(617c) Coke Elimination with High Hydrogen Yield during Reforming of Ethanol over New Ni Incorporated Mesoporous Mg-Al Due to Microwave Heating and Catalyst Synthesis Route

Dogu, T., Middle East Technical University
Gunduz, S., The Ohio State University

Due to its production possibilities from different renewable feedstock through fermentation, its low toxicity and high hydrogen content, ethanol has been considered as a potential resource for hydrogen production. Six moles of hydrogen may in principle be produced per mole of ethanol, through the steam reforming (SRE) process.

                                    C2H5OH + 3H2O ↔ 2CO2 + 6H2

However, thermodynamic limitations of water gas shift reaction, as well as the occurrence of ethanol decomposition, dehydrogenation and dehydration reactions limit the approach to this maximum yield. Coke formation through Boudouard reaction is also an important problem to be resolved in this process [1-3].

                                    2CO ↔ C + CO2

In this study, a set of new Ni-Mg incorporated mesoporous alumina materials were synthesized by impregnation and one-pot routes and catalytic performances of these materials were tested in steam reforming of ethanol (SRE) in a focused microwave reactor, as well as in a convectively heated tubular reactor. Importance of catalyst synthesis route on the product distribution and the positive effects of microwave heating on coke elimination and reactor stability were shown.

Mg incorporated mesoporous alumina catalyst support material (Mg-MA) having an ordered pore structure was synthesized following an “Evaporation Induced Self Assembly” procedure, by using P123 as the surfactant. Surface area and the mean pore diameter values of Ni impregnated Mg-MA catalyst (Ni@Mg-MA), containing Ni/Al and Mg/Al ratios of 0.1 and 0.25, were 213 m2/g and 9.7 nm, respectively.  Ni-Mg containing mesoporous alumina catalysts were also synthesized by impregnation of Ni and Mg into mesoporous alumina (Ni-Mg@MA) and also by a direct synthesis route (Ni-Mg-MA). Surface area values of these mesoporous materials were 117 m2/g and 128 m2/g, respectively. Comparison of the catalytic performances of the synthesized materials in steam reforming of ethanol, which was performed at 550 oC, showed the importance of synthesis route on hydrogen yield, product distribution, as well as on coke formation. Product distributions were shown to be closely related to the cluster size, distribution and oxidation state of nickel, which was strongly dependent on the catalyst synthesis route.  Best catalytic performance was obtained with Ni@Mg-MA, giving a hydrogen yield of about 5.1 moles hydrogen per mole of ethanol reacted. This was quite close to the thermodynamic limit of reforming and water gas shift reactions.   However, coke formation was about 60% of the weight of this catalyst within a reaction period of four hours in the convectively heated reactor.

Steam reforming of ethanol was also performed in a reactor which was heated with a focused microwave source. In this case, synthesized catalyst (0.15 g) was mixed with activated carbon (0.05 g) and charged into the quartz tubular reactor. Catalyst bed was heated by a focused microwave source, while the feed stream was at 150 oC. Catalyst-active carbon mixture absorbed only about 10 W to achieve the desired bed temperature. Complete conversion of ethanol was achieved at 550 oC in both MW and convectively heated reactors, at a space time of 1.8 s.gcat.ml-1. Hydrogen yield value obtained in the MW reactor was about the same as in the convectively heated reactor. However, performance of the MW reactor was superior from the point of view of coke elimination and reactor stability. While significant coke deposition was observed in the convectively heated reactor, almost no coke formation was observed in the focused MW reactor, within the same reaction period. Variation of product mole fractions with reaction time was also highly stable in the MW system. In fact, time-on stream tests (extending up to 24 h) performed in the microwave reactor also proved highly stable operation with very low coke formation.

Another interesting difference of the two reactor systems was related to the product distributions. In the case of convectively heated system, CO2 mole fraction (about 0.19) was higher than CO mole fraction (about 0.08), while in the case of MW system CO mole fraction (about 0.18) was higher than CO2 mole fraction (about 0.08). Higher CO2 concentration in the convectively heated reactor for the same hydrogen yield is an indication of high amount of coke formation through Boudouard reaction, which is thermodynamically favored at low temperatures. Considering the endothermic behavior of SRE reaction, temperature gradients within the convectively heated reactor are inevitable. Apparently, temperature uniformity of the catalyst bed achieved in the MW system eliminated cold spots in the bed and hence lowered coke formation at the low temperature locations. Experimental results obtained with the Co impregnated Mg-MA catalysts also supported these conclusions and proved the superiority of focused MW heating over convective heating, for coke minimization and stable operation in SRE.

 Acknowledgements: TUBITAK Project (111M338), and Turkish Academy of Sciences TUBA.


[1] Gunduz S., Dogu T., Ind. Eng. Chem. Res., 51 (2012) 8796.

[2] Arslan A., Gunduz S., Dogu T., Int. J. Hydrogen Energy, 39 (2014) 18264.

[3] Gunduz, S., Dogu T., Appl. Catal B: Env. 168 (2015) 497.