Waste ashes as catalysts for the pyrolysis–catalytic steam reforming of biomass for hydrogen-rich gas production

Waste wood pellets with a particle size of 1 mm were used as the biomass feedstock for pyrolysis reforming/steam gasification experiments. The wood pellets were produced as compressed saw dust pellets from waste wood processing by Liverpool Wood Pellets Ltd, Liverpool, UK. The proximate analysis of the wood pellets gave 75.0 wt% volatiles, 7.0 wt% moisture, 2.0 wt% ash, and 15.0 wt% fixed carbon. Elemental analysis gave 46.0 wt% carbon, 5.6 wt% hydrogen, 0.7 wt% nitrogen and 45.7 wt% oxygen (by difference).

Ash samples derived from coal, waste tyre and RDF were used as a catalyst for gas and hydrogen production during the biomass pyrolysis–catalytic steam-reforming process. The coal, waste tyre and RDF were ashed in a furnace at 700 °C for 4 h. The produced ash samples were further examined as a catalyst support using an impregnation method to produce Ni-ash catalysts. For this purpose, the coal, waste tyre, and RDF ash samples were impregnated with nickel using an aqueous solution of Ni(NO₃) ₂ ^· 6H₂O (corresponding to Ni loading of 10 wt%). The mixture of ash and Ni(NO₃) ₂ ^· 6H₂O was stirred using a magnetic stirring apparatus at 100 °C to produce a slurry, dried overnight and calcined followed by calcination at 750 °C for 3 h in an air atmosphere. Finally, the obtained catalysts were reduced at 800 °C under hydrogen.

The composition of the ash samples was determined using a Thermo Advant XP X-ray fluorescence (XRF) spectrometer. The samples were prepared as fused beads to provide a homogenous sample. A lithium borate flux was added to the ash samples and heated to 900 °C in a platinum crucible and then cast into a mould. The resultant fused bead was presented to the XRF spectrometer for analysis. The fused bead was analysed using OXSAS programme. The surface area of the produced coal, waste tyre and RDF ash samples was determined according to the Brunauer–Emmet–Teller (BET) method using a Quantachrome NOVA 2200e series apparatus. Samples were degassed at 250 °C for 3 h under vacuum prior to measurement. A Bruker D-8 diffractometer with Cu Kα radiation operated at 40 kV and 40 mA was used to obtain the X-ray diffraction (XRD) patterns of the original product ash samples and the nickel-supported ash catalysts. The XRD pattern identification was obtained using HighScore Plus software. A Hitachi SU8230 scanning electron microscope (SEM) coupled to an energy-dispersive X-ray spectrometer (EDXS) system was used to characterise and examine the surface of the tested catalysts.

It is a two-stage fixed bed reactor system consisting of a first stage involving pyrolysis of the wood pellets and a second-stage catalytic reactor containing the produced residual coal, tyre and RDF ashes. A schematic diagram of the system is shown in Fig. 1. The reactors were constructed of stainless steel with an inner diameter of 22 mm and a length of 160 mm with monitoring and control of gas flow and temperature.

The experiments consisted of the initial heating of the second-stage catalytic reactor containing 1.0 g of ash catalyst to a temperature of 800 °C and maintained at that temperature throughout the experiment. In addition, water was injected into the second-stage catalytic reactor using a syringe pump at a rate of 4.0 g h⁻¹ (steam/biomass (S/B) ratio of 2). After the catalytic reactor temperature was stabilised at 800 °C, the first-stage pyrolysis reactor containing the biomass (1.0 g) was heated at a rate of 40 °C min⁻¹ to a final temperature of 600 °C. The evolved pyrolysis gases from biomass thermal degradation mixed with the steam and underwent catalytic steam reforming over the coal, tyre and RDF ash catalysts and Ni-ash catalysts. For comparison with the ash-based catalysts, clean and washed quartz sand was used in place of the catalyst in the second-stage reactor. The sand would have none of the catalytic metals found in the ash-based catalysts. Any condensable liquid products, including condensed steam, were collected in a condenser system cooled with dry ice. Gases passing through the condenser system were collected in a Tedlar™ gas sample bag.

The product gases in the gas sample bag were analysed immediately after each experiment using three Varian CP-3380 gas chromatographs (GC). N₂, H₂, and CO gases were analysed using a 60–80-mesh molecular sieve packed column, Ar carrier gas and thermal conductivity detector (TCD). CO₂ gas was analysed on a second Varian GC with a 80–100-mesh HayeSep packed column, Ar carrier gas and TCD. CH₄, C₂H₆, C₂H₄, C₃H₈, C₃H₆, C₄H₁₀, C₄H₈ and C₄H₆ hydrocarbon gases were analysed on a third Varian GC with an HayeSep 80–100-mesh molecular sieve column, N₂ as carrier gas and flame ionisation detector (FID).

The ash samples that were produced from the coal, waste tyre rubber and RDF and used for the pyrolysis–catalytic steam-reforming experiments were characterised using XRF, surface area analysis, XRD, and SEM. Table 1 shows the metal-oxide compositional analysis of the derived combustion ashes produced by XRF. The waste tyre rubber combustion ash is dominated by the presence of zinc which is added to the tyre vulcanisation process to improve the physical properties of the rubber. In addition, silica (SiO₂) is present in high concentration, and the silica is often added as filler material for the tyre formulation. The RDF ash contains typical metal compounds that are found in waste incinerator combustion ash. RDF is prepared from municipal solid waste with the aim of raising the calorific value of the product fuel by removing metals, glass and ceramic materials. The RDF ash prepared in this work contained mainly Al₂O₃ and CaO with minor concentrations of other alkali metals. The coal ash sample consisted of mainly SiO₂, Al₂O₃ and Fe₂O₃.

Table 2 shows the surface areas and pore diameter for the ash samples and the nickel-ash catalysts. The BET surface areas of the ash samples produced from RDF, tyre rubber and coal were found to be 21.8, 13.7 and 2.5 m² g⁻¹, respectively. However, nickel catalysts supported on RDF ash showed a slightly reduced surface area than the original ash samples.

The morphologies of the produced ash catalysts and Ni-ash catalysts were characterised by scanning electron microscopy (SEM) and are presented in Fig. 2. The coal ash shows the characteristic cenospheres produced during coal combustion. The tyre rubber-derived ash and RDF-derived ash show less structure. The addition of 10 wt% nickel to the ash appears to show surface deposits of the metal.

Figure 3 shows XRD profiles of the produced ash catalysts and Ni-ash catalysts. The diffraction peaks observed around 2-Theta 44° and 52° with the ash-supported nickel catalysts were related to the presence of metallic Ni and no peaks corresponding to the NiO phase were identified. This was expected as the catalysts were reduced before the catalytic steam-reforming process. The XRD diffraction peaks observed with Ni/tyre ash were smaller compared to the XRD peaks obtained with either Ni/coal or Ni/RDF catalysts, which might indicate that the particle size of Ni is smaller for the Ni/tyre ash sample.

Table 3 shows the product yield and gas composition for the pyrolysis–catalytic steam reforming of biomass with the ash-based catalysts. In addition, experiments were carried out with clean quartz sand in place of the catalyst and the results shown in Table 3 indicate that catalytic cracking and steam reforming are taking place, even in the presence of sand. The particles of sand in the second-stage reactor is maintained at 800 °C, producing a hot surface for thermal cracking and steam reforming of the evolved pyrolysis gases from the biomass thermal degradation. In the presence of the sand, the total gas yield in relation to the feedstock biomass was 39.9 wt% which increased to 52.7 wt% when tyre rubber ash was introduced and with coal ash and RDF ash, increased to 50.3 wt% and 59.5 wt%, respectively. The metals in the ash sample produce a catalytic effect producing more product gas, particularly for the RDF ash sample. The residual char yield derived from the biomass pyrolysis in the first-stage reactor was constant at ~ 19.0 wt% for all the experiments since it was unaffected by the second-stage reactions. Consequently, the liquid yield from biomass (by difference) represented 41.1 wt% for the sand, 28.3 wt% for the tyre ash, 30.7 wt% for the coal ash and 21.5 wt% for the RDF ash. Suggesting that cracking of the higher molecular weight biomass volatile compounds to gas occurred in the presence of the ash samples compared with the non-catalytic sand.

The gas composition shown in Table 3 shows that the product gases consist of mainly CO, CO₂, H₂, CH₄ and C₂H₄ hydrocarbons. The influence of ash addition to the catalytic reactor is to increase the level of catalytic steam reforming of the hydrocarbons (CH₄ and C₂H₄) which show a decrease in concentration compared to the non-catalytic experiment (sand) and a consequent increase in H₂ production. In addition, the decrease in CO and increase in CO₂ and H₂, compared to the non-catalytic (sand) results indicates that the presence of the ash enhances the water gas shift reaction. Overall, the presence of the ash with their inherent metal content acts as a catalyst for the production of hydrogen from the biomass pyrolysis gases. The highest hydrogen gas yield of 29.73 vol% representing 7.90 mmol H₂ g⁻¹_biomass was produced in the presence of the RDF-derived ash.

Table 1 showed a high metal content in the ash, including Al₂O₃, CaO, MgO, CuO and Fe₂O₃, depending on the waste ash used. During the catalytic steam-reforming process, the metal oxides would undergo reduction to the metal via the reducing gases such as hydrogen and carbon monoxide produced during the process. The metal oxides, Al₂O₃, CaO, MgO, and Fe₂O₃, have been used as catalysts for the catalytic steam reforming of biomass [20, 21]. Magnesium, acting as a promoter in nickel-based catalysts, has also been investigated by Garcia et al. [22] where it was reported to produce a higher hydrogen yield in the steam reforming of an aqueous fraction of bio-oil. In a later report, Garcia et al. [23] also showed that magnesium containing NiMgAl₂O₅ catalyst produced high total gas and hydrogen yields for the hydrogen production from biomass catalytic steam gasification for biomass in the form of pine sawdust. The high yields of H₂ and CO₂ coupled with low yields to CH₄, C₂ hydrocarbons and CO suggesting high catalytic activity for the steam-reforming reactions of hydrocarbons and the water gas shift reaction. The presence of copper as a Ni-catalyst promotor metal for the production of hydrogen has been shown to enhance the yield of hydrogen and decomposition of methane [24]. The addition of zinc to catalytically steam reform wood sawdust biomass pyrolysis gases using a Ni/Zn/Al₂O₃ catalyst showed that the total gas yield and hydrogen yields were increased producing a maximum gas yield of 74.8 wt% and H₂ yield of 20.1 mmol H₂ g⁻¹_biomass [25]. In the absence of the Ni/Zn/Al₂O₃ catalyst, the total gas yield was only 33.0 wt% and hydrogen yield only 2.4 mmol H₂ g⁻¹_biomass. On the one hand, zinc modifies the surface structure and the surface chemistry of the catalysts by formation of zinc aluminates, and on the other hand, zinc oxide can be reduced to metallic zinc under reaction conditions, thus modifying the catalytic properties of the active phase. The presence of Zn increases the ethanol conversion to gaseous compounds as compared with the catalyst supported on the Zn-free commercial alumina. Alkali metals (K₂O, Na₂O) were also identified in the different waste ashes (Table 1) used in this study. Sodium and potassium compounds have also been reported as catalysts for hydrocarbon decomposition and improved gas quality [21]. It has been suggested that the presence of alkali metal compounds act as catalysts to increase hydrocarbon cracking and reforming and thereby increase hydrogen production from biomass [20, 26].

The wide range of metal and alkali metal catalysts present in the combustion ash samples will clearly have a catalytic effect on the catalytic steam reforming, dry reforming, thermal cracking and gasification reactions involving the evolved biomass pyrolysis gases. Other researchers have investigated the use of combustion ashes and reported a catalytic effect. Wang et al. [16] used a combustion ash produced from coal as a catalyst for the catalytic steam reforming of acetic acid and phenol and compared their results with several commonly used steam-reforming catalysts. They showed that the coal ash sample produced similar carbon conversion efficiencies and hydrogen for acetic acid to that of an Fe–Al₂O₃ catalyst. For phenol, the results were lower in terms of conversion and H₂ yield.

Table 4 shows the influence of the addition of 10 wt% nickel to the combustion ash catalysts in terms of gas yield and gas composition derived from the pyrolysis–catalytic steam reforming of biomass. In addition, comparison is again made with the yields in relation to no-catalyst (in the presence of quartz sand). There was only a small influence of nickel addition to the combustion ashes in terms of any increase in total gas yield compared to the non-nickel-impregnated combustion ash samples shown in Table 3. For example, the total gas yield increased by 6.6 wt% for the Ni-tyre ash, 4.6 wt% for the Ni-coal ash sample, whilst the Ni-RDF ash sample showed a similar total gas yield compared to the ash-only catalysts. However, the hydrogen yield in relation to the mass of biomass feedstock showed a significant increase for all the nickel-impregnated ash catalysts. For example the Ni-tyre ash catalyst increased the hydrogen yield to produce the highest yield 10.5 mmol⁻¹ H₂ from the biomass. The presence of the nickel in the ash catalysts showed that nickel metal catalytic steam reforming of the hydrocarbons (CH₄ and C₂–H₄) occurred, producing increased hydrogen yield in terms of H₂ yield as mmol g⁻¹_biomass and also a volumetric increase in hydrogen content in the gas product. The results suggest that the high metal content of the combustion ashes acts as the main catalytic effect and that addition of a further active nickel content to the catalyst produces an approximate 20% increase in hydrogen yield.

The addition of nickel to combustion ash to enhance the catalytic effect of the metals and compounds in the ash has been examined by other researchers. For example, Wang et al. [16] investigated the catalytic steam reforming of bio-oil model compounds in the form of acetic acid and phenol using a nickel-coal ash catalyst with a 15 wt% nickel loading. The conversion of acetic acid was 57.5% and for phenol conversion was 26.3% with the combustion ash, but increased to 98.4% and 83.5%, respectively, with the nickel-ash catalyst. In addition, they reported that the yield of H₂ from the catalytic steam reforming of acetic acid using the combustion ash was 50.2% and for phenol was 19.6%. However, with the addition of nickel to the ash, the yield of H₂ improved to 85.6% for acetic acid and 79.1% for phenol at a catalyst temperature of 700 °C and steam/carbon ratio of 9.2. The influence of temperature was to increase the conversion of the bio-oil model compounds and increase hydrogen yield. In a later report, Wang et al. [2] added both nickel and iron to a coal combustion ash to produce a Ni–Fe-ash catalyst for the catalytic steam reforming of acetic acid. They reported 100% conversion of the acetic acid and a hydrogen yield of 89.6% at a catalyst temperature of 700 °C.