(439e) Deactivation of Unsupported and Silica-Supported Cobalt Nanoparticles for Ethanol Steam Reforming and CO2 Methanation

Authors: 
Busca, G., University of Genova
Riani, P., University of Genova
Garbarino, G., University of Genova
Cavattoni, T., University of Genova
Deactivation of unsupported and silica-supported cobalt nanoparticles for ethanol steam reforming and CO2 methanation

Paola Riani,a,bGabriella Garbarino,b,c Tullio Cavattonid, Guido Buscab,c

a Università degli Studi Genova, DIFAR - Department of Pharmacy, Genoa, Italy

b INSTM, UdR Genova, Genoa, Italy

c Università degli Studi Genova, DICCA - Department of Civil, Chemical and Environmental Engineering, Genoa, Italy

d Università degli Studi Genova, DCCI - Department of Chemistry and Industrial Chemistry, Genoa, Italy

*Corresponding author: Guido.Busca@unige.it

  1. Introduction

Supported Cobalt catalysts, in particular Co/Al2O3 and/or Co/ZrO2-SiO2, are used industrially for the low temperature Fischer Trøpsch process producing Diesel fuel and waxes from syngases [1]. In this reaction, methane is actually the main product on molar basis. Some studies report on a remarkable activity of cobalt catalysts also for methanation of CO and CO2 (CME) [2] while others [3]ranked low cobalt catalyst for these reactions for low activity, selectivity and stability. Cobalt catalysts are also of interest for Ethanol Steam Reforming (ESR), a potentially useful process for producing renewable hydrogen from bioethanol.

Concerning catalytic activities, studies on these reactions sometimes suggest that the support plays an important role in the activation of reactants. However, we previously found that both cobalt [4] and nickel [5] and their alloy [6] may be very active for ESR also as unsupported metal nanoparticles (NPs), and as NPs supported on inert carriers [7], suggesting that the support is not necessary for reactant activation for this reaction. However, the same Ni NPs were found not active in the CME reaction.

To go deeper in our studies, we prepared unsupported and silica – supported cobalt NPs starting from different precursors and with different procedures, and we tested them in both CME and ESR. The aim was to reveal the role of preparation method and support on activity and short-term stability in these two reactions.

  1. Experimental

Data on the preparations of the catalysts under study are reported in the Table.

Notation

Support

Metal load

Precursor

Reductant

Preparation/ Separation

A

--

¥

Chloride

NaBH4

filtration

B

--

¥

Nitrate

NaBH4

Filtration/sonication

C

--

¥

Carbonyl

--

Filtration

D

SiO2 gel

20 %

Acetate

--

Wet Impregnation

E

SiO2 gel

20 %

Nitrate

--

Wet impregnation

Catalysts have been characterized by XRD, UV-vis-NIR, IR (skeletal), FE-SEM, before and after reaction. Catalytic reactions were realized in continuous fixed bed reactors. Product analyses were performed using GC, GC-MS and on-line IR.

3. Results and discussion.

Catalysts A and B revealed essentially no activity in CME in the Trange 523-773 K, although they produced some amount of CO by reverse Water Gas Shift (r-WGS). However, catalysts became rapidly deactivated also with respect r-WGS activity. Characterization data suggest that low activity is associated to the presence of impurities arising from the precursor, in particular due to boron contamination.

In contrast, catalyst C showed significant activity in CME in the temperature range, with selectivity to methane higher than to CO, but with an evident deactivation of CME upon time on stream at any temperature between 523 and 773 K. On the other hand, the catalyst showed nearly constant r-WGS activity upon time on stream, producing CO. After using it at 773 K the catalyst was almost completely deactivated with respect to CME while showing constant activity towards r-WGS, maximum yield to CO being 45 % at 773K. A very similar behavior was found for supported catalyst D, which also showed initial CME activity higher than r-WGS activity, the former however affected by fast deactivation, while the latter is nearly stable. In both cases FE-SEM data indicate that deactivation is mostly associated to the formation of encapsulating carbon. Catalyst E instead shows significant activity in CME producing up to 67% methane yield at 673 K, with the coproduction of lower amounts of CO. The activity is affected by very slow deactivation phenomena in this case, apparently associated to sintering of NPs. Both cubic and hexagonal Co is found in all spent catalysts. The lower activity and lower stability towards methanation of sample D with respect to sample E (both 20% Co/SiO2) can be tentatively attributed to the larger cobalt particle size in the former. Thus, smaller particles are more active in methanation and less subjected to deactivation by encapsulating carbon. In any case, the r-WGS activity producing CO seems to follow a parallel and independent path with respect to CME. This activity is attributed to CoO and/or Co-silicate species evident in the spent catalysts.

All catalysts under study are active in ESR, showing that boron impurities do not hinder such a reaction. All catalysts are highly selective for ethanol dehydrogenation to acetaldehyde at 523-673 K, while produce ESR predominantly at T > at 673-773 K. This suggests that high temperatures are needed for the activation of water. NPs produced from reduction with NaBH4, thus containing boron impurities, may be even more selective to hydrogen (yields > 85 %) perhaps because boron impurities may selectively poison the sites for producing methane by acetaldehyde cracking.

The active phase for ESR is represented by cubic cobalt, whose ESR activity is greater for smaller particles. Slow deactivation is mainly associated to sintering. However, these particles are also active in producing carbon nanotubes. On the other hand, the formation of carbon nanotubes does not hinder catalytic activity at the timescale of a few hour laboratory experiment.

Depending on the precursor, the catalyst may deactivate by forming larger Co-containing nanoparticles embedded on the silica/cobalt silicate surface. These particles, possibly containing also silicate species, are also not active in producing carbon nanotubes, i.e. they are unable to both activate water and favor C-C bond formation.

The data reported above show that support is not needed for catalytic activity in CME and ESR but may be important for catalyst stability. However, the silica support may also favor catalyst deactivation of cobalt NPs.

References.

[1] Busca, G.; Heterogeneous Catalytic Materials, Elsevier, 2014, pp. 314-315.

[2] Le, T.A.; Kim, M.S.; Lee, S.H.; Park, E.D.; Top. Catal. 2017 60, 714–720

[3] Kuznecova, I.; Gusca, J.; Energy Procedia 2017, 128, 255–260

[4] Riani, P.; Garbarino G.; et al. Appl. Catal. A: Gen. 2016, 518, 67–77

[5] Riani, P.; Garbarino G.; et al. J. Mol.Catal. A: Chem. 2014, 383–384, 10–16

[6]Garbarino, G.; Cavattoni, T.; et al. Catal. Lett. 2019, 149, 929-941

[7] Riani, P.; Garbarino G.; et al. J. Chem. Techn. Biotech. 2019, 94, 538-546