Catalyst for ultra-low sulfur and aromatic diel
Roberto Galiasso Tailleur*,Juan Ravigli,Samuel Quenza,Norma Valencia
PDVSA Intevep,Sede Central,Los Teques,Venezuela
Received17September2004;received in revid form10December2004;accepted13December2004
Available online23January2005offshore
Abstract
A WNiPd/TiO2ÁAl2O3new catalyst with improved desulfurization and hydrogenating capabilities was tested to produce ultra-low sulfur and aromatics diel oil.The removal of the sterically hindered sulfur and nitrogen containing polyaromatic molecules is studied.It was obrved that the new catalyst surface structure promotes the hydrogenating and ring opening of poly-alkyl-poly-aromatics that are less limiting step than that with conventional catalyst.The catalyst characterization indicated the possible formation of metallic acid sites that contribute to the ring opening reactions.The catalyst study using the extracted aromatics and the raffinated product indicated a large improvement in hydrogenation and a reduction in hydrodealkylation functions for converting the sterically hindered aromatics.The reduction in aromatics families improves diel engine emissions.
雅名#2004Elvier B.V.All rights rerved.
Keywords:Deep hydrogenation and ring opening reactions;WNiPd/TiO2ÁAl2O3catalyst;Diel engine emissions
1.Introduction
The hydrotreating process is still one of the most ud to improve diel quality.Today,to fulfill the environmental specification,it will be required a better catalyst activity and lectivity to reduce both sulfur and aromatic contents of the diel products.The tailor-made chemical mod-ification of the fuel will have a large conquence for existing refineries since significant resources need to be immediately allocated to improve the installed capacity on time.Process changes and new catalysts need to be considered together to cope with both sulfur and aromatics effects in the emissions.
A considerable amount of recent papers is dedicated to explain the benefits of new HDS catalyst and the potential modifications in the units that will help to achieve the 15ppm target[1–4].Limited amount of literature and catalyst design is focud on solving the aromatic problem, mainly when cracked diel components(LCO and LKGO)need to be incorporated into the pool.The announcement of a newest generation of HDS catalysts ems very promising in achieving more than three times better
activity than the former one.However,about two times more active catalysts are still required to reduce density and poly-aromatics content;and to improve cetane number for a premium diel fuel production.The key issue in poly-aromatic hydrogenation is the ring opening reaction that allows improving the cetane number and reduces the thermodynamic equilibrium that may control hydrogenation reaction rates.
When diel components are hydrotreated to very low sulfur levels,it becomes necessary to hydrogenate the sterically-hindered aromatics,and sulfur containing-aro-matic compounds.Poly-alkyl-diaromatic and-dibenzothio-phenes are the most refractive molecules to hydrogenate.An example of this is the4,6-dimethyl-dibenzo-thiophene as well as similar hindered poly-alkyl-naptho and-benzo-thiophenes that are recognized as the most difficult to desulfurize[5,6].Among aromatics,the2,4,6,8-tetra-methyl-naphthalene and other similar poly-alkyl-com-pounds are considered very refractory for hydrogenating reactions.Due to the source of the LCO compounds,the molecules are expected to be prent in high concentration since they are produced through cracking reactions of large
/locate/apcata *Corresponding author.Prent address:ITQ,UPV,Camino de Vera
S/N,Valencia,Spain.Tel.:+34963690446.
E-mail address:gatairo@itq.upv.es(R.G.Tailleur).
0926-860X/$–e front matter#2004Elvier B.V.All rights rerved.
doi:10.1016/j.apcata.2004.12.014
molecules in a shape lective FCC zeolite type catalyst. The kinds of molecules limit the production of near zero sulfur and very ultra-low poly-aromatics diel[7].
The prent work tries to understand the effect of a new catalyst surface structure on the reactions of aromatic and sulfur-aromatics compounds still prent in a highly desulfurized diel,as well as their role in diel emissions.
2.Experimental
Four activities have been performed in this work to identify the catalysts structure and the reaction of poly-alkyl-aromatics:
实词和虚词
Catalyst preparation and characterization.
Feed preparation and characterization.
Pilot plant test.
Product testing in a diel engine.
2.1.Catalyst preparation and characterization
The HDT catalysts were prepared by incipient wetness impregnation of a TiO2–Al2O3support treated in a steam–ammonia atmosphere[8].Such support was prepared by precipitating aluminum hydroxide and drying it at1208C and then impregnated it with titanium alcoxide(Fluka)in liquid pha at408C.After1h of contact time,the solid was decanted,and the wet powder was dried at1208C and extrudated in a1mmÂ5mm cylindrical shape using aluminum hydroxide as binder.The pellets were treated in air at4508C for4h,and then in steam at2008C with0.3bar of ammonia partial pressure during2h.After that,the TiO2–alumina support was impregnated with a water solution of ammonium-metatungstate and diamine-palladium salts (Aldrich),and dried in air at1208C for2h.This procedure was followed by a cond impregnation with nickel-nitrate (Aldrich)solution in water and by drying it i
n air at1208C. Then the solid was treated at5508C in air,for6h and subquently sulfided at3508C with diel spiked with2%of S2C for6h.
A reference catalyst(REF)was prepared in a similar way, but without Ti and Pd in it with the purpo of understanding their impact in the formulation.
Chemical characterization:The total metal analysis has been performed using atomic adsorption spectroscopy (Varian Techtron analyzer).Metals were reported in wt.% (bulk)of total metal oxides in the support(W,or Ni,or Pd, on Ti–Al).Table1.
Physical method:Surface,pore volume,and average pore diameter were measured using standard nitrogen adsorp-tion–desorption BET and mercury porosimetry methods.
2.1.1.X-ray-photoelectron spectroscopy(XPS)
Spectra were obtained in a Leybolh-Hereaus LHS-10 apparatus(Mg cathode)using50eV of power(Ref.C(1s):285.0eV).XPS method has been commonly ud to asss the metal dispersion on supports,using the peak area intensity(corrected)to measure atomic concentration.The Deffos et al.[9]method was applied here.Binding energies(eV)were analyzed for Ni(2p3/2),W(4f5/2–7/2),and Ti
(2p).In this way W,Ni,and Ti surface concentrations were measured using the peaks deconvolution and integra-tion to obtain the area.The signals are reported here as a ratio of metal on total metal in surface.In addition,the W4+/ W6+,1S/2S and Ni2+/Ni+ratios are also reported.See Table2 for dispersion values and Table3for the ratios.Palladium at this low concentration was poorly detected as a large shoulder and therefore not measured.
2.1.2.Aluminum nuclear magnetic resonance
The27Al NMR was ud to determine the structure of the support(Ti–Al)bad on the method described by Nagy et al.[10].The spectrum provides information on the different type of aluminum structures in the support (tetrahedral–octahedral coordination).Fig.1shows the results.toefl听力技巧
Pyridine adsorption:The sulfide catalyst and supports samples were treated in a specially designed reaction cell. Samples were degasd under vacuum at1Torr for20min, and then contacted with an argon stream containing0.1% pyridine and0.1%H2S,at room temperature.After that,the cell was heated to200and3008C in argon with only0.1% H2S stream to desorb the pyridine.The procedure was repeated three times.The amount of pyridine remaining in
R.G.Tailleur et al./Applied Catalysis A:General282(2005)227–235
228
Table1
Catalyst properties
Catalyst HDT REF
NiO(wt.%) 4.2 4.0
WO3(wt.%)18.518
PdO(wt.%)0.10
TiO2(wt.%)50
Al2O3Complement
Surface(m2/g)245230
V olume(cm3/g)0.520.55
Table2
Metals dispersion IMe/(I total)
Metals ratio Signal HDT1Ref
NiSy Ni(2p1/2)0.550.33
WSx W(4f5/2)10.338.22
TiO2Ti(2p) 4.70
Table3
W,Ni,S,and Ti species
Signals ratio W4+/W6+NiS/NiO S1/S2Ti1/Ti2
HDT78 6.0012
REF9770
TiO2/Al2O3–––4
the solid was measured by the area of the absorbance signal
for Pyridinion (Bro
¨nsted)and for pyridine (Lewis)measured in a FTIR (Perkin-Elmer with DTGS)equipment.The area of the main peaks at 1540,and 1440cm À1is reported in Table 4as m mol of pyridine/g of sample (assigned in accordance to Ref.[11].
2.2.Feed preparation –characterization
The Straight Run Light Gas Oil and LCO (40–60)blend,which contains 0.25%of sulfur and 65%of aromatics was obtained from Amuay Re finery in Venezuela.This feed was treated in two reactors in ries pilot plant,at 0.6MPa of total pressure,3408C of temperature and 0.5h À1of space velocity.A Tops øe 554commercial NiMo on Al 2O 3catalyst was ud diluted with 50%of inert small-diameter particles to assure proper reactor fluid dynamics [12].The H 2/HC ratio was 5(molar)and a hydrogen purity of 98%with 0.1vol.%of H 2S spiked in it.The pilot plant was operated for 1month at constant conditions to produce a $120ppm sulfur diel (called here Feed 1).Then,one part of Feed 1was trea
ted using the proprietary HDT catalyst (Product 1)while other part was treated with REF catalyst (Product 2).They were tested at three tempera-tures and three space velocities,to check the sulfur and aromatic reactivity of the remaining sulfur and aromatic compounds.Other conditions were similar to tho mentioned above.
A third part of the Feed 1was extracted with sulfuric acid.A 300kg of sample was contacted with 10kg of 98%purity sulfuric acid in a Batch type reactor for 3h at 1008C at 0.1MPa.Afterward,the reactor was cooled and 10kg of
sodium hydroxide was added at room temperature,and decanted during three days.The oiled pha was removed from the reactor and called here raf finate (Feed 2),whereas the remaining pha (sulfonated aromatics called extract)was treated with 20kg of fresh water at 1508C and 0.1MPa for 6h in the same batch reactor.Subquently,the product was cooled,and the oiled pha was removed and washed with three portions of fresh water until being sodium free.Such oiled pha was then analyzed by GC –MS method described below.The mass balance was 97%by weight for the extraction operation.Then,Feed 3was prepared by blending pure n -hexadecane (Fluka)paraf fin with the extract containing 15%by weight of the aromatics.Feeds 2and 3were both tested at three space velocities and three temper-atures using HDT and REF catalysts.The other conditions were si
milar to tho mentioned above.2.3.Product characterization
Products were characterized using a GC –MS analytical program for diel.Temperature was programmed and a silica gel type column was ud in a gas chromatograph coupled with a low-resolution mass spectrometer operating at 30eV .This special analysis was able to discriminate families of aromatics,naphthenes,and paraf fins [13].The extraction of aromatics improved the analysis of the remaining aromatics and naphthenes.In addition,a sulfur chemiluminescence ’s detector,coupled with the gas chromatograph,allowed to identify speci fic sulfur compounds like 4,6-DMDBT and other aromatics sulfur compounds.Proton nuclear magnetic resonance (1H NMR)was additionally employed to determine the average normal to isoparaf fin ratio.2.4.Diel engine testing
The NO x and PM emissions were measured using a high speed detroit diel (Series 80-1998)direct injection engine operating at 2000rpm,350kW power,1364kPa BMEP,and 1630N m torque (Table 9).With the purpo of determining NO x and PM,a micro-tunnel device to sample the gass coupled with on line micro-filter and an IR analyzer were ud.Ethyl cetane enhancer was added to the product to obtain the same cetane number value for all samples.The test objective was to prove the effect of the remaining aromatics on emissions at constant cetane number.
石家庄新东方
无忧雅思预测First,let us e what the catalytic surfaces of both catalysts look like,and then proceed with the study of sulfur and sulfur-aromatics molecules reactions.
3.Results
3.1.Catalyst characterization
Table 1shows that HDT catalyst contains sul fided W,Ni,and Pd species,as ‘‘metals ’’active pha [14],and 5wt.%of
R.G.Tailleur et al./Applied Catalysis A:General 282(2005)227–235229
Table 4
Bro
¨nsted and Lewis acidity (m mol/g)Catalysts (1490/1440cm À1)2008C 3008C B L B L HDT 0.80.550.30.34REF
水的痕1.10.450.50.33TiO 2Al 2O 3
2.00.660.80.34Al 2O 3
0.9
1
0.6
0.35
Fig.1.
27
AINMR at 104.22Hz.
TiO2in g-alumina as‘‘acidic’’function.The support was treated with steam–ammonia that produces a Ti species migration to the aluminum surface,according with XPS information discusd below.The physical properties analysis indicates that the catalyst has a130A˚of average-integrated by volume-pore diameter,in a micro-and meso-structure.
The REF catalyst has similar physical properties but the support is compod only of g-alumina.Total sulfur in sulfide catalysts was 5.8and 6.2wt.%,respectively. There was almost no carbon,nor sodium prent in the catalysts.
The XPS analysis(Tables2and3)shows that HDT catalyst has higher W dispersion and degree of sulfiding than REF one.This fact demonstrates the role of Ti and Pd on active metal dispersion.Binding energy(eV)in HDT catalyst was458.5and460.5for Ti,856.2and854.9for Ni, 34.4an
d32.5for W,and74.2for Al.In HDT,W and Ni bands shows a shift of0.4and0.3eV,respectively,and higher W4+/W6+ratios in comparison with REF catalyst (Table3).In addition,the HDT surface prents a larger amount of sulfur S1and S2bands(from162.0to162.2eV), but with lower S1/S2ratio than tho in REF catalyst[14,15]. All of that points out the prence of less reduced species in the latter one.
The XPS Ti(2p)spectra of HDT sulfide catalyst indicates the prence of one signal of poorly coordinates tetrahedral species at460.5eV,and cond one of octahedral coordinated species at458.5eV.It also has a third signal at459.3eV attributed to Ti in a pudo-spinelle or in penta-coordinated environment(27Al NMR results).This latter signal is already prent in the TiO2–alumina support.The distorted O1s signal at531.4eV shows the prence of a shoulder at530.2eV,probably associated to the new species mentioned above.There are no other signals than tho of gamma alumina,in X-ray diffraction between u are45–608,which may be attribu-ted to appreciably rutile or anasta concentration.The prence of Ti and Pd affects W and Ni dispersions and the degree of sulfiding of the cluster on surface,as it was previously demonstrated by ESR and XAFT studies [17].
英语翻译工具
Fig.1shows the27Al NMR results for HDT and REF catalysts.The OHÀspectra are not as well defined as in Si–Al support,due to the complexity of the surface where a potential W–Ni–Ti–Al cluste
r interaction may occur.In spite of that,a tetrahedral and octahedral Al organization signals can be distinguished,for both catalysts with different intensities and shifts.They have one small signal around80ppm,which corresponds to tetrahedral sites, and another three times larger around0–10ppm that is assigned to an octahedrically coordinate site.The HDT catalyst has a third shoulder around35–40ppm attributed to a pudo-spinelle structure or a pentahedral coordination [16,18].The new signal is assigned to W–Al–Ti species since it can be enhanced by redox treatment[17],and becau it does not appear,and may not be formed in REF catalyst without Ti.
Pyridine adsorption studies were performed with sulfided catalysts and with supports in prence of H2S so as to keep the reaction surface properly‘‘saturated’’.Otherwi, anomalous adsorption might be measured due to vacancies created by the H2S desorption at200and3008C.The Bro¨nsted(1544cmÀ1)/Lewis(1450cmÀ1)signals ratio at 2008C is the highest for the Ti–alumina support,followed by REF catalyst,alumina support,and HDT catalysts(3.0/ 2.4/1.4/0.9).But when the temperature augments,the ratio of acidity for the solids is modified differently.At3508C the trend is:Ti–alumina>alumina>REF>HDT(2.3/1.7/ 1.5/0.8).The impregnation of active metal reduces the Bro¨nsted acidity created by the Ti contribution while the Lewis acidity remains quite constant.It may indicate that the Bro¨nsted acidity reduction when the temperature is incread does not create addit
ional Lewis acid sites.At high temperature,the lowest Bro¨nsted acidity is in the HDT catalyst,showing a strong metal‘‘passivation’’of the surface.But in the HDT catalyst a new adsorption band appeared around1547cmÀ1,a shoulder that is similar for 200and3008C.This particular pyridine adsorption band cannot be en without a H2S atmosphere,or in oxide-HDT catalyst.This band incread with WS2content but it is not prent either in the Ti–alumina support or in REF catalyst. It ems to be associated to SH Bro¨nsted acid sites linked to metal clusters.Topsøe et al.[19]and Lauritn et al.[20] have reported the prence of similar metallic Bro¨nsted sites in the border of a NiMo/Al2O3crystal that might support our hypothesis.The effect of TiO2in the HDS reactions can be en in Refs.[21,22]for other catalysts. The interactions between Ti–W–Al are known[23]and they might be critical for the formation of‘‘metal’’acid centers.Let us now e the impact of the new active surface on the hydrogenation of the sterically hindered poly-alkyl-aromatic.
3.2.Feed characterization
Feed1,that had been partially hydrotreated and desulfurized,was characterized to evaluate the remaining sulfur and aromatic families of compounds.The amount of LCO in the initial feedstock was specially controlled due to the large contribution of the sterically hindered aromatic and sulfur-aromatic compounds.The relative influence of the compounds on the reactivity increas as the d
eepness of the hydrotreating progress and the less alkylated molecules had been converted.The highly branched aromatic may prent difficulties to be adsorbed in conventional NiMo catalyst.Feed1composition is prented in Table5,cond column.It has121ppm of total sulfur that was arbitrarily chon as sulfur target for the pre-treatment.The detailed analysis indicates the prence of small amounts of the three families of poly-alkyl-compounds:thiophene(PAT),benzo-thiophene(PABT),and dibenzo-thiophene(PADBT),plus
R.G.Tailleur et al./Applied Catalysis A:General282(2005)227–235 230
some non-identi fied sulfur compounds.They also have 15.2%of di-aromatics and 25.1%of mono-aromatics.The extract (22%of Feed 1),mainly contains diaromatic molecules that could be sulfonated in the unblocked beta and/or gamma position (ca.methyl-naphthalene in Fig.2).The extract has a high proportion of poly-alkyl-di-,and poly-alkyl-tri-ring-aromatic (PADRA,PATRA)and near zero poly-alkyl-mono-ring-aromatic (PAMRA).Some paraf fin and naphthene are also prent only due to an occlusion phenomenon effects that happened when the extract was precipitated.The SO 42Àfree extract,containing aromatics distributed all along the distillation range (180–3508C)was diluted in hexadecane.Compounds as anthracene,benzo-and di-benzo-anthracene,pyrene,chryne,perylenes,phenantrene,and benzenes are detected,but their individual
taffyconcentration is below 1%for a grand total of identi fied aromatics compound of around 10%.They are not considered here.
The raf finate still contains small amount of PADRA and PATRA (probably some of them highly sterically hindered by alkyl groups),and a high concentration of PAMRA.It also has paraf fin,poly-alkyl-mono-,di-,and tri-cyclo-paraf fin (PAMCP,PADCP and PATCP).3.3.HDS Feed 1and Feed 4
Feed 1was treated with both catalysts at the same pressure and H 2/HC ratio,whereas the residence time and
reaction temperature were changed.Fig.3shows the sulfur left in diel treated with HDT catalyst at 3408C.
It can be en that concentration curves of the remaining BT and DBT sulfur families as a function of space velocity are quite similar,as well as tho of other hindered sulfur species likes naphthothiophenes,prent in minor propor-tions.This relative shape is kept constant when the temperature is incread,indicating similar activation energies for the HDS reaction ($20kcal/mol).The analysis per families at 10ppm level has an error of Æ5%,but the total sulfur can be detected within Æ1%of error.The total sulfur content for 3408C and 1h À1of LHSV,is around 15pp
m.
The same information was obtained for the REF catalyst.The plots of unconverted sulfur concentrations (by families)as a function of residence time are depicted in Fig.4for 3408C.The sulfur concentration indicates a similar behavior than that obrved with the previous catalyst,but with less HDS activity for all the families.The total sulfur left is now 25ppm at 3408C and 1h À1of space velocity.This reprents when compared with the HDT catalyst results,almost 50%less activity.The effect of temperature is similar but the activation energy measured is lower ($19)kcal/mol).
R.G.Tailleur et al./Applied Catalysis A:General 282(2005)227–235
231
Fig.2.Place for the sulphon —in
PADRA.Fig.3.HDS
(HDT).
Fig.4.HDS (REF).
Table 5
Diel families a (concentration,wt.%)Properties
Feed 1Extract Raf finate Mass (wt.%)1002278Thiophene (ppm)
12714Benzothiophene (ppm)337825Dibenzothiophene (ppm)622505Total sulfur (ppm)12134047n +isoparaf fins 17.7222Monocycloparaf fins 13.9317Dicycloparaf fins 14.1417Tricycloparaf fins 6.508.4Mono-ring-aromatic 22.51425Di-ring-aromatic 14.6592Tri-ring-aromatic 1.480.1Cetane number (À)
38
–
41
a
he is just kiddingNaphthobenzothiophene and 10%of others indenti fied aromatic com-pounds are not included.
