Abstract
Nickel on alumina support offers opportunity for gasification of biomass for hydrogen production. In a recent contribution from our research team, (González Castañeda, D. G., et al. 2019) showed that cerium or lanthanum co-impregnation at 2 wt% with nickel may have a favorable effect for biomass catalytic gasification. However, and given an observed influence of lanthanum on the formation of small Ni crystallite sizes, five Ni/γ-Al2O3 based fluidizable La promoted catalysts were studied. Nickel-alumina catalysts promotion was effected varying La in the 0.5 and 1.0 wt% range. Once impregnation precursors loaded, they were reduced at 480 °C via an activation step. Catalysts were characterized using BET, XRD, AA, TPR, TPD, H2-chemisorption, TEM-EDX and FTIR. Catalyst performance was established in a fluidized CREC Riser Simulator, using: a) glucose as surrogate biomass, b) 600 °C, c) steam/biomass (S/B) ratio of 1, d) catalyst /biomass (C/B) ratio of 3.2 and e) 20 s reaction time. Data obtained was analyzed using an ANOVA statistical data analysis package with the 5 wt% Ni and 0.5–1 wt% La and Ce on γ-Al2O3 catalysts, prepared using a pH of 1 of impregnating solution were the best yielding 0.53–0.56 hydrogen molar fractions. These catalysts also gave a 39% reduced coke, and this while compared with the coke formed on the 2% Ce – 5 wt%Ni/γ-Al2O3 (González Castañeda, D. G., et al. 2019). This promising performance was assigned to the dominant NH3-TPD medium acidity, the high catalyst specific surface (∼140 m2/g), and the good 9% metal dispersion with 9–10 nm nickel crystallites.
1 Introduction
A large fraction of the energy demand is met nowadays worldwide using fossil fuels which are a declining source of energy, as well as they are considered to contribute to carbon dioxide and greenhouse gases emissions. Hydrogen is a very promising as alternative energy vector. This is due to its calorific value of 39.4 KWh/kg, as well as its clean combustion with no CO2 formation. Hydrogen can be produced via biomass gasification, with a neutral carbon footprint. Biomass gasification is a thermochemical process converting biomass into synthesis gas. This can be archived with the help of gasifying agents such as air/oxygen, steam and flue gases (Moilanen, Nasrullah, and Kurkela 2009; Nemtsov and Zabaniotou 2008).
Steam and dry reforming reactions (SR) and (DRM) are catalyzed by group VIII metals (non-noble and noble transition metals), with nickel being the most widely used in the industry (Rostrup-Nielsen and Bak Hansen 1993). Nickel can also contribute to reduce tars improving as a result the syngas quality (Chan and Tanksale 2014). Due to the high activity and low cost, Ni-based catalysts are extensively used for biomass catalytic gasification under different atmospheres (Ren et al. 2019). At higher temperatures, nickel enhances hydrogen yield while at lower temperatures it favors the syngas methane content.
To carry out biomass gasification, there are a variety of catalyst supports available such as Al2O3, olivine, ZrO2, TiO2, CeO2, and MgO, activated carbons (AC) and char derived from biomass among others. Acidity, specific surface area, pore structure, and electronic structure of the support are all parameters that can affect the catalyst activity (Yung, Jablonski, and Magrini-bair 2009). Alumina (Al2O3) is the most commonly used mesoporous support for Ni-based catalysts, because of their fluidizability, high surface areas (200–300 m2/g), high thermal stability to 900 °C, and great facility for forming into pellets. γ alumina is by far the most widely used alumina (Bartholomew and Farrauto 2006; Mazumder and de Lasa 2015, 2018; Sutton, Kelleher, and Ross 2001).
While preparing nickel supported on alumina, an issue observed is the formation of the nickel aluminate species (Ewbank et al. 2015). In particular and in the case of nickel prepared by dry impregnation with 2% Ni weight loading, H2 – chemisorption results indicated low dispersion, and considering TPR, ETEM, and reactivity measurements, it is concluded that surface NiAl2O4 is the primary type of nickel species (Juan-Juan, Román-Martínez, and Illán-Gómez 2009). It has been shown using XPS (Ewbank et al. 2015; Huang and Schwarz 1988), that the only nickel species on the reduced catalyst surface is NiA12O4 and this for low weight loading catalysts (<1 Ni wt. %), whereas both nickel and NiA12O4 are present on high loading catalysts (from 1 to 6 wt%). For the reduced sample, XPS can distinguish between the metallic nickel (binding energy, B.E., 852.5 eV) produced on reduction of nickel oxide and the unreduced NiA12O4 (B.E. 856.5 eV) species. Also, other authors found that 5% Ni-catalyst, after the usual pretreatment (reduction with H2 at 873 K, 20 min) is active displaying a distinctive black color. This was the case after six experiments at 873 K meaning that in this case the NiAl2O4 phase is not formed in a significant proportion (Gil-Calvo et al. 2017; José-Alonso, Illán-Gómez, and Román-Martínez 2013; Qin et al. 2015).
Low metal loaded catalysts have proven their effectiveness in dry methane reforming reactions by having high conversion and minimal deposition of coke. These catalysts have the desirable combination of metal use with reduced formation of carbonaceous deposits (José-Alonso, Illán-Gómez, and Román-Martínez 2013). Besides, nickel dispersion decreases with increasing nickel weight loading (Bartholomew and Farrauto 1976).
Regarding Ni catalysts, they display particle sintering at high temperature 800 °C, resulting in reduced catalytic activity. Sintering can, however, be controlled limiting gasification temperatures, as attempted in the present study. Furthermore, the addition of co-promoters could also benefit catalyst performance as proposed with: a) Ce or La (González Castañeda et al. 2019; Meng et al. 2015), b) Ru (Calzada Hernández et al. 2020), c) Pt, Rh, Al and Fe (Li, Hu, and Hill, 2006; Shahbaz et al. 2017).
Regarding the thermal decomposition of La-nitrates, the La(NO3)3.6H2O may convert to La-oxides via LaONO3·xLa2O3, in the 320–470 °C range. While Ce(NO3)3 form CeO2 directly above 300 °C (Ivanov et al. 2015) and its reduction can be carried out at 800 °C under H2-Ar atmosphere (Konysheva 2013).
Given the above, a 5 wt% nickel on fluidizable γ-Al2O3 is proposed in the present study. The catalyst is prepared via co-impregnation of nickel and a promoter, either lanthanum or cerium. Following this incipient wetness impregnation, the metal precursors are reduced under an oxygen free atmosphere. Prepared catalysts are evaluated in a fluidized CREC-Riser using steam glucose gasification at 600 °C. The prepared catalyst displays a large specific surface area, good metal dispersion with small metal crystallites as well as reduced acidity. It is on this basis that the prepared catalyst is anticipated to be highly active and stable with low carbon being formed, and this is a promising solution for biomass gasification.
2 Status of catalyst biomass gasification technology development
Catalytic activity in biomass gasification can be improved by adding promoters, such as Pt, Co and Cu. These metal promoters improve the reactivity given the enhanced nickel reducibility, as well as the strong promoter interaction with the catalyst support. Metal promoters also improve dispersion of the nickel metal and provide higher resistance to coke formation (Chan and Tanksale 2014).
The traditional Ni-based catalysts are easily deactivated in fixed bed reactors, as a result of being prone to sintering and/or carbon deposition at high temperature. The great ability of fluidized beds to conduct biomass gasification has been confirmed extensively as documented by (de Lasa et al. 2011).
As well, the addition of dopants such as lanthanum or cerium, can also reduce carbon deposition (Sutton, Kelleher, and Ross 2001). By controlling the amount of basic La2O3, the support acidity is reduced and the adsorption capacity of the gamma alumina is improved, which results in less coke formation (Mazumder and De Lasa 2014; Xu and Jiang 2014). Ni catalysts with low metal contents, as 1, 2.5 and 4% show a very low amount of deposited carbon (José-Alonso, Illán-Gómez, and Román-Martínez 2013). The dispersion of La3+ on alumina however, may inhibit the incorporation of Ni2+ ions into the tetrahedral vacancy of γ-alumina and increasing the ratio of octahedral Ni2+ ions to tetrahedral Ni2+ions, and thus, augmenting the reduction degree of the catalyst precursor. As a result, the Ni–La2O3/γ-Al2O3 catalyst it is anticipated to have higher catalytic activity than the Ni/γ-Al2O3 catalyst at the same nickel loading (Ren et al. 2007).
Preparation methods drastically affect the type of nickel species present on alumina supported nickel catalyst. For instance, co-impregnation method, with catalyst and promoter precursors added to the same solution, has led to better performing catalysts for steam biomass gasification. This has been shown to be the case when catalyst was prepared by sequential or successive impregnation (Tomishige et al. 2007). These authors claimed this led to catalysts with greater specific surface area and smaller crystal sizes, which means a better metal dispersion or less crystallite agglomeration (Ewbank et al. 2015; Meng et al. 2015).
The pH and metal loading in the impregnant solution are also factors affecting nickel supported on gamma-alumina catalyst, and thus require careful consideration. In this respect structure, dispersion, activity, selectivity and total carbon deposited during reaction were found to correlate with the metal weight loading (ranging from 0.9 to 6 wt%) and the initial pH of the impregnant solution (pH’s of 1, 3 and 5) (Huang and Schwarz 1987).
2.1 Reaction modelling
The biomass conversion, while using a catalyst involves many reactions (Gao et al. 2012; Moghadam et al. 2014). Primary reactions transform vaporized biomass on permanent gases, higher hydrocarbons, coke and tars and they are carried out mainly during the non-catalytic (thermal) experiments.
Secondary reactions crack the higher hydrocarbons into gases, they are enhanced with the use of catalyst.
Permanent gases react later to alter the gas composition, with this reactions being strongly influence by the catalyst present.
Thus, the resulting gaseous mixture mainly consists of carbon monoxide (CO), carbon dioxide (CO2), hydrogen (H2) and methane (CH4). This gaseous mixture is frequently designated as permanent gases.
3 Experimental methodology
3.1 Catalyst preparation
Calcination and reduction are two important steps on the catalyst precursor conversion. High calcination temperatures may lead to less reducible species and higher nickel aluminate formation and should be avoided (Juan-Juan, Román-Martínez, and Illán-Gómez 2009). Thus, in the present study a direct reduction with an atmosphere free of oxygen is considered for metal precursor conversion.
Regarding catalyst preparation, the following non-promoted and rare earth promoted nickel catalysts were prepared: 5%Ni/γ-Al2O3, 5%Ni-0.5%La/γ-Al2O3, 5%Ni-1%La/γ-Al2O3, 5%Ni-0.5%Ce/γ-Al2O3 and 5%Ni-1%Ce/γ-Al2O3. In all cases both incipient wetness impregnation (Bartholomew and Farrauto 1976) and co-impregnation methods (Meng 2015; Tomishige et al. 2007) were selected. Lanthanum nitrate and cerium nitrate were selected as promoters and Sasol Catalox® SSCa5/200γ-Al2O3 was chosen as a fluidizable mesoporous support.
Each catalyst sample was prepared using an acid solution with pH value of 1. After impregnation, nitrate metal precursors were thermally decomposed and reduced under a hydrogen flow, avoiding all contact with air.
Figure 1 presents a fixed bed quartz reactor where the catalyst reduction was carried out with 160 cm3/min of pure hydrogen being fed in the upper reactor section. Hydrogen flows down the unit with an entry quartz section packed with an 80-μm porous quartz and covered with quartz-wool to support the catalyst. The hydrogen flow contacts the catalyst containing the nickel nitrate and reduce it to obtain just metallic nickel and activating as a result catalyst is performed.
Fixed bed reactor used for catalyst precursor decomposition and reduction (Gonzalez et al. 2019).
It is considered that the furnace at temperature 480 °C, all the nitrates decompose thermally, producing nickel oxide, lanthanum oxide and cerium oxide, and when they are in contact with hydrogen, only nickel oxide is reduced to metallic nickel, but the oxides of cerium and lanthanum are not reduced to metallic cerium and lanthanum respectively due they are non-reducible under the selected catalyst preparation conditions (Konysheva 2013). A mixed phase with composition LaONO3·xLa2O3 was obtained at the applied temperature (Gobichon, Auffrédic, and Louër 1996; Ivanov et al. 2015).
In order to synthesize the materials 5%Ni/γ-Al2O3, with the impregnating solution at pH value of 1, the following steps were effected:
The alumina support (5 g) was dried during 12 h at 110 °C, to remove excess water,
An “A” solution was prepared with 1.23 g of Ni (NO3)2.6H2O (CAS: 13478-00-7) from Sigma–Aldrich and 1.29 g of deionized water,
The pH of the “A” solution was adjusted to 1, with a 0.1 M HCl,
The alumina support was placed in a flask under a 50 mmHg of vacuum and stirred for 20 min,
The resulting “A” solution was added drop-by-drop to the flask. This was done until the support pore volume was filled,
The obtained wet powder was dried for 12 h in a stove at 110 °C,
Once the alumina support was impregnated with the Ni precursor, it was reduced under a 160 cm3/min pure hydrogen flow, with the temperature raised first from room temperature to 260 °C using a 5 °C/min ramp. It was then left at 260 °C for 1 h.
After this, the catalyst sample was heated up to 480 °C at 5 °C/min and finally left at 480 °C for 10 h.
Once these steps were completed, the oven was turned “off” and the catalyst sample was cooled down for 2 h.
For the 5%Ni-x%La/γ-Al2O3 (where x = 0.5 or 1) the catalyst preparation procedure was close to the one used for the non-promoted catalysts with the following implemented changes:
For 5%Ni-0.5%La/γ-Al2O3. In this case, a “B” solution was used, 1.23 g of Ni(NO3)2.6H2O, 0.0775 g of La(NO3)3.6H2O (CAS: 10277-43-7 from Sigma–Aldrich) and 1.27 g of deionized water (solution “B”). The pH was adjusted to 1 with 0.1 M HCl,
The resulting “B” solution was added drop-by-drop to the 5 g of γ-Al2O3 under vacuum.
Furthermore, the additional reduction steps were identical to the ones used for the non-promoted catalysts. As well and for the 5%Ni–0.5%Ce/γ–Al2O3 catalyst preparation, all steps were identical to those used for the lanthanum promoted catalysts. However, in this case, Ce(NO3)3.6H2O (CAS: 10294-41-4 from Sigma–Aldrich) was used instead of La(NO3)3.6H2O.
Based on the above, and for the purpose of catalyst identification, the abbreviated notation described in Table 1 was adopted:
Various synthesized catalysts and their abbreviations.
| Catalyst | Abbreviation |
|---|---|
| 5%Ni-pH1 | 5Ni1 |
| 5%Ni-0.5%La-pH1 | 5Ni0.5La1 |
| 5%Ni-0.5%Ce-pH1 | 5Ni0.5Ce1 |
| 5%Ni-1.0%La-pH1 | 5Ni1La1 |
| 5%Ni-1.0%Ce-pH1 | 5Ni1Ce1 |
It is important to emphasize as reported in Table 1, that 0.5–1 wt% promoter loadings were used in the present study, versus the 2 wt% reported by (Gonzalez et al. 2019).
3.2 Catalyst characterization
The catalysts 5Ni1, 5Ni0.5Ce1, 5Ni0.5La1, 5Ni1La1 allowed to obtain the highest hydrogen production in the glucose gasification and were characterized using several techniques as described in the upcoming sections of this manuscript.
3.2.1 X-ray diffraction (XRD)
The X-ray powder diffraction patterns were obtained using a standard Cu Kα copper radiation in a D8 ADVANCE Bruker. Samples were scanned from 5° to 70° in the 2Ɵ degrees scale using a scanning time constant of 6.87 s. This XRD analysis allowed the identification of the present phases, in particular the crystalline phase of the γ-alumina support.
3.2.2 Atomic absorption spectroscopy
The nickel weight percent of the catalysts was determined in the Buck −210-VGP equipment. Fifty mg of each catalyst were dissolved in 60 ml of aqua regia (30 ml of HCl and 30 ml of HNO3) in a plastic container. Each sample of 50 mg was left at rest for 48 h and after that, 10 ml were taken and filled to 100 ml with deionized water. To measure the amount of nickel, a specific nickel-lamp was used at the requested wavelength 232 nm. Besides, solutions of 5, 10 and 25 ppm of nickel nitrate hexahydrated were prepared to build the calibration curve.
3.2.3 Fourier transform infrared spectroscopy (FTIR)
Pyridine FTIR was performed with a spectrometer Bruker Vector 22FTIR. The reference solution was made with a pellet of KBr. The software of the equipment was OPUS. The spectra were collected at a resolution of 4 cm−1, with a defined scan interval of 2400 to 1400 cm−1, 200 scans. A flow of nitrogen of 100 ml/min was used, with an environmental cell with injection of pyridine a 100 °C. Prior the pyridine adsorption, the samples were thermally treated at 500 °C with nitrogen flow for 2 h and they were cooling down until 100 °C.
3.2.4 Specific surface area (BET)
Catalyst specific surface areas (SSA) were determined from the nitrogen adsorption–desorption isotherms, at −196 °C and using a Micromeritics Chemisorb 2720 unit. Previously, the samples were degassed using a nitrogen flow with the standard BET procedure (Webb 2003).
3.2.5 Temperature programed reduction (TPR)
TPR runs were conducted in a Micromeritics 2720 unit (Norcross, G.A., USA) using 30 mg catalyst samples. Every sample was preheated first for line conditioning, under 50 ml/min nitrogen flow for a 30 min period. Then, the temperature was increased up to 250 °C and held for 30 min. This allowed the removal of moisture from the sample, the sample holder and the connecting lines. Following this, 10%H2/90%Ar at rate of 50 ml/min was introduced and the temperature was increased to 950 °C at 10 °C/min rate. The hydrogen uptake was recorded using a Thermal Conductivity Detector (TCD, Micromeritics, Norcross, G.A. USA).
3.2.6 Temperature programed desorption (TPD)
A Micromeritics 2720 unit (Norcross, G.A., USA) was used, 100 mg of each catalyst were pretreated with a flow of hydrogen at 600 °C for 20 min, the sample was cooled down to 100 °C under an inert gas flow. When the 100 °C temperature was reached, a 5% NH3 in He gas contacted the sample for 60 min. Then, a 50 cm3/min He carrier gas contacted the catalyst sample, with the temperature increased from 100 to 500 °C, at a temperature ramp of 10 °C/min. Desorbed NH3 was measured using a thermal conductivity detector Micromeritics, Norcross, G.A., USA).
3.2.7 H2-chemisorption
Regarding hydrogen chemisorption, a flow of Ar of 30 ml/min was employed. Several consecutive H2 pulses (10% H2 in argon) were injected into an argon carrier gas until the catalyst sample saturation was reached. Sample saturation was established when unchanged pulse areas were observed as measured by a TCD, placed at the exit of the catalyst-containing cell. Furthermore, the quantification of the total chemisorbed H2 was effected using the differences between injected pulse areas and output areas as measured by the TCD. Typically, twenty – 59 µl hydrogen pulses were required to complete a hydrogen chemisorption experiment. This chemisorbed hydrogen together with consumed hydrogen in TPR, provided the needed data for assessing the average Ni crystallite sizes (Mazumder and de Lasa 2015).
3.2.8 Transmission electron microscopy (TEM)
To determine the size and location of the nickel crystallites in the catalyst, TEM analysis was conducted using a transmission microscope TEM (Joel, Jem-2100), operating at an acceleration voltage in the range 100–200 keV. Prior to TEM measurements, the samples were dispersed in ethanol, ultrasonicated and deposited on a sample holder. The same piece of equipment was employed to do EDX analysis.
3.3 Biomass gasification
3.3.1 CREC-Riser Simulator
The experiments of biomass gasification were performed in the CREC-Riser Simulator Reactor. More details about it can be found in the literature (Patent No. 5,102,628, 1992). This is a bench-scale internal recycle batch reactor with a capacity of 53 cm3 allowing the loading of up to 1 g of catalyst.
A schematic diagram of the CREC Riser Simulator experimental setup is shown in Figure 2A. This figure illustrates the intense gas recirculation in the CREC Riser Simulator while catalysts particles are fluidized and confined inside a basket between two inconel grids. Furthermore, Figure 2B describes the various components of the CREC Riser Simulator including the reactor itself, the 4-port valve, the 6-ports valve and the vacuum chamber. Additional details regarding the operation of the CREC Riser Simulator are provided in (de Lasa 1992).
A) Schematic description of the CREC Riser Simulator reactor showing the gas recirculation path, B) Overall description of the CREC Riser Simulator and its auxiliary components (Quddus 2013).
Prior to every experiment, each catalyst, activated with pre-reduction in the reaction auxiliary equipment (at maximum temperature of 480 °C, Figure 1) was placed in the CREC Riser Simulator Reactor basket. The reactor system was then sealed, leak tested and heated to the selected reaction temperature (600 °C) under an argon atmosphere. Then, a set amount of glucose–water mixture (glucose with 99.5% of purity from Sigma-Aldrich CAS: 50-99-7 and deionized water) was injected via the reactor port.
As shown in Figure 3, once the injection was completed, the reactor pressure increased sharply during the first fraction of a second and during the remaining reaction time it was almost constant. These total pressure changes were assigned to the quick solution vaporization, primary gasification reactions, followed by much slower secondary gasification reaction.
Pressure profiles for four consecutive injections while using the 5Ni1La1 catalyst.
The produced gases were intensively mixed with the help of an impeller at 5000 rpm. This also allowed fluidization of the catalyst. Once the preselected reaction time was reached, the reaction products were evacuated from the reactor to the vacuum box for further analysis. The Personal Daq Acquisition Card recorded reactor and vacuum box pressure data as a function of run time.
Then, 1 mL gas sample was taken with a syringe and injected into the gas chromatograph. Each experiment was performed with at least five replicas to ensure reproducibility. This procedure was consistently used for both catalytic and thermal runs, with the only exception that for thermal runs, the catalyst was not loaded in the reactor basket.
Regarding gasification runs, they were carried out under the following conditions: a) 600 °C, b) 1 atm of argon, c) liquid injection of 25 μL, d) 20 s of reaction time and e) 0.050 g of loaded catalyst. This provided a catalyst/biomass (C/B) ratio of 3.2 and steam/biomass (S/B) ratio of 1.
During catalytic experiments, in-situ catalyst reduction in the CREC Riser was also performed. To accomplish this, every new catalyst sample was treated with hydrogen for 20 min at 600 °C to secure full catalyst activation and it is expected that the nickel oxide still present in the catalyst will be completely reduced to metallic nickel, as it will be shown in the results section of TPR. Following this, the catalyst was ready for the first run (injection). Completed the first run a series of consecutive runs followed.
One should notice that in all of these runs and to better understand the conditions of catalyst reactivation in between experiments, and to make sure the catalyst has the same activation conditions, the following was effected: a) contacting the catalyst for 15 min with air to burn the coke on the surface, but this has the problem the produce nickel oxide, b) followed by hydrogen for another 15 min to the reduction the nickel oxide to obtain again metallic nickel and in this way regenerate the catalyst and c) finally to purge with argon.
3.3.2 Analysis of the samples
Gas samples were analyzed in an Agilent 7890A gas chromatograph, with two capillary Agilent CP 7430 columns (Molsieve 5A and Parabond Q). These columns were connected in parallel to a thermal conductivity detector (TCD) and to a flame ionization detector (FID). Samples of 1 ml were extracted manually from the vacuum box and injected into the GC. The following temperature program was used with nitrogen as carrier gas: a) at 50 °C for 6.5 min, b) using a 15 °C/min ramp until 250 °C was reached, c) at 250 °C for 6 min.
After gasification, coke deposited on the catalyst surface was measured with Total Organic Carbon Analyzer, TOC-V-CPH Shimadzu.
3.4 Statistical analysis
The analysis of variance (ANOVA) was used to analyze the experimental data. The central idea of ANOVA is to separate the total variation in the parts participating each variation source in the experiment. A factorial design was applied to study the effect of two factors: type of promoter and load of promoter. In this way, the effect and significance of each studied variable were identified.
The F-tests in ANOVA allows to identify the factors on the dependent variable (molar fraction of each gas) and for each significant factor, what are significantly different from others. With the mean plot and interaction plot, we interpret the significant effects. With residual plots, we check the assumptions underlying the analysis of variance.
The results of ANOVA are used to verify if the spans of one average value do not overlap the spans of the average of other result, and then the variable that caused that difference in the results has a significant effect. Otherwise, the effect of that variable is not significant. Other valuable interpretation of the results of ANOVA is to identify the interaction of one variable with other (Gutierrez Pulido and Vara Salazar 2012), (Statgraphics Centurion 18 ®).
4 Results and discussion
4.1 Catalyst characterization
4.1.1 Catalyst particle size distribution and catalyst properties
Alumina fluidizable support particles determined the behavior of the catalysts studied. These were particles of Group B (Yang 2007) with a 73 μm particle size mode (Gonzalez et al. 2019), 725 kg/m3 apparent particle density. These particles fluidize well in the CREC Riser Simulator reactor. One should note that the addition of nickel being 5 wt% and of La or Ce, being 0.5 and 1.0 wt% does not significantly affect either particle size or particle density.
4.1.2 X-ray Diffraction
Figure 4 reports the XRD spectra of the reduced catalysts 5Ni1, 5Ni0.5Ce1, 5Ni0.5La1, 5Ni1La1 studied.
XRD diffractograms for the γ-Al2O3 Support and the catalysts: 5Ni1, 5Ni0.5Ce1, 5Ni0.5La1, 5Ni1La1.
X-ray powder diffraction patterns were obtained with D8 Advance, Brucker using Cu Kα radiation, filtered by Ni with a monochromatic signal, scan time constant of 2°/min. Figure 4 shows the XRD patterns for four catalysts, including that of γ-Al2O3 for comparison.
According to JCPDS 10-0425, the low intensity peaks centered at 2θ: 37.6, 45.8, 67.1 are the characteristic peaks of γ-Al2O3. The spectra for all the catalysts are similar to that of the support. It can be observed that the metals Ni, Ce and La were not detected, and their presence did not affect the phases of the gamma alumina, because they are below the detection limit (Mazumder and de Lasa 2014). It is observed that the diffraction patterns of the catalysts show a reduction in the intensities with respect to that of γ-Al2O3 peaks.
4.1.3 Specific surface area (BET)
Table 2 reports the specific surface area for various prepared catalysts. In this table, the influence of various promoters in the specific surface area is analyzed, keeping the added nickel loading at 5 wt% consistently. Gonzalez et al. (2019) reported that the specific surface area of γ-Al2O3 support prior to impregnation was 197 m2/g. Addition of 5% nickel reduces it to 126 and 115 m2/g for 5Ni4 and 5Ni1, respectively.
Superficial areas for the synthesized reduced catalysts.
| Specific surface area (m2/g) | ||
|---|---|---|
| wt % | La 1 | Ce 1 |
| 0.5 | 140 | 189 |
| 1 | 95 | 118 |
| 2* | 114* | 139* |
-
γ-Al2O3 specific surface area = 197 m2/g, 5Ni1* =115 m2/g; 5Ni4* = 126 m2/g, *Gonzalez et al. (2019).
Table 2 reports the effect of the promoters La or Ce, where the load of 1% reduces this specific surface area significantly, reporting the lowest values in the range 95–118 m2/g at pH 1. This reduction of the specific surface area is consistent with γ-Al2O3 smaller pores blocking as observed by Mazumder and de Lasa (2014) and Li, Hu, and Hill 2006. Similar results of specific area reduction were obtained by (Liu et al. 2014), suggesting that nickel species potentially block pores. Additionally, cerium provides for all the loading studied the larger specific surface area and this with respect to those for lanthanum addition.
4.1.4 Transmission electron microscopy (TEM) and energy – dispersive X-ray analysis (EDX)
Figure 5 provides typical TEM images and EDX plots for the catalysts of the present study after reduction.
TEM images of the reduced Ni catalyst A), C) and E) are micrographs for 5Ni1, 5Ni0.5Ce1 and 5Ni1La1, respectively. Meanwhile, B), D) and F) are EDX graphics for 5Ni1, 5Ni0.5Ce1 and 5Ni1La1, respectively.
It is observed in Figure 5A, the presence of nickel crystallites (black spots), with homogeneous distribution and Figure 5B (EDX) indicates the presence of nickel, oxygen and aluminum. Note, however, that the contrast between surface of Ni particles and that of the support is weak, indicating small and highly dispersed Ni crystallites (Figure 5A). Meanwhile, the micrographs of 5Ni0.5Ce1 (Figure 5C) show a better distribution of nickel crystallites than those of the 5Ni1 (Figure 5A). In Ce promoted catalysts (Figure 5C), Ni crystallites are quite apparent as shown as black dots. Furthermore, the presence of Ce is confirmed in EDX graphics (Figure 5D). Besides, for 5Ni1La1 (Figure 5E), it seems that nickel crystallites are less notorious than in the case of cerium, with the presence of lanthanum detected using EDX analysis (Figure 5F). Then, TEM-EDX confirms the presence of nickel, cerium and lanthanum on the surface of the support, which were unable to be detected using XRD.
Table 3 reports the results obtained with TEM-EDX.
EDX summary.
| Catalyst | Oxygen %wt | Aluminum %wt | Níckel %wt | Promoter %wt |
|---|---|---|---|---|
| 5Ni 1 | 51.9 | 46.4 | 1.6 | 0.0 |
| 5Ni0.5Ce1 | 60.0 | 37.5 | 1.9 | 0.6 |
| 5Ni1La1 | 80.1 | 17.2 | 2.4 | 0.3 |
It thus can be observed that the detected loadings by TEM-EDX of cerium and lanthanum were 0.6 and 0.3%, respectively. Discrepancies between the nominal and experimental values for nickel are in the range of 52–68%. Meanwhile, for the promoters, this difference for Ce is 20% and for La is 70%, respectively.
4.1.5 Determination of nickel load using atomic absorption
Table 4 reports the nickel weight percentage using the atomic absorption technique, for a nominal expected value of 5 wt% of nickel.
Weight percentage of nickel.
| Catalyst | wt% Nickel |
|---|---|
| 5Ni1 | 2.29 |
| 5Ni0.5La1 | 2.45 |
| 5Ni1La1 | 2.32 |
| 5Ni0.5Ce1 | 3.07 |
One can notice that the actual contents of nickel are smaller than the anticipated nominal values, being in all the cases this reduction the consequence of the somewhat limited incipient impregnation efficiency for depositing metals in gamma alumina supports.
Furthermore, one can also notice that the AA loadings (Table 4) are higher than those obtained with EDX (Table 3). This is consistent with AA providing an average nickel value in the entire catalyst. Meanwhile, the EDX analysis report a less trustable nickel content in the 1-μm layer close to the outer particle surface. Thus, AA analysis are more representative and reliable.
Another interesting result is the presence of nickel aluminate below 2 wt% nickel (Ewbank et al. 2015) and as well, (González Castañeda et al. 2019), did not detect nickel aluminates using a nickel–cerium co-impregnation and direct hydrogen reduction.
4.1.6 Temperature programmed reduction (TPR)
Figure 6 reports a typical TPR profile for the catalysts of the present study and those with 2% of Ce or La (González Castañeda et al. 2019), using a 10% H2 in argon flow of 50 cm3/min with a 10 °C/min ramp.
TPR profiles for the fresh untreated catalysts 5Ni1, 5Ni0.5La1, 5Ni1La1, and 5Ni0.5Ce1.
One can notice in Figure 6 that the recorded TPR, displays two characteristic peaks, with each one of these peaks centered at the temperatures, reported in Table 5.
Reduction temperatures from TPR peaks.
| Catalyst | First peak (°C) | Second peak (°C) |
|---|---|---|
| 5Ni1 | 294 | 477 |
| 5Ni0.5Ce1 | 260 | 432 |
| 5Ni0.5La1 | 260 | 472 |
| 5Ni1La1 | 242 | 453 |
| 5Ni4a | 300a | 420a |
| 5Ni2Ce4a | 380a | 550a |
| 5Ni2La1a | 380a | 520a |
In Figure 6 and Table 5, one can observe two TPR peaks. The second TPR peak was consistently recorded for 5 wt% Ni (coded 5Ni1), 0.5 wt% La, 1 wt% La and 0.5 wt% Ce, at temperatures lower than 500 °C. This allows to postulate that rare earths at these low loadings have an influence on the nickel reducibility at the origin of the second TPR peak. Also, in order to compare with the results of (González Castañeda et al. 2019), the TPR profiles with higher loading of 2 wt% of promoters, Ce and La were also included (5Ni2La1 and 5Ni2Ce4).
More specifically the catalyst 5Ni1 prepared via co-impregnation yields characteristic TPRs with: a) a first TPR peak associated to the thermal decomposition of the precursor into nickel oxide, b) a second TPR peak linked to the reduction of the nickel oxide into metallic nickel dispersed on octahedral alumina site. In this respect, the proposed catalyst preparation does not require Ni species being reduced at higher temperatures likely leading to tetrahedral alumina sites and NiAl2O4 (Li and Chen 1995; Li, Hu, and Hill 2006). Thus, the proposed direct reduction method without calcination, allows one to form metallic nickel species at lower reduction temperatures (Juan Juan 2009).
Regarding the Ni on alumina catalyst of the present study, one can record a significant concern from various authors about formation of unreactive nickel–aluminate species. Huang and Schwarz (1988) and Ewbank (2015) reported that at low Ni-loading, below 1 wt%, is difficult to reduce the formed NiAl2O4. Furthermore, in the 2-5 wt% Ni range (Huang 1988), observed the presence of both Ni and NiAl2O4. In contrast using the direct reduction preparation method of the present study, with 5%wt Ni, no NiAl2O4 was detected via TPR.
It is also interesting to observe that TPR for the catalyst 5Ni1, displays a 294 °C first peak corresponding to formed nitrous oxides from nitrate precursor thermal decomposition. This peak maximum is however observed, for the 0.5 wt% cerium loaded precursor (5Ni0.5Ce1) at 260 °C, that is lower than the 294 °C needed for 5Ni1. Thus, it is can be argued that the addition of Ce by co-impregnation weaken the interaction between Ni species and the alumina support, facilitating formation of active sites at decreased nitrate decomposition reduction temperature (Meng 2015). Figure 6 also reports a second TPR peak which can be attributed as stated above, to nickel oxide reduction (Konysheva 2013).
If one compares these low TPR peaks with those of cerium one can notice smaller TPR peaks for La, with this being true at both lower and higher La loadings. According to (Gobichon, Auffrédic, and Louër 1996), during of La(NO3)3 .6H2O thermal decomposition at 255 °C, lanthanum nitrate species dehydrate forming La(NO3)3. At this temperature, a first TPR peaks was observed for both 0.5 and 1.0% of La.
It is as well documented in the literature that the Ni(NO3)2·6H2O follows a different thermal decomposition than that of La(NO3)3·6H2O and Ce(NO3)3·6H2O. For Ni-nitrates, NO2 is generated via nickel nitrate decomposition
This postulated nitrate decomposition paths are in agreement with Ivanov et al. (2015) who claims Ce(NO3)3 forming CeO2 at 300 °C with NO2 being detected by the TCD. On the other hand, La-nitrates convert to very partially into La-oxides leading to a mixed phase LaONO3·xLa2O3 in the 320–470 °C with very limited NO2 detected by the TCD.
Figure 7 reports the variation of the maximum reduction temperatures for peaks 1 and 2, with the Ni on alumina catalyst having different lanthanum promoter loadings.
Effect of lanthanum load on the reduction temperature of peaks 1 and 2 for Nickel on alumina catalysts.
One can observe that Figure 7 shows similar trends for both first and second peaks when lanthanum concentration is increased. It appears in this respect, that a lowest temperature for both the first and second peak is observed at 1% La. On the other hand Gonzalez et al. (2019), reported a different trend for Ce. With respect to the non-promoted catalyst, the reduction temperature increases when 2% Ce was added. By other hand, when 0.5% Ce was added, the temperature was lower than that of non-promoted material (Table 5). One should notice that similar trends were observed for specific surface areas with a minimum value for both 1 wt% La and Ce.
4.1.7 Temperature programmed desorption (TPD)
Regarding the catalyst acid sites, they were measured using NH3-TPD peaks. Figure 8A reports NH3-TPD profiles for the different catalysts. One can see that the NH3-TPD profiles showed similar shapes, all of them with a single broad peak in the 100–600 °C range with maxima values in 310–320 °C. One can also observe that the 5Ni1 and 5Ni0.5La1 display almost the same NH3-TPD, meanwhile, the 5Ni1La1 and 5Ni0.5Ce1 show peaks of lower intensity, almost identical.
A) TPD profiles for different catalyst 5Ni1, 5Ni0.5La1, 5Ni1La1 and 5Ni0.5Ce1 supported on γ-alumina. B) An example of deconvolution and determination of acid sites corresponding to 5Ni1La1.
Figure 8B reports a TPD deconvolution for the 5Ni1La1. Gaussian functions were fitted to the obtained peaks and classified in three groups (Berteau and Delmon 1989): weak (25–200 °C), medium (200–400 °C) and strong (400–600 °C).
Table 6 reports the result of the peak deconvolution for 5Ni1, 5Ni0.5La1, 5Ni1La1 and 5Ni0.5Ce1 catalysts.
Distribution of acid sites (µmoles/g).
| Catalyst | Weak (25–200 °C) | Medium (200–400 °C) | Strong (400–600 °C) | Total |
|---|---|---|---|---|
| Alumina | 44 | 190 | 234 | |
| 5Ni1 | 56 | 216 | 272 | |
| 5Ni0.5La1 | 62 | 200 | 262 | |
| 5Ni1La1 | 70 | 180 | 250 | |
| 5Ni0.5Ce1 | 47 | 184 | 231 |
On the basis of Table 6, it appears that the medium strength acid sites are, in all cases, the most abundant ones. In this respect, it was also possible to observe that the nickel addition to alumina augments them in 13.7%, from 190 μmol/g for Al2O3 to 216 μmol/g for 5Ni1. Furthermore, it can also be noticed that La or Ce addition decreases the medium acid sites. This is desirable given the significant role of acid sites promoting coke formation (Mazumder and de Lasa 2015).
4.1.8 FTIR (fresh catalysts)
Figure 9 reports FTIR spectra with pyridine for the fresh catalysts after impregnation with the precursor Ni(NO3)2 and drying: 5Ni1, 5Ni0.5La1 and 5Ni0.5Ce1.
FTIR spectra of 5Ni1, 5Ni0.5La1 and 5Ni0.5Ce1, fresh and dried at 100 °C.
Figure 9 displays two bands, centered in the valleys at 1604 and 1379 cm−1, for 5Ni1. According to Chen and Song (1996) and He et al. (2015), those bands correspond to asymmetric and symmetric vibration of nickel nitrate, although their values are a little shifted to, 1620 and 1376 cm−1, respectively. One should mention that for 5Ni1, the 1379 cm−1 band is more notorious than for the other catalysts. This indicates the impregnation method was accurate and the nickel nitrate is present on the catalyst surface.
4.1.9 H2-chemisorption
Table 7 reports the hydrogen consumption with the percentage of reducible nickel.
Hydrogen consumption with the percentage of reducible nickel.
| Catalyst | H2 consumption (cc/g STP) | Reducible Ni loading (wt%) |
|---|---|---|
| 5Ni1 | 2.23 | 0.584 |
| 5Ni0.5Ce1 | 3.61 | 0.946 |
| 5Ni0.5La1 | 5.43 | 1.422 |
| 5Ni1La1 | 5.06 | 1.326 |
As it was also the case for the TPRs, reducible nickel at low temperatures was detected, even at low nickel loadings, with this being in line with Gil Calvo et al. (2017) and Qin et al. 2015. However, in the case of the present study, no irreducible aluminate species were detected at these lower nickel loading as found by Li, Hu, and Hill (2006). Regarding the 5Ni0.5La1 catalyst, it reported the highest hydrogen consumption and reducible Ni, while the 5Ni1 displayed the lowest, an important indicator of the value of adding lanthanum promoter.
Ewbank et al. (2015) state that varying only the preparation method, the interaction metal–support is affected. The controlled adsorption favors the strong metal–support interaction, meanwhile, dry impregnation, favors weak metal–support interactions. In our case, we used dry impregnation, which leaded to weak metal–support interactions and therefore the reduction temperatures are being reasonably lower.
Table 8 reports the results obtained using H2-chemisorption, nickel dispersion and Ni-crystallite size.
Nickel dispersion and Ni-crystallite size for the selected catalysts.
| Catalyst | Metal dispersion (%) | Ni-crystallite size (nm) |
|---|---|---|
| 5Ni1 | 3.40 ± 0.1 | 26.1 ± 0.70 |
| 5Ni0.5La1 | 8.12 ± 0.2 | 10.3 ± 0.65 |
| 5Ni1La1 | 9.24 ± 0.1 | 9.1 ± 0.27 |
| 5Ni0.5Ce1 | N.R. | N.R. |
| 5Ni2La1a | 2.86 ± 0.1 | 29.5 ± 1.05 |
-
a Gonzalez et al. (2019). N.R. is no reproducible.
In Table 8, one can observe that the addition of lanthanum as promoter increases the metal dispersion and decreases the nickel crystallite size. In a previous paper of our research team (Gonzalez et al. 2019), we reported results for 2% of lanthanum, and now, in this work, using lower loads of promoter, the results have improved inasmuch as the dispersions are still higher and the nickel crystallite sizes are lower, contributing to enhancing catalyst activity.
Regarding Ni crystallite, it is considered that smaller nickel crystallites are less intrusive in the γ-Al2O3 support, lessening pore blocking. For 5Ni0.5La1 (Table 2), the area 140 m2/g is bigger than that of 5Ni2La1, 114 m2/g (González Castañeda et al. 2019), which indicate that the smaller nickel crystallite sizes observed with low lanthanum increases, as expected, the specific surface area.
Table 8 also shows that for catalysts prepared with nickel and lanthanum, the dispersion of nickel crystallites varies significantly with promoter loading. Similar results were obtained with Ni and Ru (Calzada Hernández et al. 2020). Other benefits of the small nickel particles at low lanthanum loadings, is the decrease of coking as shown in Table 9. La addition appears in this respect to limit carbon nucleation, coke formation and filamentous carbon (Shang 2017), with CO2 conversion being also low (Xu and Jiang 2014).
Coke deposited on the catalysts 5Ni1, 5Ni0.5La1, 5Ni1La1 and 5Ni0.5Ce1 during the gasification of glucose, 600 °C, S/B = 1.
| Catalyst | Average weight (%) | STD |
|---|---|---|
| 5Ni1 | 0.27 | 0.022446 |
| 5Ni0.5La1 | 0.26 | 0.000253 |
| 5Ni1La1 | 0.29 | 0.000106 |
| 5Ni0.5Ce1 | 0.20 | 0.003756 |
| 5Ni4a | 0.32a | 0.001669a |
| 5Ni2La1a | 0.41a | 0.001718a |
| 5Ni2Ce4a | 0.33a | 0.004724a |
In summary, smaller crystallite sizes mean better Ni dispersion or less Ni species agglomeration (Iriondo et al. 2010), with the co-impregnation method considered as a result more favorable to form smaller Ni crystallites than consecutive impregnation methods (Meng et al. 2015).
4.2 Gasification results
Figure 10 reports the molar fractions of all the permanent gases obtained during the steam glucose gasification in the reactor CREC-Riser – Simulator, using the synthesized catalysts. The results of the catalysts with 2 % load (5Ni2La1, 5Ni2Ce1) and equilibrium model data are from Gonzalez et al. (2019).
Molar fractions of permanent gases for different catalysts reported in Table 1, at 600 °C and 20 s, S/B = 1, Cat/Biomass = 3.3. Data (*) are from Gonzalez et al. (2019).
Figure 10 shows a 0.51 hydrogen molar fraction at chemical equilibrium conditions (Gonzalez et al. 2019). This calculated hydrogen molar fraction is close to those for the catalytic runs. Regarding the H2 molar fraction for the thermal runs, it remains at 0.38 level which is certainly lower than for all catalyst tested. As well, promoters addition slightly influence H2 molar fractions with the best 0.58 H2 molar fraction observed for 5Ni0.5La1.
Furthermore, for the catalytic runs CO, CO2 and CH4 molar fractions were also close to chemical equilibrium predictions. The thermal experiments however, yielded higher CO and methane and lower H2 and CO2 than chemical equilibrium molar fractions. The observed CO molar fraction for thermal runs was 0.219 while 0.109 using nickel 5Ni1 and 0.068 with the 5Ni0.5La1. This yielded a valuable syngas with an 8.6 H2/CO ratio. These results are consistent with thermal runs dominated by primary reactions, with the catalyst having a considerable influence on the secondary reactions (Equations (1) and (2)). Secondary reactions tend to consume both methane and CO2 via dry reforming, as well as CO and CO2 via the Boudouard reaction (
4.2.1 Statistical analysis
The following sections report the variance analysis (ANOVA) of the collected experimental data and this to identify the variables which affect significantly the molar fractions of permanent gases and the interaction among the variables.
4.2.2 Analysis of hydrogen
Figure 11 reports the effect of type and load of promoter (lanthanum or cerium) on the hydrogen molar fractions.
Hydrogen molar fractions at: A) type of promoter, Ce and La, B) Different promoter loadings.
Note: 2% load data are from Gonzalez et al. (2019).
Figure 11A shows that lanthanum displays a significant larger hydrogen molar fractions than Ce. In the best-case lanthanum addition yielded 0.549 while Ce addition gave a 0,49. Figure 11B reports that increasing promoter loadings above 0.5%, invariably leads to lower hydrogen molar fractions.
Figure 12A reports the interaction between promoter type and promoter loading showing little interaction, as attested by the quasi-parallel trend of the reported lines.
Effect on hydrogen molar fractions of A) Interaction between loading and type of promoter. B) Residuals.
Note: 2% load data are from Gonzalez et al. (2019).
Figure 12B shows the residuals for repeats and this to confirm the random nature of the experimental data. The residuals random distribution confirms the correct random experimental error distribution, both for cerium and lanthanum. Thus, one can conclude that obtained data is not biased, with assumptions for repeat runs being respected.
4.2.3 Analysis of carbon monoxide
Figure 13 reports the interaction effects of promoter loading and promoter type on CO molar fraction.
Promoter effect on CO molar fraction evaluated using means and LSD (least significant difference): A) Promoter type, B) promoter loading.
Note: 2% loadings are from Gonzalez et al. (2019).
Thus, on this basis one can conclude that there is a significant CO molar fraction differences between Ce or La promoter, with no spans overlaps as shown in Figure 13A. Ce reports the biggest value. From Figure 13B, there are significant differences between 0.5 and 1.0% and also between 1.0 and 2%. The lowest value corresponds to 0.5%.
4.2.4 Analysis of methane
Figure 14 presents the effects of type and load of promoter.
Effect on CH4 molar fraction of A) type of promoter and B) load of promoter. Note: 2% load data are from Gonzalez et al. (2019).
For methane, the trend is very similar to that of carbon monoxide. There is a significant difference between the results with cerium and lanthanum (Figure 14A). Ce reports the biggest value. There is also a significant difference in the CH4 molar fractions between the catalysts with 0.5 and 1.0% promoter and also between 1.0 and 2% promoter (Figure 14B).
4.2.5 Analysis of CO2
Figure 15 presents the effect on the molar fractions of CO2 of type and load of promoter.
Effect on CO2 molar fractions of A) type of promoter and B) load of promoter. Note: 2% load data are from Gonzalez et al. (2019).
One can see that the effect of cerium is bigger than that of lanthanum and the difference is significant (Figure 15A). From Figure 15B, one can also observe that there are not significant differences among the different loads of promoters. CO2 is the only gas not affected by the load of promoters.
In this way, based on the gasification results and in the information of the catalyst characterization, it can be concluded that 5Ni0.5La1 and 5Ni1La1 were the best catalysts, producing the highest hydrogen with the lowest carbon monoxide amounts, yielding a H2/CO = 8.6 ratio. This was also valuable given this desirable high H2/CO ratio was obtained with very low methane levels. It is considered that these favorable chemical species distribution was achieved with catalyst promoting water gash shift, methane steam and dry reforming:
Furthermore, if one compares species molar fractions for the 5Ni1La1 and 5Ni0.5La1 catalysts of this study, with both equilibrium and data from (Gonzalez et al. 2019), one can observe in Figure 10, that the hydrogen molar fractions surpass reported values, as shown in Figure 10. These supra chemical equilibrium hydrogen molar fraction values, in excess to 5–14%, are considered to be the result of the postulated catalytic reactions not fully complying with chemical equilibrium.
4.2.6 Deposited coke
The following table reports the percentage of deposited coke on the used catalysts, after the experiments, calculated as (g of coke/g of catalyst) × 100%.
From Table 9, it can be observed that using with 0.5–1 wt% La or Ce loadings instead of 2 wt%, as reported in (Gonzalez et al. 2019), one can observe: a) for the La promoted catalyst there is a 37 wt% carbon reduction, when the La loading is decreased from 2 wt% to 0.5 wt%, b) for cerium, there is a 39 wt% carbon reduction when Ce loadings are diminished from 2 wt% to 0.5 wt% was reached. These findings are in line with those of Tomishige et al. 2007; Xu et al. (2009); Xu 2014, who determined carbon deposition percentages in the range of 1.4–4.3%, for the catalysts Ni/CeO2/Al2O3. For La promoter, these results are also in agreement with Mazumder and de Lasa (2015) who showed that La on Ni-alumina catalysts influence both Ni crystallite sizes and coke formation. In fact, using 0.5 and 1 wt% La catalyst for glucose gasification at 873 K, low carbon amounts were found (<0.3% of converted carbon).
In summary, the results of the present study, 0.5–1 wt% La promoted Ni on γ-Al2O3 catalysts with 140 m2/g specific surface area and 10 nm nickel crystallites, show a special promise for the biomass gasification with a resulting synthesis gas with high hydrogen content and low coke formation.
5 Conclusions
It is demonstrated that low loads of 0.5 or 1.0 wt% of lanthanum on a 5 wt% Ni/γ-Al2O3 fluidizable catalysts display a favorable effect on biomass conversion catalyst performance for hydrogen production.
It is proven that nickel and lanthanum precursor co-impregnation followed by direct reduction (free of oxygen atmosphere), leads to a synthesized catalyst with the desired surface properties.
It is shown that the prepared catalyst displays mildly an affected support specific surface area and pore size, moderate nickel crystallite sizes, good metal dispersion and medium NH3-TPD range acidity.
It is demonstrated that the prepared catalysts at the lowest La levels of 0.5 wt% exhibit in a fluidized CREC Riser Simulator excellent performance in terms of surrogate biomass gasification. These favorable performance indicators are given by the hydrogen molar fractions close to chemical equilibrium, as well as by the significantly reduced coke formation.
It is shown that a 5Ni0.5La1 catalysts leads to a highest hydrogen supra equilibrium molar fraction, with this being achieved with limited coke deposition.
It is proven using statistical analysis, that CO, CH4 and H2 molar fractions are influenced by both loading and type of promoter, with CO2 being affected by promoter type only, with negligible interaction between these factors.
Funding source: CONACYT-Mexico
Award Identifier / Grant number: CB-13-221690, 175112
Funding source: PRODEP-Mexico
Award Identifier / Grant number: 115725 UPZAC-004
Acknowledgments
This research was funded by CONACYT – Mexico-221690, Ciencia Básica – 2013. We would like to thank CONACYT-Mexico for the following scholarships awarded: (a) 175112 to ASE, (b) 586008 to DGGC, (c) 611972 to ICR, (d) 297035 to ARCH. As well, we express our appreciation to PRODEP for the 175725 (UPZAC-004) scholarship awarded to ASE. We also acknowledge the collaboration provided by Dr. Alfonso Talavera, Dr. Victor Baltazar, Dr. Gustavo Fuentes Zurita, Dr. Victor Manuel Castaño Meneses, Dr. Carlos Santolalla Vargas, C. a Dr. Juan Carlos Piña Victoria, Dr. Sergio Miguel Durón Torres, Dra. Maria Guadalupe Cárdenas Galindo and M. Sc. Mario Alberto Gómez Gallardo. Acknowledgments also go to Ms. Florencia de Lasa, who assisted with the editing of the present article.
-
Author contribution: All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.
-
Research funding: This research was funded by CONACYT – Mexico-221690, Ciencia Básica – 2013 and the scholarships awarded to ASE 175112 – CONACYT and 175725 – PRODEP (UPZAC-004).
-
Conflict of interest statement: The authors declare no conflicts of interest regarding this article.
References
Bartholomew, C. H., and R. J. Farrauto. 1976. “Chemistry of Nickel-Alumina Catalysts.” Journal of Catalysis 45 (1): 41–53, https://doi.org/10.1016/0021-9517(76)90054-3.Search in Google Scholar
Bartholomew, C. H., and R. J. Farrauto. 2006. Fundamentals of Industrial Catalytic Processes, 2nd ed. Wiley-Interscience.10.1002/9780471730071Search in Google Scholar
Calzada Hernández, A. R., D. G. González Castañeda, A. Sánchez Enriquez, H. de Lasa, and B. Serrano Rosales. 2020. “Ru-promoted Ni/γAl2O3 Fluidized Catalyst for Biomass Gasification.” Catalysts 10 (3): 1–18, https://doi.org/10.3390/catal10030316.Search in Google Scholar
Chan, F. L., and A. Tanksale. 2014. “Review of Recent Developments in Ni-based Catalysts for Biomass Gasification.” Renewable and Sustainable Energy Reviews 38: 428–38, https://doi.org/10.1016/j.rser.2014.06.011.Search in Google Scholar
Chen, J., and Q.Z. Song. 1996. Organic Spectral Analysis. Beijing China: BIT Press.Search in Google Scholar
de Lasa, H. I. 1992. Riser Simulator, US-Patent No. 5,102,628.Search in Google Scholar
de Lasa, H., E. Salaices, J. Mazumder, and R. Lucky. 2011. “Catalytic Steam Gasification of Biomass: Catalysts, Thermodynamics and Kinetics.” Chemical Reviews 111 (9): 404–33, https://doi.org/10.1021/cr200024w.Search in Google Scholar
Ewbank, J. L., L. Kovarik, F. Z. Diallo, and C. Sievers. 2015. “Effect of Metal-Support Interactions in Ni/Al2O3 Catalysts with Low Metal Loading for Methane Dry Reforming.” Applied Catalysis A: General 494: 57–67, https://doi.org/10.1016/j.apcata.2015.01.029.Search in Google Scholar
Gao, N., L. Aimin, C. Quan, Y. Qu, and L. Mao. 2012. “Characteristics of Hydrogen-Rich Gas Production of Biomass Gasification with Porous Ceramic Reforming.” International Journal of Hydrogen Energy 37: 9610–18, https://doi.org/10.1016/j.ijhydene.2012.03.069.Search in Google Scholar
Gil-Calvo, M., C. Jiménez-González, B. de Rivas, J. I. Gutiérrez-Ortiz, and R. López-Fonseca. 2017. “Effect of Ni/Al Molar Ratio on the Performance of Substoichiometric NiAl2O4spinel-Based Catalysts for Partial Oxidation of Methane.” Applied Catalysis B: Environmental 209: 128–38, https://doi.org/10.1016/j.apcatb.2017.02.063.Search in Google Scholar
Gobichon, A. E., J. P. Auffrédic, and D. Louër. 1996. “Thermal Decomposition of Neutral and Basic Lanthanum Nitrates Studied with Temperature-dependent Powder Diffraction and Thermogravimetric Analysis.” Solid State Ionics 93 (1–2): 51–64, https://doi.org/10.1016/s0167-2738(96)00498-5.Search in Google Scholar
González Castañeda, D. G., A. Sánchez Enríquez, I. Cruz Reyes, A. R. C. Hernández, and B. Serrano Rosales. 2019. “Catalytic Steam Gasification of Glucose for Hydrogen Production Using Stable Based Ni on a γ-Alumina Fluidizable Catalyst.” International Journal of Chemical Reactor Engineering: 1–17, https://doi.org/10.1515/ijcre-2019-0104.Search in Google Scholar
Gutierrez Pulido, H., and R. Vara Salazar. 2012. Analisis y Diseño de Experimentos, 3rd ed. Mexico DF.: Mc Graw Hill.Search in Google Scholar
He, S., L. Zhang, L. Mo, X. Zheng, H. Wang, Y. Luo, and S. He. 2015. “Ni/SiO2 Catalyst Prepared with Nickel Nitrate Precursor for Combination of CO2 Reforming and Partial Oxidation of Methane: Characterization and Deactivation Mechanism Investigation.” Journal of Nanomaterials 2015: 1–8, doi:https://doi.org/10.1155/2015/659402.Search in Google Scholar
Huang, Y., and J. A. Schwarz. 1987. “The Effect of Catalyst Preparation on Catalytic Activity: III. The Catalytic Activity of Ni/A12O3 Catalysts Prepared by Incipient Wetness.” Applied Catalysis 32: 45–57, https://doi.org/10.1016/s0166-9834(00)80615-6.Search in Google Scholar
Huang, Y., and J. A. Schwarz. 1988. “Effect of Catalyst Preparation on Catalytic Activity V. Chemical Structures on Nickel / Alumina Catalysts.” Applied Catalysis 36: 163–75, https://doi.org/10.1016/s0166-9834(00)80112-8.Search in Google Scholar
Iriondo, A., V. L. Barrio, J. F. Cambra, P. L. Arias, and M. B. Guemez. 2010. “Glycerol Steam Reforming over Ni Catalysts Supported on Ceria and Ceria-Promoted Alumina.” International Journal of Hydrogen Energy 35 (20): 11622–33, https://doi.org/10.1016/j.ijhydene.2010.05.105.Search in Google Scholar
Ivanov, B., I. Spassova, M. Milanova, G. Tyuliev, and M. Khristova. 2015. “Effect of the Addition of Rare Earths on the Activity of Alumina Supported Copper Cobaltite in CO Oxidation, CH4 Oxidation and NO Decomposition.” Journal of Rare Earths 33 (4): 382–90, https://doi.org/10.1016/S1002-0721(14)60430-0.Search in Google Scholar
José-Alonso, D. S., M. J. Illán-Gómez, and M. C. Román-Martínez. 2013. “Low Metal Content Co and Ni Alumina Supported Catalysts for the CO2 reforming of Methane.” International Journal of Hydrogen Energy 38 (5): 2230–9, https://doi.org/10.1016/j.ijhydene.2012.11.080.Search in Google Scholar
Juan-Juan, J., M. C. Román-Martínez, and M. J. Illán-Gómez. 2009. “Nickel Catalyst Activation in the Carbon Dioxide Reforming of Methane. Effect of pretreatments.” Applied Catalysis A: General 355 (1–2): 27–32, https://doi.org/10.1016/j.apcata.2008.10.058.Search in Google Scholar
Konysheva, E. Y. 2013. “Reduction of CeO2 in Composites with Transition Metal Complex Oxides under Hydrogen Containing Atmosphere and its Correlation with Catalytic Activity.” Frontiers of Chemical Science and Engineering 7 (3): 249–61, https://doi.org/10.1007/s11705-013-1333-6.Search in Google Scholar
Li, C., and Y. W. Chen. 1995. “Temperature-Programmed-Reduction Studies of Nickel Oxide/Alumina Catalysts: Effects of the Preparation Method.” Thermochimica Acta 256 (2): 457–65, https://doi.org/10.1016/0040-6031(94)02177-.Search in Google Scholar
Li, G., L. Hu, and J. M. Hill. 2006. “Comparison of Reducibility and Stability of Alumina-Supported Ni Catalysts Prepared by Impregnation and Co-precipitation.” Applied Catalysis A: General 301: 16–24, https://doi.org/10.1016/j.apcata.2005.11.013.Search in Google Scholar
Liu, Q., J. Gao, M. Zhang, H. Li, F. Gu, G. Xu, Z. Zhong, and F. Su. 2014. “Highly Active and Stable Ni/γ-Al2O3 Catalysts Selectively Deposited with CeO2 for CO Methanation.” RSC Advances 4 (31): 16094–103, https://doi.org/10.1039/c4ra00746h.Search in Google Scholar
Mazumder, J., and H. De Lasa. 2014. “Fluidizable Ni/La2O3-γAl2O3 Catalyst for Steam Gasification of a Cellulosic Biomass Surrogate.” Applied Catalysis B: Environmental 160–161 (1): 67–79, https://doi.org/10.1016/j.apcatb.2014.04.042.Search in Google Scholar
Mazumder, J., and H. I. de Lasa. 2015. “Fluidizable La2O3 Promoted Ni/γ-Al2O3 Catalyst for Steam Gasification of Biomass: Effect of Catalyst Preparation Conditions.” Applied Catalysis B: Environmental 168–169 (C): 250–65, https://doi.org/10.1016/j.apcatb.2014.12.009.Search in Google Scholar
Mazumder, J., and H. I. de Lasa. 2018. “Steam Gasification of a Cellulosic Biomass Surrogate Using a Ni/La2O3-ΓAl2O3 Catalyst in a CREC Fluidized Riser Simulator. Kinetics and Model Validation.” Fuel 216 (November 2017): 101–9, https://doi.org/10.1016/j.fuel.2017.11.074.Search in Google Scholar
Meng, F., Z. Li, J. Liu, X. Cui, and H. Zheng. 2015. “Effect of Promoter Ce on the Structure and Catalytic Performance of Ni/Al2O3 Catalyst for CO Methanation in Slurry-Bed Reactor.” Journal of Natural Gas Science and Engineering 23: 250–8, https://doi.org/10.1016/j.jngse.2015.01.041.Search in Google Scholar
Moghadam, R. A., S. Yusup, Y. Uemura, B. L. F. Chin, H. L. Lam, and A. Al Shoaibi. 2014. “Syngas Production from Palm Kernel Shell and Polyethylene Waste Blend in Fluidized Bed Catalytic Steam Co-gasification Process.” Energy 75: 40–4, https://doi.org/10.1016/j.energy.2014.04.062.Search in Google Scholar
Moilanen, A., M. Nasrullah, and E. Kurkela. 2009. “The Effect of Biomass Feedstock Type and Process Parameters on Achieving the Total Carbon Conversion in the Large Scale Fluidized Bed Gasification of Biomass.” Environmental Progress & Sustainable Energy 28 (3): 355–9, https://doi.org/10.1002/ep.10396.Search in Google Scholar
Nemtsov, D. A., and A. Zabaniotou. 2008. “Mathematical Modelling and Simulation Approaches of Agricultural Residues Air Gasification in a Bubbling Fluidized Bed Reactor.” Chemical Engineering Journal 143: 10–31, https://doi.org/10.1016/j.cej.2008.01.023.Search in Google Scholar
Qin, Z., J. Ren, M. Miao, Z. Li, J. Lin, and K. Xie. 2015. “The Catalytic Methanation of Coke Oven Gas over Ni-Ce/Al2O3 Catalysts Prepared by Microwave Heating: Effect of Amorphous NiO Formation.” Applied Catalysis B: Environmental 164: 18–30, https://doi.org/10.1016/j.apcatb.2014.08.047.Search in Google Scholar
Quddus, M. R. 2013. A Novel Mixed Metallic Oxygen Carrier for Chemical Looping Combustion: Preparation, Characterization & Kinetic Modeling. The University of Western Ontario.Search in Google Scholar
Ren, J., J. Cao, X. Zhao, F. Yang, and X. Wei. 2019. “Recent Advances in Syngas Production from Biomass Catalytic Gasification: A Critical Review on Reactors, Catalysts , Catalytic Mechanisms and Mathematical Models.” Renewable and Sustainable Energy Reviews 116 (October): 109426, https://doi.org/10.1016/j.rser.2019.109426.Search in Google Scholar
Ren, S.-B., J.-H. Qiu, C.-Y. Wang, B.-L. Xu, Y.-N. Fan, and Y. Chen. 2007. “Dipersion State of Nickel Ions on Gamma Al2O3 and Catalytic Activity of Derived Nickel Catalysts for Hydrogenation of Alpha Pinene.” Chinese Journal of Inorganic Chemistry 23 (6): 1021–8.Search in Google Scholar
Rostrup-Nielsen, J. R., and J.-H. Bak Hansen. 1993. “CO2 - Reforming of Methane over Transition Metals.” Journal of Catalysis 144: 38–49, https://doi.org/10.1006/jcat.1993.1312.Search in Google Scholar
Shahbaz, M., S. yusup, A. Inayat, D. O. Patrick, and M. Ammar. 2017. “The Influence of Catalysts in Biomass Steam Gasification and Catalytic Potential of Coal Bottom Ash in Biomass Steam Gasification: A Review.” Renewable and Sustainable Energy Reviews 73 (January): 468–76, https://doi.org/10.1016/j.rser.2017.01.153.Search in Google Scholar
Sutton, D., B. Kelleher, and J. R. H. Ross. 2001. “Review of Literature on Catalysts for Biomass Gasification.” Fuel Processing Technology 73: 155–73, https://doi.org/10.1016/S0378-3820(01)00208-9.Search in Google Scholar
Tomishige, K., T. Kimura, J. Nishikawa, T. Miyazawa, and K. Kunimori. 2007. “Promoting Effect of the Interaction between Ni and CeO2 on Steam Gasification of Biomass.” Catalysis Communications 8 (7): 1074–9, https://doi.org/10.1016/j.catcom.2006.05.051.Search in Google Scholar
Webb, P. A. 2003. Introduction to Chemical Adsorption Analytical Techniques and Their Applications to Catalysis. 1–12. Micromeritics Instrument Corp. Technical Publications.Search in Google Scholar
Xu, X., and E. Jiang. 2014. “Hydrogen from Wood Vinegar via Catalytic Reforming over Ni/Ce/γ-Al2O3 Catalyst.” Journal of Analytical and Applied Pyrolysis 107: 1–8, https://doi.org/10.1016/j.jaap.2013.12.003.Search in Google Scholar
Xu, J., W. Zhou, Z. Li, J. Wang, and J. Ma. 2009. “Biogas Reforming for Hydrogen Production Over Nickel and Cobalt Bimetallic Catalysts.” International Journal of Hydrogen Energy 34: 6646–54, doi:https://doi.org/10.1016/j.ijhydene.2009.06.038.Search in Google Scholar
Yung, M. M., W. S. Jablonski, and K. A. Magrini-bair. 2009. “Reviews Citations Tar Reforming and Catalytic Candle Filters Review of Catalytic Conditioning of Biomass-Derived Syngas.” Energy & Fuels 23 (4): 1874–87, https://doi.org/10.1021/ef800830n.Search in Google Scholar
© 2020 Walter de Gruyter GmbH, Berlin/Boston
