Ni incorporation in MgFe2O4 for improved CO2-splitting activity during solar fuel production

Efficacy of the sol–gel derived Ni-doped Mg-ferrites for an enhanced CO2 splitting activity is investigated. The results allied with the characterization indicate the formation of nominally phase pure Ni-doped Mg-ferrites with a coarser particle morphology. Ni-doped Mg-ferrites are further tested for multiple thermal reduction as well as CO2 splitting steps by using a thermogravimetric analyzer. The results associated with the thermogravimetric analysis confirmed that most of the Ni-doped Mg-ferrites attained a steady TR aptitude after crossing the 5th or 6th cycle. Likewise, the CS capability of all the Ni-doped Mg-ferrites accomplished consistency after 4th cycle (except for Ni0.11Mg0.88Fe2.01O4.005). The Ni0.90Mg0.11Fe2.04O4.070 showed the highest amount of O2 release (117.1 μmol/g cycle) and CO production (210.3 μmol/g cycle) in ten consecutive thermochemical cycles. Besides, Ni0.29Mg0.72Fe1.98O3.980 indicated better re-oxidation aptitude (nCO/nO2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ n_{\text{CO}} /n_{{{\text{O}}_{2} }} $$\end{document} ratio = 1.89) when compared with other Ni-doped Mg-ferrites.


Introduction
Production of solar fuels, such as H 2 or syngas, via a metal oxide (MO) based two-step solar thermochemical H 2 O/CO 2 cycle (STC) has emerged as one of the promising technologies [1][2][3]. H 2 can be used directly as a fuel [4], whereas the syngas can be converted into liquid fuels via a catalytic Fischer-Tropsch process [5]. With the help of solar fuel technology, the excessive utilization of fossil fuels and the continuous CO 2 emissions can be reduced. Besides, this technology can store the abundantly available solar energy in the form of liquid fuels such as gasoline, kerosene, and others.
In STCs, the MO plays a vital role in terms of the release and storage of the O 2 . The O 2 gets released from the MOs during a thermal reduction (TR) step. On the other hand, the MOs can regain the O 2 during the H 2 O/CO 2 (WS/CS) splitting reaction. In previous studies, a variety of MOs has been utilized for the thermochemical conversion of H 2 O and CO 2 , which mainly includes zinc oxide [6,7], tin oxide [8,9], iron oxide [10,11], CeO 2 [12,13], doped ceria materials [14,15], ferrites [16][17][18][19][20], and La-based perovskites [21][22][23]. Among the listed MOs, CeO 2 is considered as a state-of-the-art material for the STCs due to its favorable reaction kinetics and high-temperature stability.
Recently, Takalkar et al. [24,25] investigated the sol-gel derived Ni-ferrite and Mg-ferrite for the thermochemical splitting of CO 2 . The reported results indicate that both Ni-ferrite (64.4 lmol of O 2 /g cycle) and Mg-ferrite (58.7 lmol of O 2 /g cycle) were capable of releasing higher levels of O 2 than CeO 2 (48.5 lmol of O 2 /g cycle) [26] at 1400°C. The amount of CO produced at 1000°C by the Ni-ferrite (125.9 lmol of CO/g cycle) was higher and by the Mg-ferrite (79.6 lmol of CO/g cycle) was lower than CeO 2 (95.0 lmol of CO/g cycle). The obtained results show that both Ni-ferrite and Mg-ferrite possess a better TR aptitude than CeO 2 . In contrast, work needs to be done to improve the CS ability of the Mg-ferrite.
To utilize the TR capacity of both Ni-ferrite and Mg-ferrite together and to improve the CS ability of the Mg-ferrite, in this study, we have examined ternary Ni-doped Mg-ferrites (Ni x Mg 1-x Fe 2 O 4-d ) in multiple thermochemical cycles. It is believed that the incorporation of Ni in the Mg-ferrite crystal structure (partially replacing Mg ?2 by Ni ?2 ) will significantly improve the CS aptitude. The redox reactions associated with the two-step Ni-doped Mg-ferrite (NMF) based CS cycle are as follows: Ni-doped Mg-ferrites were tested by performing ten thermochemical cycles in which the TR and CS steps were carried out in the temperature range of 1000-1400°C. Obtained results indicate that most of the Ni-doped Mg-ferrites attained the redox stability after crossing the 4th or 5th thermochemical cycle. This paper reports a detailed analysis of the results obtained during synthesis, characterization, and thermochemical cycles.

Experimental section Preparation of Ni-doped Mg-ferrites
Ni-doped Mg-ferrites were synthesized by using the sol-gel approach. The stoichiometric amounts of precursors, i.e., nickel nitrate, magnesium nitrate, and iron nitrate (procured from Sigma Aldrich), were first dissolved in ethanol at room temperature. Subsequently, a predetermined quantity of propylene oxide (purchased from Sigma Aldrich) was added into the mixture. The as-prepared solution was kept undisturbed for the formation of the Ni-doped Mgferrite gel. The as-synthesized gel was aged and then dried at 120°C for 2 h. The dried Ni-doped Mg-ferrite was then crushed into a fine powder by using a mortar and pestle. Annealing of the powered Nidoped Mg-ferrite up to 1000°C for 4 h was performed by using a muffle furnace.

Characterization of Ni-doped Mg-ferrites
Crystallographic analysis of the Ni-doped Mg-ferrites was carried out by using a Powder X-ray Diffractometer (PXRD) with CuKa radiation, the irradiation time of 5 s, and Bragg angle from 25°to 70°(Panalytical XPert MPD/DY636, k = 0.15,418 nm, steps = 0.05°). The microstructural morphology and elemental/chemical composition of the Ni-doped Mgferrites were scrutinized by using a scanning electron microscope (SEM, Nova Nano 450, FEI) equipped with an energy-dispersive X-ray spectroscopy (EDS).

Cyclic CO 2 splitting test
A CO 2 splitting cyclic study which comprised of multiple TR and CS steps was performed by using a TGA (procured from Setaram SETSYS Evolution, France, Fig. 1). The TGA setup was equipped with two mass flow controllers and a vacuum gauge. The temperature cooling and heating rates were normalized at 25 K/min. The mass variation observed during the TR and CS steps was recorded by using an inbuilt Calisto processing software supplied by the Setaram. During the TGA experiments, approximately 50 mg of the Ni-doped Mg-ferrite powder was loaded on an alumina crucible and placed inside the TGA. Multiple thermochemical cycles were performed by maintaining the operating conditions as follows: a) TR at 1400°C for 60 min with an Ar flow rate at 100 ml/min and b) CS at 1000°C for 30 min with an Ar/CO 2 (50:50) mixture flow rate at 100 ml/ min. The mass variation recorded by the software was further converted into the amounts of O 2 released and CO produced by using the following equations. Additional details associated with the other essential parts of the TGA setup and the experimental procedure are listed in our previous contributions [24,25].
where n O 2 = moles of O 2 released (lmol/g); n CO = moles of CO produced (lmol/g); Dm loss = amount of loss in the mass (mg); Dm gain = amount of gain in the mass (mg); M O 2 = molecular weight of O 2 (g/mol); M O = molecular weight of O (g/mol); m NMF = mass of the Ni-doped Mg-ferrites (mg)

Results and discussion
The prime objective of this study is to identify the best combination of the  Table 1) provide further assurance of the formation of nominally phase pure Ni-doped Mg-ferrites. In addition to the phase and chemical composition, the morphology of Ni-doped Mg-ferrites was scrutinized via SEM analysis. The SEM analysis established that the variation in the Ni ?2 and Mg ?2 molar concentrations has an insignificant influence on the morphology. Ni-doped Mg-ferrites showed coarser particle morphology with a diverse particle size in the range of * 100 nm (smallest particle) to * 1 micron (largest particle  Fig. 3. After synthesizing and analyzing the physical properties, the redox reactivity of each Ni-doped Mgferrite toward the thermochemical splitting of CO 2 was investigated by using a TGA setup. Firstly, the TR and CS ability of each Ni-doped Mg-ferrite was explored by performing a single thermochemical cycle, and the TGA profiles obtained are reported in Fig. 4. As expected, Ni-doped Mg-ferrites showed a decrease in the mass during the TR step (due to the release of O 2 and formation of O 2 vacancies) and an increase in the weight during the CS step (via reoxidation reaction). These mass variations were converted into the n O 2 (by using Eq. 3) and n CO (by using Eq. 4) by each Ni-doped Mg-ferrite during 1st cycle (Fig. 5)    It is believed that during the high-temperature TR step, the chemicals left over from the synthesis of Nidoped Mg-ferrites were burned. Due to this burning, an additional drop in the mass was recorded which contributed to the false indication of a higher n O 2 . In addition to the role of unburnt chemicals, there is a chance that the re-oxidation capacity of Ni-doped Mg-ferrites was not sizable enough and hence n CO was considerably lower than n O 2 . For further exploration, the Ni-doped Mg-ferrites were tested for an additional two thermochemical cycles. TGA profiles obtained for all three cycles are presented in Fig. 6. Besides, a comparison between the three cycles based on n O 2 and n CO by Ni-doped Mg-ferrites is presented in Table 2.  15.6%, and 11.8%, respectively, in cycle 2 as compared to cycle 1. In cycle 3, however, the percentage decrease in n CO for all Ni-doped Mg-ferrites was extremely lower when compared to cycle 2. The overall results obtained indicate that Ni 0.90 Mg 0.11 Fe 2.04 O 4.070 possesses better redox reactivity toward the TR and CS reactions than the other Ni-doped Mg-ferrites. The CO production aptitude of all Ni-doped Mg-ferrites was observed to be steady in cycle 2 and cycle 3. However, it was also recorded that the O 2 releasing ability of Ni-doped   Mg-ferrites has not attained stability in the first three cycles.
A set of successive ten thermochemical cycles was carried out to understand the long-term redox reactivity and stability of Ni-doped Mg-ferrites. The experimental conditions were kept unchanged, and the TGA profiles obtained for each Ni-doped Mgferrite are reported in Fig. 7. From the results obtained until now it is clear that n O 2 by all Ni-doped Mg-ferrites during cycle 1 is not precise. Hence, to avoid publishing any misleading information, the data obtained during the 1st cycle were disregarded, and all the analysis was focused on the remaining nine cycles. Figure 7 shows the TGA profiles obtained during ten consecutive thermochemical cycles.
The TGA profiles reported in Fig. 7 provide an initial indication that the Ni-doped Mg-ferrites have attained a stable TR and CS ability after crossing cycle 5 or cycle 6. However, it was essential to verify this visual confirmation with the help of numbers. Hence, the mass variations recorded in the TGA profiles were converted into n O 2 and n CO with the help of Eqs. (3) and (4).
n O 2 by each Ni-doped Mg-ferrite (from cycle 2 to cycle 10) is presented in Fig. 8. The trends reported supported the visual confirmation provided by Fig. 7 and clearly shows that the Ni-doped Mg-ferrites have attained a steady TR aptitude after surpassing cycle 5 or cycle 6. To be precise, Ni 0. 11 4.005 . The improvement in the TR ability of doped MO depends on the interaction between the metal cations incorporated in the crystal structure. At this moment, we are not in a position to provide any specific reason for the trends reported in Fig. 8 as we have not studied the molecular interaction between the metal cations associated with the Ni-doped Mg-ferrites.
According to Fig. 9, the CS capability of all the Nidoped Mg-ferrites has attained consistency after crossing cycle 4 except for Ni 0.90 Mg 0.11 Fe 2.04 O 4.070 for     3.970 indicated the lowest re-oxidation capacity. It was further understood that the average n CO =n O 2 ratio of each Ni-doped Mg-ferrite was lower than CeO 2 (average n CO =n O 2 ¼ 1:96) and the NiFe 2 O 4 (average n CO =n O 2 ¼ 1:96). In contrast, the comparison with MgFe 2 O 4 (average n CO =n O 2 ¼ 1:35) indicates that the inclusion of Ni ?2 in the Mg-ferrite crystal structure has improved the average n CO =n O 2 ratio of Nidoped Mg-ferrites.