Received date: July 27, 2017; Accepted date: August 03, 2017; Published date: August 13, 2017
Citation: Hbaieb K, Rashid KKA, Kooli F, Liu Y, Tan SX (2017) Autothermal Reforming of Sulfur Doped Diesel Surrogate over Strontium Titanate Based Perovskites. J Material Sci Eng 6: 369. doi: 10.4172/2169-0022.1000369
Copyright: © 2017 Hbaieb K, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Visit for more related articles at Journal of Material Sciences & Engineering
Metal doped yttrium strontium titanate (Y0.08Sr0.88T1-xMxO3, M=Ni, Co, Ru) has been used for the autothermal reforming (ATR) of simulated diesel. A commercial Ni-Al2O3 has been used as a benchmarking catalyst for comparative evaluation. Activity tests using a diesel surrogate fuel containing 50 ppm and later 100 ppm sulfur showed stable performance for all catalysts over 20 hours on reactant stream. Both commercial and Ni-doped yttrium strontium titanate (YSTN) catalysts were further tested for 100 hours using the 50 ppm sulfur containing diesel surrogate. Both catalysts maintained good stability and suffered only little carbon formation. Even though the commercial Ni-Al2O3 catalyst had excellent resistance to carbon formation, the deposited carbon was mainly of graphitic type. The YSTN underwent comparatively higher carbon formation but with the formed carbon predominantly amorphous, thus easily removable upon catalyst regeneration. Unlike the Ni-Al2O3 that suffered extensive coarsening of Ni particles, the perovskite YSTN catalyst showed excellent resistance to sintering during the 100-hour stability tests.
Perovskite; Strontium titanate; Catalysis; Reforming; Hydrogen production
Research and development on fuel cell technology is quite mature and has spanned over more than one century. It is not unusual nowadays to find any major automobile company having a fuel cell vehicle that is either on the development stage or on production line. However, many difficulties and challenges are still encountered. For example, the non-flexibility of fuel cell to run on hydrocarbon fuel forced manufacturers to design the fuel cell engine to run merely on hydrogen. The latter has many constraints, such as the very low density, the flammability, the ease of leakage and the need for a primary source for its production. These constraints has made transportation, supply and storage very cumbersome and triggered many researchers and developers to introduce a reformer as an auxiliary unit to produce hydrogen on-board. Such an on-board reformer is an important component that has to compete for a space on a quite limited room for a typical transportation vehicle.
The on-board reformer necessitates the use of an effective catalyst to selectively produce hydrogen from any primary hydrocarbon fuel at high yield. As most logistic fuels are heavy hydrocarbon with large content of aromatics and include impurities, such as sulfur, that are in part very detrimental even when present in very low concentrations, very much care has to be taken to avoid deactivation and poisoning of catalyst. Deactivation of catalyst due to carbon formation is more probable the higher the carbon content of fuel. Moreover, sulfur poisoning takes place even when sulfur is present in ppm range.
It is known that Ni-based catalysts are very susceptible to carbon formation and sulfur poisoning [1,2]. Strategies to overcome carbon formation were many but with limited success. Changing the support material was one direction that many researchers have taken . For example, using ceria as support has shown better resistance to carbon formation that was attributed to the excellent oxygen storage capacity as compared with alumina. It is also believed that ceria has certain catalytic activity although not appreciable [4,5]. Other catalyst support materials were also tested such as MgO, CeO2, ZSM-5, SBA-15, etc. [6,7]. Another effective alternative to overcome carbon formation was to load the catalyst with low content of promoters and crack resistant elements such as noble metals (Pt, Pd, Ir, Rh and Ru) [8-14], potassium  and holmium . A very low content of the latter element has shown a clear suppression of carbon formation on a Ni-Alumina catalyst used for dry reforming . A further valuable alternative is to reduce Ni particles below a critical size that is as low as few nanometers . This alternative although valid, loses its potential when heating to the high reaction temperature as Ni is very prone to sintering and therefore coarsening [18,19].
It is not very far back since people have directed their attention to perovskite materials as potential reforming catalysts. The ABO3 structure has a great potential to accommodate a large variety of elements on the A or B sites. In addition, many perovskite materials can tolerate non-stochiometry (either in excess or deficiency) in both oxidizing and reducing environments. It is believed that A-site deficiency with potential combination of partial substitution of specific elements may greatly improve oxygen mobility and storage capacity. On the other hand, substitution on the B site may considerably improve activity towards hydrocarbon reforming and hydrogen selectivity. Perovskite may constitute a serious alternative to traditional metal supported catalysts due to its relatively low cost and flexibility to accommodate a large number of active metals. It has been reported that perovskite play the role of a precursor of catalytic active elements that can, under reducing environment, pop to the surface and grow to nanoparticles well dispersed at the surface .
There has been a lot of work recently on different types of perovskite as reforming catalysts [20-24]. Liu et al.  showed excellent activity and sulfur tolerance of ruthenium doped lanthanum chromite and aluminate under autothermal reforming of diesel. Mota showed that the partial substitution of Ru in LaCoO3 increased the formation of active sites on the perovskite surface, improved activity and stability and enhanced resistance to carbon formation and sulfur poisoning under oxidative reforming of diesel [26,27]. Kondakindi demonstrated an enhancement in conversion and H2 yield upon partial substitution of Ce and Co/Pd at the A and B sites, respectively of LaFeO3 perovskite for ATR of dodecane . Chen showed that the introduction of Ca in the A site of LaFe1-xNixO3 provided additional lattice oxygen that with the surface oxygen reduced carbon deposition and carbon containing intermediates and improved stability for steam reforming of ethanol . Villoria found that the use of acetate precursor and low calcination temperature in the preparation of LaCoO3 perovskite reduced crystallite size and had positive impact on hydrogen production and catalyst stability under oxidative reforming of diesel . Mawdsley showed that among few first-raw transition metal elements (Cr, Mn and Fe), Cr provided the most sulfur tolerance when used as stabilizing element in LaNiO3 for ATR of hydrocarbon fuels . Erri claimed that Ce addition improved coking resistance due to the carbon oxidation on the surface resulting from the increase in oxygen ionic conductivity .
In this paper we investigate the potential use of metal doped strontium titanate as a reforming catalyst of diesel. This material has shown great potential as an anode for a direct carbon solid oxide fuel cell (DCFC) with excellent mixed ionic-electrical conductivity (MIEC) [33-37]. It has also shown excellent activity and resistance to carbon formation when used as reforming catalyst for the ATR of dodecane even at very severe conditions . The effect of the doping element in the catalyst and the sulfur content in the fuel on the activity and catalyst deactivation is discussed.
Commercial Ni-Al2O3 catalyst of Sud-Chemie India Ltd, C11-9-02 (16 × 6 × 8 mm ring shaped) specifically developed for steam reforming of hydrocarbons was used as a benchmarking catalyst to the in-house prepared catalysts. The loading level of NiO was at 15% (12% Ni). Strontium titanium based Perovskite materials were prepared by the modified (sol-gel) Pechini method [39,40]. Titanium isopropoxide was dissolved in ethylene glycol in the ratio 1:12, diluted in ethanol and brought onto a magnetic stirrer with an integrated hotplate at a temperature of 70-80°C. Citric acid was added slowly while stirring until a clear solution was obtained. Aqueous solution of metal nitrates, prepared in the corresponding mole ratios, was slowly added to titanium isopropoxide solution while constantly stirring. The solution thus obtained was kept for some time until it became clear. The pH was then adjusted to 6-7 and the solution was left overnight while stirring for further mixing and gel formation. The resultant gel was transferred to an oven for drying and decomposition. The oven temperature was slowly and gradually heated up to 400°C with some dwell intervals to avoid swelling and instability of the gel during burn out. The dried gel was crushed, finely powdered and then calcined at 600°C for 1 hour to burn out any remaining organic substances. The resultant powder was again crushed and mixed in a mortar and further calcined at 800°C for 5 hours and then at 600°C for 10 hours before cooling to room temperature. The materials thus obtained are metal doped yttrium strontium titanate (Y0.08Sr0.88T1-xMxO3, M=Ni, Co, Ru), designated as YSTMx, with M=R for ruthenium, C for cobalt and N for nickel; the doping level, given in percent, is indicated by x.
The X-Ray diffraction technique was used to identify the phase crystallinity of the different fresh and spent catalysts. A D8 Advance diffractometer from Brucker, equipped with Ni-filter and Cu-Ka radiation, was used. Raman spectroscopy was employed to assess the presence and type of carbon deposited onto the catalysts; Laser spectra were measured using Bruker 400 spectrometer. Quantitative estimation of carbon deposition was achieved through thermogravitic analysis using a TA instrument, model SDT 600. Temperaure programmed reduction (TPR) tests were evaluated in the temperature range ambient to 800°C in a Micromeritic instrument, model Autochem II 2920 V5.02.
The activity experiments were conducted using a benchtop customized microreactor (supplied by Hi-Tech, India). The reactor tube is made of inconel having inner diameter of 10 mm. The reactor is supplied with SCADA software. The latter communicates with a sophisticated control system that is in turn communicating with the reactor control panel.
The catalyst powder was pelletized, crushed and sieved through different sieves to collect granules in the size range of 200-500 microns. The catalyst granules of weight 0.05 g were diluted with alumina media and well mixed to ensure sufficient dispersion of catalyst particles. The total volume of the diluted catalyst was 3 ml that corresponded to ~70 mm catalyst length in the reactor tube. The catalyst was loaded in the middle of the reactor tube and supported by ceramic wool. The region between the reactor top zone and the catalyst was kept vacant to ensure fast access of the reactants after being vaporized and mixed. Two thermocouples are inserted in the reactor tube through a thin centric thermowell and positioned right above and underneath catalyst bed to accurately and uniformly control the temperature of the catalyst.
The catalyst loaded in the reactor was initially reduced under hydrogen at 700°C. Then the reaction temperature was maintained at 750°C. Fuel mixture of dodecane and xylene, at 77:23 volume%, was chosen as diesel surrogate for this study. Such fuel blend simulated well jet fuel and diesel characteristics as reported elsewhere [41,42]. Sulfur was introduced in the form of dibenzothiophene (DBT), at 50 wppm concentration to test the sulfur tolerance of the catalyst.
The fuel rate was maintained at 0.3 ml/hour such that the liquid hourly space velocity was approximately 6 ml/g. cat/h. The corresponding gas hourly space velocity (GHSV) was 25,000 ml/g. catalyst/h. Steam to carbon and oxygen to carbon ratios were set to 3 and 0.32, respectively. Water and fuel were introduced into the reactor at the desired rates through two HPLC pumps. The reactant gases were introduced through separate lines using MFC controllers. All reactants were driven through spiral mixers separately before reaching mixing zone in reactor. All spiral mixers are located in an oven that can be adiabatically heated to 200°C. The reactant mixture was directed to a three zone furnace located within the oven. In the absence of a preheater, the top zone of the reactor was filled with silicon carbide inert material and heated to an intermediate temperature to both pre-heat and mix reactants before reaching the catalyst. Even though this is inconvenient and pre-heating may not be well controlled, compactness is commonly a necessity for an on-board reformer.
Dodecane free of xylene and sulfur was first fed to the reactor tube. Simulated diesel fuel was later introduced after verifying acceptable activity under dodecane. The catalytic reforming test spanned over 20 hours. Water and heavy hydrocarbon fuels are trapped in a gas liquid separator with chiller side connections (for condensation), while the gas mixture is directed to a gas chromatograph (Agilent 7890A Series GC) equipped with thermal conductivity and flame ionization detectors.
Autothermal reforming under 50 ppm sulfur-containing fuel: The activity under auto thermal reforming of diesel surrogate (77% dodecane and 23% xylene) containing 50 ppm sulfur for the different catalysts is shown in Figure 1. The hydrogen yield, shown in Figure 1, represented the moles of hydrogen produced per mole of carbon in the fuel. All catalysts showed stable and comparable performances throughout the course of testing with slightly better yield for commercial Ni-Alumina catalyst over the in-house prepared catalysts. The activity tests under sulfur free fuel are reported in details in our earlier work .
The initial drop in activity shown in Figure 1 for all catalysts is due to the switching from dodecane to the sulfur containing fuel blend. Activity drop upon sulfur introduction has also been reported elsewhere [25,43,44]. Table 1 shows the average hydrogen yield over the tested period. All quantities are comparable. The hydrogen yields ranged between 17 (YSTN10) to 18.6 (Ni-Alumina). The averaged gas product concentrations for the different catalysts are also calculated. They ranged between 64-67, 21-23, 0.5-1 and 11-13 for H2, CO2, CH4 and CO, respectively.
|20 hours under 50 ppm Sulfur||17.9||17||17.6||18.6|
|20 hours under 100 ppm Sulfur||16.8||17.5||17.8||16.7|
|100 hours under 50 ppm Sulfur||-||17.8||-||16.8|
Table 1: hydrogen yield of the different catalysts under the three different testing conditions: (1) under 20 hours of ATR for diesel surrogates doped with 50 ppm sulfur, (2) under 20 hours of ATR for diesel surrogates doped with 100 ppm sulfur and (3) under 100 hours of ATR for 50 ppm sulfur containing fuel.
The fuel pumping rate is too low and hence its accuracy is critical in making comparisons. As such, actual weight of diesel fuel was constantly measured. It was noticed that the actual fuel rate is only ~80- 90% of the set value. This implies that the actual S/C and O/C ratios are 3.3-3.7 and 0.36-0.4, respectively. This makes reaction conditions close to oxidative reforming. Under such conditions the CO2 concentration may go higher than 20%, as reported elsewhere [45-47].
Autothermal reforming under 100 ppm sulfur-containing fuel: The activities of all catalysts were almost similar and therefore it was difficult to identify the promising formulations. Hence it was necessary to intensify the severity of the reforming conditions by increasing the sulfur content from 50 to 100 ppm, keeping all other operating parameters the same. The same stability tests were run under 100 ppm sulfur containing diesel surrogate. The results are shown in Figure 2 Again the results are quite stable for all catalysts with no sign of degradation over the tested period. The initial drop in activity at the beginning of the tests is also seen (similar to the results reported above) except for the YSTR05 catalyst. The relatively higher yield for the Ni-Alumina catalyst observed in the previous test under 50 ppm containing fuel is no longer holding here. In contrast, the hydrogen yield for the Ni-Alumina catalyst is lowest when compared to those of the YST based catalysts although the difference is small.
The average quantities for the hydrogen yield are shown in Table 1. All catalysts showed similar gas product distributions. Both YSTN10 and YSTC10 catalysts showed better hydrogen yields than those of Nialumina and YSTR05. The latter catalyst showed lower performance at the beginning, but gradually increased and reached an activity comparable to that of the other catalysts after five hours on stream. The average value of hydrogen yield over the last ten hours for YSTR05 was 17.6 quite similar to those of YSTN10 and YSTC10. Eventhough the YST based catalysts showed better hydrogen yield than the Ni-Aumina catalyst, the difference is quite small.
The conditions employed were severe enough to indicate that both in-house perovskite and commercial Ni-Alumina catalysts are robust catalysts for hydrocarbon autothermal reforming. Increasing the liquid and gas hourly space velocity to 18 and 160,000 ml gcat-1 h-1, respectively, under sulfur free diesel surrogate has shown earlier stable performance for all catalysts with slight better activity for the perovskite catalysts over the commercial catalyst . As such, we drew our attention to testing the durability of the catalysts.
Durability test for ATR of 50 ppm sulfur-containing fuel: Long term durability (100 hours) testing was carried out using a sulfur (50 ppm) doped fuel blend of dodecane–xylene mixture. The tests were run for both perovskite and commercial catalysts for comparison. YSTN10 was chosen as the representative perovskite catalyst as it is doped with nickel (most matching with Ni-Alumina catalyst) and because it has the relatively worst performance among the other perovskite catalysts (highest D and G peaks for Raman spectra following activity tests under 100 ppm sulfur containing diesel surrogate). That is, if this catalyst performs well, it would be a good indication and projection for stable performance of the other catalysts. The same ratios of steam and oxygen to fuel were employed but the gas hourly space velocity was reduced to 5400 ml gcat-1 h-1. The amount of catalyst was 1 g in each test diluted by SiC inert material. Activity of both YSTN10 and Ni- Alumina catalysts are shown in Figure 3 after 100 hours run under 50 ppm sulfur containing fuel blend. Both catalysts showed high stability with no sign of degradation. YSTN10 showed slightly better activity than Ni-Alumina catalyst (Table 1). The excellent stability of both catalysts for 100 hours despite the high sulfur content is an evidence for their good potential for reforming heavy hydrocarbon fuels. Both are good sulfur tolerant catalysts.
Characterisation of catalysts
X-ray diffraction (XRD): The X-ray diffraction patterns of all catalysts are shown in Figure 4. All perovskite materials exhibited single crystalline phase. XRD spectrum of Ni-Al2O3 catalyst shows well defined characteristic peak of NiO at 2θ of 44.3°, but intensity of peaks for nickel aluminate at 2θ around 19° and 59° are too low and therefore it is a clear indication that the majority of nickel in the catalyst is present as nickel oxide, being well dispersed on the alumina support . More details on the XRD patterns of the catalysts are given elsewhere .
Figure 5 shows the XRD pattern of YSTN10 and Ni-Al2O3 catalysts following the stability test for 100 hours with fuel containing 50 ppm sulfur. The spent YSTN10 catalyst maintained its crystal structure and the observed XRD pattern is matching well with those earlier reported for strontium titanate having characteristic peaks at 2θ around 23, 32, 40, 46, 52, 58 and 68° [28,49,50]. The characteristic carbon peak at ~26° is not observed for YSTN10 thus indicating either the absence of coke deposition on the catalyst or it may be too low to get detected. This indicates stability of the catalyst with excellent coke resistance features, particularly in the context of carbon forming and removal reactions occurring simultaneously on the reforming catalysts . A tiny peak observed at 2θ ≈ 26.5° in the spent Ni-Al2O3 catalyst XRD patterns, although not distinct due to neighboring fluctuations, could be attributed to the presence of carbon.
The Scherrer equation was used to calculate the crystal size of Ni and YSTN10 particles by referring to the most intense peak of the corresponding XRD pattern. Table 2 shows these characteristic peaks and the corresponding crystal sizes for fresh and used catalysts following 100-hour durability tests. The peak for NiO at 2θ=44.3° was considered for calculating the crystallite size. The NiO particles in the Ni-Al2O3 catalyst have grown from 10.5 nm to 25.6 nm, an increase of as high as 144%. The characteristic peak at 31.75° for the YSTN10 catalyst was used for crystallite size estimation. By comparing the estimated crystal sizes for fresh and spent catalysts, it is inferred that no crystal growth took place for YSTN10 catalyst during the 100 hours-test. Thus, in contrary to Ni-Al2O3 that suffered huge Ni particle coarsening, YSTN10 has shown excellent resistance to sintering. Both the increased Ni loading in Ni-Al2O3 (compared to YSTN10 catalyst) and the smaller initial crystal size might have contributed to this low resistance to sintering in Ni-Al2O3.
|Peak max(2θ) YSTN10/NiO/Ni peak||31.75°||44.33°|
|Crystallite size (nm) YSTN10/NiO (fresh)||27.26||10.48|
|Crystallite size (nm) YSTN10/NiO (Used)||26.2||25.6|
|Crystal growth (%)||-||144|
Table 2: Crystallite size of fresh and spent catalysts.
Temperature programmed reduction: Temperature programmed reduction curves are given in Figure 6. Two distinct peaks are observed for YSTN10 and Ni-Al2O3. Low temperature peak maxima for YSTN10 and Ni-Al2O3 are 430°C and 540°C respectively. High temperature peak maxima for YSTN10 and Ni-Al2O3 are 600°C and 740°C, respectively. The extent of reduction in low temperature region was found to be higher for YSTN10 as indicated by an increase in peak area ratio (low temperature peak area / high temperature peak area) and decrease in peak maxima in TPR curves. The improved reducibility of YSTN10 could be attributed to decreased metal support interactions in it. The high temperature peak in Ni-Al2O3 corresponds to the reduction of fixed nickel oxide in the form of amorphous nickel aluminate [52,53]. Decreased metal support interaction is expected to improve the dispersion and therefore YSTN10 is expected to have a better activity and dispersion than Ni-Al2O3.
YST as such is not having any reducible metal oxides in the temperature range ambient to 800°C as indicated by the absence of peaks in the TPR curve in Figure 6.
Raman spectroscopy: Raman spectroscopy has been reported as being an established and sensitive technique for detecting carbon deposition on the catalyst [54,55]. Two carbon peaks are of interest: The G peak at a wavelength of ~1580 cm-1 and the D band at ~1350 cm-1. A single crystal graphite shows a single G peak at 1575 cm-1. In imperfect polycrystalline graphite, additional D peak is observed. The latter peak is related to disorder in the carbon’s structure. The widths of both peaks decrease with increasing degree of graphitization. Graphitization degree also increases with decreasing integrated intensity ratio ID/IG.
The Raman spectra following stability tests under 50 ppm sulfur containing diesel surrogate are shown in Figure 7. The peaks for YSTN10, YSTC10 and Ni-Alumina are broad for both D and G bands. The intensity of peaks for Ni-Alumina is higher than those of YSTN10 and YSTC10, an indication for more carbon presence on Ni-Alumina catalyst. The ratio ID/IG is 1.7, 1.3 and 1.13 for Ni-Alumina, YSTN10 and YSTC10 catalysts, respectively. Thus, the deposited carbon in these catalysts could be considered to be mostly reactive, i.e. it can be readily removed during regeneration. The YSTR05 catalyst showed relatively high G (broad) peak and no D peak. This is unexpected as ruthenium should have both high activity and resistance to carbon formation. In our previous work , ruthenium doped YST showed similar behavior although quantitatively it exhibited low carbon formation (as measured by TGA).
Raman spectra for all spent catalysts following reforming under 100 ppm containing fuel are shown in Figure 8. The YSTC10 catalyst has very low D and G peaks (almost absent), a strong indication for the high carbon formation resistance and high sulfur tolerance. The Ni-Alumina and YSTR05 catalysts also showed low peaks and thus good resistance to carbon formation. The most critical result was for YSTN10 that showed high D and G peaks and therefore worst carbon formation resistance. The type of deposited carbon can be inferred from the ID/IG ratio. The latter was 2, 2.1 and 1.8 for YSTR05, Ni-Alumina and YSTN10, respectively. Thus for all these catalysts, the amorphous reactive carbon is predominant. No ID/IG ratio was estimated for YSTC10 as both peaks were very low thus carbon formation for this catalyst is very low.
The Raman spectra for bothYSTN10 and Ni-Al2O3 catalysts following the 100 hours stability tests are shown in Figure 9. Two broad D and G peaks are evident for YSTN10 with the D peak clearly higher than the G peak which indicates that the deposited carbon is predominantly amorphous. In contrast, the Ni-Al2O3 showed very sharp and narrow G peak and much smaller D peak. This implies that the deposited carbon is of refractory character that is unfavorable for the durability of the catalyst.
Thermogravimetric analysis: The TGA measurements for catalysts following 20 hours test with sulfur doped fuel were not possible due to the small amount of catalysts used. Nevertheless, the Raman spectroscopy described in the previous section was a good qualitative estimation for the carbon deposition onto the catalysts. As the durability tests involved larger amount of catalysts, it was possible to conduct TGA experiments on YSTN10 and Ni-Alumina catalysts after these tests. The results are shown in Figure 10. Surprisingly both catalysts did not undergo any mass loss in the temperature region 300- 600°C, despite the long subjection to autothermal reforming. A small amount of mass loss ~0.84% and ~1.52% was measured thereafter for Ni-Al2O3 and YSTN10 catalysts, respectively. This indicates that the catalysts are stable over the tested period and are slightly prone to sulfur poisoning and coke deposition. The carbon deposited on Ni-Al2O3 is mainly of graphitic type as suggested from the analysis of XRD and Raman spectra of spent Ni-Al2O3 catalyst after the 100-hour durability test, whereas the carbon formed onto YSTN10 is mainly of amorphous and reactive character (Figures 5 and 9).
The transitional metals, particularly Nickel, are sensitive to sulfide formation . It is believed that the presence of sulfur can suppress carbon removal reactions and promote the formation of graphitic carbon onto catalyst active sites. That is, the formed sulfides are too stable to be removed and as such they would cover the active site surfaces and dramatically suppress or even block carbon gasification. The combined sulfide and carbon presence may then prevent surface active sites from interacting with the reactants (hydrocarbons and steam) that may lead to a major deactivation of the catalyst . However such a significant deactivation as in the case of steam reforming process is not noticed in the present studies on ATR as the deactivation of the YSTN10 and Ni- Al2O3 catalysts is minimum.
Unlike Ni-Alumina catalyst that exhibited large mass gain attributed to the Ni metal oxidation, the YSTN10 catalyst has shown little mass gain. This clearly indicated that the active sites on the surface for YSTN10 were much lower than those of Ni-Alumina catalyst. While the Ni loading in Ni-Al2O3 is as high as 12%, the total doping level of Ni in the perovskite YSTN10 perovskite is only 3%, of which a fraction was reduced on the surface. Thus, it would be worthwhile looking into promoting the growth of reduced active metal sites by either increasing the doping level of Ni into the YST lattice structure or impregnating additional nickel onto the perovskite surface. Moreover, the preparation method needs to be modified towards improving the diffusion of the lattice metal elements to the surface, by either creating B site vacancies or increasing surface area. The latter scenario was suggested by Liu and Krumpelt  as effective way of increasing number of active sites for reforming reaction. They have claimed that increasing surface area by reducing calcination temperature showed better hydrogen uptake for perovskite catalyst when subject to reducing environment under TPR tests. All of these scenarios are under current investigation.
Metal substituted yttrium strontium titanate perovskites have been proven to be promising catalysts for autothermal reforming of diesel. In comparison with commercial Ni-Al2O3 catalyst, the ceramic perovskite catalysts have shown marginal improvement in activity and better stability towards autothermal reforming under sulfur containing diesel surrogates. Even though the carbon formed on the perovskite catalyst, subsequent to 100 hours stability test, is slightly more than that on the commercial Ni-Alumina catalyst, it is mainly of amorphous character, thus can be partially removed upon regeneration. The carbon formed onto Ni-Al2O3 although little is inert and very difficult to remove. Our analysis has also shown poor resistance of Ni-Al2O3 catalyst to sintering in comparison to YSTN10 that showed no particle growth after 100 hours of testing. The perovskite catalyst has a great potential for improvement on many fronts and this will be our future effort.
This project was funded by the National Plan for Science, Technology and Innovation (MAARIFAH)-King Abdulaziz City for Science and Technology-the Kingdom of Saudi Arabia, award number 09-ENE807-05.