Co-based hydrotalcites as new catalysts for the Fischer-Tropsch synthesis process

Fuel 119 (2014) 62–69

Contents lists available at ScienceDirect

Fuel

journal homepage: www.elsevier .com/locate / fuel

Co-based hydrotalcites as new catalysts for the Fischer–Tropschsynthesis process

0016-2361/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.fuel.2013.11.014

Abbreviations: FTS, Fischer–Tropsch Synthesis; HTlc, hydrotalcite-likecompound.⇑ Corresponding author. Tel.: +39 0250314293; fax: +39 0250314300.

E-mail address: [email protected] (A. Di Fronzo).

A. Di Fronzo a,⇑, C. Pirola a, A. Comazzi a, F. Galli a, C.L. Bianchi a, A. Di Michele b, R. Vivani c, M. Nocchetti c,M. Bastianini c, D.C. Boffito d

a Dipartimento di Chimica, Università degli Studi di Milano, Via Golgi 19, 20133 Milan, Italyb Dipartimento di Fisica, Università degli Studi di Perugia, Via Pascoli, Perugia 06123, Italyc Dipartimento di Chimica, Università degli Studi di Perugia, Via Elce di Sotto, Perugia 06123, Italyd Département de Génie Chimique, École Polytechnique de Montréal, 2900, boul. Édouard-Montpetit, Montréal (QC), Canada

h i g h l i g h t s

� A series of Co-based hydrotalcites wassynthesized with a modified-ureamethod.� Fresh and used Co-based hydrotalcite

catalysts were characterized bynumerous methods.� The catalysts were tested for Fischer–

Tropsch Synthesis (FTS) using a fixedbed reactor.� CO conversion and product

selectivities are closely related to theCo amount.� FT results make Co-based

hydrotalcites new potential catalystsfor FT.

g r a p h i c a l a b s t r a c t

a r t i c l e i n f o

Article history:Received 9 October 2013Received in revised form 6 November 2013Accepted 7 November 2013Available online 22 November 2013

Keywords:HydrotalciteUrea methodFischer–Tropsch synthesisCobalt-based catalysts

a b s t r a c t

A series of ternary hydrotalcites in the nitrate form was prepared with a modified-urea method to obtainactive Co-based catalysts for the Fischer–Tropsch synthesis. An optimization study concerning theamount of cobalt in the catalysts (range 5–35 wt%.) and the reaction temperature (220–260 �C) isreported. All the samples were characterized by several methods, including XRPD, ICP-OES, SEM, TEM,FT-IR, BET, TPR and TG and tested in a fixed bed reactor. The results suggest the possibility of using syn-thetic hydrotalcites as Co-based catalysts for the Fischer–Tropsch synthesis.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction material. Syngas can be manufactured from CH4, coal or, as a

Fischer–Tropsch synthesis (FTS) is a well-known industrial pro-cess which starts from syngas (mixture of H2, CO, CO2) as a raw

new tendency, from biomass. Nowadays it is imperative to developeconomical and energy-efficient processes for the sustainable pro-duction of fuels and chemicals alternative to the ones derived frompetroleum. FTS is a well-established industrial process throughwhich this aim can be achieved. FTS has been commercially usedfor many years and is still attracting much attention as a meanof producing transportation fuels due to the variety of raw materi-als that can be used (syngas produced from coal, natural gas and

A. Di Fronzo et al. / Fuel 119 (2014) 62–69 63

biomass) [1,2]. The main industrial alternative process to convertsyngas is the methanol synthesis [3–5]. The essential target ofFTS is to produce paraffins and olefins with different molecularweights and to limit the formation of methane and CO2 [6]. FTSusually requires catalysts based on iron or cobalt. Iron catalystsare often preferred over cobalt-based ones especially when con-verting syngas with molar H2/CO ratio lower than 2 (correspondingto the stoichiometry required by the FTS reaction). This is also thetypical H2/CO ratio of syngas produced from biomass or coal [7,8].In fact, iron-based catalysts are active towards the Water Gas Shiftreaction (WGS: CO + H2O M CO2 + H2), increasing the H2/CO ratio.On the other hand when feeding a syngas mixture with a H2/CO ra-tio close to 2, cobalt catalysts are preferred due to their high selec-tivity towards heavy hydrocarbons and their low activity in WGSreaction limiting the CO2 formation. Moreover Co-based catalystsexhibit longer life-time and higher CO conversion compared withFe based catalyst [8]. Cobalt is usually find as a supported catalystin the FTS process.

In this work we propose double- and triple- layered hydroxides,also known as hydrotalcite-like compounds (HTlc) as a new kind ofFTS catalyst. HTlc can be easily prepared and essentially consist ofmixed metal hydroxides, where specific metal atoms are homoge-neously dispersed at an atomic level. HTlc are represented by theempirical formula [M(II)1�xM(III)x(OH)2]x+[An�

x/n]x�mH2O whereM(II) is a divalent cation such as Co, Mg, Zn, Ni, or Cu, M(III) is atrivalent cation such as Al, Cr, Fe or Ga; An� is an anion of chargen and m the molar amount of co-intercalated water [9].

If calcined at appropriate temperatures, the random distribu-tion of cations, characteristic of the hydroxide phase, is maintainedin the resulting mixed oxide. HTlc -based materials have been re-cently reported as good catalysts for several processes in the en-ergy field, such as hydrogen production by steam reforming ofmethanol and ethanol [10,11], photocatalytic water splitting [12]and methane reforming [13]. Up to now the only study reportedin the literature on the use of HTlc as FTS catalysts concerns theiruse as supports, in which the catalytically active metal is dispersedon the HTlc surface [14].

In the present paper a new kind of FTS catalyst, in which the ac-tive metal is part of the structural core of the HTlc, has been syn-thesized and tested. A series of Co–Zn–Al hydrotalcites, withincreasing Co content (5–35 wt%) was synthesized by the modifiedurea method [15,16] and characterized. Activity tests conducted ina fixed bed reactor resulted in satisfactory catalytic performances.Moreover, the structural and catalytic properties of these materialswere verified at FTS operating conditions and correlations betweencatalyst features and efficiency towards light and heavy hydrocar-bons selectivities were achieved.

2. Experimental part

A series of ternary HTlc, with general formula [CoxZn(1�x�y)Aly

(OH)2](NO3)y�0.5H2O, was synthesized by a modified-urea method[15]. Different volumes of the solutions of the metal nitrates, allat a concentration of 0.5 M, were mixed to obtained either aAl/(Co + Al) or Al/(Co + Al + Zn) molar ratio of 0.3, as indicated inTable 1.

Solid urea was added to the solutions, in a molar ratio of 4vs. Al. The obtained solutions were maintained at the reflux tem-perature in an open flask for 48 h. The precipitates were sepa-rated by centrifugation, washed with water, and then dried at80 �C.

The obtained materials were characterized by X-ray powder dif-fraction (XRD: PANalytical X’Pert Pro, Cu Ka radiation) operating at40 kV and 40 mA, step size 0.0170 2h� and step scan 20 s. The metalcontent of samples was determined by inductively coupled plasma

optical emission spectrometry (ICP-OES), using a Varian LibertySeries instrument.

Field emission scanning electron microscopy (SEM) imageswere obtained using a LEO1525 instrument after depositing thecatalysts onto the sample holder and sputtering coating with chro-mium. The elemental mapping of metals was obtained by using en-ergy dispersive X-ray spectroscopy (EDS) with a Bruker QuantaxEDS instrument. Transmission electron microscopy (TEM) imageswere obtained using a Philips 208 instrument.

FT-IR spectra of different samples, dispersed in KBr pellets, wererecorded at room temperature using a Bruker IFS113 V spectrome-ter. Typically, each spectrum was obtained at a resolution of1 cm�1 in the spectral region 400–5000 cm�1.

Nitrogen adsorption–desorption isotherms were determinedwith a Micromeritics ASAP 2010 instrument at �196 �C on samplesoutgassed overnight at 100 �C. The specific surface area was calcu-lated by the BET method. Micropore volume and external surfacearea were evaluated by the t-plot method. Mesopore characteriza-tion was performed using the Barrett, Joyner, Halenda method.

Conventional temperature-programmed reduction experiments(TPR) were performed using a Thermoquest Mod. TPR/D/O 1100instrument. The samples were initially pre-treated in a flow of ar-gon at 200 �C for 0.5 h. After being cooled down to 50 �C, the H2/Ar(5.1% v/v) reducing mixture was flushed through the sample at30 mL min�1 and the temperature increased from 50 to 900 �C ata constant rate of 10 �C min�1.

Water and nitrate content of the solids was determined by ther-mogravimetric (TG) analysis with a Netzsch STA 449C apparatus, inair flow, and 10 �C min�1 heating rate.

FTS was performed in a fixed bed reactor already described bythe authors in previous works [17,18]. 1 g of fresh catalyst mixedwith 1 g of a-Al2O3 as a diluting material. This diluting materialis absolutely inert for FTS both in term of activity and wateradsorption [17]. Also, since alumina is not a good thermal conduc-tor, its dilution of the catalyst allows a good control of the temper-ature inside the catalytic bed (the maximum increase of thetemperature experimentally verified by an axial thermocouplewas equal to 5 �C). The calcined catalysts were initially reducedin situ by flowing hydrogen for 4 h at 90.0 Nml min�1, 350 �C and0.8 MPa. They were then test at standard conditions by flowingsyngas (H2/CO = 2/1) at 46.8 Nml min�1 plus 5.0 Nml min�1 of N2

as an internal standard in order to ensure accurate mass balances,at 2.0 MPa and 220–260 �C. Analyses of the gas-phase products(C1–C7) were performed with an on-line micro gas-chromatograph(Agilent 3000). Liquid products were collected in a trap at 5 �C and2.0 MPa. A mass molar balance was performed for each FT run,resulting in a maximum error of ±5% on molar basis.

The characterization analyses were performed on the fresh,activated and used catalysts. ‘‘Fresh catalysts’’ indicate the samplesas prepared. ‘‘Activated catalysts’’ indicate the samples charged inthe FTS reactor and reduced using the activation procedure previ-ously described (in situ by flowing hydrogen for 4 h at90.0 Nml min�1, 350 �C and 0.8 MPa) and removed from the reac-tor. ‘‘Used catalysts’’ indicate the samples after their use in theFischer–Tropsch synthesis reaction.

3. Results and discussion

The aim of this work was to synthesize double or triple HTlccontaining respectively Co and Al or Co, Zn and Al, with high crys-tallinity and good dispersion of the metals inside the bruciticsheets, in order to have new, efficient and selective catalysts forFTS. In a recent work Zn–Al hydrotalcites in nitrate form wereobtained in a single step by adjusting the classic urea method[16]. The same synthetic conditions were used to prepare Co–Al

Table 1Details of the synthesis, composition and specific surface area (SSA), micro and mesopore volume of samples.

Sample wt.% Co % M sol.a Xb Yb SSA (m2 g�1) Vmicro (cm3/g) Vmeso (cm3/g)

Zn Co Al

Co5 5.1 55 15 30 0.10 0.33 18 <0.01 0.02Co15 16.6 35 35 30 0.32 0.29 7 <0.01 0.01Co35 35.3 – 70 30 0.67 0.33 6 <0.01 <0.01

a % M sol. = molar percentage of metals in the precipitation solution.b General formula [CoxZn(1�x�y)Aly(OH)2](NO3)y�0.5H2O.

Fig. 1. XRD pattern of the fresh samples: (a) Co5, (b) Co15, and (c) Co35.

64 A. Di Fronzo et al. / Fuel 119 (2014) 62–69

and several Co–Zn–Al HTlc in nitrate form, with increasing amountof cobalt. Fig. 1 shows the XRD patterns of the samples listed inTable 1. The interlayer distance of 8.9 Å, determined from the firstXRD reflection, is compatible with the presence of nitrate betweenthe sheets [19]. The composition of the solids are slightly differentfrom those of the precipitating solutions, in particular the solidscontain more Al. The composition and specific surface area (SSA)of the samples are reported in Table 1. Note that the SSA decreases

Fig. 2. SEM (left) and TEM (right) images of fresh

with the increase of the amount of cobalt and that all the samplesexhibit low or negligible micro- and mesopore volume.

The morphology of the synthesized HTlc was investigated bySEM and TEM. Images of Co15 and Co35, selected as representativesamples (Fig. 2 a–d), show that they are constituted by homoge-neous aggregates of hexagonal and platy particles of few hundredsnanometers of thickness and with a dimensional range between 2and 5 lm. In TEM micrographs the hexagonal morphology of HTlcmicrocrystals is more clear. The relatively large dimensions andhigh crystallinity of particles reflect a low specific surface area, inthe range of 6–18 m2 g�1 (Table 1). Composition maps of Co15(Fig. 3), obtained with coupled SEM-EDS analysis, highlight thehomogeneous Co, Zn and Al dispersion over the entire analyzedspot area, without creating single-metal domains indicating goodmetal distributions in the samples. Similar dispersion characteris-tics were obtained for all the samples studied, although not re-ported here for the sake of brevity.

TG analyses were performed in order to study the thermalbehaviour of the catalysts, leading to similar results for all the sam-ples (Fig. 4). As explained by Pérez et al [20], TG curves are com-posed of two endothermic steps: between 20 and 180 �C thesamples loose hydration water from the interlayer space, whilethe loss of constitutional water, due to condensation of hydroxylgroups starts around 200 �C together with nitrate decomposition,leading to the formation of mixed metal oxides.

The active phase in the FTS is the metallic cobalt, while the HTlcmaterials contain Co(II) ions randomly dispersed inside the brucitic

samples Co15 (a and b) and Co35 (c and d).

Fig. 3. SEM image of fresh sample Co15 and the corresponding EDS images of the metals: Al (red), Co (white) and Zn (pink). (For interpretation of the references to colour inthis figure legend, the reader is referred to the web version of this article.)

Fig. 4. TG curve of fresh Co35. Heating rate: 10 �C min�1, atmosphere: air30 ml min�1.

Fig. 5. TPR profile of the fresh samples Co5, Co15, Co35. The activation temperatureused in the FTS reactor before the catalytic test is indicated with the dashed line.

Table 2Reduction temperature and % of cobalt reduction of indicated samples.

Sample TPR

Tmax (�C) % Co Red

Co5 328 58Co15 278 54Co35 295 83

Fig. 6. XRD pattern of activated Co15. * ZnO, # Co3O4, CoAl2O4, ZnAl2O4, § CoO.

Fig. 7. FT-IR spectra of the fresh and activated Co15 and of the reference Co3O4.

A. Di Fronzo et al. / Fuel 119 (2014) 62–69 65

layers. Therefore, in order to have an active catalyst, a reductionprocedure is required to form the cobalt particles. TPR analyseswere performed to study the reduction process and select the bestconditions for the catalyst activation. Fig. 5 shows the TPR profileof the samples Co5, Co15, Co35, while Table 2 reports the reduction

temperature and the percentage of reduced Co. All the profilesexhibit two regions of reduction; the first at lower temperatures(below 400 �C), which is due to the reduction of Co3O4 while thesecond peak, above 700 �C, indicates the presence of hardly reduc-ible species. These species are probably spinel-type mixed oxides

Fig. 8. SEM (left) and TEM (right) images of activated Co15 (a and b) and Co35 (c and d).

66 A. Di Fronzo et al. / Fuel 119 (2014) 62–69

formed during the thermal treatment. TPR profiles of CoOx mixedoxides is well described in the literature [21–27]. The reductionprofile of Co3O4 in the low temperature region consists in twopeaks, corresponding to the reduction of Co3+ to Co2+ and Co2+ toCo0. According to Alvarez et al. [28], these two peaks are well sep-arated for samples with small particle size, while an intermediateparticle size causes the overlapping of the two reduction stepsresulting in a complete reduction with only one maximum at anintermediate temperature (328 �C). The TPR profiles of our HTlc(Fig. 5) are consistent with the latter case. Based on the TPR results(Table 2), the tested catalysts were activated at 350 �C for 4 h un-der hydrogen atmosphere, in order to reduce the Co ions to metal-lic Co.

Note that the sample Co15 exhibit the lowest reduction temper-ature that is significantly lower than the activation temperatureadopted in the FT process (350 �C). A low reduction temperaturemay favour the presence of catalyst in its reduced state duringthe catalytic process. The% of reduced Co evaluated by TPR is gen-erally high. However, these values strongly depend on reductionconditions (pressure, temperature ramp, gas flow, etc.) that are dif-ferent from those used in the FT reactor, as detailed in the experi-mental section.

Activated HTlc were first characterized by XRD. As an example,the XRD pattern of activated Co15 is reported in Fig. 6. The patternshows the presence of very poorly crystalline ZnO and a spinelphase, such as Co3O4, CoAl2O4, and/or ZnAl2O4 (note that mixed-oxide spinel phases show very similar XRD patterns, irrespectiveof their composition). Moreover, the reflection at 43� 2h has beenassigned to a CoO phase with very low crystallinity and no metalaggregation is observed. The sample after the reduction treatmentdoes not show metallic Co phases indicating the presence of activemetal atoms homogeneously dispersed at the nanometer level[29]. According to Jong et al. [30] and Den Breejen et al. [31] co-balt-based catalysts need large particle size of the active speciesto create optimal domains of active sites because the activity dropsfor particles smaller than 6 nm.

In order to better understand the behaviour of the materialsafter activation, FT-IR spectra of the sample Co15 (as a representa-tive sample), calcined and activated, and Co3O4 have been regis-tered (Fig. 7). The spectrum of Co15 showed the typical

absorption bands of the nitrate anion at 1377 cm�1 present inthe interlayer region of HTlc. After calcination this band disap-peared due to nitrate degradation, as indicated in TG analysis(Fig. 4) and the spectrum presented the typical absorptions ofCo3O4 spinel phase at 667 cm�1 and 566 cm�1. The activation pro-cess leads to the formation of a material with very wide bandsattributed to the Co3O4 phase.

SEM and TEM images of two activated samples are reported inFig. 8a–d. The results indicate that the material keeps the originalmorphology with hexagonal crystals of micrometric dimensions.At higher magnification (TEM images) the segregation of denseparticles homogeneously distributed on the surface and havingdimensions of about 8 nm, probably due to metallic Co, was evi-dent. The absence of diffraction peaks in the XRD pattern typicalof metallic Co phase can indicate that the Co crystalline domainsare very small (at most a few nanometers). As already describedin the recent literature, the interaction between Co and the spinelphase can stabilize the metallic nanoparticles and avoid sintering[32].

The surface area of these two activated samples was deter-mined by BET method. The values obtained are 73.4 m2 g�1 forthe sample Co15 and 78.3 m2 g�1 for the sample Co35, both higherthan the untreated materials.

The activated samples were tested at different temperatures inthe fixed bed reactor, following the procedure reported in theexperimental section. The FTS results reported in Fig. 9 and Table 3demonstrated that this new kind of catalyst, where the cobalt isincorporated into the HTlc structure rather than supported onthe HTlc, are undoubtedly active for this synthesis. This conclusionis not obvious. In fact, in the common HTlc structure, cobalt is ide-ally present as individual cobalt ions; this particular catalyticstructure is very different from those of the traditional cobalt sup-ported catalyst for FTS where Co is present in metallic form [33].

As expected, for each catalyst the activity is strongly influencedby the reaction temperature: the higher the temperature, the high-er the CO conversion, but also the selectivity towards CO2, CH4 andlight hydrocarbons is favoured by a higher temperature. The COconversion is higher for the two samples with a larger amount ofcobalt, i.e. Co15 and Co35. In particular, Co15 exhibits the highestCO conversion at all the selected temperatures. In FTS it is

Fig. 9. %Molar CO conversion for Co5 ðrÞ, Co15 ðjÞ, Co35 (N), at different reactortemperatures after 24 h of reaction.

Table 3FTS products selectivities at different reactor temperatures.

Sample Temperature (�C) Products selectivity% C2+yield

CO2 CH4 6C7 >C7

Co5 220 0.3 2.9 3.3 93.4 16.5235 0.3 5.9 10.4 83.4 19.4245 1.1 8.5 12.7 77.6 22.4

Co15 220 1.4 10.1 17.0 70.5 46.7235 8.1 26.2 47.3 18.5 49.8

Co35 220 1.4 3.9 10.3 84.4 25.0235 1.8 9.6 23.4 65.1 39.5245 5.1 25.7 65.6 10.5 47.3

6C7: all the hydrocarbons in the range C2–C7.>C7: all the hydrocarbons greater than C7.Product ‘‘i’’ selectivity=(moles C in product i)/(converted moles C) � 100.C2+yield = CO conversion � (selectivity 6 C7 + selectivity > C7) � 10�2.

Fig. 10. XRD pattern of used Co15 after catalytic runs (a) at 235 �C and 20 bar and(b) at 260 �C and 20 bar. * ZnO, # Co3O4, CoAl2O4, ZnAl2O4, § CoO.

A. Di Fronzo et al. / Fuel 119 (2014) 62–69 67

fundamental to obtain low quantities of CH4 and CO2 (undesiredproducts) to favour the formation of higher hydrocarbons. For thisreason, temperatures in the 220–235 �C range seem to be moresuitable than the higher ones.

Moreover, Table 3 shows that Co15 exhibits the highest CO con-version and the highest C2

+ total yield (without considering CH4

and CO2, see note in Table 3) also at the lowest temperature(220 �C). This result confirms that Co15 is the best performing cat-alyst obtained in this study.

It is important to highlight that the aim of this work was theevaluation of the possibility to use HTlc as a new kind of catalystfor the Fischer–Tropsch process rather than a quantitative compar-ison with other kinds of traditional FTS catalysts. Moreover, a reli-able comparison between HTlc and traditional cobalt based FTScatalysts is very difficult, due to the different structural and mor-phological features (surface area, metal dispersion, morphologicalstructure, reduction properties, porosity and so on), which are in-volved in the very complex catalytic systems for the FTS. Neverthe-less, from a qualitative point of view, it is possible to conclude thatthis new kind of catalytic materials for FTS process give resultsfully comparable with those obtained by traditional supportedcobalt. As general behaviour, it is possible to state that cobaltcatalysts are characterized by high CO conversion, high heavyhydrocarbons selectivity and low light hydrocarbons, CH4

and CO2 selectivity. The wide literature concerning traditionalCo-based catalysts is rich of different examples that, dependingon several operative parameters and preparation procedure, givedifferent FTS results, but always following the general trends

reported by the previous papers. Some exhaustive examples areshown in a recent review of Qinghong et al. [34] and Muthuet al. [35]. The results reported in Table 3 confirm this trend alsofor HTlc, where different products selectivities, typical of cobalt-based catalyst in FTS, are recognizable.

The stability of these catalysts in the operating conditionsadopted in the catalytic tests (P 6 2.0 MPa and T 6 350 �C) wasevaluated by comparative analysis between fresh samples andsamples discharged from the reactor maintained at high pressureand temperature.

In order to better understand the stability of the materials,XRD patterns of sample Co15 after the catalytic tests at two dif-ferent temperatures (235 �C and 260 �C) were collected (Fig. 10).At the lower temperature (Fig. 10a) the pattern is close to that re-corded just after the reduction process (Fig. 6). Therefore we canhypothesize that the catalyst does not change as a consequence ofthe FTS process. On the other hand, the XRD pattern of the cata-lyst recovered after the reaction at 260 �C (Fig. 10b) shows morecrystalline ZnO and spinel phase, and the strong reduction of thepeak assigned to CoO. This may indicate that during the FTS pro-cess, at high temperatures, cobalt ions crystallized into a spinelphase.

SEM images of Co5, Co15 and Co35 after the catalytic run at260 �C are shown in Fig. 11. Samples containing Zn (Co5 andCo15) reveal the presence of a nanometric phase crystallized onthe catalyst surfaces. The absence of this nanometric phase onCo35 surface (which does not contain Zn) may indicate the pres-ence of ZnO, according to the XRD pattern (Fig. 10b).

Fig. 11. SEM images of Co5 (a and b); Co15 (c and d), Co35 (e and f) after catalytic run at 260 �C and 20 bar and at different magnifications.

68 A. Di Fronzo et al. / Fuel 119 (2014) 62–69

4. Conclusions

New Co-based hydrotalcite-like compounds were synthetizedand used as catalysts in the Fischer–Tropsch synthesis. Differentamounts of cobalt and different process temperatures were inves-tigated. The catalysts under study are active in the FTS. The cata-lytic activity of the samples strictly depends on the temperature,as expected. CO conversion and process selectivity towards lightand heavy hydrocarbons are closely related to the cobalt amountin the catalysts but not in a linear way.

The reduction under a H2 flow of the Co-based hydrotalciteswas achieved in just one step at temperatures lower than 350 �C,as shown in the TPR results. Very likely, the relatively large dimen-sions and high crystallinity of these catalyst precursors favouredthe creation of optimal domains of active sites.

The aim of this work was the verification of the possibility toapply this new kind of catalysts in the FTS. The good results ob-tained will lead us in pursuing further studies about HTlc as cata-lytic materials.

Future studies may involve the investigation of the effects of thevarious parameters, such as morphology/size of crystallites, or theaddition of small amounts of promoters such as Ru in the compo-sition of the Co-hydrotalcites.

References

[1] Hadnadev-Kostic M, Tatjana J, Vulic T, Marinkovic-Neducin R, Nikolic AD, JovicB. Mg–Fe-mixed oxides derived from layered double hydroxides: a study of thesurface properties. J Serb Chem Soc 2011;76(12):1661–71.

[2] Fu T, Jiang Y, Lv J, Li Z. Effect of carbon support on Fischer–Tropsch synthesisactivity and product distribution over Co-based catalysts. Fuel Process Technol2013;110:141–9.

[3] Manenti F, Garcon-Leon AR, Bozzano G. Energy-process integration of the gas-cooled/water-cooled fixed-bed reactor network for methanol synthesis. ChemEng Trans 2013;35:1243.

[4] Manenti F, Cieri S, Restelli M, Bozzano G. Dynamic modelling of the methanolsynthesis fixed-bed reactor. Comput Chem Eng 2013;48:325–34.

[5] Manenti F, Cieri S, Restelli M. Considerations on the steady-state modeling ofmethanol synthesis fixed-bed reactor. Chem Eng Sci 2011;66(2):152–62.

[6] Zhang Q, Kang J, Wang Y. Development of novel catalysts for Fischer–Tropsch synthesis: tuning the product selectivity. Chem Cat Chem2010;2:1030–58.

[7] Pirola C, Bianchi CL, Di Michele A, Diodati P, Vitali S, Ragaini V. High loading Fe-supported Fischer–Tropsch catalysts: optimization of the catalystperformance. Catal Lett 2009;131:294–304.

[8] Hirsa M, Torres G, Krijn P. Catalysts for production of lower olefins fromsynthesis gas: a review. ACS Catal 2013;3:2130–49.

[9] Trifirò F, Vaccari A, Cavani F. Hydrotalcite-type anionic clays: preparation,properties and applications. Catal Today 1991;11:173–302.

[10] Montanari T, Sisani M, Nocchetti M, Vivani R, Delgado M, Ramis G, et al. Zinc–aluminum hydrotalcites as precursors of basic catalysts: preparation,characterization and study of the activation of methanol. Catal Today2010;152:104–9.

[11] Resini C, Montanari T, Barattini L, Ramis G, Busca G, Presto S, et al. Hydrogenproduction by ethanol steam reforming over Ni catalysts derived fromhydrotalcite-like precursors: catalyst characterization, catalytic activity andreaction path. Appl Catal A 2009;355:83–93.

[12] Lee Y, Choi J, Jeon H, Choi K, Lee J, Kang J. Titanium-embedded layered doublehydroxides as highly efficient water oxidation photocatalysts under visiblelight. Energy Environ Sci 2011;4:914–20.

[13] Rivas M, Fierro J, Guil-López R, Peña M, La Parola V, Goldwasser M.Preparation and characterization of nickel-based mixed-oxides and theirperformance for catalytic methane decomposition. Catal Today 2008;133–135:367–73.

[14] Tsai Y, Mo X, Campos A, Goodwin JG, Spivey J. Hydrotalcite supported Cocatalysts for CO hydrogenation. Appl Catal A 2011;396:91–100.

A. Di Fronzo et al. / Fuel 119 (2014) 62–69 69

[15] Costantino U, Marmottini F, Nocchetti M, Vivani R. New synthetic routes tohydrotalcite-like compounds� characterisation and properties of the obtainedmaterials. Eur J Inorg Chem 1998;10:1439–46.

[16] Bastianini M, Costenaro D, Bisio C, Marchese L, Costantino U, Vivani R, et al. Onthe intercalation of the iodine-iodide couple on layered double hydroxideswith different particle sizes. Inorg Chem 2012;51:2560–8.

[17] Bianchi CL, Pirola C, Ragaini V. Choosing the best diluent for a fixed catalyticbed: the case of CO hydrogenation. Catal Commun 2007;7:669–72.

[18] Pirola C, Bianchi CL, Di Michele A, Vitali S, Ragaini V. Fischer–Tropsch andwater gas shift chemical regimes on supported iron based catalysts at highmetal loading. Catal Commun 2009;10:823–7.

[19] Iyi N, Fujii K, Okamoto K, Sasaki T. Factors influencing the hydration of layereddouble hydroxides (LDHs) and the appearance of an intermediate secondstaging phase. Appl Clay Sci 2007;35:218–27.

[20] Pérez-Ramírez J, Mul G, Kapteijn F, Moulijn JA. A spectroscopic study ofthe effect of the trivalent cation on the thermal decomposition behaviour ofCo-based hydrotalcites. J Mat Chem 2001;11(10):2529–36.

[21] Bianchi CL. TPR and XPS investigations of Co/Al2O3 catalysts promoted withRu, Ir and Pt. Catal Lett 2001;76(3–4):155–9.

[22] Lin HY, Chen YW, Li C. The mechanism of reduction of iron oxide by hydrogen.Thermochim Acta 2003;400:61–7.

[23] Lin HY, Chen YW. The mechanism of reduction of cobalt by hydrogen. MaterChem Phys 2004;85:171–5.

[24] Sexton BA, Hughes AE, Turney TW. An XPS and TPR study of the reduction ofpromoted cobalt–kieselguhr Fischer–Tropsch catalysts. J Catal 1986;97(2):390–406.

[25] Kumar N, Payzant EA, Jothimurugesan K, Spivey JJ. Combined in situ XRD andin situ XANES studies on the reduction behavior of a rhenium promoted cobaltcatalyst. Phys Chem Chem Phys 2011;13:14735–41.

[26] Tsakoumis NE, Dehghan R, Johnsend RE, Voronov A, Van Beek W, Walmsley JC,et al. A combined in situ XAS-XRPD-Raman study of Fischer–Tropsch synthesisover a carbon supported Co catalyst. Catal Today 2013;205:86–93.

[27] Tsakoumis NE, Rønning M, Borg Ø, Rytter E, Holmen A. Deactivation of cobaltbased Fischer–Tropsch catalysts: a review. Catal Today 2010;154:162–82.

[28] Alvarez A, Ivanova S, Centeno MA, Odriozola JA. Sub-ambient CO oxidationover mesoporous Co3O4: effect of morphology on its reduction behavior andcatalytic performance. Appl Catal A 2012;431–432:9–17.

[29] Bianchi CL, Pirola C, Boffito DC, Di Fronzo A, Di Michele A, Vivani R, et al. In:Co–Zn–Al based hydrotalcites as catalysts for Fischer–Tropsch processissuepreprints of the DGMK-conference catalysis: innovative applications inpetrochemistry and refining, 2011; 2:93–8 CODEN:DGTAF7 ISSN:1433-9013.

[30] Leendert G, Bitter JH, Kuipers HPCE, Oosterbeek H, Holewijn JE, Xu X, et al.Cobalt particle size effects in the Fischer�Tropsch reaction studied withcarbon nanofiber supported catalysts. J Am Chem Soc 2006;128:3956–64.

[31] Den Breejen JP, Radstake PB, Bezemer GL, Bitter JH, Frøseth V, Holmen A, et al.On the origin of the cobalt particle size effects in Fischer�Tropsch catalysis. JAm Chem Soc 2009;131:7197–203.

[32] Rudolf C, Dragoi B, Ungureanu A, Chirieac A, Royer S, Nastro A, Dumitriu E. NiAland CoAl materials derived from Takovite-like LDHs and related structures asefficient chemoselective hydrogenation catalysts. Catal Sci Technol DOI:101.1039/C3CY00611E.

[33] Simentsova II, Khassin AA, Minyukova TP, Davydova LP, Shmakov AN,Bulavchenko OA, et al. Effect of the composition and structure of theprecursor compound on the catalytic properties of Cobalt–Aluminumcatalysts in the Fischer–Tropsch synthesis. Kinet and Catal 2012;53:497–503.

[34] Qinghong Z, Weiping D, Ye W. Review. Recent advances in understanding thekey catalyst factors for Fischer–Tropsch synthesis. J Energy Chem2013;22:27–38.

[35] Muthu KG, Gary J, Wilson DS, Burtron HD. Fischer–Tropsch synthesis: activityof metallic phases of cobalt supported on silica. Catal Today 2013;215:13–7.