Al2O3 catalyst: A longevity evaluative study

Journal of Energy Chemistry 24(2015)366–373

CO2 reforming of glycerol over La-Ni/Al2O3 catalyst: A longevity evaluative study Kah Weng Siewc , Hua Chyn Leec ,

Maksudur R. Khanb,c ,

Jolius Gimbunb,c ,

Chin Kui Chenga,b,c∗

a. Rare Earth Research Centre, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300 Gambang, Kuantan, Pahang, Malaysia; b. Centre of Excellence for Advanced Research in Fluid Flow, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300 Gambang, Kuantan, Pahang, Malaysia; c. Faculty of Chemical & Natural Resources Engineering, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300 Gambang, Kuantan, Pahang, Malaysia [ Manuscript received December 21, 2014; revised January 12, 2015 ]

Abstract This paper reports on the longevity of glycerol-dry (CO2 ) reforming over the lanthanum (La) promoted Ni/Al2 O3 catalysts. The XRD results showed that the Ni particle was well-dispersed in the presence of La promoter. In addition, via the NH3 -TPD analysis, it was found that the La promoter has reduced the acidity of Ni catalyst which may have explained the mitigation of carbon laydown. It was determined that the 3.0 wt% La-promoted Ni/Al2 O3 catalyst possessed the largest BET specific surface area of 97 m2 ·g−1 . Consequently, it yielded the best catalytic longevity performance with conversion attained more than 90%, even after 72 h of reaction duration. Significantly, it can be confirmed that the presence of CO2 during the glycerol dry reforming was essential in reducing carbon deposition, most likely via gasification pathway. This has ensured a stability of catalytic activity for a long reaction period (72 h). Key words CO2 reforming; glycerol; lanthanum; longevity; syngas

1. Introduction For more than a century, petroleum based fossil fuel has been utilized in transportation sector, power generation and industrial processes. However, petroleum-based fuel is a non-renewable source; hence it is finite [1]. Therefore, an alternative energy which is both renewable and sustainable will be attractive in lieu of anticipated shortage of liquid hydrocarbon fuel. Significantly, syngas (a gas mixture consists of H2 and CO) offers a genuine potential as a new energy source as it can be conveniently converted into liquid fuel via Fischer-Tropsch synthesis [2]. Moreover, syngas can also be employed for the productions of ammonia, methanol, fertilizer, synthetic plastic, etc. The resulting H2 gas, when purified, is even suitable for a low temperature polymeric fuel cell [3,4]. Traditionally, syngas is produced at an industrial scale via methane steam reforming which nevertheless, is a non-renewable source [5]. Moreover, steam reforming typically yields CO2 , a component of greenhouse agent via Equations (1) and (2):

CH4 + H2 O ↔ 3H2 + CO (steam reforming reaction) CO + H2 O ↔ CO2 + H2 (water-gas shift reaction)

(1) (2)

Therefore, large scale biomass utilization, i.e., glycerol (a biomass waste from cooking oil transesterification) can potentially address the aforementioned issues. In the current work, glycerol was utilized for syngas production via dry (CO2 )reforming, proposed herein as (3): C3 H8 O3 + CO2 ←→ 3H2 + 4CO + H2O

(3)

Glycerol has been touted as the most promising biomassfeedstock for energy conversion due to its availability from biodiesel industry [4]. Significantly, glycerol when dryreformed, is a greener method because this pathway consumes CO2 and requires less energy compared with the more wellknown steam reforming pathway. In our previous works, we have employed alumina supported nickel catalyst promoted by La for glycerol dry re-

∗

Corresponding author. Tel: +60-9-549-2896; Fax: +60-9-549-2889; E-mail: [email protected] This work was supported by the Ministry of Education, Malaysia through MTUN-CoE Research Grant (RDU121216) and FRGS Research Grant (RDU140112). Copyright©2015, Science Press and Dalian Institute of Chemical Physics. All rights reserved. doi: 10.1016/S2095-4956(15)60324-2

Journal of Energy Chemistry Vol. 24 No. 3 2015

forming [6,7]. Lanthanum metal was incorporated into the catalyst to improve the carbon resilience of the Ni catalyst. The results obtained, showed that 3.0 wt% La doped catalyst possessed high BET specific surface area and produced excellent catalytic reforming under different reaction conditions. Moreover, our results also have indicated that the H2 yield was 7.0% and CO yield was 4.0% (H2 /CO ratio of 1.75). Furthermore, it was found that CO2 -to-glycerol ratio (CGR) showed significant effect on catalytic activity and CGR of 1.67 at 1023 K was the best condition when dry reforming was carried out for 4 h [7]. The longevity catalytic performance however, is unknown and warrants for further evaluation. Therefore in the current paper, longevity of glycerolCO2 reforming was carried out to study the catalytic performance stability at CGR of 1.67 over a continuous 72 h of reaction duration.

367

Rigaku Miniflex II X-Ray Diffractometer was utilized for XRD analysis. The particle size of the catalyst samples was ground to less than 100 µm before the analysis. The solid samples were irradiated with nickel-filtered CuKα radiation with ˚ at 40 mA and 45 kV. The sample a wavelength (λ) of 1.542 A scanning was carried out from 10o to 80o at a scanning rate of 4 o /min. The mean size of crystallites was calculated from the Scherrer equation (cf. Equation 4): 0.9 × λ dp = (4) β × cosθ

The alumina (Al2 O3 )-supported Ni catalyst was prepared from an impregnation of Al2 O3 powder with aqueous solution of Ni(NO3 )2 ·6H2 O. Both compounds were weighed into a beaker. The solid mixture was then wetted with 40 mL distilled water and magnetic-stirred for 3 h under ambient condition. Subsequently, it was oven-dried at 373 K for 12 h. Postdrying, the catalyst was calcinated at 1073 K for 5 h, employing a heating rate of 5 K·min−1 . The catalyst was then sieved to a final particle range of 90−250 µm for the physicochemical characterization and reaction studies. For the La-promoted catalysts (La-Ni/Al2 O3 ), the aforementioned procedures were repeated whereby alumina was mixed with Ni(NO3 )2 ·6H2 O and La(NO3 )3 ·6H2 O precursors salts. In the current work, two types of La-promoted catalysts were synthesized and tested, viz., 3.0 and 5.0 wt% La.

where, λ is the X-ray wavelength, β represents the line broadening at half the maximum intensity (FWHM) in radians and θ is the Bragg angle. Thermo Finnigan TPDRO 1100 apparatus equipped with a thermal conductivity detector was employed to determine the quantitative and qualitative aspects of the desorbed NH3 for the acidity analysis. An approximately 50 mg solid catalyst sample was required. The catalyst was treated at 423 K for 15 min in N2 blanket (20 mL·min−1 ). After that, it was saturated with NH3 for 1 h for NH3 -adsorption. The purging with N2 was then carried out at room temperature for 45 min to eliminate excess NH3 that was still physisorb on the surface of catalyst. The desorption step was then carried out from 323 to 1223 K under 20 mL·min−1 of He flow, employing a heating ramp of 10 K·min−1 . The changes in NH3 concentration was recorded online. FESEM imaging was carried out using JEOL/JSM-7800F Thermal FESEM unit. The magnification employed for the morphological surface structure of the solid catalyst samples was 30 kX (100 µm). The accelerating voltage used for the experiment was in the range of 5 to 15 kV. Platinum holder was used throughout the analysis. In addition, transmission electron microscope (TEM) of the brand Zeiss Libra 200FE was used to observe the nano structure of deposited carbon on the used catalysts. The magnification range from 15 to 25 kX was employed with an accelerating voltage of 200 kV.

2.2. Catalyst characterization

2.3. Longevity studies

A thermogravimetry analyzer instrument with model Q500 was employed for all the catalyst solid-gases thermal interaction studies. Around 70 to 80 mg solid catalyst samples were required for each analysis. Specifically, the temperature thermal treatments employed were temperature-programmed calcination for freshly-dried catalysts and temperatureprogrammed oxidation (TPO) for analyzing the carbon deposition over the used catalysts. The applied temperature range was 298 to 1173 K. N2 physisorption was performed using Thermo-Scientific Surfer unit. During the analysis, solid catalyst sample weighing 0.3 to 0.4 g was transferred into a sample holder followed by an overnight degassing at 573 K for moisture and volatile impurity removal. Subsequently, the sample was transferred to the analyzer for N2 physisorption at 77 K. N2 gas with a ˚ 2 was used as the adsorbate in cross-sectional area of 16.2 A the analysis.

All the glycerol dry reforming experiments were conducted in a stainless-steel fixed bed reactor measuring 9.525 mm OD and a length of 40 cm positioned inside a split-tube furnace. The catalyst bed was supported by quartz wool. The catalytic longevity of catalyst was evaluated at 1023 K and 1 atm pressure for 72 h. Liquid glycerol at a predetermined flowrate was injected into the vaporizer located upstream of the reactor using a HPLC pump (Lab Alliance Series 1) and then downward into the reactor. Prior to the testing, the catalyst was reduced with 50 mL·min−1 STP of H2 for 2 h. The heating rate was controlled at 10 K·min−1 . All the inlet gas flow rates were regulated by Alicat Series electronic mass flow controller. The total inlet flow was fixed at a weighthourly-space-velocity (WHSV) of 3.6×104 mL·g−1·h−1 STP. Reactor outlet gases passed through a cold trap for liquid product capture and then over a drierite (CaSO4 ) bed. The exit gas was collected into 1-L Tedlar gas sampling bags.

2. Experimental 2.1. Catalyst synthesis

368

Kah Weng Siew et al./ Journal of Energy Chemistry Vol. 24 No. 3 2015

The composition of syngas produced was determined using an Agilent gas chromatograph (GC) with TCD capillary columns, HP-MOLSIV (Model No. Agilent 19095P; 30.0 m×530 µm×50.0 µm) and HP-Plot/Q column (Model No. Agilent 19095-Q04; 30.0 m×530 µm×40.0 µm). He was used as a carrier gas. Product stream flow rate was measured using a bubble meter. In order to approximate a plug flow condition and to minimize back-mixing or channelling, the ratio of catalyst bed length to catalyst particle diameter (L/dp ) was 80.0 and the ratio of inner diameter of reactor to particle diameter (dt /dp ) was 71.5. The catalytic performance was evaluated in terms of glycerol conversion to gaseous products based on atom-H balance: Xg (%) =

2FH2 + 4FCH4 × 100% 8(FC3 H8 O3 ,in )

ln

β Tp2

!

= ln

AR Ea − Ea RTp

(6)

where, β is heating rate (K·min−1 ), A is pre-exponential factor (s−1 ), Ea is activation energy (kJ·mol−1), Tp is peak temperature (K) and R is gas constant (J·mol−1 ·K−1 ).

(5)

3. Results and discussion 3.1. Catalyst characterization The calcination profiles of all the freshly-dried catalysts (unpromoted and La-promoted Ni/Al2 O3 ) showed a similar trend with a representative sample portrayed in Figure 1. It can be seen that a significant weight loss was recorded from 400 to 600 K. Beyond the 600 K, no weight loss was apparent. The decomposition of nitrate-salt precursor into oxide metal is the most likely explanation for the observation. Moreover, the results also indicate that temperature of 600 K was a minimum requirement for catalyst calcination under the air-blanket for the subsequent catalytic glycerol dry reforming reaction studies. A closer inspection of Figure 1 shows that the profiles have shifted to the higher temperature region when the ramping rates were increased (incremental from 10 to 20 K·min−1 ). For this type of behaviour, a “model-free” non-isothermal Kissinger equation (cf. Equation 6) can be employed to determine the activation energy of calcination and the results are summarized in Table 1. Figure 2 subsequently shows the resulting excellent model fittings.

Figure 1. Weight loss profile of calcination for Ni/Al2 O3 catalyst (a representative sample)

Figure 2. Kissinger profiles for La-promoted and unpromoted Ni/Al2 O3 catalysts Table 1. Kissinger modeling of non-isothermal calcination Catalysts Ni/Al2 O3 3 wt% La-Ni/Al2 O3 5 wt% La-Ni/Al2 O3

Ea (kJ·mol−1 ) 90.96 77. 03 116.58

R2 0.99 0.99 0.98

It can be seen that the Ea for the Ni/Al2 O3 , 3.0 wt% LaNi/Al2 O3 and 5.0 wt% La-Ni/Al2 O3 were 91.96, 77.03 and 116.58 kJ·mol−1 , respectively. Compared with the Ni/AlO3 catalyst, the 3.0 wt% La catalyst showed a markedly reduced Ea . However, the 5.0 wt% La sample possessed Ea magnitude that was even higher than the non-promoted catalyst. To understand the discrepancy, further physicochemical characterization was carried out. As can be seen in Table 2, N2 physisorption results reveal that the BET specific surface area of Ni/Al2 O3 catalyst was 85 m2 ·g−1 ; it was further enhanced to 97 m2 ·g−1 for 3.0 wt% La while that of the 5.0 wt% La catalyst was near similar to the unpromoted Ni/Al2 O3 . Significantly, the BET specific surface area results showed that the La functioned like a ‘spacer’ that can inhibit the agglomeration of Ni particles; hence a well-dispersion of oxide metal. This proposition has also been supported by the subsequent XRD analysis as shown in Figure 3, which shows that the NiO crystallite at 2θ of 43o was smaller with the increase in La loading. In fact, the NiO crystal size decreased in the order of Ni/Al2 O3 >3.0 wt% La-Ni/Al2 O3 >5.0 wt% La-Ni/Al2 O3 (for 5.0 wt% La, it was undetectable due to the well-dispersion). Therefore, the decrease in the BET specific surface area for the 5.0 wt% La compared with the 3.0 wt% La sample can be ascribed to the finer dispersion and smaller particle size associated with the former. This has covered the catalyst

369

Journal of Energy Chemistry Vol. 24 No. 3 2015

pores, consequently decreasing the surface area. This was also corroborated by the decreasing values of the pore volume and pore diameter with the La amount (cf. Table 2). From the findings, it can be professed that a La doping has assisted the dispersion of active metal on the alumina support and hence larger BET specific surface area. This has enabled easier calcination process for the 3.0 wt% catalyst as opposed to the non-promoted Ni/Al2 O3 catalyst. Nonetheless, beyond certain threshold, 3.0 wt% La as in the current study, smaller particle size associated with fine metal dispersion has blocked the tunnels. This may have increased the difficulty in diffusion of air during the calcination; consequently the higher Ea as shown by the 5.0 wt% La-Ni/Al2 O3 catalyst. Furthermore, the NH3 -TPD as in Figure 4 shows a single discernible peak at the interval of 400 to 800 K for all the samples which could be assigned to a weak acid site albeit with different concentration levels. Based on the results in Table 3, it can be surmised that the La-inclusion has reduced the acid site of the catalyst with the 5.0 wt% La catalyst possessing the lowest total acid site (56.0 µmol·g−1) as opposed to the unpromoted Ni/Al2 O3 catalyst with acid concentration of 1670.0 µmol·g−1 . Similar results were reported before by other authors, referenced herein [10,11]. The presence of La2 O3 yields higher surface electron density. Consequently, the protons would form a stronger bonding with these oxygen atoms which reduce the acid strength. Liu et al. has stated that even a small increment of electron density of Ni atomic sites (the sites that provide suitable Lewis acidity for carbon formation) could decrease the Lewis acidity [12]. Therefore, it is further anticipated that carbon resistance of the catalyst can be improved via La-promotion.

Figure 3. XRD patterns of La-promoted and unpromoted Ni/Al2 O3 catalysts

Figure 4. NH3 -TPD profiles of La-promoted and unpromoted Ni/Al2 O3 catalysts

Table 2. Structural properties as well as crystallite size (NiO species) of La-promoted and unpromoted Ni/Al2 O3 catalysts calcined at 1073 K Catalysts Ni/Al2 O3 3 wt% La-Ni/Al2 O3 5 wt% La-Ni/Al2 O3

BET specific surface area (m2 ·g−1 ) 85 97 86

Pore volume (cm3 ·g−1 ) 0.0343 0.0395 0.0355

Table 3. Acid site concentration of La-promoted and unpromoted Ni/Al2 O3 catalysts Catalysts Ni/Al2 O3 3 wt% La-Ni/Al2 O3 5 wt% La-Ni/Al2 O3

Acid site concentration (µmol·g−1 ) 1675.93 1667.45 55.95

3.2. Reaction studies The longevity of catalysts was evaluated over a continuous 72 h of reaction with CO2 /C ratio of 0.56 at 1023 K. The performance was probed in terms of the stability of catalytic activity i.e., exit flowrate and product compositions. The parameters of interest were (i) effects of dopant loadings whereby the 3.0 wt% La catalyst was pitted against the virgin

Average pore diameter (nm) 0.99 1.19 0.90

Crystallite size (nm) 12.8 9.1 undetectable

unpromoted Ni/Al2 O3 catalyst and 5.0 wt% La catalyst, as well as (ii) the performance of 3.0 wt% La with and without CO2 inclusion. It can be seen from Figure 5(a) that the conversion over 3.0 wt% La catalysts exhibited excellent stability judging by the conversion profile (Xg % of 100% initially, dropped to 90%) compared with the unpromoted Ni-Al2 O3 catalyst (Xg % of 95% at the onset to circa 50% at the end of 72 h reaction) and the 5.0 wt% La-Ni/Al2 O3 (Xg % of 90% at the onset to circa 50% at the end of 72 h reaction). Furthermore, the exit flowrate (cf. Figure 5b) for the 3.0 wt% La-Ni/Al2 O3 catalyst was also stable whilst the outlet flowrate for the Ni/Al2 O3 and 5.0 wt% La-Ni/Al2 O3 showed greater deterioration over the reaction time. Moreover, the H2 /CO ratio for the 3.0 wt% La promoted catalyst was stable (cf. Figure 5c). In contrast, the H2 /CO ratio for the unpromoted as well as 5.0 wt% La

370

Kah Weng Siew et al./ Journal of Energy Chemistry Vol. 24 No. 3 2015

promoted catalysts has shown a declining trend. This has confirmed that the 3.0 wt% La-Ni/Al2 O3 yielded superior longevity catalytic performance compared with the Ni/Al2 O3 and 5.0 wt% La-Ni/Al2 O3 catalysts. This can be attributed to the higher BET specific surface area owned by the 3.0 wt% La-Ni/Al2 O3 catalyst that has enabled greater access of reactants and less sensitive to carbon deposition.

The effects of CO2 presence on the stability of catalytic performance was tested by comparing the data obtained from both glycerol decomposition (CO2 /C = 0) and also at CO2 /C of 0.56 over the 3.0 wt% La-Ni/Al2 O3 catalyst at 1023 K. Compared with the previous longevity run with CO2 , the catalytic activity suffered stages of deterioration in terms of its glycerol conversion, which has dropped from 95% to 65% (30% reduction) within the 72 h reaction period as shown in Figure 6(a). Most likely, without the presence of CO2 , the deposited carbon on the catalyst surface was unable to be gasified and hence, the amount of deposited carbon would have accumulated (proven by the subsequent TPO analysis). Therefore, the rising amount of deposited carbon covered more active sites and deactivation became more severe. This has led to a continuous catalytic activity reduction for the entire reaction duration. In addition, it was obvious that this non-CO2 system yielded lower CO generation. This could be attributed to the carbon gasification pathway (cf. Equation 7) which was negligible without the presence of CO2 . CO2 + C ↔ 2CO

Figure 5. Effects of La loading on (a) glycerol conversion, (b) flowrate and (c) H2 /CO gaseous product ratio. Reaction condition: T = 1023 K, P = 1 atm, CO2 /C = 0.56, WHSV = 3.6×10−4 mL·g−1 ·h−1 STP

(7)

Figure 6. Effects of non-CO2 reaction system on (a) glycerol conversion, (b) flowrate and (c) H2 /CO gaseous product ratio. Reaction condition: T = 1023 K, P = 1 atm, CO2 /C = 0, WHSV = 3.6×10−4 mL·g−1 ·h−1 STP

Journal of Energy Chemistry Vol. 24 No. 3 2015

Moreover, the reaction flowrate also decreased with similar trend (cf. Figure 6b) whereby H2 : CO product ratio was always maintained at less than 2.0 (cf. Figure 6c). Overall, it can be summarized that the glycerol decomposition without CO2 showed lower longevity stability compared with the CO2 reforming of glycerol. 3.3. Characterization of used catalyst 3.3.1. FESEM & TEM Analyses The used catalysts, after 72 h of continuous reaction, were collected and subjected to both FESEM and TEM analyses (cf. Figures 7 to 10). All the images showed that the deposited carbon was generally comprised of graphitic filamentous type.

371

Figure 7(a) and 7(b) exhibits the FESEM and TEM images of the used catalyst for the unpromoted Ni/Al2 O3 catalyst. It can be observed that when non-promoted Ni/Al2 O3 catalyst was employed in the 72 h reaction run (1023 K, CGR 1.67), most of the deposited carbon has elongated straight nano-filaments with diameter ranging from 10 to 30 nm. According to Terrones et al. hexagon-rich cylinders as a construction of sp2 hybridised carbon can yield straight filaments. The same observation was also revealed by the TEM image that showed hollow tube formation [15]. Moreover, the observed ‘black spot’ can be attributed to the Ni particle which was attached to the filament tip. Ni particle at carbon filament tip was caused by ‘tips growth’ [13]. Significantly, the HRTEM image (cf. Figure 7c) exhibited bamboo-like compartments within the carbon filament which was the result of carbon bulk diffusion through the catalyst particle [14].

Figure 7. FESEM (a), TEM (b) and HRTEM (c) images of deposited carbon on unpromoted Ni/Al2 O3 catalyst after 72 h reaction (1023 K, CGR of 1.67)

Furthermore, the deposited carbon from the 72 h reaction over the 5.0 wt% La-Ni/Al2 O3 was mostly comprised of helical and wavy whisker type thin filaments with diameters of 10 to 30 nm. Similar to the observation for the used Ni/Al2 O3 catalyst, the FESEM and TEM images for the 5.0 wt% La-

catalyst (cf. Figure 8a & 8b) exhibited nano-filament carbon with narrow hollow tube. The Ni particle (black spot) and the bamboo-like compartment can be clearly seen from the HRTEM image (cf. Figure 8c).

Figure 8. FESEM (a), TEM (b) and HRTEM (c) images of deposited carbon on 5 wt% La-Ni/Al2 O3 catalyst (1023 K, CGR 1.67)

For the 3.0 wt% La-promoted Ni/Al2 O3 catalyst, the filamentous whisker carbon (cf. Figures 9 and 10) was found in both reaction runs, viz. 72 h reaction run at 1023 K with

CGR of 1.67 and also reaction without the presence of CO2 . Significantly, the carbon nanotubes formed in the reaction with CGR 1.67 were different from the nanotubes type that

372

Kah Weng Siew et al./ Journal of Energy Chemistry Vol. 24 No. 3 2015

was observed in the non-CO2 system. Obviously, a smooth surface single-wall carbon nanotubes was observed from the reaction with CGR 1.67 (cf. Figure 9a). TEM image of the single-wall nanotube carbon is shown in Figure 9(b). It seems that the filament has thick wall (cf. Figure 9c). In contrast, multi-wall carbon nanotubes were deposited on the catalyst within the non-CO2 system (cf. Figure 10a). It can be seen that the multi-wall carbon nanotubes had rough external surface. The TEM image of the multi-wall nanotube carbon is shown in Figure 10(b). It was also discovered that the surface of the nanotube was coated with amorphous carbon which formed the rough surface (cf. Figure 10c). It was

estimated that the single wall carbon nanotubes possessed diameter within the range of 150 to 200 nm while the multi-wall carbon nanotubes have diameter within the range of 300 to 500 nm. Moreover, the nanotubes from both reaction seems ‘curled’ or in helical shape, which was most probably due to the introduction of pentagon and heptagon pairs into the hexagonal graphite network of the tubular [15]. Nonetheless, the carbon nanotubes revealed in Figures 9 and 10 can be categorized as an open-ended type. This can be assigned to the ‘base-growth’ type mechanism which was caused by a strong interaction between the metal particles (Ni) and Al2 O3 supports [13].

Figure 9. FESEM (a), TEM (b) and HRTEM (c) images of deposited carbon on 3 wt% La-Ni/Al2 O3 catalyst (1023 K, CGR of 1.67)

Figure 10. FESEM (a), TEM (b) and HRTEM (c) images of deposited carbon on 3 wt% La-Ni/Al2 O3 catalyst (1023 K, CGR of 0)

3.3.2. Temperature-programmed oxidation of the used catalysts Based on Figure 11, it can be seen that the highest weight loss (55.0%) belongs to the used Ni/Al2 O3 catalyst which indicated that it has been deposited with highest amount of carbon. The Ni based catalysts with the highest La loading (5.0 wt% La) yielded the lowest weight loss (30.0%) indicating the lowest quantity of deposited carbon compared with the used Ni/Al2 O3 and also the 3.0 wt% La-Ni/Al2 O3 catalysts under a similar reaction condition. Nevertheless, this 5.0 wt%

La-Ni/Al2 O3 catalyst still exhibited a lower catalytic reactivity and rapid carbon deactivation. This is due to its lower BET specific surface area and smaller pore volume. Hence, even a low amount of carbon coverage would have severely deactivated the catalyst. On the other hand, it can be observed that the 3.0 wt% La-Ni/Al2 O3 catalyst employed for the non-CO2 reaction system showed higher deposited carbon compared to the reaction with CO2 /C ratio of 0.56. This indicated that the CO2 in the latter system was able to gasify the deposited carbon and managed to slow down the carbon deactivation effect, hence maintaining its productivity at a slightly higher level throughout the 72 h reaction.

Journal of Energy Chemistry Vol. 24 No. 3 2015

373

Acknowledgements Authors would like to acknowledge the financial support from the Ministry of Education, Malaysia through MTUN-CoE Research Grant (RDU121216) and the FRGS Research Grant (RDU140112). Both Kah Weng Siew and Hua Chyn Lee appreciate the Universiti Malaysia Pahang for a provision of studentship.

References [1] [2] [3] [4] [5] Figure 11. TPO results for La-promoted and unpromoted Ni/Al2 O3 catalysts after longevity runs (T = 1023 K, t = 72 h)

4. Conclusions La incorporation onto the Ni/Al2 O3 catalyst with weight loadings of 3.0 wt% has promoted the Ni metal particles dispersion while enhancing the BET specific surface area. The increment was due to the Ni particle size reduction as a result of the improved metal particle dispersion by the increasing La-doping. However, La dopant at 5.0 wt% level led to an obvious BET specific surface area reduction due to pores blockage. In addition, increasing La loading also reduced the catalyst acidity. The longevity runs revealed that the unpromoted Ni/Al2 O3 and 5.0 wt% La-Ni/Al2 O3 exhibited a more rapid carbon deactivation trend than the one over the 3.0 wt% La-Ni/Al2 O3 due to its superior physicochemical characteristics. This was supported by the TPO analysis. FESEM and TEM images of the used catalysts from this work exhibited filamentous type carbon.

[6] [7] [8] [9] [10] [11] [12] [13]

[14] [15]

Berman A, Karn R K, Epstein M. Green Chem, 2007, 9: 626 Adesina A A. Curr Opin Chem Eng, 2012, 1: 272 Eric D L, Ren T J. Energy Sust Dev, 2008, 7: 79 Lin Y C. Int J Hydrogen Energy, 2013, 38: 2678 Demirbas A. Green Energy and Technology: Methane gas hydrate. Springer, London, New York, 2010 Siew K W, Lee H C, Gimbun J, Cheng C K. J Energy Chem, 2014, 23: 15 Siew K W, Lee H C, Gimbun J, Cheng C K. Int J Hydrogen Energy, 2014, 39: 6927 Levenspiel O. Ind Eng Chem Res, 1999, 38: 4140 Hardiman K M, Trujillo F J, Adesina A A. Chem Eng Process, 2005, 44: 987 Cheng C K, Foo S Y, Adesina A A. Catal Commun, 2010, 12: 292 Foo S Y, Cheng C K, Nguyen T H, Adesina A A. J Mol Catal A, 2011, 344: 28 Liu L. [PhD Dissertation]. National University of Singapore, Singapore, 2012 Teo K B K, Charanjeet S, Chhowalla M, Milne W I. In: Nalwa H S ed. Encyclopedia of Nanoscience and Nanotechnology. American Scientific Publishers United Kingdom, 2003 Martin-Gullon I, Vera J, Conesa J A, Gonza’lez J L, Merino C. Carbon, 2006, 44: 1572 Terrones M, Hsu W K, Kroto H W, Walton D R M. Topics in Current Chemistry: Nanotubes-A Revolution in Materials Science and Electronics. Springer Verlag Berlin Heidelberg, 1999