V2O5_SCR_Deacttivate

Chemical deactivation of V 2O 5/WO 3–TiO 2SCR catalysts by additives
and impurities from fuels,lubrication oils,and urea solution
I.Catalytic studies
Oliver Kro
¨cher *,Martin Elsener Paul Scherrer Institute,5232Villigen PSI,Switzerland
Received 27March 2007;received in revised form 24April 2007;accepted 25April 2007Available online 27April 2007
Abstract
The influence of the combustion products of different lubrication oil additives (Ca,Mg,Zn,P,B,Mo)and impurities in Diesel fuel (K from raps methyl ester)or urea solution (Ca,K)on the activity and selectivity of vanadia-based SCR catalysts were investigated.Standard V 2O 5/WO 3–TiO 2catalysts coated on metal substrates (400cpsi)were impregnated with water soluble compounds of these elements and calcined at 400and 5508C,in order to investigate the chemical deactivation potential of different elements and combinations of them.
It was found that potassium strongly reduced the adsorption equilibrium constant K NH 3of ammonia.At small ammonia concentrations in the feed,only part of the active sites were covered with ammonia resulting in a reduced SCR reaction rate.At high ammonia concentrations,the surface coverage and SCR reaction rate increased,but high SCR activity at concurrent low ammonia emissions was impossible.Calcium caused less deactivation than potassium and did not affect the ammonia adsorption to the same extent,but it lowered the intrinsic SCR reaction rate.Moreover,deactivation by calcium was much reduced if counter-ions of inorganic acids were present (order of improvement:SO 42À>PO 43À>BO 33À).Zinc was again less deactivating than calcium,but the positive effect of the counter-ions was weaker than in case of calcium.The degree of N 2O production at T >5008C,which is typical for V 2O 5/WO 3–TiO 2catalysts,was not influenced by the different compounds,except for molybdenum,which induced a small increase in N 2O formation.#2007Elsevier B.V .All rights reserved.
Keywords:Selective catalytic reduction;DeNO x ;SCR catalyst;Vanadia;Deactivation;Ammonia adsorption
1.Introduction
Urea-SCR (SCR:selective catalytic reduction)is an efficient technology to reduce the nitrogen oxide emissions of Diesel vehicles [1–3].This process has been developed from ammonia SCR,which is used for several decades as a technique to remove NO x from stationary power plants [4–6].The basis of SCR is the comproportionation reaction of ammonia with NO forming harmless nitrogen:4NH 3þ4NO þO 2!4N 2þ6H 2O
The main development steps on the way from the stationary application to SCR for Diesel vehicles were to identify urea as non-poisonous storage compound for ammonia,to develop a
urea dosing strategy,which is able to cope with the varying NO x concentrations in dependency of engine load,speed and catalyst temperature,and to increase the volumetric activity of the existing vanadia-based catalysts [7].Catalysts of this type,optimized for Diesel vehicle applications,contain 1.5–2.0wt.%of V 2O 5on anatase TiO 2,promoted with $8wt.%WO 3.The vanadia concentration is higher than in catalysts for stationary applications ($1.0wt.%)to achieve high DeNO x activity.However,vanadia must be present as two-dimensional surface layer on the titania support to be selective.Too high surface concentrations result in the formation of bulk vanadia,which favour unselective oxidation reactions at higher temperatures,i.e.the formation of N 2O or the oxidation of ammonia to nitrogen.This phenomenon is more pronounced when the amount of vanadia exceeds the value that is necessary to form a monolayer on the surface.Tungsta is an important catalyst promoter,which stabilizes the anatase modification of titania,supports the spreading of the vanadia on the surface,increases
/locate/apcatb
Applied Catalysis B:Environmental 75(2008)215–227
*Corresponding author.Tel.:+41563102066;fax:+41563102323.
E-mail address:oliver.kroecher@psi.ch (O.Kro
¨cher).0926-3373/$–see front matter #2007Elsevier B.V .All rights reserved.
doi:10.1016/j.apcatb.2007.04.021
the surface acidity and improves the poison resistance to alkali metals [8].This interplay of the different constituents explains the excellent properties of V 2O 5/WO 3–TiO 2systems as SCR catalysts.
When considering a catalyst for a mobile SCR system,its durability concerning thermal degradation,mechanical attrition and chemical deactivation has to proven over the lifetime of the vehicle,which can be up to 106km for a heavy-duty truck.These requirements are of course much more severe than for stationary applications,since much higher temperatures are reached,the catalysts are exposed to thermal stress due to the varying exhaust gas temperatures and they have to withstand vibrations during operation of the vehicle.Moreover,SCR catalysts onboard of Diesel vehicles may deactivate by the action of water,which condenses in the catalytic converter after engine start,and by different substances in the exhaust gas,which deposit on the catalyst surface.The origin of these substances are mainly fuel and lubrication oil additives,e.g.detergents,dispersants,viscosity improvers etc.,which are responsible for a series of important product properties
(Table 1).These additives are converted to inorganic compounds in the combustion chamber or emitted unchanged into the exhaust gas from the turbocharger lubrication system.Besides these intended additives,impurities,such as the sulphur in fuel and lubrication oils,also play a role in the emissions of Diesel engines.Table 2lists the concentrations of emission-relevant elements in lubrication oils and their most likely combustion products [9].The most emission-relevant elements in lubrication oil additives are sulphur,calcium,zinc and phosphorus,which are emitted as CaCO 3(or CaO),SO 2,ZnO and H 3PO 4from the combustion chamber.Some lubrication oils contain magnesium in addition to or instead of calcium,others contain boron instead of phosphorus compounds and also molybdenum-based additives are some-times used.Magnesium,emitted as MgCO 3or MgO,being less basic than CaCO 3or CaO,respectively,is expected to be less deactivating.It is interesting to note that a zinc-free lubrication oil could be developed for heavy-duty Diesel engines,although zinc dithio dialkyl phosphate is claimed to be a very important multifunctional additive,reducing engine wear,preventing oxidation and bearing corrosion.For a discussion of the functionality of the different additives refer to Keller et al.[10].The increasing part of raps methyl ester (RME)in Diesel fuel,produced by esterification of raps oil by mainly potassium hydroxide,is a potential source of potassium,which is known to be a very strong poison for vanadia-based SCR catalysts.The urea solution,used as reducing agent in the SCR process,poses another problem,since it may contain small amounts of alkali elements.Finally,engine abrasion produces traces of iron,copper,aluminium,chromium etc.,which are also found in the exhaust gas.
In general,only a small part of the compounds in the exhaust gas is expected to deposit on the catalyst surface under usual operating conditions,but the major part will slip through the smooth converter channels in form of small aerosols.However,this is not valid for the urea solution in case of a short distance between urea injection and SCR catalyst,which favours the capture of urea droplets at the catalyst entrance.
The deactivation of SCR catalysts onboard a Diesel heavy-duty truck has already been investigated in a road test and about
Table 1
Lubrication oil and fuel additives in Germany in 1997from [9]Additive (tons)
Engine oils
Fuels Gasoline
Diesel Detergents/dispersants 28,60012,600
1,750
Viscosity improvers
19,700Multifunctional additives
(zinc dithio dialkyl phosphate)4,900Antioxidants
4,300a
1,200
Metal deactivators Foam inhibitor Corrosion inhibitor 400150Cetan number improvers 3,700Antiwax additives 7,100Lubrication improver 1,000
Subtotal 57,50014,20013,700
Total
85,400
a
Sum of the four additives.
Table 2
Concentration of emission-relevant elements in lubrication oils and their most likely combustion products from [9]Element a
Concentration in lubrication oil Combustion products
Light-duty (wt%)
Heavy-duty (wt%)Carbon 11.6611.09C !gases (CO 2,CO,HC b ),PM c Sulfur 0.220.30S !PM (SO 42À),gases (SO 3,SO 2)Calcium 0.190.30Ca !PM (sulfate,carbonate,oxide)Zinc
0.100.12Zn !PM (oxide,pyrophosphate)Phosphorus 0.100.11P !PM (oxide,Zn 3(PO 4)2)Magnesium 0.080.11Mg !PM (sulfate,oxide)Nitrogen 0.070.06N !gases (NO,NO 2)Silicon 0.0020.002Si !PM (oxide)
Total
12.42
12.09
a Oxygen and hydrogen not included.
b HC:hydrocarbons.c
PM:particulate matter.
O.Kro
¨cher,M.Elsener /Applied Catalysis B:Environmental 75(2008)215–227216
10%reduction of the DeNO x activity was found after 535,000km,which was tentatively assigned to phosphorus deposits on the catalyst[11].However,the superposition of the effects of thermal degradation,mechanical attrition,pore blocking,different poisons in various amounts and concentra-tion profiles along the length of the converter and in the catalyst layer prevents any clear conclusion on the pure chemical deactivation effect of the different exhaust gas components. Although some literature can be found on the chemical deactivation of SCR catalysts,they mainly deal with tungsta-free model SCR catalysts and only emissions specific for powerplant application were included,such as halogens, different alkali metals,alkaline earth metals,arsenic,lead, sulphur and phosphorus[12–15].Arsenic was found to be the most severe poison[13],which is fortunately irrelevant for Diesel vehicles.Many studies deal with the strong poisoning effect of alkali metals on vanadia catalysts by decreasing the acidity and preventing the effective activation of ammonia on the surface[16–20].
In the present study,we were interested in the pure chemical effect of the emissions specific for Diesel exhaust gas on practice-oriented V2O5/WO3–TiO2SCR catalysts,thereby excluding any other deactivation mechanisms,such as pore blocking.In thefirst part of the work,the influence of the different exhaust gas constituents on the catalytic performance were investigated.The results of the characterization of selected deactivated catalysts as well as a mechanistic explanation of the observed effects are reported in the second part of this publication[21]and a summary of the project can be found in[22].
For the deactivation experiments,samples of V2O5/WO3–TiO2SCR catalysts coated on metallic substrates were dip-coated with different concentrations and combinations of the elements typical for Diesel exhaust gas emissions in form of salt solutions,calcined and tested for their catalytic perfor-mance.Since some additives reach the SCR catalyst undecomposed via the turbocharger lubrication system without being combusted,supplementing experiments were carried out with the original organometallic additive compounds in order to investigate if they interact in a similar way with the catalyst as the metal salts.
2.Experimental
2.1.Catalysts
Standard metal substrates with a cell density of400cpsi,a length of21mm and a diameter of21mm(V=7.2cm3)were coated by Wacker Chemie with1.3g of a V2O5/WO3–TiO2 catalyst developed for SCR onboard of heavy-duty Diesel vehicles.The formulation is based on a V2O5/WO3–TiO2 catalyst,which was originally developed in combination with a coating process at the Paul Scherrer Institute.For verification of the test results or continuative experiments,the detailed recipe can be found in[7].
Due to the manufacturing process,two large channels remain in the centre of the modules,which were sealed with inert ceramicfibres in order to prevent any uncontrolledflux in the reactor.The volume of these channels was subtracted from the gross volume of the modules resulting in a net reactor volume of7.0cm3.Prior to the tests the catalysts were degreened at5508C for50h.
2.2.Analytics
The catalyst modules were tested in a laboratory quartz reactor at a GHSV of52,000hÀ1.The model exhaust gas (364l N/h)composed of10%O2,5%H2O,1000ppm NO,0to $2000ppm NH3,and balance N2.All components were dosed with gas or liquid massflow controllers,whereby water was evaporated in a heated box.
High resolution FTIR spectroscopy was used to analyze the reaction products NO and NO2with a detection limit of3–5ppm and N2O and NH3with a detection limit of1ppm.The nitrogen balances could always be closed within1–2% accuracy,taking into account the formation of nitrogen with a NH3:NO stoichiometry of1:1according to the standard-SCR reaction:4NH3+4NO+O2!4N2+6H2O.
The catalysts were characterized by ammonia slip–DeNO x curves,which were recorded going from450to2008C in508C steps.For every temperature,ammonia dosage is increased stepwise and after reaching steady-state conditions,the DeNO x value as well as the ammonia emission are measured.For evaluation and interpretation of the data,mainly the DeNO x at 10ppm ammonia slip and DeNO x at a=1.2ða¼NH3
in
=NO inÞwere used.
2.3.Doping of catalysts
Most of the experiments were carried out with0.4mol%of the doping element based on the sum of titanium,tungsten and vanadium,which is equal to20mol%based on vanadium.That allows us to describe the influence of the single doping elements independent of the atomic mass.The chemicals used for the aqueous impregnations are listed in Table3.
The tared catalyst module is dipped for2s in the aqueous solution.After that,the module is shortly blown out to remove excess impregnation solution without alteration of the solution uptake.Moreover,the short impregnation time prevents chromatographic and adsorption effects as well as washing out of the catalyst layer.The catalyst module,after weighing,is dried at908C andfinally calcined at4008C for5h.After the measurement of the catalyst performance,all catalysts were calcined for a second time at5508C for5h and tested again.
For the supplementing experiments on the direct influence of the undecomposed lubrication oil additives,Fuchs Europe Schmierstoffe provided a zinc-containing and a zinc-free mixture of commercial organometallic lubrication oil additives dissolved in hexane.The relative molar composition of mixture FES105-1129based on calcium(Ca=1)was Ca=1,Zn=0, P=0.45,S=1.2,resulting in a molar concentration on the catalyst of0.32mol%Ca and0mol%Zn.The corresponding composition of mixture FES105-1130was Ca=1,Zn=0.24, P=0.45,S=1.2,resulting in0.31mol%Ca and0.07mol%Zn
O.Kro¨cher,M.Elsener/Applied Catalysis B:Environmental75(2008)215–227217
on the catalyst.Prior to the first measurement,both impregnated catalyst modules were sulphated with 100ppm SO 2at 4008C for 5h.3.Results
parison of undoped and doped SCR catalysts Ammonia slip–DeNO x curves proved to be a suitable graphical representation for the assessment of the quality of SCR catalysts [7,23].Fig.1exemplarily shows the shape of these curves for one of the undoped catalyst samples after degreening at 5508C for 50h.For such a highly active SCR catalyst,the curves run through the lower right corner for temperatures between 350and 4508C,i.e.high DeNO x values are reached at very low ammonia emissions.For low temperatures,the catalyst activity decreases,resulting in ammonia emissions at lower DeNO x values.At temperatures of >=3508C,small variances in the catalytic activity of the different samples hardly affect the slip–DeNO x curves and differences are within the measuring accuracy of +/À1%DeNO x .At 2508C or below,the curve shape is much more sensitive towards the sample activity but the reproducibility is
reduced.The large ammonia storage capacity of the catalysts at low temperatures results in a waiting time of up to 10min until steady-state conditions are reached after concentration changes.Small,but unavoidable fluctuations in the reactor temperature and water dosage during this period causes small distortions of the slip–DeNO x curves.Thus,the curve shapes at the intermediate temperature of 3008C are best suited for the purpose of comparison,since it is sensitive regarding the catalyst activity but quickly equilibrated and hardly affected by fluctuations of the water feed concentration.
A further reduction of the amount of raw data is possible by extracting only the DeNO x at 10ppm ammonia slip from the curves,representing realistic operating conditions in an SCR system onboard of a Diesel vehicle,and the DeNO x values for constant ammonia overdosage with a =1.2,providing infor-mation about the maximum SCR activity of the catalysts.Although the 32laboratory catalyst samples from Wacker were coated with very similar amounts of active mass,the samples showed a significant variance in activity of up to 20%.The mean DeNO x values at 10ppm NH 3slip were 49.9%with a standard deviation of 7.8%at 2508C,87.3%with a standard deviation of 3.4%at 3008C,and 97.0%with a standard deviation of 0.7%at 4508C.Due to the large deviations in activity of six catalyst samples from the mean value,these catalysts were not included in the study.Thereby,the standard deviation of the remaining 26samples was significantly reduced,which were used for the deactivation experiments.N 2O formation was as expected small with values between 10and 12ppm at 4508C (0ppm at 2508C and 3008C).
The significant differences in the catalytic activity of the undoped samples prevented a direct comparison of the deactivated catalysts.Therefore,a relative DeNO x was defined as:
rel :DeNO x 10ppm ¼
DeNO x cat x deact :DeNO x cat x new :
(1)
with rel :DeNO x 10ppm ¼relative DeNO x at 10ppm ammonia slip (–),DeNO x cat x deact :¼DeNO x of deactivated catalyst X at 10ppm ammonia slip (%),and DeNO x cat x new :¼DeNO x of new catalyst X at 10ppm ammonia slip (%).
The relative DeNO x at a =1.2was calculated accordingly.By multiplying these relative
DeNO x values with the mean
Fig.1.Ammonia emissions as function of DeNO x for a typical V 2O 5/WO 3–TiO 2catalyst coated on metal support designed for SCR applications onboard of Diesel vehicles.The catalyst sample was degreened at 5508C for 5h before testing.(*)2008C,(*)2508C,(4)3008C,(~)3508C,(Â)4008C,(&)4508C.
Table 3
List of doping elements used in this study Element Precursor
Formula Quality Supplier Potassium (K)
Potassium carbonate K 2CO 3Puriss.Fluka Potassium sulfate
K 2SO 4Pro analysi Merck Potassium hydrogensulfate KHSO 4
Pro analysi Merck Molybdenum (Mo)Ammonium molybdate
(NH 4)6Mo 7O 24Á4H 2O Ultra (>99%)Fluka Phosphorus (P)Orthophosphoric acid (85%)H 3PO 4Pro analysi
Merck Sulphur (S)Sulphur dioxide SO 2
1%in N 2(>99.9%)Carbagas Zinc (Zn)Zinc acetate Zn(CH 3COO)2Á2H 2O Puriss.Fluka Boron (B)Orthoboric acid H 3BO 3
Puriss.Fluka Calcium (Ca)Calcium acetate Ca(CH 3COO)2Áx H 2O 96%Merck Magnesium (Mg)
Magnesium acetate
Mg(CH 3COO)2Á4H 2O
Purum
Fluka
O.Kro
¨cher,M.Elsener /Applied Catalysis B:Environmental 75(2008)215–227218
DeNO x value of all undoped samples,normalized DeNO x values were obtained,which are independent of the scattering activities of the undoped catalyst samples.
The deactivation of the catalyst is best described by the decrease of the mass based reaction rate constant,which can be calculated for afirst-order reaction in a plug-flow reactor according to:
k mass¼ÀVÃ
W
lnð1ÀXÞ(2)
with k mass=mass based reaction rate constant[cm3/(g s)],V*= volumeflow rate at T reactor,p reactor(cm3/s),W=catalyst mass (g),and X=NO conversion(–).
All reaction rate constants k mass were calculated from DeNO x at a=1.2,which was found to be equal to the maximum attainable DeNO x for almost all catalysts,i.e.ammonia is present in excess on the catalyst surface and pseudofirst-order conditions are achieved.Relative k mass values were obtained
analogously to rel.DeNO x
10ppm .
k mass is only with the intrinsic activity of the catalyst at low temperatures up to2508C.For highly active catalysts,such as used in this study,the SCR reaction increasingly takes place in the upper catalyst layer already at 3008C or higher.Thus,the deactivation of the catalysts might be partly compensated by deeper diffusion of the reactants and, thus,a better exploitation of the catalyst[24].However,the evaluation of the results showed only a moderate increase of the relative k mass for higher temperatures.The relative k mass values for250and3008C should,therefore,be representative of the intrinsic activity loss of the catalyst.
3.2.SCR activity after impregnation with water
In a preliminary investigation,it was tested,if the catalyst activity is affected by water and if water-free solvents have to be used for following investigations.Therefore,the catalysts werefirst tested as delivered and after a pre-treatment at5508C for50h.Then,the catalysts were dip-coated with water and tested again.For the dip coating,the catalysts were immersed into deionized water for2s,shortly blown out and dried in airflow at908C.The loss in catalyst mass was$19mg,which is$2%of the original mass.The blown-out solution had a deep yellow colour.Finally,the catalysts were calcined at4008C for 5h.Subsequently,a second calcination was done at5508C for 5h in order to intensify the solid-state reactions.
The catalyst activity increased by the thermal pre-treatment at5508C for50h as shown in Table4.The subsequent dipping in water and the calcination at4008C had almost no influence on the catalyst with a tendency to slightly smaller activities. However,the second calcination at5508C for5h increased the activity again.In general,the differences were very low and near the measuring accuracy and reproducibility.
N2O formation was generally low and only detectable at higher temperatures(Table4).Dipping in water caused a slight increase of the N2O formation,but still on a reasonable level.The second calcination at5508C increased the selectivity a little bit again,resulting in a slight decrease in N2O formation.
3.3.Doping with single elements
Atfirst,catalyst modules were doped with0.4mol%of the single elements boron,phosphorus,zinc,calcium,magnesium and potassium.Boron was added as aqueous ortho boric acid (H3BO3),phosphorus as ortho phosphoric acid(H3PO4), potassium as carbonate(K2CO3)and zinc,calcium as well as magnesium as acetate(Zn(CH3COO)2,Ca(CH3COO)2, Mg(CH3COO)2).During thefirst calcination at4008C for5h, the compounds decomposed in different ways.Calcium and magnesium acetate decomposed to the corresponding carbo-nates(CaCO3,MgCO3),whereas zinc acetate was likely converted to zinc oxide(ZnO),as zinc carbonate itself decomposes already at1208C.Potassium carbonate decom-posed to potassium oxide(K2O)during calcination.
Fig.2shows the DeNO x values at10ppm ammonia slip for 4508C and3008C of the samples in the undoped state,after doping followed by thefirst calcination at4008C and after the second calcination at5508C.The addition of ortho boric acid caused no deactivation for this concentration.Doping with phosphorus resulted in a moderate decrease of activity,i.e.k mass was lowered to80–85%of the original value.At4508C,this had only a small influence on DeNO x(3–4%decrease),while DeNO x was10%lower at3008C and20–25%lower below 3008C.For molybdenum,no deactivation was observed at high temperatures,but DeNO x slightly decreased at200and2508C. The addition of molybdenum increased the N2O formation by a factor of2–3to40ppm at4508C.The addition of zinc caused about50%reduction of the catalyst activity k mass.At high temperatures DeNO x at10ppm ammonia slip was reduced to 80–90%of the original value.At lower temperatures,this
Table4
Influence of testing procedure on DeNO x and N2O formation
T(8C)DeNO x at10ppm NH3slip(%)N2O at10ppm NH3slip(ppm)
New50h at
5508C Dipping in water
+5h at4008C
+5h at
5508C
New50h
at5508C
Dipping in water
+5h at4008C
+5h at
5508C
4509194929346119 400899696961122 350809393940010 30059767778
25030434242
20011151516
O.Kro¨cher,M.Elsener/Applied Catalysis B:Environmental75(2008)215–227219
decrease was even more pronounced.By the addition of calcium,the activity of the catalyst dropped to 40%of its original value.Doping with magnesium resulted in a clear deactivation over the entire temperature range.Its deactivation potential was;however,lower than calcium.The addition of potassium caused a very strong deactivation of the catalyst to below 10%of its original value.
After the first measurement,the catalysts were calcined a second time at higher temperature (5508C for 5h)in order to intensify the solid-state reactions of the dopants with the catalysts and tested again.The catalyst with boron showed again the same activity.The activities of the other catalysts increased by 10–20%,i.e.the deactivation was partly restored.However,this effect was not very pronounced and the DeNO x values only slightly changed in the range of a few percent.No increase in N 2O formation could be observed for any of the doped catalysts,except for the molybdenum-containing sample.In general,the N 2O formation was affected in parallel to the deactivation.
Doping with 0.4mol%potassium resulted in an already strong deactivation,which prevented the reliable assessment of its deactivation potential (Fig.2).By lowering the potassium doping from 0.4to 0.2%clearly higher DeNO x values could be obtained for low temperatures,but at 4508C still only 15%of the original values were measured (Fig.3).For potassium content of only 0.1%,a moderate catalyst deactivation was achieved.The DeNO x values measured with this catalyst
ranged between those for the catalyst with 0.4%Zn and 0.4%Ca.
3.4.Effect of counter-ions
The influence of counter-ions from different inorganic acids on the primary effect of the cations was tested by doping catalysts with the same potassium concentration in form of three different potassium salts (K 2CO 3=basic,K 2SO 4=neu-tral,KHSO 4=acidic).In case of calcium,direct impregnation with calcium sulphate was avoided due to its low solubility.Therefore,the catalyst was impregnated with calcium acetate and treated with 100ppm SO 2in the feed gas at 4008C for 5h,which resulted in the formation of CaSO 4on the catalyst.Moreover,calcium was tested as phosphate and borate in order to compare these compounds with calcium sulphate,which was expected to be relatively inert on the SCR catalysts.Magnesium sulphate was likewise prepared by treatment with SO 2and magnesium phosphate was also tested.Zinc was combined with phosphate and borate.The concentration of each element was 0.4mol%,except for the in-situ prepared sulphates.A simple 1:1ratio was chosen for the cations and anions since the real concentration ratios of the different components in the exhaust gas is unknown and it is not clear,how they meet on the surface and what are the exact reaction products (e.g.zinc
phosphate
parison of catalysts deactivated by different concentrations of potassium.DeNO x at 10ppm NH 3slip (a)after calcination at 4008C for 5h and (b)after calcination at 5508C for 5
h.
Fig.2.Catalysts doped with single elements.DeNO x at 10ppm NH 3slip (a)
after calcination at 4008C for 5h and (b)after calcination at 5508C for 5h.
O.Kro
¨cher,M.Elsener /Applied Catalysis B:Environmental 75(2008)215–227220
Zn 3(PO 4)2or zinc pyrophosphate Zn 2P 2O 7).Moreover,this
ratio allowed us to impregnate calcium as well as zinc with phosphate in one step,since the corresponding hydrogen phosphates are sufficiently soluble for the preparation of the impregnation solution.
The deactivation potential of potassium is so strong,that the addition of a second component had hardly any effect in our experiments (Fig.4).DeNO x decreased to below 10%of its original value for the catalyst with potassium sulphate.The catalyst with potassium hydrogensulfate was only slightly better,but DeNO x at 10ppm ammonia slip did not exceed 20%of the value before doping.No appreciable influence was observed for phosphorus or boron.
The deactivation of the calcium containing catalysts strongly depends on the acidity of the corresponding acid of the second doping element.CaSO 4showed no deactivation effect at high temperatures,but at low temperatures a significant activity loss was observed,comparable to that of calcium phosphate.The addition of orthoboric acid only slightly improved the calcium-doped catalyst.
The sulphated as well as the phosphated magnesium doped samples showed a clearly higher activity than the sample without counter-ions,however,the improvement was much less pronounced than in the case of calcium.The activity of the catalyst with sulphated magnesium
is again reduced by the second calcination at 5508C,whereas for the phosphate-containing catalyst it recovers somewhat.
The combination of zinc with phosphorus resulted in a distinctly lower deactivation at low and intermediate tempera-tures than doping with zinc alone.At 400and 4508C both catalysts are comparable.After the additional calcination at 5508C the Zn/P catalyst showed a slight decrease in DeNO x at 400and 4508C.The catalyst with zinc and boron was quite similar to that with zinc alone.After calcination at 4008C,somewhat smaller DeNO x values are obtained than with zinc,but after calcination at 5508C they are virtually the same.The activity of the catalysts Zn/B and Zn are clearly improved by the calcination at 5508C,whereas the catalyst Zn/P remain more or less on the same level,so that all catalysts are getting quite similar.
3.5.Doping with combinations of cations and anions In addition to the doping with single cations and the doping with combinations of single cations with different counter-ions,we examined the influence of more than one cation or two anions on the catalyst performance (Fig.5
).
Fig.4.Effect of counter-ions on the deactivation potential.DeNO x at 10ppm NH 3slip (a)after calcination at 4008C for 5h and (b)after calcination at 5508C for 5
h.
Fig.5.Effect of combinations of elements.DeNO x at 10ppm NH 3slip (a)after calcination at 4008C for 5h and (b)after calcination at 5508C for 5h.The samples Ca/P/SO 4and Ca/Zn/P/SO 4were prepared by impregnation of V 2O 5/WO 3–TiO 2catalysts with mixtures of commercial lubrication oil additives in hexane.
O.Kro ¨cher,M.Elsener /Applied Catalysis B:Environmental 75(2008)215–227221。