Journal of Catalysis250(2007)
331–341
/locate/jcat
Deactivation of Ni catalysts during methane autothermal reforming
with CO2and O2in afluidized-bed reactor
Zhaoyin Hou∗,Jing Gao,Jianzhong Guo,Dan Liang,Hui Lou,Xiaoming Zheng
Institute of Catalysis,Department of Chemistry,Zhejiang University(xixi campus)Hangzhou310028,China
Received28March2007;revid22June2007;accepted27June2007
Available online7August2007
Abstract
A ries of different-sized Ni catalysts(4.5–45.0nm)were prepared and ud for methane autothermal reforming with CO2and O2in a fluidized-bed reactor.It was found that the activity and sta
bility of Ni catalysts depend strongly on the particle size and the operating space velocity. Small sized Ni is more active and stable at space velocity<54,000h−1.Characterizations disclod that methane decomposition rate decreas with the enlarging Ni particle size,and some of the surface carbons(derived from methane decomposition)are inactive in CO2atmosphere.As the methane decomposition rate slows on larger Ni particles and at higher space velocity to ensure complete conversion of the oxygen,surface Ni will be gradually oxidized by remaining O2,leading to Ni deactivation.吃药可以吃鸡蛋吗
元明粉别名
©2007Elvier Inc.All rights rerved.
Keywords:Methane;Reforming;Ni catalyst;Deactivation;Space velocity
广义的教育
1.Introduction
The transformation of methane and carbon dioxide,the cheapest carbon-containing materials and the most problematic greenhou gas,into more valuable compounds has long at-tracted the attention of rearchers.Among published results, the catalytic CO2reforming of CH4to synthesis gas[Eq.(1)] has been investigated comprehensively over the past few years, as has been well summarized in recent reviews[1–5].The syn-thesis gas generated from CO2reforming of CH4has a l
ow H2/CO ratio and thus is suitable for the Fischer–Tropsch syn-thesis of higher hydrocarbons and for the synthesis of oxy-genate products.Rostrup–Nieln’s group[6,7]and Rucken-stein’s group[1,5,8]have made significant contributions to im-proving the understanding of steam and CO2reforming over Ni catalysts in both experimental rearch and theoretical studies:
(1) CH4+CO2→2CO+2H2, H298K=247kJ/mol.
Most of the group VIII metals are more or less active toward this reaction,and Ni has been popularly explored for its high
*Corresponding author.Fax:+8657188273283.
E-mail address:zyhou@zju.edu(Z.Hou).activity and low price[1–5,9].Due to the inherent inertness of methane and CO2,high temperature(typically,800–900◦C) is needed to achieve a meaningful yield.Under such vere conditions,carbon deposition and/or sintering of the metal par-ticles occur on the surface of the catalyst,and sintering further accelerates carbon deposition becau large metal enmbles stimulate coke formation[8].
Another drawback is that CO2reforming of methane is a highly endothermic reaction(247kJ/mol).The
addition of oxy-gen to the reforming reactants is an effective method of sup-plying heat,becau partial oxidation of methane to synthesis gas is an exothermic reaction[Eq.(2)].This combined CO2re-forming and partial oxidation of methane,also called methane autothermal reforming[MATR],has drawn significant inter-est in recent years in alternative routes for the conversion of methane to synthesis gas[Eq.(3)][10–16]:
(2) CH4+(1/2)O2→CO+2H2, H298K=−38kJ/mol, CH4+x CO2+((1−x)/2)O2→CO+2H2,
(3) H298K=(285x−38)kJ/mol(0<x<1).
In published papers,it is suggested that MATR is carried out in two parate reaction zones in thefixed-bed reactor—in the first zone,part methane is combusted into CO2and steam to en-
0021-9517/$–e front matter©2007Elvier Inc.All rights rerved. doi:10.1016/j.jcat.2007.06.023
332Z.Hou et al./Journal of Catalysis250(2007)331–341
sure the complete conversion of the oxygen in feed,producing a hot(>1300◦C)stream,and in the cond zone,the unconverted methane is reformed to synthesis gas by CO2and steam.A sig-nificant temperature gradient in the catalyst bed formed,which ultimately resulted in the thermal sintering and deactivation of catalyst[10,17,18].It is of great practical interest to overcome this limitation of the MATR process infixed-bed reactor.
Tomishige[11–13]and Zheng[15,19,20]reported their studies on the MATR process in afluidized-bed reactor,and they suggested that the high rates of heat transfer and high stability of operation be obtained.Fluidization has a favorable effect on the inhibition of carbon deposition,which is probably becau the catalyst particles are circulated between the oxidiz-ing zone and reducing zone and carbon gasification proceeds readily in the oxidizing zone.Moreover,catalyst can maintain a suitable level of reducibility duringfluidization that enhances the conversion of methane[11–13,15,19,20].
Recently,we have found that the activity and stability of Ni catalysts during the MATR influidized-bed reactor depend strongly on the particle size of Ni and the operating space veloc-ity.Larger-sized Ni catalysts(average Ni particle size>16nm) are inactive at higher space velocity(90,000h−1)and deacti
vate rapidly(at low space velocity,18,000h−1).Whereas small-sized ones(average Ni particle<9.5nm)are more active and stable at the space velocity<54,000h−1.Characterizations of the spent catalysts have found that neither carbon deposition nor metal sintering would form on the deactivated catalysts.That is, the deactivation mechanism of Ni catalysts during the MATR process in afluidized-bed reactor cannot be explained by tho achievements of CO2reforming of methane.More fundamen-tal studies are still needed to reveal the deactivation mechanism of Ni catalyst in the MATR process in thefluidized-bed reactor. The aim of this work is to correlate the particle size of Ni and the operating space velocity with its catalytic performance and deactivation.For this purpo,MATR at different space veloc-ities,methane decomposition(in both continuous and puld methaneflow)and the reactivity of carbon deposited during methane decomposition toward CO2and O2were determined on different-sized Ni catalysts.
2.Experimental
2.1.Catalyst preparation and characterizations
Ni catalysts sized from 4.5to45.0nm were prepared via direct impregnation of[Ni(en)3]2+(en,ethylenediamine), [Ni(EDTA)]2−(EDTA;ethylenediaminetetraacetic acid),nick-el(II)acetylacetonat
e dihydrate(99+%),nickel(II)acetate,and Ni(NO3)2·6H2O onto a spherical SiO2support(special prod-uct forfluidized reactor,S BET=330m2g−1,average diam-
eter0.25–0.38mm,Qingdao,China).Nickel(II)acetylaceto-nate dihydrate(99+%),nickel(II)acetate,and Ni(NO3)2·6H2O were purchad from Acros Organics(Beijing Branch,China). [Ni(en)3]2+and[Ni(EDTA)]2−were prepared as described pre-viously[21–23].The loading amount of Ni was controlled at 5wt%of the support.The precursors were dried at80◦C in vacuum and calcined at800◦C in stagnant air for4h.
The morphologies of the catalysts were characterized by X-ray diffraction(XRD)and transmission electron microscopy (TEM).XRD was obtained in D8Advance(Bruker,Germany) equipment using nickel-filtered Cu Kαradiation at40kV and 40mA.Diffraction data were recorded using continuous scan-ning at a rate of0.02◦/s,step0.02◦.Average Ni particle size was calculated according to Scherrer–Warren equation. TEM images were obtained using an accelerating voltage of 200kV(TEM,JEOL-2020F).Samples werefirst ground to powder,reduced in hydrogen,and disperd on Cu grids in tetrachloromethane under supersonic-wave shaking.
2.2.MATR on different-sized Ni catalysts
MATR was carried out in a quartzfluidized-bed reactor (id=12mm).As noted previously[24],the minimumfluidiza-tion velocity(ωmf)and the maximumfluidization velocity(ωt) could be calculated as
(4)ωmf=
d2p(r s−r g)g
1650μg
(m/s)
and
(5)ωt=
d2p(r s−r g)g
18μg
(m/s),
where d p and r s are the diameter and the density of catalyst,and μg and r g are the viscosity and density of the reactant mixture. In this experiment(d p=0.25–0.38mm,r s=470–500kg/m3, r g=0.32kg/m3andμg=37.5µPa/s),the total feed gasflow rate must be controlled between78and3070ml/min to ensure the efficientfluidization of catalysts particles in the reactor at 700◦C.
CH4(99.99%),CO2(99.9%),and O2(99.9%)were intro-duced into the reactor controlled by three ts of massflow controller(Brooks5850E)at a molar ratio of CH4:CO2:O2= 10:4:3.Catalyst isfirst reduced in H2at700◦C for1h.The effluent gas was cooled in an ice-water trap and analyzed us-ing an online gas chromatograph(Shimadzu,GC-8A)with a packed column(TDX-01)and a thermal conductivity detector. All spent catalysts were further characterized by XRD and TG-DTA(PE-TGA7)at50–900◦C in airflow.
2.3.The reactivity of surface carbons on different-sized Ni catalysts
The reactivity of surface carbons toward CO2and O2was in-vestigated via coking reaction(CH4temperature-programmed decomposition[CH4-TPDe])and followed temperature-pro-grammed oxidation with CO2(CO2-TPO)and O2-TPO.Before the experiments,catalysts werefirst r
educed in H2flow at 700◦C for1h,then cooled to50◦C in Ar.CH4-TPDe was per-formed in10%CH4/Ar(50ml/min)from50to800◦C at a ramp of15◦C/min and held at800◦C for20min.Then the catalyst bed was cooled to50◦C in Ar,and CO2-TPO was per-formed in10%CO2/Ar(50ml/min)from50to800◦C at a ramp of15◦C/min.Finally,after the catalyst bed was cooled to50◦C in Ar,concutive O2-TPO was carried out in aflow
Z.Hou et al./Journal of Catalysis250(2007)331–341333
of10%O2/Ar(50ml/min)from50to800◦C at15◦C/min. All gas in effluent were detected by a mass analyzer(Om-niStar GSD301,Switzerland)and recorded as functions of tem-perature.Parallel experiments were performed simultaneously, and carbons formed in CH4-TPD and residual carbons after CO2-TPO were detected by TG-DTA from50to900◦C in air flow.The morphologies of the surface carbons(formed in CH4-TPDe)were characterized by TEM.
2.4.Methane activation on different-sized Ni catalysts
Methane activation and catalytic conversion on different-sized Ni catalysts in various atmospheres(CH4,CH4/CO2,and CH4/O2)were investigated via pul-injected surface reaction. CH4,CH4/O2(2:1),or CH4/CO2(1:1)pul(total volume, 500µL with250µL CH4and Ar in balance)was
乒乓球女运动员injected into the reduced catalysts(20mg)with a6-port gas sampling valve under a stream of Ar carrier gas(100mL/min).The reaction temperature was kept at700◦C.All gas in effluent were de-tected,the turnover frequency of methane(TOF,defined as mol of CH4converted/mol of Ni atom per cond)was calculated on the basis of methane conversion.
2.5.Reducibility of different-sized Ni catalysts after CH4 decomposition
The reducibility of different-sized Ni catalysts was detected via concutive CH4–O2–H2pul experiments in the same equipment as described in Section2.3.Six methane puls (250µL per pul)werefirst injected into the reduced catalyst (20mg),followed by three oxygen puls,and then hydrogen puls in a stream of Ar carrier gas(100mL/min).The reaction temperature was kept at700◦C,and all gas in effluent were recorded.
3.Results
3.1.The dimensions of Ni on SiO2
Table1summarizes the TEM and XRD analysis results of the Ni/SiO2catalysts prepared from different Ni precursors. The mean size of nickel crystallites was detected by XRD and calculated fro
m the broadening of the Ni(111)according to the Scherrer–Warren equation.The calculated average Ni particle sizes incread from4.5to45.0nm in the Ni/SiO2catalysts prepared via[Ni(en)3]2+and nickel nitrate.In the Ni/SiO2cat-alysts prepared from[Ni(EDTA)]2−,nickel(II)acetylacetonate, and nickel(II)acetate,the calculated average Ni particle sizes were6.0,8.5,and16.0nm,respectively.
Figs.1A–1E show the typical TEM images of Ni/SiO2cat-alysts prepared from[Ni(en)3]2+,[Ni(EDTA)]2−,nickel(II) acetylacetonate,nickel(II)acetate,and nickel nitrate.The his-tograms of the particle size distribution are inrted in the top-right corner of the TEM images.Highly disperd Ni par-ticles were detected in tho Ni/SiO2catalysts prepared from [Ni(en)3]2+and[Ni(EDTA)]2−,and the detected Ni particles were concentrated at5.1and7.0nm,respectively.The average Table1
The dimensions of Ni on SiO2from different Ni precursors a
Ni precursors Ni diameter(nm)
TEM XRD b Aqueous solution of[Ni(en)3]2+5.14.5 Aqueous solution of[Ni(EDTA)]2−7.06.0 Ethanol solution of nickel(II)acetylacetonate9.48.5 Aqueous solution of nickel(II)acetate16.116.0 Aqueous solution of Ni(NO3)251.245.0 a Sample was reduced in H
2at700
◦C for1h.如何描写人物
b Calculated according to Scherrer–Warren equation.
particle size of Ni(summarized in Table1)was calculated as
粗壮的反义词是什么(6)
d TEM=
n i d3i
n i d i
,
where n i is the number of particles having a characteristic di-ameter d i(within a given diameter range).
In Ni/SiO2catalysts prepared from nickel(II)acetate and nickel nitrate,a lower density of Ni particles on the support should be expected.However,the silica support appeared to be almost completely cov
ered by Ni particles(Fig.1D),suggest-ing that the Ni could be unevenly distributed within the silica spheres,with some parts heavily loaded with Ni and others bearing only few Ni particles.
3.2.MATR on different-sized Ni catalysts
Fig.2shows the activity of different-sized Ni/SiO2catalysts for MATR influidized-bed reactor at700◦C and a space veloc-ity of18,000h−1.The detected conversions of CH4and CO2 were higher and more stable on the smaller-particle Ni cata-lysts(<9.5nm)than that on the larger-particle ones(>16nm). Moreover,the oxygen in the feed was consumed completely on the smaller-particle Ni catalysts.The highest CH4and CO2con-versions were77.0and67.9%,respectively.But on the45.0-nm Ni catalyst,the initial conversions of CH4and CO2were only 47.2and27.5%,and the catalyst deactivated rapidly.The resid-ual oxygen in effluent incread during the deactivation process.
At higher space velocity(54,000h−1),the detected conver-sion of methane on4.5-nm Ni catalyst decread rapidly from 71.0to66.4%in thefirst4h and remained constant in the following38h(Fig.3).But the45.0-nm Ni catalyst deacti-vated completely in only2h.When the space velocity was enhanced to90,000h−1,the detected conversion of methane on the4.5-nm Ni catalyst decread continuously from67.2to 37.2%in48h on stream,and the45.0-nm Ni catalyst was inac-tive at this space velocity.
Fig.4shows the durability of the4.5-nm Ni catalyst for MATR in afluidized-bed reactor at lower space velocity.It re-mained stable during100h on stream(at9000h−1),with the detected conversion of CH4decreasing only slightly,from81.3 to80.8%.At18,000h−1,the detected conversion of CH4de-cread from77.0to70.0%in thefirst26h,but it recovered in the following16h and remained stable(72.0%)for the last 38h.
334Z.Hou et al./Journal of Catalysis250(2007)331–341
Fig.1.TEM image and particle distribution of Ni/SiO2catalysts prepared from different Ni precursors:(A)[Ni(en)3]2+,(B)[Ni(EDTA)]2−,(C)nickel(II)acetyl-acetonate,(D)nickel(II)acetate,and(E)Ni(NO3)2.
Z.Hou et al./Journal of Catalysis 250(2007)331–341
335
Fig.2.Activity of the different sized Ni catalysts for MATR in fluidized bed reactor at 18,000h −1:(F )4.5,(2)6.0,(Q )8.5,(×)16.0,and (×+)45.0
nm.
Fig.3.Activity of 4.5nm and 45.0nm sized Ni catalysts at higher space ve-locity:(2)and (F )4.5nm sized Ni at 54,000and 90,000h −1,(Q )and (")45.0nm sized Ni at 54,000and 90,000h −1
.
发烧头晕Fig.4.The durability of the 4.5nm sized Ni catalyst for MATR in fluidized bed reactor:(F )9000and (2)18,000h −1.
TG-DTA and XRD analysis of the spent catalysts found that neither carbon deposition nor metal sintering formed.Ni was in high dispersion and metallic state in the spent 4.5-nm catalyst
as
Fig. 5.XRD spectrum of different sized Ni catalysts:(a) 4.5nm fresh,
(b)4.5nm spent (6h),(c)45.0nm fresh,and (d)45.0nm spent (6
h).
Fig.6.MS signals of effluent gas during CH 4-TPDe:(a)on the 45.0nm sized Ni catalyst and (B)on the 4.5nm sized Ni catalyst.
that in the fresh sample.But NiO obviously formed in the spent 45.0-nm Ni catalyst (Fig.5),suggesting that Ni was oxidized during time on stream.This might be the direct reason for the deactivation of the larger-particle Ni catalysts.
336Z.Hou et al./Journal of Catalysis 250(2007)
分数除法教学设计331–341
Fig.7.TEM and EDX analysis of the 45.0nm sized Ni catalyst after CH 4-TPDe:(A)TEM image and (B)
EDX analysis (in circle).Spectrum processing:no peaks omitted;quantization method:cliff lorimer thin ratio ction;processing option:all elements analyzed (normalid),number of iterations =1.
3.3.The reactivity of surface carbons on different-sized Ni catalysts
Figs.6A and 6B show the typical MS signals of effluent gas during the coke reaction via CH 4-TPDe on the 45.0-and 4.5-nm Ni catalysts,respectively.Three CH 4decomposition peaks can be en at 410–800◦C on the 45.0-nm Ni catalyst,and four peaks can be en at 370–800◦C on the 4.5-nm Ni cat-alyst.The peaks can be assigned to different kinds of carbons formed as described in earlier work [25].It is interesting to note that the initial CH 4decomposition temperature decread from
410◦C (on the 45.0-nm Ni catalyst)to 370◦C (on the 4.5-nm Ni catalyst),and the amount of deposited carbons decread from 12.9to 1.7mmol /g cat (Table 2).
TEM images and EDX analysis of the carbon-covered Ni catalysts are shown in Figs.7and 8.Large amounts of en-capsulating carbon and whisker carbon [26–28]formed on the 45.0-nm Ni catalyst,completely enwrapping the Ni particles (Fig.7A).EDX analysis (in the circled area)of the enwrapped Ni particles found that the surface C/Ni ratio reached 96.3/3.7(Fig.7B).In contrast,the 4.5-
nm Ni catalyst showed high dis-persion (Fig.8A),with a surface C/Ni ratio of 9.6/2.1(Fig.8B).
