Analysis of Ni species formed on zeolites, mesoporous silica and alumina supports and their catalytic behavior in the dry reforming of methane
The presented investigation is focused on the analysis of Ni species formed on microporous (zeolites MFI and FAU) and mesoporous materials (Al-MCM-41 and SBA-15) and alumina supports and their catalytic behavior in the dry reforming of methane. The paper lays emphasis on the relationship between the catalytic behavior of Ni-based catalysts and their textural/structural properties. Ni-based catalysts were prepared by wet impregnation (11 wt% of Ni) followed by calcination in air and reduction in hydrogen. The properties of Ni-based catalysts were also compared prior and after the catalytic tests. The critical role was played by the high value of the specific surface area and the high strength of the interaction between the Ni species and the support, which both determined the high dispersion and stability of metal Ni0 particles. Ni–Al–MCM-41 and Ni–SBA-15 showed the values of the conversion of CO2 and CH4 above 90% (stable during 12 h). Slightly lower values of the conversion of CO2 and CH4 were observed over Ni–Al2O3 (also stable during 12 h). In contrast to these materials, Ni–MFI and Ni–FAU exhibited the worse metallic Ni0 particles dispersion and very bad catalytic behavior.
KeywordsDry reforming Methane Nickel Catalysis Particle size
Supported noble metal catalysts (Rh, Ru, Ir, Pt and Pd) have efficient catalytic performance and low sensitivity to carbon deposits, but their high price and unavailability prevent their industrial application .
Ni-based catalysts represent a promising alternative. This is mainly due to their low cost and high availability. Several papers and reviews dealing with the comparison of varied supports of Ni species were published (e.g. [6, 7, 8]). There is a common agreement that DRM is a structure sensitive reaction, therefore Ni particle size and the metal-support interaction play the crucial role in the activity and the stability of Ni-based catalysts in the DRM reaction. Although high catalytic activity has been reported for various Ni-based catalysts in the DRM reaction, the deactivation of Ni species by coke deposits represents a problem which has to be solved . Recently, the choice of suitable support [9, 10, 11, 12], the addition of promoting metals (e.g. La  or Ce ) and the conditioning of catalyst treatment  have brought new possibilities of the development of active and stabile Ni-based catalysts.
Both Ni–Al2O3 and Ni–SiO2 were successfully published as suitable catalysts used in DRM. When alumina or silica is chosen as supporting materials of Ni species, their thermal treatment results in the formation of NiAl2O4 spinels or Ni phyllosilicates (Si4Ni3O10(OH)2 or Si2Ni3O5(OH)4) [6, 16, 17]. NiAl2O4 spinels and Ni phyllosilicates are subsequently reduced to metal Ni0 particles, which are well dispersed on the supports [18, 19]. Recently, zeolites [20, 21, 22, 23, 24, 25, 26, 27, 28] and mesoporous silica materials [29, 30, 31, 32, 33] have been intensively studied as attractive supports of Ni species for DRM. The advantage of zeolites is their high sorption affinity and capacity for CO2 [24, 34]. While commonly tested Ni-alumina successfully works in DRM at temperatures higher than 700 °C, Ni-zeolites have been reported to be significantly active in DRM even at temperatures lower than 700 °C (e.g. 20% conversion of methane over Ni–MFI at 500 °C [24, 26]). However, the most active Ni-zeolite based catalysts concurrently show the highest amount of formed carbon deposits . The advantages of mesoporous MCM-41 and SBA-15 materials are large surface areas and pore dimensions in units of nanometers that lead to the well dispersed metallic Ni0 particles and prevent sintering effect . Nevertheless, there were also reports of the destruction of the structure of mesoporous silica based materials at higher temperature . Simultaneously, the papers are mainly focused on improving of the catalytic activity of Ni-mesoporous silica based catalysts by addition of other metals (Co, La, Mg) [30, 36, 37].
The aim of this study is the direct comparison of microporous materials (zeolites MFI and FAU), mesoporous silica (SBA-15), mesoporous aluminosilicate (Al–MCM-41) and alumina as the supports of Ni species for DRM. All studied Ni-based catalysts were prepared by the wet impregnation method with the constant Ni loading (11 wt% of Ni) that was chosen based on our previous research dealing with Ni-alumina catalysts . In detail, we focused on the relationship between the catalytic behavior of Ni-based catalysts and their structural/textural properties and on the deactivation of individual catalysts.
MFI (ACS materials, >400 m2 g−1), FAU (Zeolyst, >750 m2 g−1), Al-MCM-41 (ACS materials, >800 m2 g−1), SBA-15 (ACS materials, >480 m2 g−1) and Al2O3 (NanoScale, >250 m2 g−1) were used as supporting materials. Ni–MFI (Si/Al 12.5), Ni–FAU (Si/Al 2.6), Ni–Al–MCM-41 (Si/Al 25), Ni–SBA-15 (silica) and Ni–Al2O3 catalysts with 11 wt% Ni were prepared by impregnating of support with solution of ethanol and Ni(NO3)2·6H2O. After impregnation, the materials were dried at room temperature overnight and calcined at 700 °C in the flow of air for 5 h. Such prepared catalysts were denoted as “as-prepared” Ni-based catalysts.
Before catalytic experiments, all as-prepared Ni-catalysts were in situ treated with a mixture of H2 (5%) in Ar (50 mL/min) for 1 h at 750 °C. Such treated catalysts were denoted as “reduced” catalysts.
As-prepared catalysts were characterized by XRD, H2-TPR, BET, SEM and TGA. The reduced and the used catalysts were characterized by N2-adsorption, XRD, H2-TPR, TGA and SEM. In order to prevent the re-oxidation of metallic Ni particles in such catalysts, the catalysts were characterized immediately after their reduction or catalytic testing.
Here 0.89 is the shape factor for spherical particle, D means the crystallite size (nm), λ is X-ray wavelength of used X-ray radiation (0.154056 nm), β is FWHM value (rad) and θ is the diffraction angle (rad).
H2-TPR profiles of the as-prepared and reduced Ni-based catalysts were monitored by the AutoChem II 2920 (Micromeritics, USA). A quartz microreactor was charged with 100 mg of catalysts. The reduction was carried out from room temperature up to 1050 °C, with a heating rate of 10 °C min−1 in the flow of 5% H2/Ar gas (50 cm3 min−1). The changes in the hydrogen concentration were monitored by TCD detector.
The BET surface area and pore diameter distribution were determined for as-prepared and reduced Ni-based catalysts according to the N2 adsorption isotherms using an ASAP 2020 instrument (Micromeritics, USA) and evaluated by MicroActive Software (Micromeritics, USA). Before each adsorption measurements, the sample was degassed to allow a slow removal of the of pre-adsorbed water at low temperatures. This was done to avoid potential structural damage of sample due to the surface tension effects and hydrothermal alternation. Starting at ambient temperature, the sample was degassed at 110 °C (temperature ramp of 0.5 °C min−1) until the residual pressure of 1 Pa was attained. After that, the sample was heated at constant temperature 110 °C for 1 h, followed by subsequent increase of temperature from 110 °C up to 200 °C (temperature ramp of 1 °C min−1). The sample was degassed at this temperature under a turbo molecular pump vacuum for 8 h. The specific surface area was calculated according to BET method. The total surface area was determined by means of t-plot using Harkins–Jura equation for calculation of adsorbed layer thickness. The pore volume and pore size distribution were determined by NL DFT approach by using the “N2 @ 77 K” model for cylindrical pores and oxide surface.
SEM images were measured by a JSM-7500F microscope with cold field emission source (JEOL, Japan).
The thermal behavior of used catalysts was characterized by thermogravimetric analyzer TA SDT Q600. The experiments were carried out under a dynamic atmosphere from 25 to 900 °C (heating rate of 10 °C min−1) in the flow of air (flow rate 100 cm3 min−1). The sample mass was 5 mg. Results of measurements were evaluated online using SDT Q600 software.
The DRM reaction was carried out in a fixed-bed tubular reactor. 40 mg of catalyst was placed between two layers of quartz wool. The as-prepared catalysts were first in situ treated with a mixture of H2 (5%) in Ar (50 mL min−1) during 1 h at 750 °C. After that, the catalyst was cooled down under reducing atmosphere to 60 °C. Then the catalyst was put in the contact with the reaction mixture and heated up to 750 °C with a heating rate of 1 °C min−1. At 750 °C, the reaction was monitored for 12 h. The CH4 and CO2 reactants were mixed at a volumetric ratio of 1 and they were diluted in helium (10:10:80 in vol.). The gas hourly space velocity (GHSV) was 150,000 L kg−1 h−1. The analysis of reactants was done by a gas chromatograph (Varian CP-3800) with a thermal conductivity detector (TCD) and two in series columns (Molecular Sieve, 5A, Porapak Q).
Characterization of calcined Ni-based catalysts
Textural data of studied Ni catalysts
After 5 h calcination at 700 °C in air
After 1 h reduction in H2/Ar (5%) at 750 °C
Chen et al.  and Saha et al.  pointed out the relation between the ratio of pore volume and specific surface area (Vmicro/SBET) and the dispersion of Ni particles. Although it should be stressed that there is no obvious connection between the value of the specific surface area and the pore volume, it is interesting that both groups reported the direct connection between the dispersion of Ni species, the catalytic activity and the value of Vmicro/SBET ratio. Both groups reported that higher value of Vmicro/SBET ratio results in higher dispersion of nickel particles on the support and subsequently in high catalytic activity in dry reforming of methane [29, 38]. In our case (Table 1), Ni–Al2O3 (0.18 nm), Ni–SBA-15 (0.15 nm), Ni–Al–MCM-41 (0.11 nm) exhibited significantly higher values of Vmicro/SBET ratios in comparison to Ni–MFI (0.051 nm) and Ni–FAU (0.038 nm) catalysts.
The H2-TPR profiles of Ni–MFI, Ni–FAU, Ni–Al–MCM-41 and Ni–SBA-15 catalysts are significantly more complicated. In principal, the H2-TPR profile of any Ni-based material could be separated into several regions. The reduction peaks below 500 °C have been ascribed to weakly interacting NiO particles on the external surface of the supporting material [29, 46]. The reduction peaks between 500 °C and 600 °C are usually ascribed to the reduction of strongly interacting NiO species present on the internal surface of the support [16, 39, 40, 41, 47]. Reduction peaks with the maximum above 600 °C are mostly assigned to the reduction of Ni2+ species in the form of cationic sites or compound-like surface complexes as Ni aluminates or Ni phyllosilicates, depending on the character of supporting material [6, 16, 17, 48].
The H2-TPR profile of Ni–MFI exhibited (i) the reduction peaks with a maximum at 280 and 450 °C, both reflecting the presence of NiO particles with varied strength of their interaction with zeolite matrix and (ii) the reduction peak at 640 °C, probably reflecting the presence of Ni2+ cationic cites or Ni phyllosilicates . The broad reduction peak between 800 and 1100 °C could be assigned to the dihydroxylation of zeolite matrix . The H2-TPR profile of Ni–FAU exhibited dominant reduction peaks at 540 and 610–640 °C with a shoulder up to 900 °C. These reduction peaks could be attributed to the presence of NiO species, Ni2+ cationic cites and Ni phyllosilicates, but it is hard to separate them. In addition, we calculated the theoretical maximum amount of Ni that could be presented as Ni2+ at the cationic sites in the case of MFI and FAU zeolites assuming the maximum theoretic value of Ni/Al molar ratio to be 0.5. These values were 3.6 wt% Ni in Ni–MFI with Si/Al 12.5, and 13.6 wt% Ni in Ni–FAU with Si/Al 2.6. From that point of view, used Ni–FAU and Ni–MFI zeolites with 11 wt% of Ni represented the materials with the amount of nickel which could be either over the maximum of ionic exchange level (Ni–MFI) or below that level (Ni–FAU). However, the formation of NiO particles with varied strength of their interaction with zeolite matrix means that any Ni–MFI and Ni–FAU did not contain the exclusive presence of Ni species at the cationic sites.
In contrast to Ni–MFI and Ni–FAU, the reducibility of Ni–Al–MCM-41 and Ni–SBA-15 was shifted to higher temperatures. H2-TPR profiles of both Ni–Al–MCM-41 and Ni–SBA-15 materials exhibit an intensive reduction peak with a maximum between 700 and 730 °C, which should correspond to the reduction of Ni phyllosilicates [16, 17]. The lower intensity reduction peak with a maximum at 470 °C could be ascribed to the reduction of strongly interacting NiO [29, 46]. With respect to XRD, the Ni phyllosilicates, expected to be present on Ni–Al–MCM-41 and Ni–SBA-15, have to be present in amorphous phase.
Characterization of reduced Ni-based catalysts
As DRM was carried out over reduced forms of Ni-based catalysts (reduction proceeds at 750 °C in the flow of 5% H2/Ar for 1 h), the reduced forms of Ni-based materials were characterized by N2-adsorption, XRD, H2-TPR and SEM.
The reduction of Ni–MFI and Ni–FAU led to the decrease of specific surface area (both ca 27%), internal micropore surface area (both ca 27%) and volume of micropores (42% for Ni–MFI and 77% for Ni–FAU) in contrast to as-prepared Ni-based catalysts. On the other hand, a marginal decrease of the specific surface area was observed after the reduction of Ni–Al–MCM-41 and Ni–Al–SBA-15 materials. In the case of Ni–Al2O3, the reduction also led to the decrease of the specific surface area (ca 27%), but there was still quit high specific surface area available to the dispersion of nickel species.
Average metal Ni0 particles size in the catalysts calculated from the diffraction peak at 2θ = 44.5° in XRD patterns using the Scherrer equation
Relative amount of metal Ni0a (%)
Ni particle size (nm)
Contrary to these micro and mesoporous materials, the reduction of Ni–Al2O3 did not lead to the catalyst with maximum amount of metallic Ni0 particles. It is already evident from the H2-TPR profile of as-prepared Ni–Al2O3, which exhibited one reduction peak with the maximum at 865 °C (Fig. 2). In agreement, the H2-TPR profile of reduced Ni–Al2O3 (at 750 °C in the flow of 5% H2/Ar for 1 h) exhibits a residual, less intensive reduction peak with a maximum at 863 °C (see Fig. 3, right side). Comparing the total areas of H2-TPR profiles of as-prepared and reduced Ni–Al2O3, it was evaluated that only 70% of the total amount of nickel presented on as-prepared Ni–Al2O3 was reduced into the form of metallic Ni0 particles after 1 hour reduction at 750 °C in flow of 5% H2/Ar. The absence of diffraction lines of some residual non-reduced Ni species (Fig. 1, see lines denoted as b) could be clearly explained by the simultaneous absence of such diffraction lines even in diffractograms of as-prepared Ni–Al2O3 (before reduction). Although some small part of residual “non-reduced” nickel species could be subsequently reduced during the reaction, it should be stressed that the reaction conditions of DRM are less reducing than hydrogen at the same temperature .
DRM and the parameters affecting the activity of Ni-based catalysts
It is evident that the highest conversions of CO2 and CH4 were observed for Ni–Al–MCM-41 and Ni–SBA-15 catalysts. The conversions of both reaction gases simultaneously reached values at around 80% at 600 °C and even at around 95% at 750 °C. The conversions of CO2 (95%) and CH4 (92%) were almost constant for 12 h reaction at 750 °C. Ni–Al2O3 exhibited slightly lower conversions of CO2 (90% at 750 °C) and CH4 (87% at 750 °C), but both these values were also stable for the studied 12 h. The slightly lower activity of Ni–Al2O3 in contrast to Ni–Al–MCM-41 and Ni–SBA-15 catalysts (Fig. 5) could be simply explained by the lower amount of metallic Ni0 particles presented in the reduced Ni–Al2O3 than in reduced Ni–Al–MCM-41 and Ni–SBA-15 materials, as it was mentioned above (Table 2).
On the other hand, Ni–MFI and Ni–FAU zeolites exhibited significantly worse catalytic behavior in comparison with Ni–Al–MCM-41, Ni–SBA-15 and Ni–Al2O3. For Ni–MFI, initial conversions of CH4 and CO2 reached 64 and 72% at 750 °C. After that, these values slowly decreased during 12 h of the reaction. Ni–FAU was already deactivated during increasing reaction temperature (500 °C). The decrease in the activity of Ni–FAU followed by the increase of the pressure drop subsequently led to the total suppression of the reaction.
The values of the initial conversions of CH4 and CO2 were clearly connected with the dispersion of the metallic Ni0 particles, i.e. with the values of the average size of metallic Ni0 particles. It is evident that the value of the average size metallic Ni0 particles originated from the textural properties of supports, more specifically the high value of specific surface area leading to the high dispersion of nickel species. Reduced Ni–Al–MCM-41, Ni–SBA-15 and Ni–Al2O3 catalysts exhibited high conversions of CH4 and CO2 as well as high values of specific surface areas that led to the low average size of formed metallic Ni0 particles (9–11 nm, Table 2). In contrast, reduced Ni–FAU and Ni–MFI zeolites contained large metallic Ni0 particles (23 and 29 nm, Table 2). It is generally accepted that such large metallic Ni0 particles are not efficient in the dry reforming reaction . Metallic Ni0 particles were formed exclusively after the reduction of Ni-catalysts. It is quite impossible to describe the position of metallic Ni particles formed on the individual supports, but the difference in the size of the formed metallic Ni particles is evident (Table 2). Luengnaruemitchai et al. reported the highest catalytic activity of Ni–Y zeolite with 7 wt% Ni (materials with the Ni loading from 3 to 7 wt% were studied), it seems that the 11 wt% of Ni that we used in this work is already beyond the optimal value of Ni loading suitable for zeolites (even in the case of Ni–FAU). Although it is not possible to determine this, it can be expected that large part of Ni species were located on internal surface of zeolite materials. The reduction of such Ni species led to the formation of metallic Ni0 particles, which could subsequently block the micropores of zeolite materials as was observed in the case of Ni–FAU.
Comparing the catalytic behavior of Ni–FAU (Si/Al molar ratio 2.6) versus Ni–MFI (Si/Al molar ratio 12.5) and the catalytic behavior of Ni–Al–MCM-41 (Si/Al molar ratio 25) versus Ni–SBA-15 (pure silica), it was not affected by the value of Si/Al molar ratio. Thus, the presence of well dispersed metallic Ni0 particles was more important in DRM regardless their formation on pure silica or aluminosilica support.
Characterization of Ni-based catalysts after the reaction and the parameters affecting the stability of Ni-based catalysts
Fig. 1c shows the XRD patterns of used Ni–Al–MCM-41, Ni–SBA-15, Ni–Al2O3, Ni–MFI and Ni–FAU catalysts after catalytic test. First, the diffraction lines corresponding to the presence of metallic Ni0 particles (2θ = 44.5°, 52° and 76.5°) were evaluated to determine the average size of metallic Ni0 particles after the reaction (calculated from diffraction line 2θ = 44.5°). The change in the average size of metallic Ni0 particles during the reaction is critical in the discussion of sintering effects . It is important to stress that Ni–Al–MCM-41, Ni–SBA-15 and Ni–Al2O3 exhibited marginal change of the average size of Ni0 particles during 12 h of reaction (Table 2). This is in agreement with the preserved values of the conversions of CH4 and CO2 (Fig. 5). In contrast, a significant (ca 50%) increase of the average size of Ni0 particles was observed for used Ni–MFI and Ni–FAU zeolites. The average size of Ni0 particles increased from 29 to 45 nm for Ni–MFI and from 23 to 34 nm for Ni–FAU. This change in the average size of Ni0 particles was clearly associated with the dramatic decrease in the conversions of CH4 and CO2 (Fig. 5). It also should be mentioned that new diffraction lines at 2θ = 37°, 43° and 63° appeared in the diffraction pattern of used (after reaction) Ni–FAU. These lines correspond to the presence of re-oxidized NiO particles. Nevertheless, their formation could be connected with the untimely stop of the reaction, which may originate from the increase of pressure drop as it was mentioned above (Fig. 5).
Second, new diffraction lines appeared in the diffraction patterns of all used Ni-based catalysts at 26° and 43.1°. These diffraction lines could be attributed to the presence of crystalline carbon . These diffraction lines were the most evident in diffraction pattern of used Ni–Al–MCM-41 and Ni–SBA-15. The lower intensities of these lines were observed in the case of Ni–Al2O3 and Ni–MFI materials. XRD pattern of Ni–FAU did not observed any diffraction lines corresponding to the crystalline carbon.
The SEM images (Fig. 3, right side column) show the presence of a filamentous (crystalline) structure of carbon deposits on Ni–MFI, Ni–Al–MCM-41, Ni–SBA-15 and Ni–Al2O3 catalysts after the catalytic test. In contrast, the SEM image of Ni–FAU does not show the presence of crystalline carbon structure, but very large Ni particles are present on external surface of the material. The size of metal Ni0 particles is significantly higher in used Ni–FAU than in reduced Ni–FAU. These observations are in a good agreement with the XRD results.
A relatively high amount of filamentous carbon on Ni–Al–MCM-41 (Si/Al 25) and Ni–SBA-15 catalysts (in contrast to other studied catalysts) did not result in any change of the average size of metallic Ni0 particles and the values of CH4 and CO2 conversions. This observation is in agreement with the fact that the ability to keep the size of Ni0 particles (sintering effect) during the DRM reaction depends on the type of carbon formed during the reaction. While the formation of filamentous carbon have been reported to be associated with no significant change of the size of Ni0 particles as well as with the low decrease of activity [53, 54, 55, 56], the formation of amorphous carbon (not formed on the above mentioned catalysts) has been reported on large Ni0 particles during DRM [53, 57, 58]. Simultaneously, it was reported that it is the amorphous carbon and not the crystalline form which caused the blocking the catalyst and consecutive suspension of the reaction .
In contrast to Ni–Al–MCM-41, Ni–SBA-15, Ni–MFI and Ni–FAU, the DTG profile of Ni–Al2O3 shows broad and very low intensity peaks. Nevertheless, besides the presence of a peak at 500–700 °C (filamentous carbon ), a low intensity peak at 400–550 °C could be attributed to the presence of amorphous carbon deposits , usually associated with the decrease of activity. Although the formation of a low amount of amorphous carbon deposits is detected on Ni–Al2O3, the values of CH4 and CO2 conversions stood unchanged during a 12-h reaction. This observation is probably caused by the simultaneous reduction of unreduced NiAl2O4 species under the reduction conditions of the reaction mixture, which keep the amount of accessible active metallic Ni0 particles constant during monitored time .
In the case of Ni–MFI, the exclusive formation of filamentous carbon was also observed. However, a dramatic decrease of the activity as well as the increase of the average size of metallic Ni0 particles was observed. The DRM reaction even stopped due to the critical pressure drop during the increasing of temperature over Ni–FAU, although neither any meaningful amount of carbon deposits nor an overly high size of metallic Ni0 particles was detected subsequently. This could be explained by a contribution of one or both of the following phenomena. First, the reaction could result in the subsequent change of textural properties of Ni–MFI and Ni–FAU, which subsequently could lead to the increase of the average size of metallic Ni0 particles. Second, the large metallic Ni0 particles formed on Ni–MFI and Ni–FAU, predominately formed from weakly interaction NiO species, are not stable during the reaction and they could be easily sintering to much larger Ni0 particles. In the case of Ni–FAU, carbon deposits and very large Ni0 particles may be suspected to be localized only on upper thin layer of the catalyst bed, which caused the blocking of the reactor for the gas flow. But because of the homogenization of the whole volume of the used catalyst, the consecutive results of the characterization methods were misleading. Previously, we reported that the metallic Ni0 particles with the similar average size of Ni0 particles (25 nm) formed on NiCe–Al2O3 from Ni species strongly interacting with the support was resistant to the increase of their diameter during the reaction . Thus, it could be suggested that the ability to keep the size of Ni0 particles during the DRM reaction also depends on the type of original Ni species present on as-prepared Ni-catalysts and their interaction with the support.
This manuscript describes the relationship between the structural, textural and catalytic properties of Ni–Al2O3, Ni–SBA-15, Ni–Al–MCM-41, Ni–MFI and Ni–FAU in the dry reforming of methane. The values of the CH4 and CO2 conversions depended on the values of the average size of metallic Ni0 particles, which originated from the textural properties (i.e. high value of specific surface area, or high values of Vmicro/SBET ratio as it was reported by Chen et al.  and Saha et al.  and the reducibility of Ni species).
Reduced Ni–Al–MCM-41, Ni–SBA-15 and Ni–Al2O3 catalysts exhibited high values of CH4 and CO2 conversions as well as a low average size of formed metallic Ni0 particles originating from the high values of the specific surface areas. The lowest value of the average Ni0 particle size was determined in Ni–Al2O3 (9 nm), followed with Ni–Al–MCM-41 (11 nm) and Ni–SBA-15 materials (11 nm). The slightly lower activity of Ni–Al2O3 in contrast to Ni–Al–MCM-41 and Ni–SBA-15 catalysts was connected with a lower amount of metallic Ni0 particles presented in the reduced Ni–Al2O3 than in the reduced Ni–Al–MCM-41 and Ni–SBA-15 materials caused by lower reducibility of Ni–Al2O4 species (only 70% efficiency of the reduction at 750 °C). In contrast, the reduced Ni–FAU and Ni–MFI zeolites contained large metallic Ni0 particles (23 and 29 nm).
Large metallic Ni0 particles were preferentially formed from more easily reduced, weakly interacting NiO particles presented on the external surface of the supporting materials. In contrast, the formation of small metallic Ni0 particles preferentially required the presence of Ni aluminates or Ni phyllosilicates, which were reduced more slowly at higher temperatures.
It is important that Ni–Al–MCM-41, Ni–SBA-15 and Ni–Al2O3 exhibited marginal change of the average size of Ni0 particles and the values of CH4 and CO2 conversions during 12 h of the reaction. Although carbon deposits were formed on Ni–Al–MCM-41 (Si/Al 25) and Ni–SBA-15 (silica) catalysts, it was filamentous type of carbon, whose formation did not result in the decrease of activity and the sintering effect. In addition, the mesoporous structure of the Ni–Al–MCM-41 catalyst was preserved after the reaction. In contrast, a significant (ca 50%) increase of the average size of Ni0 particles was observed in used Ni–MFI and Ni–FAU zeolites.
The ability to keep the size of Ni0 particles during the DRM reaction could be suggested to be connected with the contribution of one or both of the following phenomena. First, the reaction could result in the change of textural properties that subsequently could lead to the increase of the average size of Ni0 particles. Second, the ability to keep the size of Ni0 particles during the DRM reaction depended on the interaction of Ni species with the support. Metallic Ni0 particles formed on Ni–MFI and Ni–FAU, predominately formed from weakly interacting NiO species, were not stable during the reaction and they could sinter easily to much larger Ni0 particles. On the other hand, the Ni0 particles formed from Ni aluminates or Ni phyllosilicates at high temperature were more resistant to size change during the reaction.
The authors gratefully thank to the European Social Fund in the Czech Republic for financial support of the project ‘Router’ Development of Research Teams at the University of Pardubice (Project No. CZ.1.07/2.3.00/30.0058). We also thank the Ministry of Science and Education of Spain for financial support (Projects ENE2011-24412 and CTQ2014-60524-R).
- 8.Zhang J, Zhang XF, Tu M, Liu WF, Liu HO, Qiu JS, Zhou L, Shao ZG, Ho HL, Yeung KL (2012) Preparation of core (Ni base)-shell (Silicalite-1) catalysts and their application for alkali resistance in direct internal reforming molten carbonate fuel cell. J Power Sources 198:14–22. doi:10.1016/j.jpowsour.2011.09.070 CrossRefGoogle Scholar
- 14.Smoláková L, Kout M, Čapek L, Rodriguez-Gomez A, Gonzalez-delaCruz VM, Hromádko L, Caballero A (2016) Nickel catalyst with outstanding activity in the DRM reaction prepared by high temperature calcination treatment. Int J Hydrog Energy 41(20):8459–8469. doi:10.1016/j.ijhydene.2016.03.161 CrossRefGoogle Scholar
- 20.Izquierdo U, Barrio VL, Bizkarra K, Gutierrez AM, Arraibi JR, Gartzia L, Banuelos J, Lopez-Arbeloa I, Cambra JF (2014) Ni and Rh-Ni catalysts supported on Zeolites L for hydrogen and syngas production by biogas reforming processes. Chem Eng J 238:178–188. doi:10.1016/j.cej.2013.08.093 CrossRefGoogle Scholar
- 55.de Sousa FF, de Sousa HSA, Oliveira AC, Junior MCC, Ayala AP, Barros EB, Viana BC, Filho JM, Oliveira AC (2012) Nanostructured Ni-containing spinel oxides for the dry reforming of methane: effect of the presence of cobalt and nickel on the deactivation behaviour of catalysts. Int J Hydrog Energy 37(4):3201–3212. doi:10.1016/j.ijhydene.2011.11.072 CrossRefGoogle Scholar