H₂ Production from Pyrolysis-Steam Reforming of Municipal Solid Waste and Biomass: A Comparative Study When Using the Self-Derived Char-Based Catalysts

Abstract

This study employed a two-stage fixed-bed pyrolysis-reforming reactor to investigate H₂ production behaviors from municipal solid waste (MSW) and biomass with their self-derived catalysts under different operating parameters. The self-derived catalysts are prepared by mechanically mixing pyrolysis-derived chars with CaO and iron powders. The main results are as follows: (1) The higher oxygen content in biomass facilitates oxidative dehydrogenation reactions, enabling in situ generation of H₂O, which results in a higher H₂/CO ratio for biomass compared to MSW under steam-free conditions. (2) There are optimal values for the reforming temperature and steam-to-feedstock ratio (S/F) to achieve best performance. In the presence of steam, MSW generally exhibits superior H₂ and syngas production performance to biomass; (3) Both MSW char (MSWC)- and biomass char (BC)-based catalysts showed satisfied H₂ production and tar cracking performance at 850–900 °C, and the MSWC-based catalyst demonstrated better catalytic activity than the BC-based catalyst due to its higher contents of several active metals. In addition, the iron powder can be recycled easily, proving the effectiveness of the self-derived convenient and cheap catalysts.

1. Introduction

Producing H₂ from waste is a popular topic in the context of the carbon-neutral era. Thermochemical H₂ production from biomass or municipal solid wastes (MSWs) is a green H₂ pathway that utilizes high temperatures to break down large organic molecules into small gas molecules such as H₂ and CO. Among the various technical routes, the most common approach involves pyrolysis followed by catalytic reforming of volatiles [1]. With H₂ being the primary target product, steam is typically added during the reforming process to achieve higher H₂ yields. A typical reaction is the water–gas shift reaction (Equation (1)), which utilizes H₂O to convert CO in the gas into CO₂ while simultaneously generating H₂. Meanwhile, the introduction of steam can effectively inhibit carbon deposition on the catalyst, thereby maintaining its catalytic activity.

Table 1. Comparison of H₂ production performance from different feedstocks and catalysts.

Feedstock	Catalyst and Operation Parameters	Catalytic Performance	Ref.
Water hyacinth	9 wt. % Ni/sepiolite + SiC Tp 1 = 650 °C; Tr 1 = 800 °C; Steam/Feedstock = 0.	H2:CO:CO2:CH4 = 77.2:7.8:10.7:4.3 YH2 = 1225 L kg−1feedstock Yliquid = 16 wt. %	[1]
HDPE	14 wt. %NiO/CaAl2O3 Tp = 500 °C; Tr = 700 °C; Steam/Feedstock = 4.	H2:CO:CO2:CH4:C2+ = 71.5:10.5:16.9:0.7:0.4 YH2 = 4231 L kg−1feedstock Yliquid = 0.06 wt. %	[2]
Wood sawdust	Fe-ZnO/Al2O3 = 1:1 Tp = 500 °C; Tr = 800 °C; Steam/Feedstock = 0.	H2:CO:CO2:CH4:C2+ = 40.0:24.8:24.7:8.1:2.4 YH2 = 236 L kg−1feedstock	[3]
SMWP (Stimulated mixed waste plastics)	Fe-Ni/MCM-41 (Fe/Ni = 10:10) Tp = 500 °C; Tr = 800 °C. mfeedstock = 3.4 g Steam flow rate = 2 mL h−1	H2:CO:CO2:CH4:C2+ = 46.7:32.2:1.9:6.2:12.9 YH2 = 1129 L kg−1feedstock	[4]
LDPE	Tire char Tp = 500 °C; Tr = 1000 °C. mfeedstock = 9.2 g Steam flow rate = 8 mL h−1	H2:CO:CO2:CH4:C2+ = 49.7:30.4:13.0:5.1:1.8 YH2 = 3261 L kg−1feedstock	[5]
Bio-oil	3Ni9Co/Ce-Zr-O Tr = 850 °C; Moisture content of bio-oil = 57.52 wt. %.	H2:CO:CO2:CH4 = 57.65:28.32:11.01:3.02 YH2 = 72.15 wt. % (in relation to mass of feedstock)	[6]
Raw bio-oil	Ni/La2O3-αAl2O3 Tr = 700 °C; S/C = 6.	H2 Concentration = 66 vol% YH2 = 83 wt. % (in relation to mass of feedstock)	[7]
Phenol (bio-oil model compound)	Rh/MgCeZrO Tr = 700 °C; Steam/Phenol = 80:1	H2:CO:CO2 = 9.4:0.4:4.3 Phenol conversion = 70 wt. %	[8]
Acetic acid	Rh/CeZrO2 Tr = 761 °C; Steam/acetic acid = 2:1	H2 selectivity = 70% Acetic acid conversion = 100%	[9]
Polypropylene	Co/Al2O3 Tp = 500 °C; Tr = 850 °C; Steam/Feedstock = 1.33.	H2/CO = 3:1 YH2 = 2940 L kg−1feedstock (pyrolysis stage: YH2 = 1781 L kg−1feedstock)	[10]

¹ Tp: pyrolysis temperature; Tr: reforming temperature.

presents the H₂ production performance via the pyrolysis-reforming process when different feedstocks and catalysts were adopted. It shows that plastic-based feedstocks exhibit significantly higher H₂ yield than biomass-based materials. Meanwhile, increasing water input generally enhances H₂ yield. The higher reforming temperature led to improved gas production, and the reforming temperature typically needs to exceed 700 °C to achieve satisfactory H₂ production performance. The catalysts employed are generally noble metals or transition metals supported on porous materials; for example, Ni supported on Al₂O₃ is frequently adopted. The catalyst maintenance contributes to an important fraction of the running cost for H₂ production.

H₂O + CO ⇄ CO₂ + H₂

(1)

The catalysts for reforming reactions are generally composed of metallic elements and porous supports. Both the doped metal and the catalyst support play crucial roles in the activity of the catalyst [11,12]. Taking biomass as an example, oxygen-containing compounds generated from biomass pyrolysis first dissociated on the active metal sites to form adsorbed carbon species. Simultaneously, H₂O was adsorbed and dissociated on the support to form -OH species, which then migrated to the metal sites or the metal–support interface to react with carbon species, producing CO_x and H₂ [11,13]; both the active metal and the support affected the performance of the whole process. The carbon-supported catalyst adopts porous carbon as a carrier which can adsorb and decompose macromolecular organic compounds. Particularly, carbon materials featuring abundant micropores and high specific surface areas play a significant role in macromolecule decomposition [14]. However, the decomposition of macromolecules at micropore sites is typically accompanied by condensation reactions, leading to carbon deposition [14,15]. With the introduction of steam, the gasification of deposited carbon regenerates the micropores [14].

As an inexpensive and readily available commonly used active metal, Ni promotes methanation reactions at low temperatures [16]. However, at medium to high temperatures (500–800 °C), it serves as a suitable catalyst for steam reforming reactions, exhibiting high reactivity in the cleavage of C-C and C-H bonds, achieving an H₂ yield of over 70 vol% [2]. Nevertheless, Ni catalysts are prone to deactivation due to carbon deposition. Additionally, the tendency of Ni to sinter at high temperatures (>500 °C) reduces its surface area during the reaction, further promoting carbon deposition. Therefore, achieving high dispersion of Ni particles on the catalyst surface and facilitating carbon conversion are key strategies to inhibit carbon deposition [13]. Metal oxides such as Al₂O₃ are common catalyst supports, but its anti-coking performance is poor. Ammonia temperature-programmed desorption (NH₃-TPD) studies on Al₂O₃, CeO₂-Al₂O₃, and La₂O₃-Al₂O₃ reveal that Al₂O₃ is the most acidic support [17]. Acidic supports can promote dehydration reactions, for example, compounds like ethanol in bio-oil dehydrate at acid sites to form ethylene, and the polymerization of unsaturated hydrocarbons promotes carbon deposition [13,17]. This confirms that the carbon deposition effect is related to the acidity of the support [13,17].

Lanthanide metal oxides are often used to modify Al₂O₃ supports. For instance, CeO₂ enhances the reducibility of the catalyst and improves the dispersion of metal particles due to its ability to store and release oxygen. Similarly, La₂O₃ also improves the dispersion of active metal particles and exhibits excellent anti-sintering properties, resulting in more stable catalytic performance for the modified catalysts [17]. Bizkarra et al. modified Ni/Al₂O₃ catalysts by incorporating CeO₂ and La₂O₃, finding that the addition of CeO₂ more significantly improved catalyst performance and increased hydrogen yield than La₂O₃ [17]. Valle et al. proposed that after incorporating La₂O₃ into the Al₂O₃ support, the enhanced H₂O adsorption capacity led to the gasification of oxygen-containing compounds adsorbed on Ni particles, significantly mitigating catalyst deactivation [18].

Additionally, impregnating the Ni/CeO₂-Al₂O₃ catalyst with the noble metal Rh further enhanced its activity and stability, and the strong interaction between Rh and Ni resulted in a gas composition closer to thermodynamic equilibrium [17]. Santamaria et al. found that the Ni/MgO-Al₂O₃ dual-support catalyst exhibited high activity and stability in steam reforming reactions, attributed to the formation of highly stable MgAl₂O₄ spinel, which influenced the interaction between Ni and Al₂O₃, thereby improving the dispersion of the Ni active phase and the catalyst’s reducibility [19]. In addition, alkali and alkaline earth metal elements (AAEMs) are reported to enhance the adsorption of H₂O on the catalyst and accelerate the water–gas shift reaction [20].

In addition to Ni, Chen et al. investigated the H₂ production performance of a Co-Fe/ZSM-5 bimetallic catalyst in the aqueous phase reforming of bio-oil [21]. They found that, compared to using Co alone as the loaded metal, the presence of the Co-Fe alloy suppressed methanation reactions while simultaneously enhancing the water–gas shift reaction, achieving a high H₂ yield of 81 wt. % (in relation to mass of feedstock) and a carbon conversion rate of 85% [21]. The transition metal Fe has been proven to exhibit high catalytic activity in H₂ production from organic pyrolysis [22]. An et al. confirmed that the presence of Fe in catalysts can significantly enhance H₂ selectivity [23].

All the above catalysts need extensive maintenance during operation, and they are expensive to manufacture. This study focuses on the H₂ production performance of self-derived char, a byproduct derived from municipal solid waste (MSW) and biomass, which serves as a self-sacrificing-type reforming catalyst for volatiles during steam reforming. The MSW-derived char (MSWC) and biochar (BC) catalysts are studied in this work. To improve the activity of the char, CaO is added as it is commonly employed as an in situ CO₂ sorbent to shift the equilibrium toward H₂ production, a method known as sorption-enhanced hydrogen production (SEHP) [24,25]. Fe powder is simply mixed with chars by mechanical grinding of the mixture. Therefore, cheap catalysts consisting of pyrolytic char and mechanically mixed iron/CaO powders are employed, eliminating the complex preparation procedures and maintenance during the process, as part of char can be converted into H₂ simultaneously and iron can be recovered by magnetic separation.

In subsequent applications, the multi-impurity crude syngas remains applicable for SOFC or ICE applications. Mauro et al. [26] demonstrated an integrated system combining syngas utilization with both SOFC and ICE technologies.

2. Results and Discussions

2.1. Catalyst Performance in the Absence of Steam

Since the catalyst used in this study is a carbon-based material, its inevitable consumption under steam atmosphere will influence the catalytic performance. To eliminate the interference of the steam gasification of carbon on the H₂ production, and to provide a benchmark to evaluate the role of the steam supply, pyrolysis and catalytic reforming experiments of MSW and biomass were first conducted in the absence of steam, with the results shown in Figure 1. The data demonstrate that under steam-free conditions, MSW as feedstock yields significantly higher syngas production than biomass, but with no advantage in the H₂/CO ratio.

Compared to the non-Fe-added case, the incorporation of 10 wt. % iron powder in MSWC increased the syngas yield by 13.94% while reducing C₂₊ content in syngas by 59.17 vol%. However, the H₂/CO ratio showed negligible change (only 2.74% increase). This indicates the following: (i) The presence of Fe particulates creates additional metallic sites on the support, enhancing the dissociation of macromolecular organics; (ii) The absence of H₂O molecules in the system limits the supply of -OH groups for H₂ generation, thereby constraining the improvement in the H₂/CO yield ratio.

The enhanced H₂/CO ratio observed in the biomass–BC system may originate from the following: (i) BC has superior initial pore structure, characterized by higher specific surface area and greater microporosity, which facilitates the decomposition of macromolecules. As demonstrated in previous studies [14], aromatic compounds in biomass pyrolysis volatiles undergo significant decomposition on charcoal surfaces; (ii) Abundant oxygen moieties (C=O, -COOH) in biomass induce oxidative pathways (RH → R• + H₂O) [27], with the in situ-generated H₂O participating in steam reforming (C_nH_m + H₂O → CO_x + H₂) at active sites [11].

Notably, in all tested cases with CaO or/and Fe powder addition, no tar was collected in the condenser, demonstrating that all catalysts, by mechanically mixing MSWC and BC with CaO/Fe powder, exhibit strong activity in decomposing tar compounds without the participation of steam.

2.2. H₂ Production Performance and Quality of Syngas in the Presence of Steam

2.2.1. Influence of Feedstock Type

As expected, no tar was collected in the condenser after introducing steam during the reforming step. Figure 2 presents the gas composition of syngas produced from MSW and biomass. It is observed that biomass yields higher concentrations of CO and CO₂ and lower levels of C₂₊ hydrocarbons compared to the MSW feedstock.

Figure 3 displays the gas yield, H₂ yield, and H₂ concentration of the syngas gas from MSW and biomass under different reforming temperatures and steam/fuel (S/F) ratios. Overall, MSW exhibits better gas production performance than biomass. Using MSW as feedstock, the maximum H₂ yield (2022 mL g⁻¹_MSW) and H₂ concentration (71 vol%) were achieved at Tr = 850 °C, S/F = 1.4, while for biomass feedstock, the maximum H₂ yield and concentration occurred at Tr = 900 °C, S/F = 1.2 (1357 mL g⁻¹_biomass, 51 vol%). The reasons are analyzed as follows: MSWC has a higher catalytic activity, while biomass char suffers more significant active site loss due to stronger AAEMs migration: Although both biomass and MSW pyrolysis chars contain various catalytically active AAEMs [28,29], their catalytic effects differ substantially due to distinct chemical forms. Mei et al. conducted comparative studies on volatile reforming using MSW char (MSWC) and biomass char (BC), revealing the superior catalytic activity of MSWC [15]. This difference originates from the fundamentally distinct nature of the dominant AAEMs in each char type: in MSWC, AAEMs primarily exist as acid-soluble inorganic species that actively induce surface active site formation, thereby enhancing catalytic activity [15]; while in BC, AAEMs predominantly occur in water-soluble and organically bound forms that demonstrate weaker capability in breaking π-bonds in aromatic rings. These AAEMs are more prone to migration during gasification, leading to the loss of catalytic active sites, then the modification of carbon structural due to AAEMs’ migration further degrades BC’s activity [15].

1.

MSW contains both biomass and waste plastics, and there exists a synergistic effect during their co-pyrolysis, which is manifested in improved reaction kinetics and reduced energy barriers. Multiple studies have analyzed the synergistic effects of biomass-plastic co-processing, finding that co-processing yields higher H₂ production compared to individual processing of either biomass or plastics alone, while also increasing the calorific value of the produced gas and the CGE of the reaction [30,31,32]. Kiran et al. discovered that secondary reactions between plastic volatiles and biomass volatiles are the primary contributors to the co-pyrolysis synergy; their interaction promotes the transfer of H from plastics and O from biomass, thereby enhancing volatile reforming reactions and improving reaction kinetics [31]. Rahul et al. also observed that co-pyrolysis of plastics and biomass improves kinetics and reduces activation energy due to this synergy [30]. Therefore, MSW containing both plastics and biomass demonstrates superior H₂ production performance compared to biomass alone.

2.

The pyrolysis volatiles from biomass contain significantly more oxygenated compounds than those from MSW, making biomass-derived volatiles more challenging to reform [33]. And the inherent AAEM species in biochar demonstrate weaker activity in breaking π-bonds within aromatic rings [33], which explains why MSW exhibits greater hydrogen production potential than biomass.

2.2.2. Influence of the Operating Parameters

Figure 3 illustrates that the H₂ production performances of MSW and biomass improve with the overall increase in steam quantity, especially at a lower temperature of 850 °C. However, Figure 3 indicates that at 900 °C, the H₂ yield may show a declining trend with increasing steam input: the H₂ concentrations decrease with increasing steam/feedstock ratio at 900 °C for MSW, although the corresponding H₂ yields are increasing. That is because more CO is produced by char gasification at 900 °C via the following reactions:

C + H₂O ⇄ CO + H₂

(2)

C + CO₂ ⇄ 2CO

(3)

The bigger mass loss of MSWC proved the above reactions happened. For the biomass, at the excess of H₂O supply at the higher temperature of 900 °C, the water–gas shift (WGS, ΔH_300K = −41 kJ/mol) reaction is inhibited as it is an exothermic reaction. In addition, excessive H₂O generates an overabundance of ·H radicals, which penetrate the carbon matrix (CM) of BC and promote the aromatization of the BC [34]. This leads to structural ordering of the CM, reducing the catalytic activity of BC and making it more resistant to gasification [15]. In addition, it is also reported that alkali and alkaline earth metals (AAEMs) in char catalysts migrate at elevated temperatures, reducing the number of active catalytic sites on the char surface.

The above phenomena suggest that there exists an optimal combination of reforming temperature and S/F ratio (Steam/Feedstock), which is consistent with the findings of Mondal’s study [35]. For MSW, the S/F ratio of 1.4 at 850 °C is superior for obtaining the highest H₂ concentration and yield; while for biomass, the S/F ratio of 1.2 at 900 °C is the best condition.

Zhang et al. [36] also found that the H₂ concentration reached its maximum when the volumetric ratio of steam to volatiles was 4. Jia et al. [37] stated in their study that excessive steam addition does not necessarily improve performance; the negative effects of excessive steam can be also caused by the energy consumption of the excess steam [37,38]. In addition, Sun et al. [33] observed that biochar exhibits preferential selectivity for removing aliphatic components from tar, maintaining a 100% removal rate for long-chain alkanes throughout the reforming process. However, steam addition at higher temperature significantly increases the content of oxygenated aromatics, thereby increasing the resistance for their cracking and an increased overall reforming difficulty at higher temperatures. Zhang et al. [39] observed that as temperature increases, low-ring-number compounds in volatiles progressively transform into polycyclic compounds. All of these can explain the reduced H₂ yield and lower concentration at 900 °C for S/F = 1.4.

As mentioned above, since the latent heat of vaporization of water has a significant impact on energy utilization efficiency, this paper adopts another index (η₁) as a criterion for evaluating hydrogen production efficiency:

$${\mathsf{\eta }}_1={\displaystyle \frac{{\mathrm{E}}_{\mathrm{H}2}}{{\mathrm{E}}_{\mathrm{H}2\mathrm{O}}}}$$

where E_H2 is the higher heating value of the produced H₂. E_H2O is the enthalpy of the steam under the reaction conditions.

The values of η₁ under various operating parameters are shown in

Table 2. η₁ under different operating parameters.

Feedstock	Operating Parameters	EH2/kJ	EH2O 1/kJ	η1	Egas 2/EH2O
MSW	Tp = 550 °C, Tr = 850 °C, S/F = 0.8	11.15	3.32	3.36	5.341
	Tp = 550 °C, Tr = 850 °C, S/F = 1.2	18.54	4.98	3.72	5.317
	Tp = 550 °C, Tr = 850 °C, S/F = 1.4	25.82	5.81	4.44	5.332
	Tp = 550 °C, Tr = 900 °C, S/F = 0.8	17.68	3.40	5.20	9.242
	Tp = 550 °C, Tr = 900 °C, S/F = 1.2	19.38	5.10	3.80	7.313
	Tp = 550 °C, Tr = 900 °C, S/F = 1.4	20.68	5.95	3.47	6.752
Biomass	Tp = 550 °C, Tr = 850 °C, S/F = 0.8	5.54	3.32	1.67	3.347
	Tp = 550 °C, Tr = 850 °C, S/F = 1.2	9.68	4.98	1.95	2.790
	Tp = 550 °C, Tr = 850 °C, S/F = 1.4	12.91	5.81	2.22	3.224
	Tp = 550 °C, Tr = 900 °C, S/F = 0.8	12.86	3.40	3.78	7.592
	Tp = 550 °C, Tr = 900 °C, S/F = 1.2	17.33	5.10	3.40	5.198
	Tp = 550 °C, Tr = 900 °C, S/F = 1.4	17.09	5.95	2.87	3.389

¹ E_H2O is calculated by mass_H2O * (2260 + 4.18 × (100 − 25) + 2.1 × (Tr − 100)). ² E_gas is the higher heating value of the produced gas.

.

From Table 2, it can be observed that when using Ƞ₁ as the evaluation criterion, the optimal conditions for H₂ production are Tr = 900 °C and S/F = 0.8, whether for MSW or biomass. At 850 °C, η₁ increases with the increasing S/F ratio, but at 900 °C, η₁ decreases with the increasing S/F ratio. On the other hand, the E_gas/E_H2O ratio also decreases at 900 °C.

At 900 °C, the WGS reaction is inhibited, and the iron catalyst can be oxidized by H₂O, so the C + H₂O ⇄ CO + H₂ is not so effective as happened at 850 °C when the S/F ratio increased. So, part of the steam is wasted. To obtain higher energy efficiency and H₂ yield, Tr = 900 °C, S/F = 0.8 or Tr = 850 °C, S/F = 1.4 can be recommended.

2.3. Catalyst Changes and Recovery

After the reaction, approximately more than 50 wt. % of the char is consumed, which indicates that the gasification of char is one of the primary reactions during reforming and it also acts as a route to recover the energy of the char. Figure 4 shows the comparisons of the XRD patterns of the fresh and the spent catalyst 2 and 3. It shows that MSWC-based catalyst 2 contains richer AAEMs, which enhances the adsorption and dissociation of H₂O and the consumption of char through the water–gas shift reaction. Additionally, in spent catalyst 2, a newly formed Fe-Ni alloy is founded. The formation of the Fe-Ni alloy may be one of the factors contributing to the superior H₂ production performance of MSWC compared to catalyst 3. The effective NiMn₂O₄ is also found; these minerals contribute to the activity of catalyst 2. The XRD patterns of catalyst 3 showed negligible changes before and after the reaction, with a much simpler crystal composition, as shown in Figure 4b.

By comparing the SEM images (Figure 5), it is evident that fresh catalyst 2 displays a well-organized and intact carbon structure, with CaO and Fe hidden within the carbon framework. In contrast, spent catalyst 2 clearly reveals the emergence of Fe and CaO on the surface during to the loss of the carbon matrix, and the elemental distribution has also changed, as evidenced by the mapping data in Table 2. This indicates that after the reaction, the char structure undergoes significant gasification, demonstrating that catalyst 2 is a self-sacrificial catalyst. Figure 5 also shows the migration and agglomeration of Fe particulates during the reaction, especially for catalyst 3; when the carbon carrier was partially gasified, the anchor sites for metallic particles were destroyed, causing the agglomeration of metal particles [15], which ultimately resulted in its limited H₂ yield.

For catalyst 3, it is obviously more porous in the beginning, while after the reaction the porous structure is disrupted, as shown in Figure 5, and Fe particulates become apparent agglomerated on the catalyst surface, harming its activity, which is also why at 900 °C, η₁ index decreased as S/F ratio increased. Therefore, the consumption of biomass char is less pronounced compared to MSWC, and the increase in the Fe proportion on the surface is not as significant as that observed for catalyst 2 (

Table 3. EDS mapping data of catalysts.

Catalyst	Percentage by Mass
	C	Fe	Ca
Fresh Catalyst 2	78.32 wt. % 1	0.45 wt. %	6.47 wt. %
Spent Catalyst 2	23.75 wt. %	7.05 wt. %	34.51 wt. %
Fresh Catalyst 3	85.27 wt. %	0.08 wt. %	0.35 wt. %
Spent Catalyst 3	48.81 wt. %	0.63 wt. %	6.43 wt. %

¹ wt. %: percentage by weight.

), which suggests that the biomass char participated to a lesser extent in gasification reactions.

As iron powder is the most valuable material for preparing the catalyst, its recovery is very necessary. By simple magnetic separation, Fe particulates can be recovered from the spent catalysts.

Table 4. Iron powder recovery from the spent catalysts.

Catalyst	Percentage by Original Mass
	Operating Conditions	Recovered Fe
Spent Catalyst 2	Tr = 850 °C, S/F = 0	250 wt. %
Spent Catalyst 3	Tr = 850 °C, S/F = 0	98 wt. %
Spent Catalyst 2	Tr = 850 °C, S/F = 1.4	219 wt. %
Spent Catalyst 3	Tr = 850 °C, S/F = 1.4	75 wt. %

provides the recovery ratio of the Fe powder. It shows that Fe powder can be largely recovered; for catalyst 2, the recovered magnetic powder is more than twice of the original mass, which is because the new formation of many magnetic minerals such as Fe-Ni alloy, they can be recovered when using magnetic separation. To use catalyst 2, the cost for preparing catalyst will be greatly decreased, so in practice, even if biomass is adopted for H₂ production, catalyst 2 is recommended as this catalyst can reduce the running cost while the biochar can be saved for other purposes.

3. Materials and Methods

3.1. Experimental Materials

3.1.1. Characterization of Feedstock

The MSW samples used in this study were collected from a waste incineration plant in Shanghai (China). The samples were air-dried in a greenhouse for one week, followed by manual component separation. Each MSW component was shredded individually into pieces of 10–20 mm, then the shredded samples were blended extensively before pyrolysis according to their ratio shown in

Table 5. Characterization of MSW and biomass.

Properties	Items	MSW	Biomass
Proximate analysis/(wt. %, ad. 1)	Moisture	3.80	6.40
	Ash	14.00	1.41
	Volatile	72.73	81.70
	Fixed carbon	9.47	10.49
Ultimate analysis/(wt. %, ad. 1)	C	53.94	47.01
	H	8.12	5.60
	N	0.26	0.90
	S	0.10	0.00
	O (diff.) 2	19.78	38.68
Physical composition/(wt. %, ad. 1)	Kitchen waste	12.90
	Paper	22.90
	Fiber	9.20
	Plastics	42.40
	Wood	2.50
	Residue	10.20

¹ ad: air-dry basis. ² O: obtained by mass balance.

. For the biomass experiments, air-dried Prunus cerasifera was used. The compositional analysis of MSW samples, as well as the proximate and ultimate analyses of both MSW and biomass, are presented in Table 5.

3.1.2. Characterization of Fresh Catalysts

The catalysts used in this experiment were based on chars derived from the pyrolysis of MSW and Prunus cerasifera biomass at 550 °C. To enhance catalytic performance while maintaining cost-effectiveness and easy availability, 10 wt. % iron powder (400 mesh) and 10 wt. % CaO powder (200 mesh) were mechanically mixed with the chars. The characterization of MSWC and BC is presented in

Table 6. Physicochemical properties of MSWC and BC samples from N₂ physisorption analysis.

Char	SBET 1 (m2 g−1)	Sp, micro 2 (m2 g−1)	Vp, micro 3 (cm3 g−1)	Vp 4 (cm3 g−1)	AD 5 (nm)
MSWC	13.85	7.22	0.0030	0.0324	9.36
BC	166.66	154.76	0.0636	0.0860	2.06

¹ S_BET: the BET surface area; ² S_{p, micro}: micropore area; ³ V_{p, micro}: micropore volume; ⁴ V_p: the total pore volume. ⁵ AD: the average pore diameter measured by N₂ absorption/desorption.

. Figure 6 indicates that MSWC contains more macropores or mesopores, and its adsorption isotherm exhibits an H3-type hysteresis loop, suggesting an irregular pore structure. In contrast, BC primarily consists of micropores, with its adsorption isotherm displaying an H2a-type hysteresis loop, which is commonly observed in certain ordered three-dimensional mesoporous materials. The XRF results in

Table 7. Major elemental composition (wt. %) of ash samples analyzed by XRF ¹.

Catalyst	Na2O	MgO	Al2O3	SiO2	K2O	CaO	TiO2	Fe2O3	NiO	ZnO
MSWC	0.962	3.012	3.727	18.299	0.967	18.224	2.715	5.504	0.150	0.025
BC	0.086	0.758	0.233	0.997	1.813	6.160	0.049	0.528	0.000	0.029

¹ Unnormalized data.

indicate that both MSWC and BC contain abundant AAEMs and possess porous structures, making them suitable to be carriers for catalysts. The Mn element was not detected by XRF due to its limited detection sensitivity for trace-level concentrations (<100 ppm), whereas the presence of NiMn₂O₄ was clearly identified by XRD owing to its superior resolution and precise phase identification capability.

3.2. Experimental Setup

A schematic diagram of the two-stage fixed-bed pyrolysis-steam reforming reactor is shown in Figure 7; the inner diameter of the reactor used in the experiment was 40 mm. N₂ was used as a purge and carrier gas during the process at a flow rate of 20 mL min⁻¹. First, the lower-stage heater is activated and heats the catalysts placed on the porous fixed-bed in the lower stage to 850 °C/900 °C at a heating rate of 20 °C min⁻¹; the catalyst bed height was 5 mm. As the S/F ratio increases from 0.8 to 1.4, GHSV changes from 2472 h⁻¹ to 2490 h⁻¹ for 850 °C, and from 2583 h⁻¹ to 2602 h⁻¹ for 900 °C. After the temperature and N₂ flow rate stabilize, the upper-stage feedstock is heated to 550 °C at a heating rate of 26.5 °C min⁻¹. When the upper stage reaches 210 °C, the injection pump is initiated with a water flow rate of 0.02–0.035 mL min⁻¹ (corresponding to different S/F ratios) to supply steam and simultaneously begin gas collection, which continues for 40 min. The condensable components are collected in the condensation system. All experiments were repeated three times and the carbon balance (Table S1) validates the results.

3.3. Characterization and Measurement Methods

The non-condensing gas products were analyzed by using gas chromatography (7820 A, Agilent, Santa Clara, CA, USA).

The absorption/desorption isotherms (N₂, −196 °C) were determined using a gas absorption analyzer (ASAP 2460, Micromeritics, Norcross, GA, USA) to characterize the pore structure of the chars. Brunauer–Emmett–Teller (BET) and Barrett–Joyner–Halenda (BJH) methods were used to determine the specific surface area, S_BET, and the pore size distribution of the chars, respectively [15].

Elemental analysis (EA) was performed using an elemental analyzer (Thermo Fisher Scientific Flash 2000, Waltham, MA, USA). Approximately 1–2 mg of sample was combusted at 950–1200 °C under oxygen atmosphere, and the resulting gases were separated by gas chromatography (GC) and quantified via thermal conductivity detection (TCD). Sulfanilamide was used as the calibration standard.

The X-ray diffraction (XRD) patterns of chars were recorded by Rigaku Smartlab, Japan, with scanning speed of 2°/min within the range of 10–80° (2θ). The X-ray source was equipped with a copper anode target (Cu Kα, λ = 1.5418 Å).

Microstructural characterization was performed using a field-emission scanning electron microscope (FE-SEM) (ZEISS GeminiSEM 360, Jena, Germany) equipped with an energy-dispersive X-ray spectroscopy (EDS). Samples were sputter-coated with Au to enhance conductivity, and images were required at accelerating voltages of 1–15 kV with probe currents of 50 pA–20 nA, using secondary electron (SE). Elemental distribution was analyzed at 15 kV/5 nA, collecting spectra for 5–10 min per map.

X-ray fluorescence (XRF) analysis was performed using a wavelength-dispersive XRF spectrometer (Panalytical Axios, Almelo, The Netherlands).

4. Conclusions

In this paper, a catalyst is prepared by mechanically mixing pyrolytic chars (MSWC and BC) with 10 wt. % of iron powder and 10 wt. % of CaO powder, eliminating the need for complex preparation procedures and representing a cost-effective and highly effective catalytic system. Moreover, after the experiment, at least 76 wt. % of the iron powder could still be recovered from the spent catalyst, which further enhances the economic viability of the catalyst.

In the presence of steam, the maximum H₂ concentrations in the product gas were 70.89 vol% (reforming temperature 850 °C, S/M = 1.4) from MSW and 53.32 vol% from (reforming temperature 850 °C, S/M = 1.2) biomass. MSW demonstrated superior H₂ production performance compared to biomass, attributable to the following factors: (i) MSW-derived char exhibited higher catalytic activity, whereas biomass char suffered more significant depletion of active sites due to enhanced migration of AAEMs; (ii) MSW contains waste plastic components with higher calorific values, and their co-pyrolysis created synergistic effects that improved reaction kinetics for H₂ production; and the pyrolytic volatiles from MSW were predominantly aliphatic hydrocarbons, while biomass pyrolysis yielded a substantial amount of aromatic compounds. The former were more readily cracked to produce H₂, and the water-soluble AAEMs in biomass char exhibited inherently weaker capability in breaking π-bonds in aromatic rings.

There exist optimal reforming temperatures and S/F ratios to achieve the highest H₂ yields and concentrations in the syngas; they are 850 °C and 1.4 for MSW and 900 and 1.2 for biomass. Both MSWC and BC are largely gasified during reforming but MSWC loses more carbon. The iron powder can be recovered from the spent catalyst by simple magnetic separation, and catalyst 2, based on MSWC, is recommended for this process.

Overall, this study developed a highly effective catalyst for producing hydrogen-rich syngas, which holds significant potential as a feedstock for methanol, sustainable aviation fuel (SAF), and other high-value fuels. However, the current catalyst cannot completely eliminate methane and other undesirable hydrocarbons. Also, the used catalyst is just recycled but its lifespan was not checked and should be a focus of work in the future.