Introduction
The rapid depletion of fossil fuels and associated environmental issues such as global warming and climate change are becoming global concerns [1]. Carbon dioxide (CO2) is deemed as the principal greenhouse gas (GHG) [2,3]. It constitutes approximately 82% of GHG that contribute to global warming [4]. Based on the report from the United States Environmental Protection Agency, fossil fuel combustion is the largest source of emissions accounting for 80.9% of the total CO2 emissions in 2014 [5]. It is estimated that the increase in global temperatures by 1.9°C and sea level expansions of 3.8 m [6,7] may result in the extinction of 15 to 40 threatened species worldwide [8]. However, the global energy demand is continuously increasing at an alarming rate year after year [9]. In 2017, fossil fuels accounted for 81% of the global electricity generation compared to other sources of energy [10]. Therefore, the global search for alternative renewable energy sources as a replacement for conventional fossil fuels has become a necessity.
Among the alternatives, hydrogen (H2) is considered a practical approach to generate electricity in the 21st century sustainably. In addition, H2 has the highest energy density (122 MJ/kg) among existing fuels. Its energy yield is approximately 2.75 times higher than that of most hydrocarbons (HCs) [11]. However, H2 does not occur naturally on the earth but commonly exists as part of other substances in nature such as water, alcohol, natural gas, biomass, coal, and hydrocarbon [12]. Consequently, it can only be obtained from H2-containing resources through chemical reaction processes. The diversity of sources makes H2 a promising energy carrier for the future [12,13]. By employing H2 gas, the crises of supply disruption and the impact of GHG emissions associated with conventional fossil fuel-based energy systems can be avoided, as depicted as Fig. 1. The reason for this is that H2 utilization generates only water vapor as a by-product with zero GHG emissions during fuel cell application [14,15]. For these reasons, great efforts should be committed to exploiting the production of H2.
In recent years, numerous technologies including thermochemical conversion [11,16–18], electrolysis [19,20], and photolysis [21,22] are under development for H2 production. Natural gas steam reforming has been used in the industry over the years. The polymer membrane electrolyte (PEM) electrolyzer technology has also been developed for commercial applications. However, electrolysis consumes the most energy among these technologies [23]. Besides, the low efficiency of photolysis-based H2 production rate currently makes it a commercially unfeasible technology. Among these possible options, H2 production from biomass gasification is considered an economical and promising technology due to its carbon neutrality, environmentally friendly, sustainable, and renewable characteristics [24–26]. Therefore, biomass thermochemical conversion can be practised in the near term and is deemed as a potential technology in the long-term [27]. However, significant tar formation along with raw syngas is a serious impediment to the development and deployment of biomass gasification [28]. Hence, it is necessary to solve the tar problem of gasification to make the method more attractive for commercialization.
Typically, the tar content in syngas produced from biomass gasification ranges from 0.5 to 100 g/Nm3 [29,30]. However, the tolerance limit of tar in syngas for various applications is 1, 5, and 100 mg/Nm3 in fuel cells, gas turbines, and internal combustion engines, respectively [30]. Tar condenses at low temperatures, subsequently resulting in syngas end-used or process-related problems which typically include blockages and corrosion in downstream filters, fuel lines, engine nozzles, and turbines [31,32] as well as bad odour issues around the gasification plant. Furthermore, the formation of tar represents a decrease in conversion efficiency since biomass is converted to tar instead of syngas.
To date, several strategies aimed at raw syngas purification, including physical separation such as wet scrubbing [33], filtration [34], and electrostatic precipitation [35] have been attempted. Although physical separation considerably removes tar from raw syngas, it has great potential to create secondary pollution. Since tar is a complex mixture of HCs, it is more practical to convert tar into valuable H2 gas through thermochemical processes such as catalytic reforming, thermal, and catalytic cracking [36]. Hence, the physical removal and further reduction/oxidation of tar are essential to improve the H2 production with minimal wastes.
Steam reforming is a promising technique that provides a conversion mechanism for liquid HCs [12]. It offers a higher concentration of H2 in the reformate which is about 70% to 80% (vol.) on a dry basis compared to other reforming technologies (40%–50% (vol.)) [37]. In addition, it produces about 100000 Nm3/h of H2 gas on an industrial scale [38]. The resulting cost of H2 by steam reforming is $3.38/kg H2 or the equivalent of $1.55/gal for gasoline [39]. Besides, dry reforming and autothermal reforming of tar are also currently investigated by some researchers but the study of these technologies is still in an early stage.
A catalyst is any chemical substance that lowers the activation energy to accelerate the chemical reaction rate of steam reforming without being consumed in the process [40]. In recent years, several supported metal catalysts have been developed and utilized for laboratory-scale steam reforming of tar. Typically, the non-noble transition metals (Ni [41], Fe [42], and Co [43]) and noble metals (Pt [44], Ru [45], and Pd [46]) are adopted as the active metal in steam reforming catalysts. Besides, metal oxide [24], rare earth oxide [47], olivine [42], calcined rocks [48], and clay minerals [49] are adopted as support in steam reforming catalysts.
So far, numerous supported catalysts have been extensively developed and investigated for tar steam reforming. However, a relevant review in this field is currently lacking in the literature. Therefore, it is worthwhile to critically review the current research, challenges, and prospects of supported catalysts for the steam reforming of tar generated from biomass gasification reported over the past five years. The objective of the present paper is also to provide a state-of-the-art overview of catalytic reforming studies of gasified biomass tar for H2 production.
Gasified biomass tar
Biomass is a sustainable and renewable organic source that consists of agricultural, forestry, municipal solid, and animal waste residues [50]. The utilization of biomass as feedstock for gasification is environmentally benign compared to non-renewable resources. Therefore, biomass gasification has become a popular technology for H2 production [51]. However, the formation of tar along with syngas is a major drawback of biomass gasification [28].
Figure 2 presents the biomass gasification route and the proposed technique for improving syngas production by steam reforming of tar discharged from the gasifier. The biomass gasification process is controlled in an atmosphere with the presence of a gasifying agent (air, steam or a combination of these) and high temperatures ranging from 600°C to 1000°C [52]. Tar is separated from the raw syngas using a wet scrubber, which subsequently undergoes a steam reforming reaction to produce more syngas. Next, the syngas is purified by the CO shift reactor and pressure swing adsorption to obtain pure H2 gas as a fuel for electricity generation and H2 vehicle application. However, a portion of the syngas is sent to the industry for chemical and liquid fuel manufacturing.
Tar consists of various condensable HCs ranging from monocyclic to polycyclic aromatic HCs along with primary oxygenated to tertiary deoxygenated HCs [53]. Moreover, nitrogen-polycyclic aromatic HCs are also generated during biomass gasification such as isoquinoline, pyridine, and quinoline [53,54]. The formation of tar is dependent on several gasification parameters including the selected feedstock [55], gasifier [56,57], reaction time [58,59], temperature [58,59], and gasifying agent [60].
Yu et al. [55] examined the formation of tar during biomass gasification. The findings show that lignin exhibits a higher yield, which causes the formation of more complex tar components compared to cellulose and hemicellulose. Furthermore, an updraft gasifier produces a larger quantity of tar (12000 mg/Nm3) compared to a downdraft gasifier (100–150 mg/Nm3) [56,57]. According to Berrueco et al. [58] and Erkiaga et al. [59], the total tar yield decreases with the increase in reaction temperature and time. In terms of gasifying agent, tar formation can be significantly reduced by introducing steam and O2 during gasification [60]. The reason for this is that O2 accelerates the destruction of primary tar and conversion of phenolic to aromatic compounds [61]. Furthermore, the presence of steam prevents the polymerization reaction during gasification [60].
Figure 3 illustrates the composition of tar derived from gasification of various biomass feedstocks in the literature [62–65]. To facilitate sampling and analysis, tar can be categorized into five (5) classes according to its molecular weight (see Table 1). Due to the condensation characteristics of classes 1, 4, and 5 tars, fouling and clogging remains a problem in the downstream process. However, classes 2 and 3, tars are typically responsible for catalytic deactivation through compete on for active sites on the catalysts [53].
Tab.1 Tar classification based on molecular weight [53] |
| Class | Description | Properties | Example |
|---|---|---|---|
| 1 | GC-undetectable | Heaviest tars, condensable at high temperature | – |
| 2 | Heterocyclic aromatic HC | Highly water-soluble | Pyridine, phenol, cresols, quinoline, isoquinoline and dibenzophenol |
| 3 | Light aromatic HC (1 ring) | Do not pose a condensation and solubility related problem | Toluene, ethylbenzene, xylenes, styrene |
| 4 | Light polycyclic aromatic HC (2–3 rings) | Condensable at low temperature even with low concentration | Indene, naphthalene, fluorine, phenanthrene and anthracene |
| 5 | Heavy polycyclic aromatic HC (4–7 rings) | Condensable at high temperatures even with low concentration | Fluoranthene, pyrene, chrysene, perylene and coronene |
Steam reforming of gasified biomass tar
Among the reforming methods, steam reforming has the highest efficiency. Hence, it is a deep-rooted conversion technology that produces H2 rich gas using HCs as feedstock at high temperatures from 700°C to 900°C and pressures from 0.3 to 2.5 MPa in the presence of metal-based catalyst [66]. The main target of steam reforming is to obtain a high H2 yield with the minimum CO content [67,68]. For example, using toluene as a feedstock, the H2/CO ratio produced by steam, dry, and autothermal reforming is 1.57, 0.29, and 0.71, respectively. Therefore, steam reforming is the most preferred process for integration into tar removal techniques and conversion tar into valuable H2 rich gas.
During steam reforming, numerous parallel reactions (see Table 2 and Fig. 4) occur simultaneously. As a result, the competing processes result in the formation and distribution of different products, namely, H2, CO, CO2, and CH4. The two main reversible reactions involved in steam reforming are the strongly endothermic reforming reactions (Eq. (1) for HCs and Eq. (2) for oxygenated HCs) [24,69], followed by a moderately exothermic water-gas shift reaction (Eq. (3)). In addition, the steam reforming of tar is typically accompanied by coke formation (Eq. (7)) on the catalyst surface, which can result in deactivation [70,71]. However, the catalyst deactivation can be prevented by carefully controlling the ratio of H2O and CO2 through Eqs. (8) and (9) where C is the deposited carbonaceous species [70–72].
Tab.2 Possible reactions involved in gasified biomass tar steam reforming process |
| Reaction | /(kJ·mol−1) | Refs. |
|---|---|---|
| Steam reforming of hydrocarbon Steam reforming of oxygenated hydrocarbon | >0 >0 | [24,69] [24,69] |
| Water-gas shift | −41 | [72,73] |
| Dry reforming | >0 | [48,74] |
| Hydrodealkylation | <0 | [74,75] |
| Methane steam reforming | 206.9 | [74,75] |
| Carbon formation | <0 | [70,71] |
| Boudouard reaction | 172 | [70,72] |
| Carbon gasification | 131 | [70,71] |
Catalysts development of gasified biomass tar steam reforming
The current research on supported catalysts designed for the steam reforming of gasified biomass tar declared over the past five years is reviewed and discussed in this section. The catalyst types discussed in this section include Ni-based, other metal based-, promoted, alloy, supported, perovskite and hydrotalcite catalysts. The steam reforming processes of gasified biomass tar are summarized in Table 3.
Tab.3 Summary of catalytic gasified biomass tar steam reforming processes |
| Active metals | Supports | Preparation methods | Tar model compound | Operating conditions | Catalytic performance/% | Remarks | Ref. |
|---|---|---|---|---|---|---|---|
| Ni (10% (wt.)) | Activated char | Impregnation | Toluene, naphthalene | T = 800°C, S/C= 2, GHSV= 8000 h−1 | Tar conv. = 92–100, H2 comp. = 66–67 | Structural damage and surface area deterioration were observed on spent catalyst | [41] |
| Ni (10% (wt.)) | Al2O3, MgO, CaO | – | Toluene, phenol, naphthalene, pyrene | T = 450°C, S/C= 5, GHSV= 1900 h−1 | Tar conv. = 80–100, H2 yield= 2–13 | Ni/Al2O3 and Ni/CaO had an unstable behavior in H2 yield | [79] |
| Ni (20% (wt.)) | Al2O3 | Impregnation | Phenol, toluene, Furfural, methyl naphthalene, ndene, anisole | T = 750°C, S/C= 3 | C conv. = 90, H2 yield= 14.3 | H2 yield is much lower than the potential H2 yield (63%) calculated from stoichiometry; O2 contributed the largest constitute of reformate followed by CO and H2 | [24] |
| Cu (1% (wt.)) | Calcined scallop shell | Incipient wetness impregnation | Tar derived from cedar wood gasification | T = 700°C, Catalyst= 2 g, Water= 0.09 mL/min | H2 yield= 60 mmol/gcarbon | Existence of Ca(OH)2 on catalyst improved the basicity of catalyst and anti-coking ability | [84] |
| Ru (1% (wt.)) | 12SrO-7Al2O3 | Physical mixing, impregnation | Toluene | T = 600°C, S/C= 2, W/F = 7 g h/mol | Tar conv. = 80 | Ru(PPh3)3Cl2 recognized as a better Ru precursor for a high catalytic activity as compared with RuCl3nH2O | [87] |
| Pt (1.5% (wt.)) | Al2O3, CeO2/Al2O3 | Incipient wetness impregnation | Toluene | T = 700°C, Steam/toluene= 40 | Tar conv. = 80–95, H2 comp. = 65–68, H2/CO= 6.5–8.5 | Doping of CeO2 decreased the selectivity to CO but increased the selectivity to CO2; Pt/CeO2/Al2O3 produced a higher H2/CO | [44] |
| Ba/Ni, Sr/Ni, Ca/Ni (2.28+ 5% (wt.)) | LaAlO3 | Pechini method /Impregnation | Toluene | T = 600°C, S/C= 2, WHSV= 27.1 h−1 | C conv. = 28–44, H2 yield= 26–41 | Toluene conversion and H2 yields increased drastically by the addition of alkaline-earth metals | [97] |
| Ni/Ru-Mn (16+ 0.6+ 2.6% (wt.)) | α-Al2O3 | Incipient wetness impregnation | Toluene | T = 600°C, S/C= 25, GHSV= 10000 h−1 | C conv. = 100, H2 comp. = 68.1 | Formation of filamentous carbon which leads to reactor clogging and pressure drop was observed on spent catalyst surface | [45] |
| Fe (10 wt.%) Fe-Ni (5+ 5 wt.%) | Olivine | Thermal fusion | Toluene | T = 850°C, S/C= 0.93, WHSV= 0.88 h−1 | Tar conv. = 98, H2 yield= 88-98 | Tendency of carbon formation of Fe/olivine was slightly higher than Fe-Ni/olivine | [42] |
| Pt/Ni (0.85+ 5 wt.%) | La0.7Sr0.3AlO3−δ | Pechini method/Impregnation | Toluene | T = 600°C, S/C= 8.9, GHSV= 12000 h−1 | C conv. = 59.1, H2 yield= 52.7 | Pt/Ni was the best impregnation order lead to a high H2 yield and a high tolerance to coking | [111] |
| Ni, Co (10 wt.%) Ni/Co (5+ 5 wt.%) | ZrO2 | Impregnation | Phenol | T = 600°C, S/C= 9, WHSV= 115.56 h−1 | Tar conv. = 33–53, H2 yield= 24–51 | Bimetallic catalyst exhibited better catalytic activity than monometallic catalysts | [117] |
| Ni (20 wt.%) | Lignite char, Al2O3 | Ion exchange | Toluene | T = 650°C, S/C= 2 | H2 yield= 512–1125 mmol/g-Ni | Lignite char is readily gasified and not suitable service as catalyst support for steam reforming | [120] |
| Ni (10 wt.%) | Activated carbon, olivine, Al2O3 | Incipient wetness impregnation | Toluene | T = 600°C, S/C= 2, LHSV= 0.87 h−1 | C conv. = 18–100 | The large surface area and microporous structure of activated carbon support contributed to a fine Ni particle distribution and consequently lead to a high catalytic activity | [122] |
| LaNi0.5Mn0.5O3 | Pechini method | Toluene | T = 700°C, S/C= 3, HSV= 20000 h−1 | Tar conv. = 100, H2 comp. = 42 | Catalyst required high reduction temperature (up to 1000°C) | [133] | |
| V (3 wt.%) | Mg/Al | Co-precipitation /Impregnation | Toluene | T = 500°C, S/C= 2, WHSV= 16.6 h−1 | C conv. = 77.5, H2 comp. = 57 | A higher V content presented a better activity in toluene conversion while a lower V content produced a higher H2 composition of reformate | [138] |
Notes: GHSV— gas hourly space velocity; WHSV—weight hourly space velocity; LHSV— liquid hourly space velocity; W/F—time factor (catalyst weight/toluene molar flow rate); C conv.—carbon conversion; Tar conv.—tar conversion; H2 comp.— H2 composition |
Ni-based catalysts
From an industrial standpoint, the application of Ni-based catalyst is more practical because of its economic feasibility and marked performance in C-H, C-C, and O-H bonds cleavage [24,76]. However, it is prone to deactivation by sintering [77,78] and coke formation [78–80] on active sites. The performance of Ni/red cedar activated char derived from various precursors in steam reforming of toluene and naphthalene/toluene mixtures was evaluated by Qian and Kumar [41]. It was found that nickel nitrate derived catalyst was more active in toluene conversion than those derived from nickel acetate. In this case, both Ni precursors showed a similar textural property of the catalyst in terms of specific surface area, pore volume, and pore size. This indicates that the better performance of nickel nitrate derived catalyst is mainly attributed to its smaller metallic size and higher Ni dispersion.
Moreover, the nickel nitrate precursor is more readily reduced to the metallic state, which is important in bond cleavage during steam reforming [81]. In another study, Park et al. [82] also reported that the catalysts derived from nickel nitrate were more stable and active than the catalysts derived from nickel chloride or nickel sulphide. Although they compared the Ni precursor for catalyst preparation, no explanation was given on how the Ni precursor affected the characteristic of the catalyst, which subsequently influenced the catalytic activity.
Vivanpatarakij et al. [79] conducted a series of tar steam reforming experiments (toluene, phenol, naphthalene and pyrene) at different Ni loadings (10%–20% (wt.)) and Ni-based catalyst supports including Al2O3, MgO, and CaO. Among the support materials used, Al2O3 exhibited a higher reactivity in carbon conversion and H2 production. Based on Ni loading, it was found that stable and near complete conversion occurred for all tar components (except naphthalene) when 20% (wt.) of Ni was loaded on Al2O3. Furthermore, they reported that the tar reforming ability was in the order: naphthalene<pyrene<phenol<toluene.
Kim et al. [83] stated that the formation of spinel NiAl2O4 in a higher Ni loading catalyst promotes the formation of carbon on the catalyst and consequently deactivates the catalyst. Artetxe et al. [24] conducted an experiment on the steam reforming of tar compounds (toluene, phenol, anisole, methylnaphthalene, furfural, and indene) over the Ni/Al2O3 catalyst with a 5%–40% (wt.) of Ni content. They claimed that the optimum Ni loading was 20% (wt.) beyond which the catalytic activity did not increase with further increase in Ni-content in the catalyst. This result can be associated with the formation of larger Ni particles, reduction in the specific surface area, and unreacted carbon on the catalyst surface. From the experiment conducted by Artetxe et al. [24], it can be noticed that oxygenated hydrocarbon have a higher reactivity toward conversion and H2 production compared to an aromatic hydrocarbon, although oxygenates promote the coke deposition on the catalyst.
Other metal-based catalysts
Kaewpanha et al. [84] investigated the steam reforming of tar from cedar wood gasification over Cu/calcined scallop shell catalysts with different Cu loadings (0.5%–5% (wt.)). They observed that the highest H2 yield was achieved at a Cu loading of 1% (wt.), while a further increase in Cu loading deteriorated the catalytic performance. This was attributed to the existence of portlandite (Ca(OH)2) on the 1% (wt.) Cu/calcined scallop shell catalyst, which improved the basicity of catalyst and the efficiency of coke suppression. Besides, the CaO contained in the calcined support was reported as CO2 sorption to shift the water-gas shift (WGS) thermodynamic equilibrium to the H2 product [85,86]. Therefore, a high Cu loading favored the formation of calcium copper oxide that reduced the amount of CaO on support, consequently reducing the H2 yield from WGS reaction.
The effect of Ru precursor and pre-treatment conditions (N2 or H2 atmosphere) on steam reforming of toluene was investigated by Iida et al. [87]. Two types of precursors, Ru(PPh3)3Cl2 by physical mixing and RuCl3 nH2O by impregnation, were employed to prepare the 12SrO-7Al2O3 supported catalysts. It was observed that the catalyst synthesized by physical mixing of Ru(PPh3)3Cl2 and N2 pre-treatment before the steam reforming process exhibited a higher toluene conversion. This finding was explained by the lower melting point of Ru(PPh3)3Cl2 (159°C), which further enhanced dispersion over the surface of catalyst support during calcination. Concerning pre-treatment, an inert atmosphere was favored for the reduction of Ru to avoid the sintering of Ru particles compared to the rapid reduction by H2 stream.
The effect of CeO2 promoter on toluene steam reforming over Pt/Al2O3 was evaluated by Castro et al. [44]. The main products of the experiments were H2, CO and CO2. The results indicated that the Pt/CeO2/Al2O3 catalyst had a lower selectivity toward CO and a higher selectivity toward CO2 compared to the Pt/Al2O3 catalyst. This is due to the oxygen vacancies in CeO2, which promoted the WGS reaction and oxidation of CO [88,89]. Although CeO2 promoted catalyst exhibited an excellent activity and a H2 selectivity, the particles of CeO2 and Pt were unstable, aggregated, and subsequently deactivated over the reaction time [90,91]. Therefore, future studies on the modification of Pt-CeO2 catalyst are required to stabilize the CeO2 and Pt particles for more effective catalytic activities.
Promoted catalysts
Promoters are typically employed to modify the support structure of catalysts. This enhances the surface area available for catalytic reaction, the catalytic activity per unit surface area, and stability against unwanted side reactions [92]. Oh et al. [45] performed steam reforming of toluene over Ru-Mn promoted Ni/Al2O3 catalyst. The Ni/Ru-Mn/Al2O3 catalyst showed a complete carbon conversion at 600°C, with H2 as the highest fraction of reformate. The reason for this is that the addition of Mn promoter favored the dehydration of the reactant and promoted the C-O and C-C bonds rupture, which resulted in increased H2 selectivity [93]. Furthermore, the addition of Ru promoter ensured that the spent catalyst had no significant change in Ni crystallite size by sintering, whereas the amorphous carbon deposition was nearly undetected [45,94]. In addition, the high intrinsic kinetics of carbon gasification on Ru and the low carbon solubility in the bulk of Ru prevent carbon growth during steam reforming [95,96].
Toluene steam reforming was conducted by Higo et al. [97] to study the promoting effect of alkaline-earth metals (Sr, Ba, and Ca) on Ni/LaAlO3 catalyst. The authors reported that the promoting effect is in the order of Ba>Ca>Sr in terms of H2 yield and carbon conversion. Conversely, the coke formation of catalysts exhibited an opposite trend to the catalytic activity. The catalysts with strong basicity had a high catalytic activity and a high resistance against coke deposition during steam reforming [98–100]. In addition, the alkaline-earth metals neutralized the acidity of catalysts and consequently improved their anti-coking ability by facilitating carbon gasification with steam. Apart from the basicity property, Ba (which demonstrated the greatest promoting effect) also improved the reducibility of NiO by adsorbing and providing OH derived from steam to Ni active site for carbon decomposition [101].
Bimetallic or alloy catalysts
As reported, Ni catalysts are prone to coke deposition [79,102] and agglomeration [77,103]. Recently, bimetallic catalyst has been utilized for steam reforming due to its positive effect on metal interactions either through geometric or electronic effects [104]. Therefore, alloying Ni with other metals may improve the coke resistance, stability, and robustness of the catalyst [105–107]. Meng et al. [42] comparatively investigated the activity of Fe-, Fe2O3-, and Fe/Ni-based olivine catalysts for toluene steam reforming. Among the catalysts, Fe-Ni/olivine had a higher H2 yield and resistance to coke formation owing to the synergistic effect of the Fe-Ni alloy particles [108]. Fe has a higher oxygen affinity and provides an oxygen atom to the Ni species to facilitate the decomposition of deposited carbon [109,110]. However, Fe2O3/olivine had a slightly higher toluene conversion due to its main oxidized component (Fe2O3), which was more readily reduced than MgFe2O4.
Mukai et al. [111] reported that Pt/Ni/La0.7Sr0.3AlO3−δ had a high catalytic performance and a low coke deposition for H2 production via toluene steam reforming. They also stated that the catalytic activity of the catalyst without pre-reduction was almost identical to its pre-reduction variant. This indicated that the additive Pt formed the Pt-Ni alloy structure and permitted rapid reduction of NiO by enhancing the spillover of atomic H on NiO surface through rapid dissociation of H2 [112,113]. Hence, the synergy between Pt with high reducibility and Ni with a high reforming activity leads to a better performance of the bimetallic catalyst in steam reforming compared to the monometallic catalyst. Mukai et al. [111] also found that the Pt-Ni alloy structure with a better metal dispersion reduced the risk of coke deposition on the catalyst surface, as corroborated by other researchers [114–116].
Nabgan et al. [117] developed a bimetallic NiCo/ZrO2 catalyst for phenol steam reforming. They observed that the bimetallic catalyst showed a better activity for phenol conversion and H2 yield compared to the monometallic catalyst. Based on their study, the Ni/ZrO2 catalyst was deactivated mainly by moisture adsorption and coking. However, by introducing Co to the Ni/ZrO2 catalyst, the adsorption of moisture was subsequently inhibited. Besides, they also suggested that the bimetallic catalyst neutralized the acidity of catalyst, whereas the monometallic catalyst promoted the acid properties of the catalyst. Moreover, a study on NiCo/Al2O3 by Luo et al. [118] revealed that an increase in Co loading resulted in a high H2 selectivity and a low CH4 production during glycerine steam reforming. The synergy between Ni and Co also improved the anti-metal sintering ability of catalyst due to the formation of Ni-Co alloy, stable solid solution, and strong metal-support interaction [119].
Catalyst support
Cao et al. [120] conducted a steam reforming of toluene over Ni/Al2O3 and Ni/lignite char catalysts at 650°C. They stated that the catalytic performance and deactivation were related to the support. The Ni/Al2O3 had a stable performance for 5 h without particle sintering but with a low amount of carbon deposited on the catalyst. Conversely, Ni/lignite char was rapidly deactivated by Ni particle aggregation. This is attributed to the structural destruction of carbon support by steam gasification. Therefore, Ni/lignite char catalyst is not suitable for steam reforming although it had an excellent anti-coking ability. Although the Ni/Al2O3 has stable performance, in this case, the Ni/Al2O3 catalyst is always reported with rapid deactivation. For example, Park et al. [121] found that Ni/Al2O3 showed a serve catalytic deactivation after 30 h of reaction. Therefore, in this case, 5 h of the catalytic test failed to provide adequate proof for the stability of the catalyst.
Liu et al. [122] evaluated the activity of Ni-based catalyst with different supports for the steam reforming of toluene. It was found that activated carbon provided a large surface area and a microporous structure to generate a catalyst with fine Ni particles and a high Ni dispersion. On the other hand, the low surface area of olivine causes the sintering of Ni during calcination and steam reforming reaction. Although Al2O3 has a larger surface area, it is prone to deactivation by coke deposition compared to activated carbon and olivine. They concluded that the Ni/activated carbon catalyst exhibited the highest carbon conversion and stability. Unlike the findings of Cao et al. [120], the carbon in the lignite char was readily gasified during steam reforming. However, the gasification of the activated carbon support was very low but did not diminish its catalytic activity during steam reforming. The Al2O3 and olivine supported catalysts were, however, not as effective as activated carbon supported catalyst, particularly the olivine supported catalyst.
Ni/Al2O3 and Ni/olivine catalysts initially showed a lower activity, which was ascribed to the incomplete reduction of Ni oxide [122]. Furthermore, the strong interaction of the active metal and support severely hindered reduction. For example, the formation of spinel NiAl2O4 after calcination had a high resistance against reduction and may be reduced at a temperature of above 800°C [123,124]. In this case, the Ni/olivine catalyst was calcined at 1000°C followed by a reduction at 700°C. As reported by other researchers [125–127], a high-temperature calcination (900°C–1400°C) causes the replacement of Mg in the olivine lattice by Ni. The strong Ni-olivine interaction requires high temperatures for full reduction of NiO species (900°C–950°C) [125–127]. Due to the smaller pore volume and surface area, the Ni/olivine catalyst experienced a fast deactivation by pore clogging. However, the acidic support of Ni/Al2O3 catalyst promoted the formation of coke on the acid sites through the oligomerization of the toluene molecule [44].
The various textural, acid-basic, and potential metal-support interaction properties of supports are a crucial element in the preparation of catalysts with a high stability and activity [128,129]. The temperature of calcination and reduction which dominate the transformation of metal precursor to active metallic phase is also an essential factor for catalyst preparation. Thermogravimetric analysis and temperature-programmed reduction techniques are proposed to determine the suitable temperature for calcination and reduction, respectively. Other characterization techniques can also be integrated with these thermal analyses to investigate the physical and chemical changes of catalyst.
Perovskite and hydrotalcite catalysts
Perovskite (ABO3, where A= alkaline earth metal or lanthanide; B= transition metal such as Mn, Ni, Co, and Cu) could either be a catalyst or a support for the active metal. Typically, the A-site metal has a powerful effect on stability while the B-site metal represents the primary active site [130]. Typically, perovskite has an excellent thermal stability, a well-defined structure, and mobile oxygen, which are beneficial to steam reforming reaction [131,132].
The perovskite catalyst LaNi0.5Mn0.5O3 was developed and investigated by Quitete et al. [133] for toluene steam reforming. The catalyst accomplished the complete conversion of toluene at 700°C and an S/C ratio of 3, which is close to industrial-scale values. Despite its low specific surface area (1.9 m2/g), LaNi0.5Mn0.5O3 showed an active and stable behavior in toluene steam reforming for 22 h of reaction. As reported [98,134,135], Ni-based perovskite catalysts are relatively stable and resistant to deactivation by coking.
Hydrotalcite, also known as an aluminum-magnesium layered double hydroxide, is a naturally occurring nanostructured anionic clay. It has high thermal stability and provides a large surface area for uniform dispersion of active metals [136,137]. Compared to alumina supported catalysts, hydrotalcites are more resistant to metallic sintering and coke formation [137].
Mitran et al. [138] studied the effect of vanadium loading (0.9%–3% (wt.)) on the steam reforming of toluene. They observed that V/MgAl catalysts that contain polyvanadate species are more active during steam reforming of toluene. However, the catalysts with isolated species possess a higher selectivity for H2 production. For a higher loading of V in the catalyst, the monometric VOx (i. e., V-O-support) is the predominant species that can be reduced at higher temperature [139,140]. For a lower loading of V in the catalyst, the polymetric VOx (i. e., V-O-V, which has less interaction with support) is the pronounced species that can be reduced at a lower temperature [139,140]. Other studies also claim the addition of V promotes the WGS activity [139,141], but the H2 constitution of reformate does not show a linear correlation with its loading [141] due to the side reaction of steam reforming.
Table 4 shows the comparison of perovskite and hydrotalcite catalysts with conventional supported catalysts (discussed in Section 4.4). It can be concluded that perovskite and hydrotalcite catalysts have a more complex structure and better properties compared to conventional supported catalysts. The unique structure of perovskite improves the dispersion of transition metals and offers a stable interaction within the perovskite lattice [142]. However, calcined hydrotalcite with a small crystal size offers a highly specific surface area for good dispersion of active sites [143]. Besides, the redox property of perovskite with high mobility of lattice oxygen promotes the oxidation of deposited coke [144]. Therefore, perovskite and hydrotalcite catalysts have a better performance in terms of catalytic activity, coke suppression, and thermal stability [144].
Tab.4 Comparison of perovskite and hydrotalcite catalysts with conventional supported catalysts |
| Catalyst | Perovskite | Hydrotalcite | Conventional supported |
|---|---|---|---|
| General formula | ABO3 where A= alkaline earth metal; B= transition metal | Mg6Al2CO3(OH)16·4(H2O) | Metal oxide, oxides mineral, carbonaceous material |
| Examples of Ni based catalyst | LaNi0.5Mn0.5O3, La0.9Ni0.2Mg0.1Al0.8O3 | Ni/MgAl | Ni/Al2O3, Ni/olivine, Ni/lignite char |
| Structure | Crystal structure, nonstoichiometric oxygen | Brucite-like structure, where Mg2+ attached with OH- ions to form octahedral structure | – |
| Synthesis method | Complex, i.e., citrate method, solution combustion techniques | Complex, i.e., urea hydrolysis, sol-gel method, microwave treatment | Easier, i.e., impregnation, precipitation |
| Thermal stability | Higher | Higher | Lower especially carbonaceous material |
| Resistance against coke deposition | Stronger | Stronger | Weaker |
Parametric effect on steam reforming of gasified biomass tar
From previous literature studies, the most important factors for tar steam reforming are temperature, S/C ratio, and space velocity or space time [145,146]. In this section, the effect of these factors on steam reforming efficiency is presented. The favored conditions of gasified biomass tar steam reforming for H2 production are summarized in Tables 5 to 7.
Tab.5 Favored temperature for H2 production in steam reforming of gasified biomass tar |
| Favored temp. /°C | Tar model | Catalyst | Metal loading /% (wt.) | Other operating condition | Catalytic performance/% | Remark | Ref. |
|---|---|---|---|---|---|---|---|
| 700–900 | Toluene, toluene/naphthalene | Ni/MgO/Al2O3 | 10.0 | S/C= 1.5, GHSV= 20000 h−1 | Tar conv. = 89–100, H2 comp. = 22–30 | CO2 and CO are the main products at lower and higher temperatures, respectively; large cyclic HCs have a higher thermal stability | [74] |
| 800 | Toluene | Ru/α-Al2O3, Ni/α-Al2O3 | 2.0 | S/C= 3.57, WHSV= 10000 h−1 | C conv. = 96–98, H2 comp. = 69–76 | Ru is more stable and has a higher activity toward tar conversion than Ni | [121] |
| 800 | Phenol | Ni/g-Al2O3 | 10.0 | S/C= 13, WHSV= 0.444 h−1 | C conv. = 57, H2 comp. = 14 | The low conversion may be due to the use of the unreduced catalyst | [31] |
| 650–750 | Toluene | Ce0.4Ni0.6AlO3, La0.2Ni0.8AlO3 | 5.8 | S/C= 1.5, WHSV= 23068 h−1 | Tar conv. = 100, H2 comp. = 25–30 | The present of CeO2 allows the full conversion of toluene at a lower temperature | [150] |
Notes: WHSV—weight hourly space velocity; C conv.—carbon conversion; Tar conv.—tar conversion; H2 comp.—H2 composition |
Tab.6 Favored S/C ratio for H2 production in the steam reforming of gasified biomass tar |
| Favored S/C ratio | Tar model | Catalyst | Metal loading (wt.%) | Other operating condition | Catalytic performance (%) | Remark | Ref. |
|---|---|---|---|---|---|---|---|
| 3.5–5.0 | Toluene | Ni/olivine, Ni/Ce/olivine, Ni/Ce/Mg/olivine | 3.0 | T = 790°C, GHSV= 782 | C conv. = 59–93, H2 comp. = 60–66 | Ni/Ce/Mg/olivine had a more stable performance at a low S/C ratio | [158] |
| 2.0 | Benzene | NiO/ceramic foam | 3.5 | T = 750°C, WHSV= 5.6 | C conv. = 85.4, H2 comp. = 60 | The H2 selectivity is not affected by S/C ratio in this case | [156] |
| 8.0 | Phenol/ethanol | Ni/Cu/sepiolite clay | 20.0 | T = 650°C, WHSV= 3.2 | C conv. = 75, H2 yield= 73 | The limit of S/C ratio is not achieved since carbon conversion showed an increased trend | [159] |
Notes: GHSV—gas hourly space velocity; WHSV—weight hourly space velocity; C conv.—carbon conversion; H2 comp.—H2 composition |
Tab.7 Favored space velocity or space time for H2 production in steam reforming of gasified biomass tar |
| Favored WHSV/h−1 | Tar model | Catalyst | Metal loading/% (wt.) | Other operating condition | Catalytic performance/% | Remark | Ref. |
|---|---|---|---|---|---|---|---|
| 5000 | Toluene | Ru/α-Al2O3 | 2.0 | T = 700°C, S/C= 1.43 | Tar conv. = 87.4, H2 comp. = 61.5 | Carbon conversion showed a decreased trend with space velocity, indicating that the adsorption-limited was not achieved | [121] |
| 10000 | Toluene | Ni/coal | 10.6 | T = 400°C, S/C= 15 | H2 yield= 62 | H2 yield was stabilized above 40000 h−1, implying that the adsorption-limited was achieved | [163] |
| 0.1–0.4 | Toluene, benzene, phenol | Ni/Mg/Al, Ni-Fe/Mg/Al | 12.0 | T = 600°C, S/C= 1.67 | Tar conv. = 100%, H2/CO= 4.4–5.6 | Ni-Fe alloy has a better performance than Ni-based catalyst although Ni based catalyst showed a higher H2/CO ratio of reformate | [109] |
Notes: Tar conv.—Tar conversion; H2 comp.—H2 composition |
Temperature
The reaction temperature has a significant impact on steam reforming of gasified biomass tar and the catalyst used. At a higher temperature, the endothermic steam reforming reaction is enhanced, thereby resulting in thermodynamic equilibrium displacement of the exothermic WGS reaction. Consequently, maximum conversion occurs, resulting in higher yields of H2 [49,147,148].
Josuinkas et al. [74] examined the influence of temperature on the steam reforming of benzene, toluene, and 10%(wt.) naphthalene/toluene over Ni/MgO/Al2O3 catalyst. It was found that the most favorable temperature for complete conversion of benzene and toluene was 700°C. However, naphthalene was more resistant to conversion even at high temperatures. In the steam reforming of naphthalene/toluene, toluene was inhibited by naphthalene and the favorable temperature for complete conversion increased to 900°C. This finding is in good agreement with Qian et al. [41] and Jess [149]. Since naphthalene and heavier HCs (e.g., 2–4 rings) typically constitute around 33%(wt.) in gasified biomass tar [62], the less reactivity of heavier HCs reforming should be taken into account. Although toluene could completely be converted at 900°C when mixed with naphthalene, the temperature is too high which limits the application in the industry in terms of cost and safety. Therefore, more broad research is required to determine more effective catalyst and suitable operating conditions for steam reforming of tar.
Park et al. [121] examined the steam reforming of toluene over Ru- and Ni-based catalysts supported on α-Al2O3 at different temperatures. They pointed out that the H2 selectivity and toluene conversion of both catalysts increased dramatically with elevating reaction temperature. Artetxe et al. [31] reported similar observations. Similarly, Quitete and Souza [150] investigated the steam reforming of toluene over perovskite catalysts from 400°C to 800°C. They observed that Ce0.4Ni0.6AlO3 had a better catalytic performance and achieved complete toluene conversion at 650°C whereas La0.2Ni0.8AlO3 exhibited complete conversion at 750°C. This can be attributed to the high oxygen-release capacity of CeO2, which strongly inhibits coke deposited on the catalyst, consequently reducing the risk of catalyst deactivation (Fig. 5) [151,152].
Steam to carbon ratio
The S/C ratio is one of the crucial factors in steam reforming of gasified biomass tar due to its impact on coke formation, feedstock conversion, and H2 yield [153,154]. The increase in the S/C ratio reduces coke formation by facilitating the gasification of deposited carbon, as verified by Tao et al. [155]. The authors performed an elemental analysis, which indicated that the carbon content of the spent catalyst decreased at higher S/C ratios. However, excess steam negatively influenced the tar conversion since water molecules competed with tar at adsorptive active sites. This was verified by Gao et al. [156] who conducted benzene steam reforming over Ni/ceramic foam catalyst with a distinct S/C molar ratio (0–3). Furthermore, the high S/C ratio is not feasible for industrial application because of the associated cost of high power consumption for steam generation and steam separation from the reformate. Excess steam also absorbs the heat within the reactor, which leads to a fluctuation in the reaction temperature [157]. On the other hand, insufficient steam supplement causes the steam reforming and WGS reactions cannot achieve their state of completion, consequently resulting in the low conversion and H2 production [157]. Therefore, the S/C molar ratios used in the reaction should be higher than the stoichiometric value of the reaction
Zhang et al. [158] studied the effect of S/C ratio on toluene steam reforming over Ni-based catalysts and concluded that H2 composition increased as S/C ratio increased. However, the result was not consistent with the findings of Gao et al. [156], which showed a stable trend of H2 composition at a higher S/C ratio. Similarly, Zhang et al. [158] also observed that the highest carbon conversion was obtained at an S/C ratio of 5 with Ni and Ni/Ce based olivine catalysts while at an S/C ratio of 3.5 with Ni/Ce/Mg/olivine catalyst. The addition of Mg to Ni/Ce/olivine enhanced the performance and coke resistance at the S/C ratio of 3.5. This is due to its basicity and the presence of well-stabilized NiO-MgO solid solution. Liang and coworkers [159] investigated the effect of different S/C ratios on steam reforming of phenol-ethanol over bimetallic Ni-Cu/sepiolite catalyst. The results demonstrated that H2 yield and carbon conversion increased with rising S/C ratios. This is because sufficient steam drives the tar reforming and a high water partial pressure favors the equilibrium of WGS reaction shift toward H2 production [121,159].
Space velocity and space-time
Catalytic steam reforming is influenced by the space velocity and space-time, which reflects the contact time of tar on the active sites of the catalyst [160,161]. The variables are significant to avoid catalyst from wastage during industrial applications. When the adsorption sites of a catalyst are limited, the chemical reaction rate is less sensitive to space velocity and space-time [31]. With regard to space-time, tar conversion increases as space-time are increased, which indicates that the active sites of the catalyst increases or the contact time of the tar-active sites increases [162].
Park et al. [121] studied the steam reforming of toluene over Ru/α-Al2O3 with a space velocity ranging from 5000 to 30000 h−1. They observed that carbon conversion and H2 production were inversely proportional to the space velocity. In another study, the variation of H2 yield from toluene steam reforming over Ni/coal with distinct space velocity range was analyzed by Kim et al. [163]. The H2 yield decreased with increasing space velocities. Koike et al. [109] analyzed the space-time effect on the catalytic performance of tar steam reforming over Ni and Ni-Fe alloy catalyst supported on the Mg-Al hydrotalcite-like material. The results indicated that tar conversion increased with increasing space-time. However, the H2/CO ratio decreased with increasing space-time. At a higher tar conversion, the residual of H2O decreased, thereby limiting the contribution of WGS reaction and H2 selectivity.
Challenges of steam reforming of gasified biomass tar
Although great efforts have been devoted to steam reforming catalyst research, several challenges still exist, including catalyst deactivation, material selection for reactor fabrication, engineering economics and operational cost for high temperature (700°C to 900°C) reactions. Besides the operating costs, the high temperature also affects issues such as catalyst deactivation [164] and sintering [31,77].
Catalyst deactivation
Typically, Ni-based catalysts are deactivated after a long reaction time, which strongly limits their industrial application [165]. However, the most common deactivation problems include ① coke formation which blocks the active site and encapsulates the metal particles; ② active metal sintering which decreases the exposed surface area of the active site [166,167].
Coke formation
At lower temperatures (<400°C), the reverse of Boudouard reaction (Eq. (8)) and reverse of carbon gasification (Eq. (9)) facilitate the formation of coke on catalyst surface [164]. During steam reforming, coke can deposit on catalyst either as amorphous (Fig. 6(b)) or filamentous carbon (Fig. 6(c)). Note that the former leads to the deactivation caused by the hindrance of the active site, whereas the latter does not significantly contribute to deactivation but causes reactor clogging and pressure depression [74,121,168].
Hu et al. [102] observed that both amorphous (12.76 mg/gcatalyst) and filamentous coke (16.84 mg/gcatalyst) existed on spent Ni/Al2O3 catalyst. These results were verified by temperature-programmed oxidation (TPO) analysis, as presented in Fig. 7. Besides, the addition of Fe contributed to coke suppression by increasing the coverage of oxygen species on the catalyst surface. A similar phenomenon has also been observed on the Ni-Fe/SBA-15 catalyst by Kathiraser et al. [169]. In the case of Ru/SrCO3-Al2O3 catalyst, Lida et al. [170] proposed that the amount of coke deposited was remarkably higher with decreasing SrCO3/Al2O3 loading ratio. This is because of the decrease in coke suppression agent, SrCO3 which possesses a highly reactive hydroxyl functional group. As reported by Karnjanakom et al. [103], the coke formation rate of the Ni/MCM-41 catalyst was sequentially increased with the Ni content (5%–40% (wt.)). The authors showed that the ethylene glycol assisted method enhances the resistance of Ni/MCM-41 to coke deposition compared to the co-impregnation method.
Active metal sintering
The presence of excess steam and high temperature accelerates the sintering rate of the active metal in catalysts [31,77]. For instance, Ni-based catalysts are susceptible to sintering by agglomerating metallic Ni particles since the steam reforming temperature (700°C–900°C) is typically higher than its Tammann temperature (691°C, above which Ni sintering will readily take place) [171,172].
Oemar et al. [135] reported that apart from the coke deposition, the metal sintering of LaNiO3 catalyst also decreased its catalytic activity (Fig. 8). The Ni crystallite size of spent LaNiO3 was increased considerably by 34.7% after 8 h of reaction. Park et al. [121] stated that the deterioration of Ru/α-A2O3 resulted from Ru sintering, which yielded a bulky crystalline structure that was detrimental to further catalytic reactions. Quitete et al. [173] concluded that NiO/CaAl12O19 and NiO/LaAl11O18 with a low Ni loading (approximately 6%(wt.)) experienced shrinkage in Ni crystallite size after reduction. Furthermore, the catalysts experienced a rapid sintering-related deactivation (increased above 150% of Ni crystallite size) compared to the high Ni loading catalyst (14%(wt.)) [173]. In contrast, Karnjanakom et al. [103] mentioned that the high loading of Ni in Ni/MCM-41 catalyst (20%–40%(wt.)) promoted not only the coke deposition rate but also the sintering.
Catalyst poisoning
The presence of impurities in the biomass-derived syngas such as sulfur-, nitrogen-, and chlorine containing compounds could poison the catalyst in the downstream tar reforming process [174,175]. Of the impurities, sulfur is the most common poison, causing severe deactivation of steam reforming catalysts. However, apart from the H2S, the effect of other impurities on the steam reforming of tar has not been extensively investigated. Nitrogen-containing compounds such as nitrogen oxides (NOx) could deactivate catalyst by oxidizing the active metal [175]. On the other hand, hydrogen chloride (HCl) poisons the catalyst due to the chemisorption of HCl on a metal site followed by metal sintering [174,176]. This process leads to an irreversible deactivation of catalyst and causes an abatement of catalytic activity.
Generally, about 20–200 ppm of sulfur-containing compounds, mainly hydrogen sulphide (H2S), is contained in the syngas derived from biomass gasification [177–179]. Metal catalysts, especially Ni catalyst, are susceptible to sulfur poisoning due to the strong dissociative chemisorption of sulfur on the metal site [174,180,181]. For example, H2S chemisorbs on the Ni site, which alters the atomic surface structure and forms inactive nickel sulphide (Eq. (10)) [182,183], consequently reducing the accessibility of active sites for tar. In addition, the presence of H2S also promotes the coke formation during steam reforming of tar [133,184]. The reason for this is that the formation of inactive metal sulphides suppresses the reforming reaction while the tar cracking reaction continues to take place, leading to the coke formation that derived from the carbonaceous product of tar cracking [184]. However, the sulfur poisoning can be prevented only at high pressures (20–30 bar) and temperature (>900°C) but those conditions are less favorable to the industrial application [185].
Type and material of the reactor
Catalytic steam reforming is considered as a promising approach for addressing tar formation and improving the H2 production during biomass gasification. Steam reforming is prone to challenges, which must be eradicated to ensure higher productivity and effective commercialization. Typically, tar contains heavy polycyclic aromatic hydrocarbon with a high thermal stability [74,154]. The steam reforming process requires a high operating temperature (700°C–900°C). However, the high temperature and internal pressure make the reformer tube susceptible to creep cracking. In principle, the control of high temperatures poses a tough challenge, which increases the operational or capital costs of power consumption, reactor material, engineering, and installation [186,187].
Currently, nickel- or iron-based oxide dispersion strengthened alloys that withstand an extreme temperature of 1100°C are available but relatively expensive [188]. Inconel is nickel-chromium-based superalloys used extensively for reformer tubes in steam reforming [189]. Moreover, refractory metals like tungsten and molybdenum exhibit extremely poor oxidation resistance but provide a high-temperature endurance capability [190]. Superplastic ceramic materials are also one of the acceptable candidates for a high temperature but require further development of appropriate joining methods [188].
Several types of reactors have been employed for steam reforming. The fixed bed reactor is the most common and simplest type for industrial-scale H2 production. The drawbacks of the fixed bed reactor consist of significant radial and axial temperature gradients that lower the bed effective thermal conductivity [191]. Micro-channel reactor with well-coated catalysts provides a high surface area to volume ratio, which leads to a better heat and mass transfer within the reactor. However, the problem of this reactor is the low durability of the coated catalyst [192]. Besides, the requirement of a well-defined catalyst with a regular shape and much smaller particles limits the application of commercial catalysts [193]. Membrane reactors produce the high purity of particular gas but its fabrication cost is relatively high and the mechanical resistance is low [193]. Therefore, a suitable type and high-temperature resistance material must be considered in reactor design to balance the trade-off between the safety and economy of the reactor.
Dry reforming and its challenges
Typically, 10% to 30% (vol.) of CO2 is released from the biomass gasification [194]. Therefore, CO2 reforming of tar toward hydrogen gas, so-called dry reforming of tar (Eq. (4)), was recently developed by some researchers [195–198]. Dry reforming reduces the CO2 emission, improves the carbon conversion, and eliminates the cost of a steam generation [198]. However, the equilibrium of gas production is generally affected by the reverse WGS reaction (Eq. (10)). This results in the lower H2/CO ratio of produced gas and is less favorable for hydrogen production [199]. Dry reforming also suffers rapid catalyst deactivation by carbon deposition via CO disproportionation (Eq. (11)). Besides, the adsorption of CO and CO2 on the catalyst can decelerate the dry reforming rate [200].
Rached et al. [197] investigated the addition of Ce and La into Ni-Al catalyst for dry reforming of toluene. Ce and La promote the CO2 adsorption, hinder the CO disproportionation, and consequently reduce the carbon deposition on the active metal [201]. They reported that Ce promoted catalyst had a more pronounced CO2 adsorption effect and a better resistance against coke formation on the catalyst. Bao et al. [198] evaluated the performance of Co/MgO catalyst in dry reforming of tar at 570°C. They claimed that the activity and stability of the catalyst increased with the loading of Co (5%–15%(wt.)). With the evidence of TGA and XRD analysis of spent catalyst, coke deposition and sintering-related deactivations are negligible in this case. However, they found that the catalyst deactivation was mainly attributed to the oxidation of metallic Co by CO2 during the dry reforming reaction.
Autothermal reforming and its challenges
To address the catalyst deactivation and intensive energy consumption of steam reforming, oxygen (O2) gas is introduced to promote the cracking reaction and hinder the carbon deposition [202,203]. This process is the so-called autothermal reforming (ATR) or oxidative steam reforming. The involvement of partial oxidation (POX) in ATR also increases the yield of H2. By combining the endothermic steam reforming (Eq. (1)) and exothermic POX (Eq. (12)) [204], external heat supply and indirect heat exchangers are not required by ATR [46]. The reason for this is that, by increasing O2 to a level, the energy consumption via steam reforming is balanced by the energy generated from the oxidation [205]. Hence, the overall reaction is theoretically autothermal or self-sustaining.
The dominant reactions during ATR are steam reforming, POX, and WGS reactions. ATR also shows a lower coke formation on the catalyst as compared to steam reforming. The oxidative environment allows a portion of deposited coke oxidized by oxygen to produce CO (Eq. (13)). Thus, catalyst can be used for prolonged periods without deactivation [206]. However, ATR either requires a costly O2 purification system for pure O2 feeding or treats the product gas with a diluent (N2 from the air) [205]. However, introducing of air instead of O2 is typically reported as reducing the heating value of the process.
Wang et al. [202] integrated the La0.8Sr0.2Ni0.8Fe0.2O3–δ (LSNF) perovskite catalyst with BaBi0.05Co0.8Nb0.15O3–δ (BBCN) hollow fiber membranes for ATR of toluene (see Fig. 9). They found that the integration offered a higher conversion and a lower carbon deposition than using catalyst only. This is because the oxygen permeable membrane transports lattice oxygen to the steam reforming reaction, consequently provides another form of reforming agent for POX. In this research, the air is introduced to the reactor. The permeable membrane allows only oxygen diffuses through the membrane to reach the catalyst region via oxygen vacancies and electronic defects (see Fig. 10). Therefore, the purity of permeated oxygen can achieve 100%.
Future focus and prospect
Since steam reforming is operated at a high temperature, future research on better reactor material and configuration is necessary. In addition, scale-up procedures with highly improved operating temperature control are required to enhance the cost, efficiency, and safety of steam reforming of gasified biomass tar [207]. The CO2 emissions can be reduced by integrating CO2 sorbent (e.g. CaO based material) into steam reforming. This is termed the sorption enhanced steam reforming (SESR), which has been examined at the laboratory but not in commercial scale [208–210]. Therefore, further investigation and definite proofs-of-concept for SESR are required to improve the performance of scale-up and industrial scale applications [211,212]. Besides, it is necessary to develop a continuous reaction-regeneration of the SESR system for extended operation time.
The composition of biomass tar is complex, in which, each component has a different influence on reforming efficiency, gaseous product distribution, and catalyst deactivation [24,74,79]. Although recent research studies reported superior catalytic performance in laboratory steam reforming of biomass tar model, the oxygenated and polycyclic aromatic HCs in the real biomass tar could promote coke deposition and lower the activity of the catalyst, respectively. Besides, a matrix of complex reactions among the intermediate products and different tar compositions may occur. It is difficult to predict the catalytic process mechanism during steam reforming of biomass tar. Therefore, it is necessary to conduct research on catalytic performance for steam reforming of real biomass tar to make sure that the coke resistance ability and the activity of the catalyst is adequate to deal with the complex real biomass tar. Besides, a chemical looping system for continuous tar reforming with simultaneous catalyst regeneration is also an important issue for industrial practice.
Moreover, an advanced catalyst with a better catalytic activity, stability, selectivity, and economic feasibility is required for catalytic research and industrialized steam reforming of tar [70,213]. Furthermore, an investigation on the synergetic effect and mutual interaction mechanism of active metal, support, and promoter are recommended in future research, especially, of alloy, hydrotalcite, and perovskite-type catalysts [214,215]. To make the developed catalyst more practical in industrial scale, optimization of tar steam reforming parameters over a newly developed catalyst, for maximum throughput with minimum raw material and energy consumption, merits further examination [145,146,216]. Apart from the conventional steam reforming approach, new technologies such as dry reforming and ATR are desired to be exploited thoroughly and intensively.
To achieve the sustainability of this technology, there are three pillars that must be met, namely, social, environmental, and economic [217]. In the social aspect, the H2 produced by steam reforming offers alternative energy and reduces the dependence of fossil fuel. In the environmental aspect, steam reforming converts the hazardous tar produced from gasification into H2 rich gas. The economic aspect indicates that steam reforming is more cost-effective compared to other tar elimination methods because it converts tar into more valuable products and incurs less cost compared to other reforming technologies. Therefore, the improvement of H2 production by steam reforming of biomass tar complies with the pillars of sustainability. However, the sustainability of this technology can be further improved by CO2 capture, the application of low-cost element for catalyst preparation, and modified catalyst for high H2 selectivity.
Conclusions
Coking and metallic sintering reactions always occur in steam reforming of gasified biomass tar. Sometimes, catalyst poisoning also occurs if the gaseous impurities are introduced along with the tar. Despite the significant achievements in catalysis research for tar steam reforming, most catalysts still lack the characteristics to ensure a high feedstock conversion, H2 selectivity, and the resistance to deactivation. In particular, the investigation of steam reforming of tar along with the raw gaseous products produced from gasification is lacking in the scientific literature. Besides, other desirable qualities, including economic and environmental feasibility, also need to be taken into consideration in effective steam reforming. To obtain an effective catalyst and efficient H2 production, a combination of suitable catalyst formulation along with proper reactor design and operating conditions plays a significant role. The recent research indicates that promoted, alloy, perovskite, and hydrotalcite-type catalysts have a high potential to enhance the catalytic performance. These catalysts reportedly exhibit a high catalytic activity, a high selectivity toward H2, a high stability, and an extended lifetime in tar conversion. Therefore, further research is required to extend the knowledge of this class of catalysts and their operating conditions. On the other hand, the further assessment of the alternative reforming process such as dry reforming and ATR associated with their limitation including costly oxygen purification and catalyst deactivation is essential.