Support-Active Phase Interaction in Oxidized and Reduced NiFe-Based Bifunctional Oxygen Carriers for Biomass Chemical Looping Gasification

Abstract

The rational design of oxygen carriers (OCs) is critical for enhancing biomass chemical looping gasification (BCLG) performance. This work systematically investigated the effects of different supports (Al₂O₃, ZrO₂, TiO₂, SiO₂) on the performance of NiFe-based OCs with oxidation and catalytic reforming functions. The gasification reactivity and support-active phase interaction of OCs in both oxidized and reduced states were evaluated. XRD, H₂-TPR, XPS, and SEM techniques were employed to characterize the phase composition, synergistic interactions, and surface morphology. The results showed that NiFeAl exhibited the optimal gasification performance in both oxidized and reduced states, achieving a syngas (H₂ + CO) yield of approximately 1.4 m³/kg (dry walnut shell). NiFeAl featured a higher Fe binding energy, abundant cavity structures, and the uniform dispersion of Ni and Fe on Al₂O₃, which confirm the formation of an appropriately strong Ni-Fe-Al ternary system. In contrast, NiFeZr suffered from the higher CO₂ yield, attributed to the over-oxidation caused by the weak interactions. NiFeTi and NiFeSi had lower syngas yields due to their poor reducibility induced by excessively strong interactions. This work verifies that moderate support-active phase interactions in OCs are optimal for BCLG.

1. Introduction

Biomass is a carbon-neutral renewable resource with abundant sources, and its utilization can not only reduce CO₂ emissions but also enhance energy security [1]. Syngas from biomass gasification is a critical feedstock for green methanol and bio-jet fuel synthesis, and plays a vital role in carbon emission reduction for the shipping and aviation industries. However, existing biomass gasification technologies face several challenges, including the high cost of air separation units and elevated tar and CO₂ contents. Biomass chemical looping gasification (BCLG) has emerged as a promising alternative, adopting a dual-reactor configuration (fuel reactor and air reactor), as shown in Figure 1. In this system, oxygen carriers (OCs) circulate between the two reactors to deliver oxygen, enabling the production of high-quality syngas using only air. As the circulating medium, OCs perform three key functions: replacing pure oxygen supply, regulating thermal management, and promoting tar conversion [2]. Notably, the performance of OCs, including redox activity, catalytic performance, and long-term stability, is a core factor affecting the overall efficiency of BCLG. In the fuel reactor (FR), the reduced OC does not immediately return to the air reactor (AR) but remains to participate in biomass gasification, mainly exerting catalytic functions such as cracking and reforming. Hence, the OC should exhibit good gasification performance even in its reduced form.

Figure 1. Schematic diagram of biomass chemical looping gasification technology.

Artificially synthesized OCs are generally composed of active phases and inert supports. The choice of support material imposes a profound influence on OCs’ performance [3]. Commonly used inert supports include Al₂O₃, ZrO₂, TiO₂, SiO₂, MgAl₂O₄, etc. Wang et al. [4] found that MnFe₂O₄ reacts with SiO₂ during redox cycles, forming MnSiO₃/Mn₂SiO₄ and leading to OC deactivation. Interactions between ferrites and iron silicates, formed by local reactions between coal ash and Fe oxides, also directly affect the redox kinetics of OCs [5]. TiO₂ may react with Ni to form highly stable NiTiO₃ with sluggish reaction kinetics [6,7]. ZrO₂ undergoes a monoclinic-to-cubic phase transition at about 1170 °C, which may deteriorate the structural stability of OCs rather than reinforce it [3]. DFT calculations by Feng et al. [8] suggest that Ni-ZrO₂ interactions can weaken the Ni-O bond strength. Furthermore, Al₂O₃-supported Cu-based OCs exhibit good anti-sintering behavior but reduced oxygen transport capacity [9,10]. NiAl₂O₄ formed from NiO-Al₂O₃ effectively suppresses Ni sintering, and the resulting metallic Ni⁰ shows enhanced catalytic activity [11,12]. Hu et al. [13] investigated Fe₂O₃ supported on Al₂O₃, ZrO₂, TiO₂, and SiO₂, and found that Fe₂O₃/Al₂O₃ achieved the highest gas yield, syngas purity, and carbon conversion efficiency. Fe₂O₃/TiO₂ and Fe₂O₃/ZrO₂ showed favorable CO productivity, whereas Fe₂O₃/SiO₂ exhibited higher carbon conversion. Overall, support-active phase interactions can enhance catalytic activity and mitigate sintering, but may also negatively affect redox kinetics and oxygen transfer efficiency.

The design of OCs also has a significant impact on tar formation during the BCLG process. Extensive efforts have therefore been devoted to tar mitigation. Steam-assisted chemical looping co-gasification was explored to lower tar yield and promote the production of H₂-rich syngas (CO + H₂) [14,15]. Yan et al. [16] modified LaCoO₃ perovskite at A/B sites with Sr/Fe, obtaining effective gas yields of 1.304 and 1.188 Nm³/kg with CO₂ concentrations of ~18%. Xia et al. [17] introduced A-site deficiencies into LaFeO₃, achieving an effective gas yield of ~1.22 Nm³/kg and a CO₂ concentration of ~15% via chemical looping steam gasification (CLSG). Nevertheless, these routes commonly suffer from undesirably high CO₂ selectivity, mainly caused by the promoted water-gas shift (WGS) reaction.

In contrast, catalytic in situ tar removal is recognized as an effective strategy for BCLG. Ni-based OCs are widely used for tar catalytic conversion [18,19,20]. Soheil et al. [21] found that Ni supported on CaZr_xTi_1−xO₃ perovskite exhibited better Ni dispersion and stronger metal-support interaction than conventional Ni-Al₂O₃ systems, thus enhancing catalytic activity and oxygen mobility. Ni promotes C-H bond cleavage in tar, favoring H₂ production. Fe doping further improves tar cracking activity and hydrogen yield [22], while Ni facilitates lattice oxygen migration to enhance redox performance [23]. However, carbon deposition and Ni sintering still present critical challenges to OCs’ stability [24]. Oxidation-reforming bifunctional OCs have been recognized as a promising strategy for high-efficiency biomass-to-syngas conversion [25]. In our previous work, we also developed a NiFeAl bifunctional OC by supporting Ni-Fe oxides on Al₂O₃, achieving an effective gas yield of 1.2–1.3 Nm³/kg with gas purity exceeding 95% [26]. However, the influence of different supports on the oxidation and reforming dual functionalities of OCs has not been systematically investigated.

Moreover, as mentioned above, dual-bed reactors (FR and AR) are commonly used in BCLG. After regeneration in the AR, OCs enter the FR and are gradually converted from the oxidized state to the reduced state. Accordingly, OCs exist as a mixture of oxidized and reduced phases in the FR. This is distinctly different from conventional reforming catalysts, which usually require pre-reduction. It thus imposes a strict requirement: OCs must rapidly exhibit excellent reforming activity even in the oxidized state upon entering the FR, ensuring high gasification reactivity in both states. However, studies focusing on the performance differences between oxidized and reduced oxygen carriers remain insufficient.

In this work, a series of supported NiFe-based bifunctional OCs with oxidation-reforming capabilities were synthesized, and the effects of different supports (Al₂O₃, ZrO₂, TiO₂, SiO₂) on their performances under both oxidized and reduced states were systematically investigated. XRD, H₂-TPR, XPS and SEM were also employed to characterize the synergistic interactions within the Ni-Fe-support systems. The results can provide valuable guidance for the rational design of OCs with favorable oxidation and catalytic reforming functionalities.

2. Results and Discussion

2.1. Effect of Supports on the Gasification Performance

2.1.1. Oxidized OCs

Figure 2 illustrates the gasification performance of various oxidized OCs and the pyrolysis results of individual walnut shell samples at 900 °C. From Figure 2a, NiFeAl-O exhibits the highest effective gas yield (1.39 m³/kg), indicating its excellent gasification performance. Meanwhile, NiFeAl-O has the lowest production of CH₄, CO_2, and C₂–C₃ hydrocarbons, which are significantly lower than that obtained under pyrolysis conditions. The results reveal that NiFeAl-O possesses a higher reforming agent (CO₂/H₂O) utilization efficiency, which enhances the tar and CH₄ reforming reactions and promotes the conversion of carbon and hydrogen into effective syngas. This confirms that NiFeAl-O exhibits remarkable synergistic effects during the coupled oxidation-reforming process.

Figure 2. Gasification performance of oxidized NiFeX. (a) Gas yield; (b) gas composition.

The CH₄ yield of NiFeZr-O is lower than that under pyrolysis, indicating that it has slight reforming capability. However, its significantly higher CO₂ yield relative to pyrolysis implies over-oxidation, which arises from its high lattice oxygen activity and poor selectivity. NiFeTi-O shows a lower effective gas yield and higher CO₂ yield. Notably, NiFeTi-O produces a higher CH₄ yield but a lower H₂ yield than that of NiFeAl-O, indicating inferior reforming activity. This can be attributed to the poor reducibility of NiFeTi-O, which hinders the effective exposure of Ni⁰ active sites and thus limits efficient catalysis of the reforming reaction shown in Equation (1). The CH₄, CO₂, and C₂–C₃ hydrocarbons yields of NiFeSi-O are similar to pyrolysis, confirming negligible oxidation and reforming functionalities.

CH₄/Tar + CO₂/H₂O → CO + H₂

(1)

As shown in Figure 2b, the concentration of effective gas produced by NiFeAl-O exceeds 95%, whereas the concentrations of CO₂ and CH₄ are below 5% and 1%, respectively. For the other OCs, the CH₄ concentration ranges from 3% to 15%, and the CO₂ concentration ranges from 8% to 17%. As can be seen from Figure 2a,b, the H₂ yields and concentration differ significantly among the various oxygen carriers, with NiFeAl exhibiting the best performance. Combined with the previous analysis of CH₄, CO₂, and C₂–C₃ hydrocarbons, it can be inferred that this is mainly due to the superior reforming ability of NiFeAl.

Figure 3 shows the carbon conversion and CO selectivity of various oxidized OCs and pyrolysis. As shown in Figure 3a, the carbon conversion of all oxidized OCs is consistently higher than that under pyrolysis, indicating that these OCs promote the conversion of carbonaceous macromolecules into low-carbon compounds. Among them, NiFeZr-O exhibits the highest carbon conversion, which can be attributed to the over-oxidation reaction shown in Equation (2). From Figure 3b, NiFeAl-O displays the highest CO selectivity, while NiFeZr-O shows the lowest, indicating that NiFeAl-O has the optimal partial oxidation performance.

H₂/CO/CH₄/Tar + O_* → CO₂ + H₂O

(2)

Figure 3. Carbon conversion and CO selectivity of oxidized NiFeX. (a) Carbon conversion; (b) CO selectivity.

2.1.2. Reduced OCs

As mentioned in the introduction, in the BCLG process, both oxidized and reduced OCs coexist in the FR. Therefore, besides the performance of the oxidized state, the reactivity of the reduced state is also important. Moreover, as shown in Section 2.1.1, the oxidized forms of NiFeZr, NiFeTi, and NiFeSi exhibited poor gasification performance and limited reforming activity. To further investigate whether the inferior reforming performance is associated with the presence of lattice oxygen, the OCs were pre-reduced and their performance was evaluated. The gasification performance of various reduced OCs at 900 °C is shown in Figure 4. According to Figure 4a, the effective gas yields of NiFeAl-R, NiFeZr-R, and NiFeTi-R reached 1.44, 1.37, and 1.28 m³/kg, respectively. The observed results can be attributed to the excellent reforming activity of the pre-reduced OCs, for the following reasons: (1) The effective gas yield increases significantly, accompanied by a pronounced decrease in the yields of CH₄, CO₂, and C₂–C₃ hydrocarbons. (2) The pre-reduction treatment eliminates lattice oxygen involvement in the gasification reactions and the influence of oxidation reaction can thus be excluded. (3) Biomass-derived methoxy (–OCH₃) and carboxyl (–COOH) groups typically decompose into CO₂ during the cracking reaction shown in Equation (3) [27]. However, present experimental data shows no increase in CO₂ yield relative to pyrolysis, indicating that cracking reaction is also weak.

Tar → H₂ + CO + CO₂ + hydrocarbons (CH₄, C₂H₆, C₂H₄, etc.) + C

(3)

Figure 4. Gasification performance of reduced NiFeX. (a) Gas yield; (b) gas composition.

The H₂ yield of NiFeTi-R is lower than that of NiFeAl-R and NiFeZr-R, while the CH₄ yield remains relatively higher, indicating insufficient reforming activity of NiFeTi-R. This can be attributed to the inadequate exposure of Ni⁰ active sites. Notably, the effective gas yield of NiFeSi-R is relatively low, and the yields of CH₄, CO₂, and C₂–C₃ hydrocarbons are comparable to those under pyrolysis, confirming that NiFeSi-R exhibits no significant reforming activity. Figure 4a,b show that while the CO yields of NiFeAl, NiFeZr, and NiFeTi are comparable, the H₂ yields decrease gradually and the CH₄ yields increase, confirming that NiFeAl has better reforming activity.

Table 1 summarizes the gasification performance of oxidized and reduced OCs at 900 °C, including gas yield, carbon conversion, and CO selectivity. The data in Table 1 show that the effective gas yield of NiFeAl remains approximately 1.4 m³/kg in both the oxidized (NiFeAl-O) and reduced (NiFeAl-R) states, demonstrating its excellent and stable performance regardless of state. The slightly higher CO₂ yield of NiFeAl-O confirms that lattice oxygen participates in the gasification reaction. However, no obvious over-oxidation reaction is observed relative to pyrolysis, indicating moderate and well-controlled oxygen activity in NiFeAl.

Table 1. Gasification performance results of NiFeX at 900 °C.

	Oxidized OCs				Reduced OCs
	NiFeAl	NiFeZr	NiFeTi	NiFeSi	NiFeAl	NiFeZr	NiFeTi	NiFeSi
H2 (m3/kg)	0.71	0.47	0.28	0.23	0.74	0.66	0.58	0.25
CO (m3/kg)	0.68	0.56	0.46	0.37	0.70	0.71	0.71	0.38
CH4 (m3/kg)	0.00	0.04	0.09	0.11	0.01	0.03	0.07	0.11
CO2 (m3/kg)	0.06	0.21	0.14	0.08	0.02	0.01	0.04	0.08
C2C3 (m3/kg)	0.00	0.00	0.01	0.04	0.00	0.00	0.00	0.04
H2 (%)	48.9	36.4	28.3	27.6	50.2	46.9	41.1	29.3
CO (%)	46.7	43.5	46.6	44.7	47.6	50.1	50.4	44.2
CH4 (%)	0.3	3.4	8.9	13.5	0.6	2.1	5.2	12.7
CO2 (%)	4.1	16.5	14.7	9.4	1.6	0.9	3.0	9.5
C2C3 (%)	0.0	0.2	1.5	4.8	0.0	0.0	0.3	4.3
Carbon conversion (%)	76.8	84.6	74.6	65.6	-	-	-	-
CO selectivity (%)	91.9	72.5	76.0	82.6	-	-	-	-

By contrast, the effective gas yield of NiFeZr-R increases by 38% compared with NiFeZr-O, accompanied by a sharp decrease in CO₂ yield. This reveals that the lattice oxygen in NiFeZr-O is overly active and tends to cause over-oxidation via Equation (2). Only when lattice oxygen is depleted (i.e., NiFeZr-R) does the sample favor reforming reactions, highlighting that NiFeZr only performs well in its reduced state, with unsatisfactory partial oxidation behavior. For NiFeTi-R, although the syngas yield is improved, the CH₄ yield is not effectively suppressed, indicating incomplete reforming reactions and still-inferior catalytic performance. In the case of NiFeSi, the gas yields barely differ between NiFeSi-O and NiFeSi-R, confirming that NiFeSi is inefficient in both oxidation and reforming reactions.

2.2. XRD Characterization

Figure 5a shows the XRD patterns of fresh NiFeX samples. It reveals that NiFeAl exhibits only a distinct α-Al₂O₃ phase, corresponding to PDF#10-0173 with a hexagonal crystal structure. No obvious nickel-containing species are detected, which is likely attributed to the low NiO loading (1 wt.%). Interestingly, no distinct iron-containing species are observed either, even with a Fe₂O₃ loading of 10 wt.%. This indicates that iron species may be incorporated into the Al₂O₃ lattice to form a solid solution.

Figure 5. XRD patterns of different materials. (a) Fresh NiFeX; (b) comparison of NiFeAl and Al₂O₃ support.

To explore the absence of Fe₂O₃ diffraction peaks in fresh NiFeAl, the XRD patterns of fresh NiFeAl and Al₂O₃ are compared in Figure 5b. The Al₂O₃ diffraction peaks in NiFeAl shift slightly toward lower angles relative to pure Al₂O₃, which can be ascribed to the incorporation of Ni and Fe species into the Al₂O₃ lattice. Given that the ionic radii of Ni²⁺ (0.069 nm) and Fe³⁺ (0.055 nm) are larger than that of Al³⁺ (0.0535 nm), the diffraction angle θ decreases according to Bragg’s law. The unit cell parameters of Al₂O₃ and NiFeAl are also calculated accordingly. For pure Al₂O₃, the parameters are determined as a = 4.76142 Å, b = 4.76142 Å, c = 13.00108 Å, with a unit cell volume of 255.26 ų. After Ni and Fe doping, the unit cell parameters of the Al₂O₃ phase in NiFeAl-O changed to a = 4.77942 Å, b = 4.77942 Å, c = 13.0447 Å, and the unit cell volume increases to 258.06 ų. This confirms that a portion of Ni and Fe ions enter the Al₂O₃ unit cells, resulting in lattice expansion.

From Figure 5a, the fresh NiFeZr displays diffraction peaks assigned to ZrO₂ (PDF#37-1484) and Fe₂O₃ (PDF#33-0664), respectively. However, no Fe-Zr composite crystalline phase is detected, suggesting that Fe does not combine with ZrO₂ to form a new phase. Accordingly, the Fe-O-Zr interaction may be relatively weak. The fresh NiFeTi shows characteristic peaks of TiO₂ (PDF#21-1276) and Fe-Ti composite oxides. The standard PDF patterns for Fe₃Ti₃O₁₀ and Fe₂TiO₅ overlap considerably and are difficult to distinguish. The formation of Fe-Ti composite oxides indicates a strong interaction between Fe and Ti species. Fresh NiFeSi exhibits diffraction peaks corresponding to SiO₂, Fe₂O₃, and Fe_2.95Si_0.05O₄. Notably, Fe₂O₃ exists in two distinct crystal systems: rhombohedral (PDF#33-0664) and hexagonal (PDF#39-1346). The presence of Fe_2.95Si_0.05O₄ verifies a strong interaction between Fe and Si species.

The XRD patterns of NiFeX in the oxidized, reduced, and spent states are shown in Figure 6. From Figure 6a, the intensity of the Al₂O₃ peaks for NiFeAl exhibits no obvious variation among the oxidized, reduced, and spent states. However, the oxidized NiFeAl shows distinct diffraction peaks at 30.66°, 36.10°, and 63.96° (marked by red circles), while these peaks are absent in the reduced and spent samples. This can be attributed to the phase transformation induced by the reduction of Ni/Fe oxides.

Figure 6. XRD patterns of NiFeX in different states. (a) NiFeAl; (b) NiFeZr; (c) NiFeTi; (d) NiFeSi.

As shown in Figure 6b, no Fe₂O₃ or NiFe₂O₄ phases are detected in the reduced and spent NiFeZr samples. This can be attributed to the reduction of the oxygen carrier by reducing gases such as H₂, CO, and CH₄, which confirms the excellent reducibility of NiFeZr. From Figure 6c, Ni⁰ is observed in reduced NiFeTi, whereas no iron oxides are observed. In spent NiFeTi, the FeTiO₃ phase (PDF#29-0733) is present while Ni⁰ is absent. This indicates that the gasification process induced the conversion of Fe³⁺ to Fe²⁺ but failed to achieve deep reduction of NiFeTi. As shown in Figure 6d, both oxidized and spent NiFeSi contain three phases: SiO₂, Fe₂O₃, and Fe_2.95Si_0.05O₄, whereas the reduced NiFeSi only exhibits two phases: SiO₂ and Fe_2.95Si_0.05O₄. This indicates that its reduction requires relatively harsh conditions. Such strong interactions suppress the gasification performance, which is consistent with the lower effective gas yield achieved over NiFeSi.

From the viewpoint of crystal phase evolution, when Fe species combine with the support to form new phases that cannot be effectively reduced during gasification, the material exhibits inferior gasification performance. This confirms that an excessively strong Fe-support interaction of OCs is unfavorable in the BCLG process.

2.3. H₂-TPR Characterization

The H₂-TPR profiles of fresh NiFeX samples are presented in Figure 7. The main reduction peaks of NiFeAl appear at 862 °C, and the peak area is smaller than those of NiFeZr and NiFeTi. As shown by the hydrogen consumption data in Table 2, the amount of reducible lattice oxygen is limited, suggesting that its lattice oxygen activity is moderate and resistant to over-oxidation. NiFeZr shows a single reduction peak at 616 °C, indicating high lattice oxygen activity. The reduction process is nearly complete at approximately 700 °C, and its hydrogen consumption is about three times that of NiFeAl, implying that most lattice oxygen can participate in gasification at 900 °C. Although this favors the exposure of reforming sites, it promotes the over-oxidation reaction shown in Equation (2) to produce excessive CO₂, which is consistent with the gasification results in Section 2.1.

Figure 7. H₂-TPR profiles of fresh NiFeX samples.

Table 2. Total H₂ consumption of different OCs in H₂-TPR experiments.

OCs	NiFeAl	NiFeZr	NiFeTi	NiFeSi	NiAl	FeAl	NiZr	FeZr	NiTi	FeTi	NiSi	FeSi
Quantity (mmol/g)	0.371	1.031	1.160	0.063	0.103	0.150	0.100	0.968	0.195	1.033	0.081	0.117

As shown in Figure 7, NiFeTi exhibits three distinct reduction peaks corresponding to different lattice oxygen active sites. Since the gasification experiments is conducted at 900 °C, the lattice oxygen in the high-temperature region of NiFeTi cannot participate in the reaction, whereas the reducible oxygen species at medium and low temperatures tend to induce over-oxidation. Consequently, due to insufficient exposure of Ni⁰ sites, the generated H₂O and CO₂ cannot be efficiently consumed for tar reforming. This phenomenon is clearly supported by the gasification results in Section 2.1, where higher CH₄ and CO₂ yields and a lower effective gas yield are observed. NiFeSi shows no obvious reduction peak and has the lowest hydrogen consumption, indicating that NiFeSi is the most difficult to reduce among the four OCs. This result is consistent with the XRD observation that the Fe_2.95Si_0.05O₄ phase in NiFeSi remains stable before and after the reaction. Its lattice oxygen is hardly released, and the reforming sites cannot be exposed, thus restricting both its oxidation capacity and reforming performance.

To further investigate the interactions between Ni-Fe species and the supports, monometallic NiAl, FeAl, NiZr, FeZr, NiTi, FeTi, NiSi, and FeSi samples are synthesized and characterized by H₂-TPR, as illustrated in Figure 8. According to Figure 8a, NiAl exhibits a weak reduction peak at approximately 500 °C, corresponding to the reduction of NiO, indicating that only a small fraction of NiO exists in a free form. As is well known, for supported metal oxide catalysts, the low-temperature and high-temperature TPR peaks correspond to weak and strong interactions among components, respectively [28]. Thus, the peaks at 646 °C and 907 °C represent weak and strong Ni–O–Al bonding, respectively. Notably, the peak area at 907 °C is significantly larger than that at 646 °C, suggesting that a great amount of NiO species interact strongly with Al₂O₃. FeAl shows reduction peaks at 450 °C, 677 °C, and 1000 °C. The peak at 450 °C corresponds to the reduction of Fe₂O₃ to Fe₃O₄ [29,30], while the peaks at and above 600 °C are attributed to the stepwise reduction of Fe₃O₄ → FeO → Fe [31,32]. Accordingly, the peaks at 677 °C and 1000 °C can be assigned to weak and strong Fe-O-Al bonding, respectively.

Figure 8. H₂-TPR profiles of OCs in different systems: (a) Ni-Fe-Al; (b) Ni-Fe-Zr; (c) Ni-Fe-Ti; (d) Ni-Fe-Si.

NiFeAl displays reduction peaks at 470 °C, 719 °C, and 860 °C. Compared with FeAl, the medium-temperature reduction peak of NiFeAl shifted upward from 677 °C to 715 °C, while the high-temperature reduction peak shifted downward from 1000 °C to 862 °C. This indicates that the addition of Ni enhances the weak Fe-O-Al bonding but suppresses the formation of FeAl₂O₄. Similarly, compared with NiAl, the high-temperature reduction peaks of NiFeAl appear at lower temperatures, suggesting that the incorporation of Fe weakens the Ni species–Al support interactions [33], thus making part of Ni more easily reducible. This leads to a larger total amount of reducible Ni/Fe oxides, thereby improving the overall redox behavior of the material. As shown in Table 2, the H₂ consumption of NiFeAl exceeds the sum of that for NiAl and FeAl. It enables more lattice oxygen in the OC to participate in the gasification process.

From Figure 8b, NiZr exhibits two reduction peaks at 410 °C and 830 °C, which can be assigned to the reduction of NiO and the strong Ni-O-Zr bonding, respectively. FeZr shows two reduction peaks at 503 °C and 620 °C. The peak at 503 °C is attributed to weak Fe-O-Zr bonding, while the peak at 620 °C corresponds to the stepwise reduction of Fe₃O₄ → FeO → Fe. No strong Fe-Zr interactions are observed. NiFeZr exhibits only one reduction peak at 616 °C. Although the maximum reduction temperatures of FeZr and NiFeZr are similar, the addition of Ni clearly narrows the temperature range of FeZr reduction peaks and enhances the peak intensity. This indicates that Ni improves the lattice oxygen activity and thus accelerates the reduction of Fe₃O₄ → FeO → Fe. The highly active lattice oxygen may cause over-oxidation during gasification and promote the formation of CO₂, which is consistent with the gasification performance results.

As shown in Figure 8c, the reduction peaks of NiTi at 642 °C and 767 °C correspond to the weak and strong Ni-O-Ti bonding, respectively. For FeTi, a small reduction peak appears at 674 °C, with the main peak at 827 °C and an incomplete peak at 1000 °C. Unlike FeAl and FeZr, FeTi shows no discernible low-temperature reduction peak, suggesting that the conversion of Fe₂O₃ to Fe₃O₄ is not significant. Overall, the reduction peaks of FeTi are shifted to higher temperatures, indicating strong interactions between Fe and Ti oxides, which may facilitate the formation of new phases.

NiFeTi exhibits reduction peaks at 642 °C, 774 °C, and 955 °C. The peak at 642 °C is associated with the weak Ni-O-Ti bonding. However, its intensity deviates from expectations, suggesting that this peak may partially originate from the reduction of Fe oxides. Although both FeTi and NiFeTi also display three reduction peaks, those of NiFeTi are located at lower temperatures. Specifically, significant reduction of FeTi initiates at approximately 720 °C, whereas reduction of NiFeTi begins at around 520 °C. This indicates that the addition of Ni weakens the Fe-O-Ti bonding, thereby promoting the reduction of Fe oxides. In terms of hydrogen consumption shown in Table 2, the value for NiFeTi is approximately equal to the sum of those for FeTi and NiTi. It indicates that while Ni lowers the reduction temperature of Fe oxides, it does not increase the total amount of reducible Fe species. Compared with the H₂-TPR results of NiFeAl, NiFeTi shows a significant reduction peak above 900 °C, indicating incomplete reduction under the reaction conditions of this work. This incomplete reduction hinders the participation of lattice oxygen and suppresses the formation of Ni⁰ reforming active sites, resulting in the inferior gas production performance of NiFeTi relative to NiFeAl.

Figure 8d indicates that the reduction peak of NiSi at 458 °C corresponds to free NiO. Similar to NiAl and NiTi, the reduction peak at 682 °C is attributed to the weak Ni-O-Si bonding, while the strong Ni-O-Si bonding is reflected in the reduction peak at about 800 °C. Consistent with the above analyses of FeAl, FeZr, and FeTi, the reduction peak of FeSi at 473 °C corresponds to the reduction of Fe₂O₃ to Fe₃O₄, while the reduction peaks at 670 °C, 732 °C, and 1000 °C are attributed to the Fe-O-Si bonding. The high-temperature reduction peak of FeSi indicates a strong interaction between Fe and Si, which may suggest the formation of new phases such as Fe_2.95Si_0.05O₄, as observed in the XRD results.

The reduction peak of NiFeSi at 483 °C is related to the reduction of free NiO and the transformation of Fe₂O₃ to Fe₃O₄. By comparing the profiles of FeSi and NiSi, it can be seen that the addition of Ni significantly decreases the high-temperatures reduction peak area of FeSi, suggesting that the reduction of part of the Fe oxides is inhibited, possibly due to interactions with SiO₂. However, no significant shift in the reduction peaks to higher temperatures is observed. From the comparison of hydrogen consumption in Table 2, it is evident that Ni significantly suppresses the reduction behavior of Fe oxides in NiFeSi.

Combined with the H₂-TPR profiles and gasification performance, excessively strong internal interactions within OCs will impair their reducibility. This not only hinders the effective release of lattice oxygen but also suppresses the formation of reforming active sites. Conversely, weak internal interactions lead to highly active lattice oxygen, which easily induces over-oxidation and thus is also unfavorable for effective gas production. Notably, when the reduction peaks of OCs fall within the range of 600–900 °C, moderate interactions between components are more conducive to the BCLG reaction.

2.4. XPS Characterization

Figure 9 displays the XPS results of various OCs. As shown in Figure 9a, distinct variations in Fe 2p peak intensity are observed among the four OCs. These variations may be attributed to the following two reasons: (1) differences in the surface Fe concentration or dispersion; (2) photoelectron escape attenuation and electrical conductivity changes induced by the support materials. Besides the Fe 2p_3/2 and Fe 2p_1/2 peaks centered at 711–712 eV and 724 eV, respectively, the characteristic satellite peaks at ~719 eV and ~734 eV indicate the dominant presence of Fe³⁺ species [34,35]. The peak at ~712 eV can be deconvoluted into two peaks at ~711 eV and ~713 eV, corresponding to composite Fe³⁺ (affected by Ni species [36]) and Fe-Al interactions, respectively. Notably, NiFeAl exhibits a higher Fe 2p_3/2 binding energy (BE), suggesting a lower electron cloud density around Fe atoms and significant electron transfer. This electronic state benefits the gasification reaction by facilitating H₂O adsorption and boosting the steam reforming of tar (e.g., toluene) [37], which ultimately improves the yield of effective syngas.

Figure 9. XPS spectra of fresh NiFeX: (a) Fe 2p; (b) Ni 2p_3/2.

From Figure 9b, the Ni 2p_3/2 spectra show a main peak at 855–856 eV, which is attributed to Ni²⁺ and accompanied by a satellite peak at ~862 eV. The peak at 855 eV can be deconvoluted into two components at ~854.6 eV and 855.5 eV, corresponding to Ni-Fe interactions and Ni-support interactions, respectively [38,39,40,41]. Both Ni and Fe in NiFeAl display higher BEs, implying lower electron cloud density and greater electron transfer from the active metals to the support. This confirms the presence of stronger synergistic interactions among Ni, Fe, and Al in NiFeAl. It also illustrates the corresponding relationship between the synergistic effect and gasification performance, and further validates the superiority of NiFeAl.

2.5. SEM Characterization

The surface morphology of the fresh NiFeX is analyzed by SEM, as shown in Figure 10. The four OCs exhibit distinct surface characteristics. NiFeAl displays agglomerated irregular nanoparticles. Although particle agglomeration is observed, it possesses a dense array of mesoscale cavities, which enhance the diffusion efficiency of reactants and products, and then improve overall mass transfer efficiency. However, NiFeZr, NiFeTi, and NiFeSi present micrometer-sized particles that reflect the characteristic dimensions of the support. Small agglomerates, likely Ni/Fe oxides, are distributed on the support surface. The pore structure of these three OCs is significantly less developed than that of NiFeAl.

Figure 10. SEM images of fresh NiFeX: (a-1) NiFeAl; (a-2) enlarged area of NiFeAl; (b-1) NiFeZr; (b-2) enlarged area of NiFeZr; (c-1) NiFeTi; (c-2) enlarged area of NiFeTi; (d-1) NiFeSi; (d-2) enlarged area of NiFeSi.

EDS-mapping characterization is used to further analyze the elemental distribution of each component, as shown in Figure 11. The results demonstrate a uniform distribution of Ni and Fe in NiFeAl, suggesting strong interactions among Ni, Fe, and Al. This can be attributed to the solid-solution dispersion of Ni and Fe species, which is consistent with the XRD results. In contrast, Ni and Fe in NiFeZr and NiFeTi exhibit an uneven distribution with obvious element aggregation, which correlates well with the existence of multiple crystalline phases observed in the XRD patterns. Additionally, Ni and Fe show a relatively uniform distribution on the SiO₂ surface in NiFeSi.

Figure 11. SEM-EDS mapping of fresh NiFeX.

The spent OCs were also examined by SEM and elemental mapping (Figures S1 and S2, respectively). As shown in the SEM images, all OCs largely retained their original morphology. NiFeAl maintained a well-preserved nanoparticle-aggregated structure with medium-sized cavities and a particle size of approximately 300 nm. NiFeZr, NiFeTi, and NiFeSi still mainly presented micrometer-sized support particles, with NiFeTi and NiFeSi showing irregular small particles on the support surface. The corresponding mapping results (Figure S2) demonstrate that the Ni and Fe elements remained uniformly dispersed in spent NiFeAl, while spent NiFeZr, NiFeTi, and NiFeSi showed non-uniform distributions.

2.6. Reaction Mechanism

Based on the gasification performance evaluation and characterization results, this study proposes the reaction mechanism of NiFeX, as illustrated in Figure 12. As shown in Figure 12a, Ni and Fe species are uniformly dispersed within the Al₂O₃ support, giving rise to a moderate ternary interaction among the components. The oxygen activity is thus well balanced, effectively avoiding over-oxidation of H₂ and CO. Moreover, the Fe species exhibits a relatively stronger electron-deficient state, which promotes water vapor adsorption and enhances the reforming reaction between tar (e.g., toluene) and steam [37], leading to high-quality syngas. Figure 12b shows that the interaction between ZrO₂ and the NiFe species is relatively weak, resulting in high oxygen activity. This favors complete oxidation, producing more CO₂ and H₂O and consequently lowering the utilization efficiency of carbon and hydrogen atoms. Figure 12c indicates that the Ni and Fe species interact more strongly with the TiO₂ support, which leads to poor reducibility and inferior reforming activity. As a result, a significant amount of CH₄ and a small amount of C₂–C₃ hydrocarbons are detected. Figure 12d reveals that Fe species form a stable Fe_2.95Si_0.05O₄ phase with the SiO₂ support, making the oxygen carrier difficult to reduce. Consequently, both oxidation and reforming activities are poor, resulting in a high yield of C₂–C₃ hydrocarbons.

Figure 12. Schematic diagram of the reaction mechanism of NiFeX: (a) NiFeAl; (b) NiFeZr; (c) NiFeTi; (d) NiFeSi.

3. Experiments and Methods

3.1. Materials

Walnut shell (obtained from Zishu Mineral Products Co., Ltd., Shijiazhuang, China) was used as a typical biomass sample, and its proximate and ultimate analyses are listed in Table 3. The raw walnut shell particles were crushed in a grinder (sourced from Taisite Instrument Co., Ltd., Tianjin, China) and sieved to obtain the 20–40 mesh fraction. The sample was then dried at 105 °C for 24 h to obtain dry biomass.

Table 3. The ultimate and proximate analyses of the walnut shell.

Proximate Analysis (wt %)				Ultimate Analysis (wt %, d)					LHVd
Mad	Vd	Ad	FCd *	C	H	N	S	O *	MJ/kg
9.43	82.87	0.46	16.67	47.40	7.02	0.13	0.03	44.96	18.57

Note: M—moisture, V—volatile, A—ash, FC—fixed carbon, LHV—lower heating value, d—dry basis, ad—air dry basis, *—by difference.

NiFe-based bifunctional OCs supported on different inert materials (Al₂O₃, ZrO₂, TiO₂, SiO₂) were synthesized via the equal-volume impregnation method. The properties of the supports are as follows: Al₂O₃ (industrial grade, specific surface area 145.17 m²/g, sourced from a local supplier in Zibo, China); SiO₂ (purity > 98%, specific surface area 300–400 m²/g, sourced from a local supplier in Qingdao, China); ZrO₂ (purity 99%, Shanghai Macklin Biochemical Technology Co., Ltd., Shanghai, China); TiO₂ (purity ≥ 98%, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China). Firstly, the water absorption capacity of different supports was assessed. Secondly, the corresponding amounts of Ni(NO₃)₂·6H₂O (purity ≥ 98%, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) and Fe(NO₃)₃·9H₂O (purity ≥ 98.5%, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) were dissolved in deionized water and the resulting solution was impregnated onto the supports. The samples were then aged for 12 h, dried at 80 °C for 6 h, and calcined at 1200 °C for 6 h with a heating rate of 2 °C/min in a muffle furnace (Hefei Kejing Materials Technology Co., Ltd., Hefei, China). The detailed preparation procedure has been reported in our previous work [26]. The prepared NiFe-based OCs were denoted as NiFeX, where X represents the inert support (e.g., NiFeAl). To differentiate between oxidized and reduced states, the oxidized and reduced samples were labeled NiFeX-O and NiFeX-R, respectively. Additionally, monometallic OCs including NiAl, FeAl, NiZr, FeZr, NiTi, FeTi, NiSi, and FeSi were prepared by the same method to better clarify the interactions between different components.

3.2. Experimental Methods

The gasification experiments were carried out in a fixed-bed reactor, as illustrated in Figure 13. Initially, the OCs (4.4 g) were placed into a quartz tube (bed height ca. 3–5 cm), and walnut shell samples (0.43 g) were loaded into the feeder at the top of the tube. The furnace was then heated to 900 °C at a rate of 10 °C/min. Argon was used as the carrier and purge gas, flowing through the system at 32 mL/min. For experiments evaluating the performance of reduced OCs, the sample was firstly reduced in 30% H₂ during the heating process and held at 900 °C for 30 min. After the system stabilized, the walnut shell was added into the quartz tube and placed above the oxygen carrier. The walnut shell was first pyrolyzed, and the resulting pyrolysis gas then came into contact with the oxygen carrier and underwent oxidation and reforming reactions. Due to the limited solid–solid contact and slow solid–solid reaction rate, it can be assumed that the pyrolysis char did not participate in the reaction with the oxygen carrier, and only the pyrolysis gas was involved. The product gas was collected for 10 min under continuous Ar flow (32 mL/min). The syngas composition was analyzed using a gas chromatograph (GC8850, Lunan Ruihong, Tengzhou, China). To further compare and evaluate the reforming performance of the OCs, pyrolysis experiments of walnut shell samples were also performed under the same procedure. In contrast to gasification, no OCs were used during the pyrolysis experiments, while all other experimental conditions remained identical. Each gasification experiment was performed in two independent replicate runs. The error bars in Figure 2 and Figure 4 represent the deviation of gas yields between two independent repeated experiments.

Figure 13. Fixed-bed gasification experimental system diagram.

The crystal phase structure of the OCs in oxidized, reduced, and post-reaction states were analyzed using XRD (SmartLab9kW, Rigaku Corporation, Tokyo, Japan). The reducibility and intercomponent interaction of the samples were investigated by H₂-TPR (BELCAT II, MicrotracBEL Corp., Osaka, Japan). XPS (Thermo Scientific ESCALAB QXi, Waltham, MA, USA) was employed to characterize the elemental valence states and chemical environments of the samples. All powder samples were prepared under ambient conditions and then rapidly transferred into the XPS vacuum load-lock chamber. After the sample chamber was evacuated to high vacuum, the samples were immediately introduced into the XPS analysis chamber to avoid any further air exposure, where most of the physically adsorbed CO₂ and H₂O were removed. SEM (Merlin Compact, Carl Zeiss AG, Oberkochen, Germany) was used to observe the microscopic morphology of the prepared OCs, while EDS (Oxford X-Max, Oxford Instruments, Oxfordshire, UK) was used to evaluate the dispersion of the active elements on the supports.

3.3. Definition and Calculation

The total syngas yield per unit mass of dry walnut shell was defined by Equation (4):

$${\mathrm{G}}_{\mathrm{V}}={\displaystyle \frac{\mathrm{t}\times {\mathrm{F}}_{\mathrm{A}\mathrm{r}}\times \left(1-{\mathrm{V}}_{\mathrm{A}\mathrm{r}}^0\right)\times {10}^{-6}}{{\mathrm{V}}_{\mathrm{A}\mathrm{r}}^0\times m\times {10}^{-3}}}$$

(4)

where ${\mathrm{G}}_{\mathrm{V}}$ was total syngas yield (m³/kg, at 25 °C, 1 atm), ${\mathrm{F}}_{\mathrm{A}\mathrm{r}}$ was the flow rate of argon (32 mL/min), ${\mathrm{V}}_{\mathrm{A}\mathrm{r}}^0$ was argon concentration in syngas (vol %), t was gas collection time (10 min), m was the mass of dry walnut shell (0.43 g).

The volume concentration of different components in syngas was defined as Equation (5):

$${\mathrm{V}}_{\mathrm{i}}={\displaystyle \frac{{\mathrm{V}}_{\mathrm{i}}^0}{{\displaystyle \sum {\mathrm{V}}_{\mathrm{i}}^0}}}$$

(5)

where ${\mathrm{V}}_{\mathrm{i}}$ was the volume concentration (%), i was the detected components in syngas (H₂, CH₄, CO, CO₂, C₂H₆, C₂H₄, C₃H₈, and C₃H₆), ${\mathrm{V}}_{\mathrm{i}}^0$ was the original gas concentration measured by GC8850 (vol %).

The components yield, effective gas yield, CO selectivity, and carbon conversion were defined by Equations (6)–(9):

$${\mathrm{G}}_{\mathrm{i}}={\mathrm{G}}_{\mathrm{V}}\times {\mathrm{V}}_{\mathrm{i}}$$

(6)

$${\mathrm{G}}_{\mathrm{e}}={\mathrm{G}}_{{\mathrm{H}}_2}+{\mathrm{G}}_{\mathrm{C}\mathrm{O}}$$

(7)

$${\mathrm{S}}_{\mathrm{C}\mathrm{O}}={\displaystyle \frac{{\mathrm{G}}_{\mathrm{C}\mathrm{O}}}{{\mathrm{G}}_{\mathrm{C}\mathrm{O}}+{\mathrm{G}}_{{\mathrm{C}\mathrm{O}}_2}}}\times 100\mathrm{\%}$$

(8)

$${\mathsf{\eta }}_{\mathrm{C}}={\displaystyle \frac{(\sum {\mathrm{G}}_1+\sum {\mathrm{G}}_2\times 2+\sum {\mathrm{G}}_3\times 3)\times \mathrm{M}\times 273}{{\mathrm{V}}_{\mathrm{m}}\times 298\times {\mathrm{C}}_{\mathrm{w}}}}\times 100\%$$

(9)

where ${\mathrm{G}}_{\mathrm{i}}$ was the yield of each gas component (m³/kg), ${\mathrm{G}}_{\mathrm{e}}$ was the effective gas yield (m³/kg), ${\mathrm{S}}_{\mathrm{C}\mathrm{O}}$ and ${\mathsf{\eta }}_{\mathrm{C}}$ were the CO selectivity and carbon conversion, respectively (%). G₁ represented the yields of CH₄, CO, CO₂ (m³/kg), G₂ corresponded to the yields of C₂H₆ and C₂H₄ (m³/kg), G₃ denoted the yields of C₃H₈ and C₃H₆ (m³/kg); M and V_m were the molar mass of carbon and molar volume of the gas, respectively. C_w was the mass of carbon contained in the dry walnut shell samples (kg).

4. Conclusions

In this study, the effects of different supports (Al₂O₃, ZrO₂, TiO₂, SiO₂) on the performance of NiFe-based bifunctional OCs were investigated. The biomass gasification reactivity and support-active phase interaction of various OCs in both oxidized and reduced states were evaluated. The main conclusions are as follows:

(1)

NiFeAl exhibits the optimal gasification performance in both oxidized and reduced states, with effective gas (CO + H₂) yields of 1.39 m³/kg and 1.43 m³/kg (95% and 97% concentrations, respectively). NiFeZr shows good reforming reactivity in the reduced state but severe over-oxidation in the oxidized state. NiFeTi has inferior catalytic reforming performance in oxidized states while NiFeSi is inefficient in both oxidized and reduced states.

(2)

No detectable Fe-Zr oxides generate in NiFeZr, with iron oxides retained as Fe₂O₃ and NiFe₂O₄ phases. These weak interactions lead to highly active lattice oxygen and subsequently over-oxidation. Complex Fe-Ti and Fe-Si oxides generate in NiFeTi and NiFeSi, respectively. These excessively strong interactions impair their reducibility and result in poor oxidation and reforming performance.

(3)

NiFeAl has abundant cavity structures, and enhanced binding energies of Ni and Fe, along with a uniform dispersion of Ni and Fe species on Al₂O₃. Partial incorporation of Ni and Fe into the Al₂O₃ lattice also suppresses iron oxide segregation. The results verify the robust Ni-Fe-Al ternary interactions and further confirm that moderate interactions between the support and active phase in OCs are optimal for biomass chemical looping gasification.