The Trade-Off between Combustion and Partial Oxidation during Chemical Looping Conversion of Methane

Abstract

The chemical looping reforming and combustion of methane have attracted increasing interest as processes for clean energy and syngas production, with potential to reduce carbon dioxide emissions. Previous literature on the development of oxygen carriers evidenced the effects that oxygen availability exerts on the selectivity of the oxidation reaction. In the present paper, we evaluate the performance of chromite sand (Chro), cerium dioxide (CeO₂), and mixed cerium–copper oxide (Ce–Cu) as oxygen carriers for either reforming or combustion according to their oxygen availability. The oxides are tested in 2 to 5 min reduction intervals in a CH₄/N₂ mixture (5, 10 and 20% vol.) followed by regeneration in O₂/N₂ (3, 5, or 21% vol.), with redox cycles conducted either at 850 °C or 950 °C. The obtained rank of selectivity towards complete CH₄ combustion is Ce–Cu > CeO₂ > Chro. Another relevant finding is the role of the degree of carrier conversion in promoting partial or total oxidation. In particular, the selectivity towards CO₂ markedly decreases at increasing carrier conversion, disclosing new strategies for process design and optimization by controlling the carrier conversion degree.

1. Introduction

Chemical looping combustion is an emerging technology allowing for the inherent separation of CO₂ in flue gases [1,2].

It is accomplished by using a suitable oxygen carrier, usually a metal oxide with multiple oxidation states (e.g., FeO_x, MnO_x, NiO_x), and it has the important advantage of limiting NO_x formation thanks to the flameless behavior and well controlled process temperature [3]. Chemical looping can also be effectively used for the reforming and partial oxidation of methane, with high selectivity toward H₂ and CO depending on the nature of the carrier [4]. Cu and Ni oxides generally favor complete oxidation, while cerium oxide is more suitable for partial oxidation to CO. In addition to selectivity, other crucial issues to consider for oxide selection in chemical looping operation are coke deposition and carrier lifetime, in particular, for fluidized bed systems, where attrition phenomena are more relevant [2]. Furthermore, the safety and disposal of the spent carrier are also fundamental aspects that need to be explored. In this respect, the use of Ni compounds is decreasing due to concerns about their toxicity [5].

The trade-off between combustion and partial oxidation is somehow connected to the prompt availability of oxygen-rich sites in the microstructure of the carrier. In the case of cerium dioxide, for example, it has been repeatedly observed that while the production of carbon monoxide is overall favored compared to combustion, carbon dioxide can be nonetheless formed at the start of the reaction due to the high availability of surface oxygen species [6,7]. The availability of surface oxygen, related to the energy barrier for oxygen vacancy formation, the amount of total oxygen available for exchange (oxygen carrying capacity), the morphology and the surface chemistry, particularly oxygen vacancy concentration, all greatly affect the selectivity of oxidation [8]. For example, it has been suggested that the mechanism of methane oxidation on cerium dioxide changes as surface oxygen is depleted, with the activation of methane occurring preferentially on oxygen anions at the start of reaction, while on more reduced surfaces activation occurs also on oxygen vacancy sites [9]. Oxygen vacancies have also been observed to play a critical role in determining selectivity for reforming over combustion for iron oxygen carriers [10], with increased vacancy concentration favoring the partial oxidation of methane and lowering the energy barrier for C–H bond cleavage. In the present paper, we further investigate this shift in selectivity, which needs to be carefully evaluated to properly design the chemical looping process.

Natural chromite sand, made of a mixture of chromium and iron oxide mixed with other species such as alumina, silica, and others, is commonly used in metallurgy as foundry sand and as a source of chromium [11,12]. It has been previously investigated, as a catalyst, for use in fluidized bed combustion for CO abatement providing limited advantages [13], whilst copper-modified chromite proved more effective [14]. Its potential application as an oxygen carrier has, to the author’s knowledge, so far not been investigated at depth in the literature. However, due to its inclusion of high concentrations of oxidation active iron oxide species (especially in low-grade chromite sands) and its high melting point, these materials may deserve further investigation. Copper oxide is a well-known catalyst for low-temperature combustion as well as for the decomposition of methane [15], but it is not commonly used for reforming processes. The use of CeO₂ for reforming processes has been investigated via several aspects, in particular the redox kinetics and carbon deposition at different temperatures and residence times [7,16,17]. Chemical looping combustion with CuO or CeO₂ has also been tested in fluidized bed apparatus [18,19,20], providing good results in terms of the material’s lifetime, mechanical resistance, and regeneration efficiency.

The aim of the present research is to understand the potential use and the chemical behavior of different natural and synthetic oxygen carriers under changing operating conditions, moving from chemical looping combustion to partial methane oxidation. The influence of equivalence factor, temperature and conversion time is reported and discussed to provide further insights into the design of the process and the materials.

2. Materials and Methods

2.1. Materials

Three types of oxygen carriers have been selected: natural chromite sand (Cromitec 400, hereafter defined Chro), CeO₂ and CeO₂/CuO (samples named Ce–Cu) granules.

Commercial powders of CeO₂ (PIKEM, Wilnecote, UK) and CuO (Merck, Darmstadt, Germany) have been uniaxially pressed around 90 MPa, obtaining 20 mm diameter pellets subsequently crushed and sieved to 0.60–0.84 mm and thermally treated for 20 min in air at 900 °C.

For all the prepared materials O₂ capacity was computed based on reducible species content. The main properties of the carriers are reported in

Table 1. Properties of oxygen carriers.

	Chro	CeO2	Ce–Cu
Size (mm)	0.20–0.40	0.60–0.84	0.60–0.84
Density (kg/m3)	4170	7220	6930
O2 capacity (mmol/g)	0.90	1.45	2.98
Major phases
CeO2, wt. %	-	>99	68
CuO, wt. %	-	-	32
Cr2O3, wt. %	47	-	-
FeO, wt. %	26	-	-
Al2O3, wt. %	15	-	-
MgO, wt. %	10	-	-
SiO2, wt. %	1	-	-

.

Cerium–copper carrier exhibits the highest O₂ capacity value while chromite has the lowest, the oxygen capacity being related to FeO alone; therefore, this excludes the possibility that Cr₂O₃ can be reduced at the process temperature.

Samples of the carriers have been characterized by electronic microscopy FE-SEM (Zeiss SIGMA, Carl Zeiss Microscopy GmbH, Jena, Germany, D) and X-ray powder diffractometry (XRD) using a Bruker D8 Advance (Bruker–Karlsruhe, Karlsruhe, Germany, D) diffractometer.

2.2. Experimental Apparatus

The schematic of the experimental rig used for chemical looping tests is shown in Figure 1. The reactor is a 10 mm ID quartz tube equipped with a ceramic distributor in the bottom. The reactor is installed inside an electric furnace (Carbolite 1200, Carbolite LTD, Hope Valley, UK). Bronkhorst mass-flowmeters are used to supply gas streams from compressed gas bottles (O₂, N₂, CH₄). A Pollutek GAS-3100P continuous gas analyzer (Pollutek Gas Analysis, Lubbeek, Belgium) has been employed for O₂, CO, CO₂, CH₄ and H₂ detection.

In a typical experiment, a mass of 14 to 16 g of oxygen carrier is loaded in the reactor to ensure the same bed volume for the three oxygen carriers (≈5 cm³), with a layer of ceramic wool on the top of the bed to limit fluidization and avoid the entrainment of particles outside the reactor. The residence time of the gas stream in the fixed bed was lower than 0.1 s. For the half cycle of reduction, streams of 1 NL/min of CH₄ at 5, 10 and 20 vol.% in N₂ were selected to study the effect of methane concentration on reaction selectivity, while for carrier regeneration, streams of O₂ at 3, 5, or 21 vol.% in N₂ at a total 1 NL/min flow rate were evaluated. The main factor in the choice of oxygen concentration during regeneration was the necessity to avoid overheating triggered by the exothermic reaction without loss in regeneration efficiency.

Reduction experiments were carried out by varying the time between 2 and 10 min, while for regeneration the reactor was kept on stream until a breakthrough in the inlet oxygen molar fraction was obtained.

2.3. Data Elaboration

The CH₄ conversion, ξ_CH4, is evaluated by the integration of the difference between the inlet and outlet methane flow rate divided by the methane fed during the reduction step of length t_c = 2, 5 or 10 min (Equation (1)). The outlet molar flow rate is calculated using N₂ balance as a reference.

ξ_CH4 = ∫(q_in × y_CH4,in − q_out × y_CH4,out)dt/(q_in × y_CH4,in × t_c)

(1)

The CO and CO₂ selectivity (η_CO, η_CO2) are computed (Equations (2) and (3)) based on the converted CH₄ by integration of their molar fraction profiles,

η_CO = ∫(q_out × y_CO,out)dt/(q_in × y_CH4,in × t_r)

(2)

η_CO2 = ∫(q_out × y_CO2,out)dt/(q_in × y_CH4,in × t_r)

(3)

Carbon deposition is calculated from carbon balance by considering the amount of CO and CO₂ released during the regeneration stage, also calculated by integrating the molar fraction profile. Carbon selectivity can therefore be obtained as

η_C = ∫(q_out × y_CO,out + q_out × y_CO2,out)dt/(q_in × y_CH4,in × t_c)

(4)

The equivalence factor of a single run corresponds to

e = n_{O2 av,i}/n_{O2 stoich, CH4}

(5)

where n_{O2 stoich, CH4} indicates the moles of oxygen necessary for stoichiometric total combustion of the total fed methane, while n_{O2 av,i} is the total amount of oxygen which can be released from the carrier i.

n_{O2 stoich, CH4} = 2(q_in × y_CH4,in × t_c)

(6)

The reactions considered for carrier reduction are

2CeO₂ → Ce₂O₃ + 1/2O₂

(R1)

Fe₂O₃ → 2FeO + 1/2O₂

(R2)

CuO → Cu + 1/2O₂

(R3)

Therefore, for cerium dioxide, the amount of available oxygen was calculated considering complete reduction to Ce₂O₃:

n_{O2 av, CeO2} = 1/4(m_CeO2/M_CeO2)

(7)

For chromite, the amount of oxygen was calculated considering the FeO content of chromite only (Table 1), the reduction of the other oxides contained in chromite being thermodynamically hindered at the investigated temperature. The maximum releasable O₂ content derives from converting FeO to Fe₂O₃ and then reducing it back to FeO (Equation (8)), ω_i being the generic mass fraction of the component i. The reduction of FeO to metallic Fe was excluded,

n_{O2 av, Chro} = 1/4(m_Chro × ω_FeO)/M_FeO

(8)

While the thermodynamic formation of completely reduced Fe would be feasible, metallic iron was not observed in the XRD patterns of reduced samples. Similarly, Leion et al. also excluded the total reduction of iron when discussing the use of ilmenite as an oxygen carrier [21].

For the Ce–Cu carrier, both reactions R1 and R3 were considered,

n_{O2 av, Ce–Cu} = 1/4(m_Ce–Cu × ω_CeO2)/M_CeO2 + 1/4(m_Ce–Cu × ω_CuO)/M_CuO

(9)

The oxygen capacity (Table 1), reported as mmol of O₂ per gram, is also used to evaluate the conversion degree of the carrier, ξ_carrier, obtained by evaluating the total oxygen released during the reduction step through oxygen balance for CO, CO₂, O₂ and H₂O species, H₂O production being estimated via hydrogen balance.

ξ_carrier = ∫{q_out × [0.5 × y_{CO, out} + y_CO2,out + y_O2,out] + 0.5[2 × q_in × y_CH4,in − q_out × (2 × y_CH4,out − y_H2,out)]}dt/n_{O2,av,carrier}

(10)

where “carrier” indicates Chro, CeO₂, or Ce–Cu.

Experimental errors are mainly caused by the intrinsic transient character of the looping operation and are estimated in ±10% of the reported data.

3. Results

3.1. Reaction Test Results

Figure 2 shows the volumetric fraction profiles of CH₄, CO, CO₂, H₂ and O₂ for both oxidation (t = 1 ÷ 3 min) and regeneration (t = 4 ÷ 9 min) on the Chro carrier for a 2′ reduction step in 10% methane followed by 6′ regeneration in 3% O₂. A 1′ flushing with N₂ was performed in the passage from oxidation to regeneration and vice versa. It is clearly visible how the volume fractions change as reaction proceeds, with CO₂ production decreasing during the reduction step, whilst the H₂ and CO volume fractions increase. Similar profiles are also observed over CeO₂, while for the Ce–Cu carrier, CO and H₂ production were found to be almost completely absent, even for longer reduction steps. Considering an equal mass of carriers, the Ce–Cu carrier theoretically offers the highest oxygen availability compared to the other two tested (see Table 1). Also, the Ce–Cu carrier is the only one that shows the release of gaseous oxygen due to thermal dissociation at the investigated temperature, thus providing oxygen in a larger amount for reaction with methane.

Figure 3 shows the comparison among Chro, CeO₂ and Ce–Cu carriers in terms of CH₄ conversion for tests carried out at 850 and 950 °C with reduction time of 2 or 5 min. It is clear that, in terms of maximum methane conversion, the rank of the carrier is Ce–Cu > CeO₂ > Chro, as a consequence of the difference in O₂ capacity and availability. Oxygen is the most widely available in the Ce–Cu carrier, both in terms of the quantity of available oxygen (Table 1) and of the easiness of oxygen release, as it is the only investigated material that releases gaseous oxygen at the investigated reaction temperature. Therefore, it is not surprising that methane conversion is greatest for this material, as the combustion reaction is not strictly surface-mediated, but can also involve fully gaseous species. It is worth noting that this Chemical Looping Oxygen Uncoupling (CLOU) mechanism [22] is also likely responsible for the similar values of ξ_CH4 at 850 and 950 °C for Ce–Cu, owing to the capability of this carrier to make available gaseous molecular O₂ imposing a less strict limit on process kinetics compared to the strictly surface-mediated oxidation on the other carriers, with the presence of metallic Cu further favoring the reaction by providing active sites for methane C–H bond cleavage.

On the contrary, the very low values of ξ_CH4 obtained for Chro and CeO₂ carriers at 850 °C moved the investigation preferably to the temperature of 950 °C. Furthermore, Chro carrier exhibited greater instability in performance, particularly after undergoing cycles at 950 °C. After undergoing reaction at 950 °C, the material appeared more effective even at a lower temperature: the observed average ξ_CH4 for 2 min partial oxidation steps at 850 °C was 1.7 ± 0.2% for fresh material, while it increased to 7.4 ± 0.4% after the material underwent reaction at a higher temperature. It is likely that the material undergoes structural changes during operation at higher temperatures. Once extracted from the reactor, Chro carrier was found to be partly sintered, and this could be the reason for the differences in performance.

Table 2. CH₄ conversion, CO selectivity and CO₂ selectivity for Chro, CeO₂ and Ce–Cu carriers (y_CH4 = 0.10, 950 °C).

	Time, min	e	ξCH4	ηCO	ηCO2
Chro	2	0.81	0.13	0.12	0,29
	5	0.32	0.37	0.45	0,00
CeO2	2	1.25	0.19	0.24	0.78
	5	0.50	0.25	0.58	0.54
Ce–Cu	2	2.48	0.60	0.00	1.00
	5	0.99	0.83	0.00	1.00

reports the selectivity and yield data for Chro, CeO₂ and Ce–Cu at 950 °C and y_CH4 = 0.10. The equivalence factor changes largely due to the difference in oxygen capacity of the three carriers. The greater availability of oxygen in Ce–Cu clearly appears, resulting in conversion up to 0.83 and the total combustion of methane regardless of the total reduction time. The catalytic activity was retained over repeated cycles (>10) of reduction and oxidation for all samples.

For Chro and cerium dioxide, selectivity and conversion change throughout the reaction.

Figure 4 shows CH₄ conversion, and CO and CO₂ selectivity, at different CH₄ concentrations, as well as reduction time over CeO₂ carrier. The trends in the changes of these variables are mutually consistent: ξ_CH4 decreases as the mole fraction of CH₄ increases, corresponding to a lower equivalence factor in the whole test. Selectivity towards partial oxidation also increases with a longer reduction time.

The selectivities of CO and CO₂ clearly exhibit opposite trends when increasing the mole fraction of CH₄ for both reduction times. Therefore, the conditions required to achieve total combustion, over CeO₂, are those corresponding to a short reduction time and low y_CH4, i.e., a higher equivalence factor.

Experimental data, for all carriers, have also been evaluated to allow for consideration of carbon deposition. Coke formation was appreciable only for experiments carried out at 950 °C, where the presence of carbon lowered the selectivity towards CO₂ and CO.

Table 3. Carbon (coke) selectivity (%) in Chro, CeO₂ and Ce–Cu at different y_CH4 and oxidation times (T = 950 °C).

yCH4	0.05	0.05	0.10	0.10	0.20	0.20
Time, min	2	5	2	5	2	5
Chro	3.5	0.7	1.3	11.9	-	-
CeO2	7.7	2.4	4.4	1.1	1.0	0.7
Ce–Cu	0.8	0.1	0.5	1.4	4.6	2.6

Mean standard deviation ±0.3.

reports the carbon selectivity, η_C, for these tests, which is always below 10%, with the exception of the test with Chro at y_CH4 = 0.10 and t = 5′ (η_C = 11.9%), which was affected by a partial agglomeration of the fixed-bed particles. The accumulated carbon was readily converted during carrier regeneration, without causing apparent problems to carrier regeneration.

Figure 5 displays the instantaneous selectivity towards CO and CO₂ for tests carried out with CeO₂ carrier at 950 °C and increasing y_CH4. The reported behavior is consistent with the changes in operating conditions and the transient character of the test. CO₂ selectivity was the highest at the initial time, i.e., at maximum equivalent factor (e → +ꝏ), and declined with decreasing e, corresponding to the progressive depletion of oxygen sites in the carrier. The increase in the mole fraction of CH₄ leads to a shift in the CO₂ curves towards lower values due to the lower equivalence factor and the lower availability of oxygen. Congruently, the selectivity behavior of CO is perfectly symmetrical with that of CO₂.

3.2. Characterization of the Samples

Oxygen carriers, before and after use in the reactor, have been characterized by SEM and XRD analyses.

Figure 6 displays the SEM images of the granules at different magnifications. Low-magnification images (Figure 6a,b) demonstrate that for the CeO₂ and Ce–Cu carrier, the granules remain almost unchanged before and after use in the reactor. No fine particles were formed during reaction, which is reasonable, as the tests were carried out in a fixed bed without abrasion and rather limited thermal stress. In general, all granules are dense with well-defined external surfaces. Therefore, we have confirmed the good mechanical stability of both synthetic CeO₂ and Ce–Cu carriers. The morphology of Chro carrier granules recalls the cubic-octahedral symmetry of the material with isometric/rounded granules with smooth surfaces and without microstructures, a bit like crystalline faces. Some particle agglomerates are also evident in the case of chromite use (Figure 6b). In this regard, a rather large agglomerate was recovered after unloading the chromite bed, while the other two materials did not present similar problems.

The XRD analysis (Figure 7) proves that before and after the test, the CeO₂ and Ce–Cu samples were very similar, without substantial crystallographic–structural and crystalline difference. Only a certain difference can be seen in terms of crystallite size. The Ce–Cu carrier exhibits some peaks of CuO (pristine oxide) and Cu₂O, as a consequence of the reduction step and incomplete regeneration. Conversely, before and after the test, chromite samples exhibit large differences with clear oxidation phenomena and the presence of Fe₂O₃ peaks in the one subjected to chemical looping. Some spinel structures between Fe and Chro oxides also appear for the used Chro carrier. Also, a higher baseline for chromite is present due to the lower degree of crystallinity of this natural material.

Ceria samples do not exhibit a substantial presence of any additional phases or Cerium oxide with different oxidation states (e.g., Ce₂O₃), and they appear quite pure and composed of a single phase according to XRD analysis. Even after undergoing several reaction steps, the material does not display any relevant change, either from the chemical or structural point of view (e.g., entry or loss of oxygen), with no apparent change in cell parameters. Both the spacing and intensity of reflections are completely coherent before and after treatment. This finding suggests very good stability and regeneration efficiency.

Ce–Cu samples are also very similar before and after reaction, without substantial crystallographic–structural and crystalline difference in the present CeO₂, before and after treatment. The only changes in Ce–Cu carrier are exhibited concerning the copper phase, with some peaks of Cu₂O (cuprite) appearing together with CuO (tenorite, as pristine oxide) peaks in the spent material, as a consequence of the reduction step and incomplete regeneration. Concerning this regeneration process, some traces of the presence of an intermediate oxidative phase, where copper is present with both oxidation numbers as Cu₄O₃ (paramelaconite, where Cu¹⁺₂Cu²⁺₂O₃), have been found [23]. Even a certain difference could be appreciated in terms of crystallite size before and after treatment.

The chromite sample comes from mineral natural sand collecting several spinel phases with a complex stoichiometry (Mg,Fe)(CrAl)₂O₄, Fe(Cr,Al)₂O₄, AlFe₂O₄, etc., according to the main elements already found in the chemical analyses. The higher baseline of the patterns is due to a fluorescence effect of XRD in the presence of transition metals (Mn, Co, Ni, Fe, Cr, and others). Only a few traces of accessory residual and refractory mineral phases have been found (e.g., corundum, magnetite, olivine).

4. Discussion

Table 4. CH₄ conversion, CO and CO₂ selectivity in Chro, CeO₂ and Ce–Cu at different y_CH4 values and conversion degrees of the carrier (T = 950 °C).

	yCH4	0.05	0.05	0.10	0.10	0.20	0.20
	ξcarrier	0.10	0.20	0.10	0.20	0.10	0.20
Chro	ξCH4	0.18	(-)	0.21	0.14	(-)	(-)
	ηCO	0.15	(-)	0.39	0.58	(-)	(-)
	ηCO2	0.85	(-)	0.36	0.35	(-)	(-)
CeO2	ξCH4	0.29	0.28	0.22	0.21	0.17	0.16
	ηCO	0.47	0.33	0.59	0.74	0.63	0.80
	ηCO2	0.53	0.77	0.44	0.27	0.39	0.24
Ce–Cu	ξCH4	0.73	0.73	0.74	0.74	0.74	0.74
	ηCO	0.00	0.00	0.00	0.00	0.00	0.00
	ηCO2	1.00	1.00	1.00	1.00	0.87	0.88

provides a comparison of methane conversion and selectivity as the input mole fraction and carrier conversion vary for the three materials investigated. The cells in which there are high values (0.50) of selectivity in CO₂ and CO have been highlighted in dark and light grey, respectively. Despite some inconsistencies due to the limits of the experimental technique, especially regarding short conversion times, it can be noted that the initial phases of reaction (ξ_carrier = 0.1) favor the total oxidation of methane with the production of CO₂, while the progress of the carrier conversion (ξ_carrier = 0.2) leads to the partial oxidation and formation of CO. For the values investigated, Ce–Cu always promotes total combustion with the formation of CO₂ and CO present only in traces. This can be attributed to the high O₂ capacity of the carrier, as well as its CLOU behavior due to the release of gaseous oxygen. These results could be linked to the different activities of oxygen sites in the carrier [8], depending on their ease of accessibility.

The microstructural analysis showed different behaviors among the three carriers (Figure 6c,d). The granules of CeO₂ retain their surface characteristics well even after exposure to high temperature and redox atmosphere. On the contrary, chromite gives rise to an evident modification of the exposed surface with fractures and the formation of an amorphous layer, while the crystalline structure remains unchanged in the internal core of the granule (Figure 6d). An intermediate behavior occurs for Ce–Cu granules, where some formations of sintered micro-granules are evident (Figure 6d) upon the exposure to high temperature and chemical reaction, probably due to the less refractory character of copper oxide with respect to ceria.

Conversely, before and after the test, chromite samples exhibit large differences with clear oxidation phenomena, with a presence of peaks from neo-formed Fe₂O₃ (particularly lying on the surface of the granules where the oxidation process was more intense) in the one subjected to chemical looping, together with a rearranging of chemical compositions of the phases. Iron oxide phases appear to be segregated on the grain surface. This is consistent with phase diagrams for Fe–Cr–O systems reported in literature [24,25].

The CuO–CeO₂ oxygen carrier displayed the highest oxygen availability, while the chromite carrier showed limited activity towards methane.

Overall, the good selection of reaction step duration is crucial to ensuring the desired process selectivity for all oxygen carriers. Previous literature has suggested that the mechanism of reaction of methane on cerium dioxide changes as the material is reduced [9], and so does the selectivity. Methane was shown to be activated on surface oxygen sites over oxidated cerium dioxide, which are also active in the further oxidation of syngas. As the surface oxygen sites are depleted, methane activation starts occurring on the formed oxygen vacancies, while the lower surface oxygen availability stops the further formation of CO₂ and H₂O. This must be considered when discussing process optimization, and the extent of reduction reached by the carrier can be used to modulate reaction selectivity. Limiting reaction time allows one to reduce coke deposition, but should be accomplished with care, as an overly strict limitation in the reaction step duration can cause a noticeable overall loss in selectivity towards partial oxidation.

In the case of cerium dioxide, the present paper clearly displays the presence of a lower threshold in the extent of carrier conversion that needs to be surpassed if one is to observe a significant prevalence of partial oxidation when compared to complete combustion.

A carrier conversion lower than 10% (Figure 5) in the observed reaction conditions leads to a very high influence of total combustion on process yield. The previous literature on chemical looping over cerium dioxide showed the need for short reaction cycles to avoid excessive carbon deposition [26,27], but this tradeoff between complete and partial oxidation also needs to be addressed when selecting cycle length.

Chromite also displayed a similar behavior to CeO₂ carrier, with a first phase of reaction where complete oxidation is favored followed by a second step with prevalent syngas formation. In this case, the first phase of complete combustion is most likely attributed to the reduction of surface Fe₂O₃ to Fe₃O₄, with the further reduction to FeO being responsible for the generation of syngas, as previously reported in the literature for iron-based oxygen carriers [28]. In general, chromite sand displayed low activity and yields for both chemical looping reforming and combustion, despite its relevant content (26%) of iron species. While iron species demonstrated activity related to methane oxidation, as shown by the XRD patterns, chromium species were shown to not be active in the experimental conditions investigated, and the material displayed significant aggregation and carbon deposition during reduction, leading to an overall unsatisfying performance of the material.

Finally, in the case of the Ce–Cu process, step duration showed no effect on the selectivity towards either complete or partial oxidation for these short reaction times, as oxygen uncoupling favored complete combustion independently of the extent of carrier reduction. The Ce–Cu carrier thus appears to be unsuited for reforming, but shows excellent performance for combustion. For the Ce–Cu carrier, the selection of an ideal reduction step duration would thus mostly be determined by the rate of carbon deposition compared to the rate of oxidation as carrier conversion increases. Previously, Saddiq et al. investigated the kinetics for the oxidation of liquefied petroleum gas (propane/butane mixture) on 10% CuO/CeO₂ powder at 800 °C, and observed that a first-order kinetic model best described the reforming reaction, thanks to the high availability of oxygen, while a shrinking core model best described the reaction rate of pellets of the same material due to reduced oxygen release [29]. Nonetheless, high oxygen release was observed in both cases. He et al. also investigated the use of a 50/50 wt. % CuO/CeO₂ oxygen carrier for methane reforming [30]. They observed an almost complete selectivity towards partial oxidation above 850 °C, coupled with a very high carbon deposition, but this is not consistent with our observed results. In our experiments, the Cu–Ce oxygen carrier was almost completely selective towards complete oxidation both at 850 °C and 950 °C, and carbon deposition became relevant only for the longer-running tests at higher methane concentration. This discrepancy in results could be partly explained by the lower methane concentration and shorter reaction time used in our experiments compared to their results. The high selectivity towards complete combustion observed for our results can also be related to the preparation method employed for the Ce–Cu carrier. He et al. [30] prepared their sample through the co-precipitation method, which may allow a more intimate mixing of the two phases, allowing for cerium dioxide to play a major role in determining process selectivity. Our low carbon deposition results are closer to the observations made by Elgarni et al. [31], who also observed limited carbon deposition for CeO₂-supported CuO. Both Elgarni et al. [31] and Tijani et al. [32] observed a reduction in overall oxygen exchange capacity for CeO₂-supported Cu carriers when increasing the operation temperature above 900 °C, which would coincide with an increase in carbon deposition, but this is not immediately evident in our experimental results, as only limited aggregation was observed in our case for the Ce–Cu carrier.

5. Conclusions

The chemical looping conversion of methane was investigated in a lab-scale fixed-bed reactor. Three oxygen carriers with different oxygen capacities, one exhibiting CLOU behavior, were used. The materials showed different reactivity and oxygen transport capacity depending on their chemical composition, crystallographic phases, and microstructure. The rank of the studied carriers in methane conversion was Ce–Cu > CeO₂ > Chro. The Ce–Cu carrier was very effective in achieving the full oxidation of methane, whilst CeO₂ allowed either full or partial oxidation. Carbon deposition on CeO₂ was appreciable only for tests carried out at 950 °C. Natural chromite was unsatisfactory because of the absence of any contribution of the prevalent Cr₂O₃ phase.

The study of the process has also shown that the operating conditions and the control of the carrier conversion degree can alternatively lead to the partial (high CO selectivity) or total (high CO₂ selectivity) oxidation of CH₄. Indeed, the selectivity towards carbon monoxide also increases with a longer reduction time. The degree of carrier conversion, tunable by an effective switching strategy between the two cyclic phases of reduction and regeneration, is therefore a key parameter in guiding the process towards the desired products.

The further development of the research will consider steady-state operation, preferably in a circulating fluidized bed where the residence times of the gas and solids in reducing and oxidizing reactors can be easily modified.