continuous flow of chilled water (Julabo FC 1600T). The carrier and protective Ar gas flow rates were measured and monitored continuously with the help of mass flow controllers.
As a starting point of the TGA experiments, * 50 mg of Ce 0.9 A 0.1 O 2-d powder was placed inside the heating furnace with the help of a platinum (100 ll) crucible. Before each thermochemical experiment, the residual air residing in the hollow space of the furnace was purged with inert Ar. The space of the furnace was then filled with a protective Ar gas with (100 ml/min). The thermal reduction (TR) of the Ce 0.9 A 0.1 O 2-d materials was carried out in the presence of inert Ar (at 1400°C for 60 min), whereas the CS reaction was conducted by utilizing a gas mixture containing 50%CO 2 ? 50%Ar (at 1000°C for 30 min). Before analyzing the RR of the Ce 0.9 A 0.1 O 2-d materials, a blank TGA experiment was performed by using an empty platinum crucible. The data obtained during the blank experiments were subtracted from the actual TAG experiments (performed by using the Ce 0.9 A 0.1 O 2-d materials) to eliminate the thermal buoyancy effect. The mass loss ðDm loss Þ recorded during each TR step and the mass gain (Dm gain ) observed during each CS step was attributed to the n O 2 released and n CO produced by each Ce 0.9 A 0.1 O 2-d material as follows:

Results and discussion
The synthesized Ce 0.9 A 0.1 O 2-d materials were characterized by using PXRD to identify their phase composition. The wide-angle X-ray diffraction patterns of the co-precipitation synthesized Ce 0.9 A 0.1 O 2-d materials are presented in Fig. 2 Fig. 3. The comparison between the elemental compositions of the as-synthesized and calcined powders and the exact chemical composition of each Ce 0.9 A 0.1 O 2-d material is listed in Table 1. The results obtained via the EDS analysis complements the findings associated with the PXRD analysis and confirms the formation of nominally phase pure Ce 0.9 A 0.1 O 2-d materials.
The particle morphology of the Ce 0.9 A 0.1 O 2-d materials was studied via SEM analysis. Exemplified SEM images of CeSr, CeBa, CeCa, and CeSn are presented in Fig. 4. The overall structure of all four images looks very similar, which further indicates  that the doping of different cations, i.e., Li, Mg, Ca, Sr, Ba, and Sn does not have any significant impact on the particle morphology of Ce 0.9 A 0.1 O 2-d materials. The particles are roughly spherical and seem agglomerated. The average particle size for the SEM images obtained was calculated by using the ImageJ software, and it is listed in Table 1. As per the numbers listed, it is evident that all the Ce 0.9 A 0.1 O 2-d materials have an average particle size in the range of 150-200 nm. Characterized Ce 0.9 A 0.1 O 2-d materials were tested toward thermochemical CS reactions by using a TGA setup. The variations in the mass of the samples were recorded by the Calisto software (embedded in the TGA setup). As an example, a total mass loss of 0.260 mg during the TR step conducted at 1400°C, and a total mass gain of 0.128 mg during the CS step performed at 1000°C was recorded for the CeMg material. These mass variations were converted into the n O 2 released (163.0 lmol/g) and n CO produced (160.0 lmol/g) by the CeMg materials by using Eqs. (1) and (2). The ratio of n CO =n O 2 for CeMg = 0.981 (considerably lower than the theoretical ratio = 2). A similar trend was noticed in the case of other Ce 0.9 A 0.1 O 2-d materials.
A higher release of O 2 during the TR step as compared to the lower production of CO during the CS step was unusual. As per the published literature [40,41], the loss in the mass of reactive samples recorded during the first TR step can be attributed to  O 2 release plus the release of the volatile components. Some of the chemicals used during the synthesis of the reactive samples remained unburnt during the calcination step (1000°C). These chemicals were burned and released from the sample during the TR step (1400°C). By following the results reported in the published literature [40,41], it was concluded that the n O 2 released by all the Ce 0.9 A 0.1 O 2-d materials, calculated as per the mass loss recorded during the first TR step, was deceptive. Hence, to avoid misrepresentation, the data obtained during the first cycle were not considered hereafter. The RR of the Ce 0.9 A 0.1 O 2-d materials was examined by performing four consecutive cycles (by excluding cycle1). The variations in the mass recorded for cycle 2 to cycle 4 are presented in Fig. 5. The mass variations recorded for cycle 2 to cycle 4 were converted into the n O 2 released and n CO produced and reported in Table 2. The Ce 0.9 A 0.1 O 2-d materials first compared with each other based on the n O 2 released in each cycle. The data listed in Table 2 clearly show that the n O 2 released in cycle 2 by all the Ce 0.9 A 0.1 O 2-d materials was higher than that of cycle 3. According to the numbers obtained, the n O 2 released by CeLi, CeMg, CeCa, CeSr, CeBa, and CeSn in cycle 2 was higher by 1.4 lmol/g, 3.2 lmol/g, 1.8 lmol/g, 1.1 lmol/g, 4.2 lmol/g, and 0.7 lmol/g than cycle 3, respectively. Likewise, the n O 2 released by CeLi, CeMg, CeCa, CeSr, CeBa, and CeSn materials in cycle 4 was lower by 2.0 lmol/g, 4.1 lmol/g, 1.7 lmol/g, 1.1 lmol/g, 2.1 lmol/g, and 1.9 lmol/g as compared to cycle 3, respectively. The data reported for the CS steps have a story similar to the TR step. The CO production capacity of the CeLi, CeMg, CeCa, CeSr, CeBa, and CeSn materials was decreased in cycle 4 by 4.7%, 3.3%, 8.9%, 5.5%, 4.5%, and 18.8% as compared to the CO production realized in cycle 2.
The results obtained in four consecutive cycles indicate that the TR and CS ability of all Ce 0.9 A 0.1 O 2-d materials reduced with the rise in the number of thermochemical cycles. Performing four cycles was an initial check, and hence, a detailed analysis of the co-precipitation synthesized Ce 0.9 A 0.1 O 2-d materials (at least by performing ten cycles) was essential. Therefore, the long-term RR of all the Ce 0.9 A 0.1 O 2-d materials was examined by performing a set of 10 consecutive cycles. Again, to avoid the misrepresentation, the data obtained in the first cycle were not considered in this analysis. The TGA profiles obtained for all the Ce 0.9 A 0.1 O 2-d materials from cycle 2 to cycle 10 are presented in Fig. 6. A close look at the TGA plots shows an indication of the attainment of stable RR by the Ce 0.9 A 0.1 O 2-d materials at around cycle 4 to cycle 5. However, a detailed comparison in terms of numbers was a must.
The n O 2 released by the Ce 0.9 A 0.1 O 2-d materials during each thermochemical cycle is presented in Fig. 7. The data presented show that all the Ce 0.9-A 0.1 O 2-d materials indicate a stable release of O 2 from cycle 4 to cycle 10. For example, the CeSr released 60.7 lmol/g, 60.1 lmol/g, 60.8 lmol/g, and 60.4 lmol/g of O 2 in cycle 4, cycle 6, cycle 8, and cycle 10. Likewise, the CeMg is also capable of releasing a stable amount of O 2 in the range of 68.0-68.8 lmol/g from cycle 4 to cycle 10. Similar to the O 2 releasing capacity, the CO production aptitude of all the Ce 0.9 A 0.1 O 2-d materials was also noticed to be steady from cycle 4 to cycle 10 (Fig. 8). For example, CeLi produced 126.3 lmol/g, 126.0 lmol/  g, 126.6 lmol/g, and 125.9 lmol/g of CO in cycles 4, 6, 8, and 10. The results obtained indicate that the initial four cycles were needed to stabilize the material properties and molecular interaction between the cations to achieve a constant O 2 release and CO production.
Besides the interaction between the metal cations, the morphology and crystal structure also play a vital role in the stability of the redox materials. The Ce 0.9 A 0.1 O 2-d materials obtained after performing the first, fifth, and tenth thermochemical cycles were characterized via SEM and PXRD. The findings acquired via PXRD analysis shows that the phase composition of all Ce 0.9 A 0.1 O 2-d materials remained unaffected during multiple thermochemical cycles. The exemplified PXRD peaks for CeMg and CeBa are reported in Fig. 9a, b. The SEM analysis indicates that due to the high-temperature sintering, significant growth in the particle size of Ce 0.9 A 0.1 O 2-d materials was recorded after cycle 1. It was also understood that the material morphology of the reacted Ce 0.9 A 0.1 O 2-d materials obtained after cycle 5 and cycle 10 was identical to morphology observed after cycle 1. The results obtained via SEM analysis shows that the morphology of all Ce 0.9 A 0.1 O 2-d materials was    stable from cycle 2 to cycle 10. As an example, the SEM images for the CeMg and CeBa obtained after cycle 1, cycle 5, and cycle 10 are presented in Fig. 10a, b. Figure 11 represents the comparison between the Ce 0.9 A 0.1 O 2-d materials based on average n O 2 released, n CO produced, and n CO =n O 2 ratio (from cycle 2 to cycle 10). In terms of n O 2 released, the Ce 0.9 A 0.1 O 2-d materials can be arranged as: The average n O 2 released by all the Ce 0.9 A 0.1 O 2-d materials was very close to each other. It was observed to be in the range of 60-70 lmol/g cycle, except for CeSn, for which it was considerably higher (107.6 lmol/g cycle). Similar to the average n O 2 released, the average n CO produced by the CeSn materials (180.5 lmol/g cycle) was relatively higher than the other Ce 0.9 A 0.1 O 2-d materials. Based on their average n CO production capacity, the investigated Ce 0.9 A 0.1 O 2-d materials can be ranked in the following order: Overall, as presented in Fig. 11, the incorporation of Li, Mg, Ca, Sr, Ba, and Sn has considerably improved the RR of Ce 0.9 A 0.1 O 2-d materials as compared to the pure ceria.
The CeSn and CeCa were observed to be the top and bottom-most choices for the CS. Although the CeSn materials seem to the best choice in terms of average n O 2 released and n CO produced, the average n CO =n O 2 ratio was considerably lower than most of the Ce 0.9 A 0.1 O 2-d materials. As shown in Fig. 11, the Ce 0.9 A 0.1 O 2-d materials can be categorized in the following order based on their re-oxidation ability: These results indicate that the CO production ability for the CeSn can be further increased by improving the n CO =n O 2 ratio. Our research group is currently exploring the Ce x Sn 1-x O 2-d (where x = 0.05-0.5) materials to find out the best candidate for the thermochemical WS and CS.

Summary and conclusions
By applying the co-precipitation of the hydroxide method, the Ce 0.9 A 0.1 O 2-d materials were synthesized by doping the alkali (Li), alkaline earth (Mg, Ca, Sr, Ba), and post-transition (Sn) metal cations in the ceria cubic structure. The nominally phase pure Figure 10 SEM images of CeMg a1 after cycle 1, a2 after cycle 5, a3 after cycle 10, and CeBa-a1 after cycle 1, a2 after cycle 5, a3 after cycle 10. Figure 11 Comparison between the Ce 0.9 A 0.1 O 2-d materials based on average n O2 released, n CO produced, and n CO =n O2 ratio (from cycle 2 to cycle 10).
composition of each Ce 0.9 A 0.1 O 2-d material was confirmed via the PXRD and EDS analysis. The SEM analysis indicates that the doping of different cations, i.e., Li, Mg, Ca, Sr, Ba, and Sn does not have any significant impact on the particle morphology of Ce 0.9 A 0.1 O 2-d materials (average particle size in the range of 150-200 nm). The long-term RR of the Ce 0.9 A 0.1 O 2-d materials was investigated in ten successive thermochemical cycles. The O 2 releasing and CO production capacity of each Ce 0.9 A 0.1 O 2-d material were observed to be steady from cycle 4 to cycle 10. In terms of average n O 2 released from cycle 2 to cycle 10, the Ce 0.9 A 0.1 O 2-d materials can be arranged as: CeSn On the other hand, based on their average n CO production capacity from cycle 2 to cycle 10, the investigated Ce 0.9 A 0.1 O 2-d materials can be ranked in the following order: CeSn material showed the highest n O 2 release (107.6 lmol of O 2 /g cycle) and n CO production (180.5 lmol of CO/g cycle) as compared to the CeO 2 and the remaining Ce 0.9 A 0.1 O 2-d materials. responsibility of author(s). The authors also gratefully acknowledge the Center of Advances Materials (CAM) at Qatar University for carrying out the XRD analysis and the Central Laboratory Unit (CLU) for services related to electron microscopy and EDS.

Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licen ses/by/4.0/.