catalyst activity as indicated by the positive estimated parameters of S and T which
agrees with experimental data as well Positive interactions exist between potassium
and sulfur sodium and sulfur however negative interactions exist between sulfur and
and poisons on catalysts surface chemistry including BET surface areas average pore
activity These investigations from lab-prepared catalysts together with commercial
Addition of tungsten sulfate and poisons alters insignificantly the surface areas
increase NO reduction activity and this increase is associated with the increased
number of Broslashnsted acid sites as indicated by ammonia adsorption The change in
activity is not related to (indeed is either inversely correlated or uncorrelated with)
Lewis acid site population and correlates closely with the Broslashnsted acid site
concentration The most specific evidence for this conclusion comes from the strong
131
correlation between NO activity and Broslashnsted-acid ammonia adsorption peak areas
and the lack of correlation with for example Lewis-acid peak area Therefore
Broslashnsted acid sites participate much more actively than Lewis acid sites in catalytic
reduction of NO with observed activity commonly being proportional to Broslashnsted
acid site concentrations However Broslashnsted acid sites alone do not provide NO
reduction activity for example 9WO3Ti with abundant Broslashnsted acid sites shows
zero SCR activity In this sense the acid sites themselves are not the active centers
but rather play a supporting role In addition to increase Broslashnsted acid site
concentration vanadia species on titania surfaces provide by far the greatest NOx
reduction activity although some minor amount of activity occurs on titania surfaces
Vanadia shows capability to provide a large quantity of Broslashnsted acid sites on
titania surface as indicated in Table 510and Figure 525 5 wt vanadia catalyst
shows a strong ammonia chemisorption on Broslashnsted acid sites with a IR peak area of
600 while 9 WO3Ti provides an ammonia adsorption IR peak area of 252 on
Broslashnsted acid sites On the other hand vanadia actively oxidizes SO2 to SO3 due to its
redox activity therefore vanadia content usually remains below 1 wt for
commercial SCR catalysts to minimize SO2 oxidation (Bartholomew and Farrauto
2006 Bartholomew 1997) The literature consistently reports that vanadia appears as
a highly active redox agent in catalysts (Bartholomew and Farrauto 2006 Liu et al
2005 Wachs 2005 Wachs et al 1996 Wachs et al 2005 Wachs and Weckhuysen
1997 Wang and Wachs 2004) and data from this investigation agrees that its
function in 1V2O5-9 WO3TiO2 is probably as a redox site rather than a Broslashnsted
acid site Tungsten and sulfate species on the other hand provide most of the
Broslashnsted acid sites Individual tungsten and sulfate species react in conjunction with
vanadia species to increase adjacent adsorbed ammonia site concentrations and hence
132
observed reactivity Consequently a dual-site reaction mechanism appears most likely
for the NOx reduction by ammonia where a redox site (vanadia sites) function
together with adjacent Broslashnsted acid sites (tungsten andor sulfate species) as the
active centers during the SCR reaction cycle
Thus far there has been no definite indication about the specific Broslashnsted acid
sites in the literature (Wachs 1997) vanadia tungsten and sulfate species all being
nominated as candidates The current investigation indicates that Broslashnsted acid sites
do not need to associate with a specific type of surface species so long as the surface
sites can provide Broslashnsted acid and that all three sites contribute to the surface
concentrations of ammonia Topsoslashe et al (1995) also recommended a dual acid-redox
catalytic cycle for NO reduction The current investigation conceptually agrees with
most of Topsoslashersquos dual active sites concepts and provides more details Specifically
many species can provide Broslashnsted acid sites with 1 wtvanadia being among the
least efficient and the primary function of 1 wt vanadia is further clarified which is
predominantly a redox site Given that adsorbed ammonia predominately comes from
non-vanadia sites and that redox appears to happen almost exclusively at vanadia
sites the reaction center is most likely the interface between vanadia and the catalyst
substrate
Moreover our adsorption investigations also suggest the edge between vandia
and titania could be the active center Vanadia sulfate and NO species interact with
surface OH group on titania Vanadia provides the active sites and sulfate species
increase the activity Vanadia and sulfate both suppress NO adsorption on catalyst
surface and sulfate contents decreases with increasing vanadia concentrations
Consequently all of the above three species NO V and S affect the SCR reaction
and they compete with each other for surface OH sites on TiO2 One explanation for
133
the observation is that instead of vanadia species being the active center the edge
between the vanadia and Broslashnsted acid sites could be the active center
A general view based on our investigation invovles strongly adsorbed ammonia
species on a catalyst surface at acid sites activation at the edge between redox sites
and acid sites and then reaction with gas phase NO through an Eley-Rideal reaction
mechanism Figure 547 illustrates the details which is similar to Topsoslashersquos reaction
mechanism routine (Topsoslashe et al 1995) Moreover the interpretation of this
investigation points out that not only vanadia as suggested by Topsoslashe but also
tungsten and sulfate could provide Broslashnsted acid sites Moreover the synergy
between the oxidation rate on redox site and the transformation rate of adsorbed
ammonia on Broslashnsted acid sites to the active center seems critical for the SCR
reaction rate
Figure 547 Scheme illustrating the cycle of the SCR reaction over vanadiatitania
catalyst based on mechanism proposed by et al (Topsoslashe et al 1995)
134
135
Chapter 6 Conclusions and Recommendations
Vanadia supported on titania material represents the predominant commercial
SCR catalyst applied to reduce NOx with NH3 from boilers burning coal-biomass and
coals Although SCR of NO is efficient deactivation of vanadia catalyst represents a
potential major problem in industrial applications contributing to the cost increase
and applying difficulties Therefore a series of activity tests and surface chemistry
investigations including BET surface area and average pore diameter measurements
NOx reduction activity estimations and surface chemical composition analyses on
both commercial and laboratory samples and ammonia and nitric oxide adsorption
and sulfation on lab-prepared samples demonstrated the reaction and deactivation
mechanism of vanadia catalyst for coal-biomass and coal-firing boilers
61 Principal Conclusions
1 Activity investigations on commercial exposed commercial M1and M2 samples
show catalysts deactivate with extended exposure to flue gases from both coal
firing and coal-biomass co-firing boilers The activity loss of coal exposure
samples differs from that of biomass-exposed samples Kinetic investigations
combined with surface characterization suggest that foulingpore plugging
dominates the deactivation mechanism for vanadia catalyst during coal-fired
boilers while poisoning is significant but not dominant Poisoning is more
significant during biomass-coal firing
136
2 Based on NH3-NO coadsorption tests and investigation of tungsten NO
reduction activity chemisorbed ammonia reacts with gas-phase or weakly
adsorbed nitric oxide on a dual redox-acid active center through an Eley-Rideal
mechanism Vanadia provides redox sites and Broslashnsted acid sites (supplied
mostly by tungsten and sulfate species and less by vanadia species) provide the
primary acid sites The reduction of nitric oxide with ammonia occurs possibly
at the edge of redox and acid sites and in any case involves adsorbed ammonia
primarily from adjacent Broslashnsted acid sites reacting with vanadia
3 Alkali metals potassium (K) and sodium (Na) as well as an alkaline earth metal
calcium (Ca) poison vanadia catalysts with alkali metals being stronger poisons
than alkaline earth metals K Na and Ca deactivate vanadia catalysts by
neutralizing or displacing Broslashnsted acid sites and by decreasing acidity of
Broslashnsted acid sites Poison strengths scale with basicities
4 Broslashnsted acid sites correlate strongly with activity while Lewis acid sites are
uncorrelated or inversely correlated Broslashnsted acid sites can form on tungsten
sulfate sites andor vanadia species
5 Sulfate species form on catalysts exposed to typical industrial SO2-laden flue
gas In situ FTIR analyses of sulfation on vanadia catalysts combined with XPS
measurement on sulfated samples indicate that titania instead of vanadia sites
interact with suflate species and vanadia sites (unsulfated) remain as the redox
center Comparisons of intrinsic kinetic constants of fresh and sulfated 1
V2O5TiO2 at temperatures from 524 -564 K and the sulfation effect on
ammonia adsorption suggest that sulfate species assist vanadia sites catalytically
by providing more Broslashnsted acid sites
137
6 Tungsten greatly increases vanadia catalyst NO reduction activity (by about
250 in typical systems) although individual tungsten species possess no
catalytic activity for NOx reduction This large increase originates from the large
increase in Broslashnsted acid site population associated with tungsten addition
Tungsten also significantly mitigates the effects of alkali and alkaline earth
poisoning though these basic compounds represent potent poisons to even
tungsten-laden catalysts
62 Unique Contributions
The following represent the most unique and original contributions of this work
compared to the existing literature
1 Applied results from better controlled lab-scale analysis to fundamentally
explain observations from industrially exposed commercial catalysts The results
indicate that deactivation of commercially exposed vanadia catalysts is a
combination of channel plugging pore plugging masking and poisoning
Different mechanisms dominate in deactivation of different catalysts exposed to
the same flue gas at the same time and the same catalysts exposed to different
flue gas with varying time None of these mechanisms is insignificant but the
physical mechanisms are possibly the most severe in coal-based systems
2 Conducted First ndash time ever in situ sulfation investigations on 0-5wt
V2O5TiO2 under both dry and wet conditions establish a pronounced role of
sulfur in SCR activity
3 Provided evidence that titania largely sulfates during SCR reactions and
vanadia sulfates to a much lower extent if at all
138
4 Illustrated that sulfation increases NOx reduction activity by introducing more
Broslashnsted acid sites on catalyst surfaces without changing the bond strength
(acidity) of these sites
5 Discovered that tungsten possesses no NOx reduction activity but contributes
greatly to the number of Broslashnsted acid sites and hence the activity of vanadia-
based catalysts
6 Clarified that vanadia sulfur and NO compete for the same surface sites which
are OH groups on titania
7 First ndashtime ever statistically analyzed and concluded that interactions between
some poisons and sulfation and between sulfation and temperature are
significant Interactions between poisons are generally not significant
63 Recommendations for Future Research
The current investigation examined industrial fresh and exposed commercial
monolith catalyst activities and used a model adapted from the literature to calculate
kinetic constants This model appears as an apparent but not an intrinsic kinetic
activity calculation due to many simplifications A more accurate model including
more fundamental descriptions of catalyst properties (surface reactions effects of
composition gradients in the catalyst (eg poisons) multiple pore distribution model
and radial bulk flow instead of a homogeneous bulk flow) would improve the
scientific interpretation of these data However the current model contains the
amount of complexity appropriate for a CFD code
Sulfation investigations demonstrated (for the first-time ever) in situ FTIR
evidence of sulfation sites on a vanadia catalyst surfaces and the effect of sulfation
was also thoroughly investigated both spectroscopically and kinetically on post-
139
sulfation samples More meaningful data could arise from in situ situations where the
SO2 gas appears in the reactant gas which is more representative of conditions
observed in commercial boilers
The alkali metals potassium (K) and sodium (Na) as well as the alkaline earth
metal calcium (Ca) poison vandia catalysts The contaminated samples applied in our
poisoning investigation were prepared by an incipient impregnation method
Deposition of poison metal compounds (chloride or sulfate compound) on to a
catalyst surfaces may better simulate the actual deactivation situations that occur in
industrial SCR reactors
Oxidation state shifting of vanadium atom during the SCR reaction could be
monitored by in situ Raman spectroscopy This in situ investigation would provide
oxidation state information for vanadia species during SCR reaction which should
provide abundant and critical highlights about the vanadia species rolefunction
during the SCR reaction cycle for example the redox capability Moreover Raman
spectroscopy would reveal the vanadia species catalytic mechanism as well as catalyst
deactivation mechanism in a more direct way
Calculation of a turn over frequency (TOF) either in apparent or intrinsic
calculations for kinetic investigations would provide more mechanistic information
than traditional kinetics Calculating TOF requires active metal surface area
measurement In this case the vanadia surface area would need to be measured which
could be measured by oxygen chemisorption
140
141
REFERENCES
Amiridis M D I E Wachs G Deo J M Jehng and D S Kim Reactivity of V2O5 catalysts for the selective catalytic reduction of NO by NH3 Influence of vanadia loading H2O and SO2 Journal of Catalysis 161 (1) 247-253 (1996)
Amores J M G V S Escribano G Ramis and G Busca An FT-IR study of ammonia adsorption and oxidation over anatase-supported metal oxides Applied Catalysis B-Environmental 13 (1) 45-58 (1997)
Bartholomew C H and R J Farrauto Fundamentals of industrial catalytic processes Hoboken NJ John Wiley c2006 xxiii 966p (2006)
Bartholomew R J F a C H Fundamentals of industrial catalytic processes Blackie academic amp professional (1997)
Bartok W and A F Sarofim Fossile fule combustion A source book New York John amp Wiley (1991)
Beretta A E Tronconi G Groppi and P Forzatti Monolithic catalysts for the selective reduction of NOx with NH3 from stationary sources Chemical Industries (Dekker) 71 (Structured Catalysts and Reactors) 121-148 (1998)
Briand L E O P Tkachenko M Guraya X Gao I E Wachs and W Gruenert Surface-Analytical Studies of Supported Vanadium Oxide Monolayer Catalysts Journal of Physical Chemistry B 108 (15) 4823-4830 (2004)
Busca G F Bregani and G Rocchini Inquinamento 34 40 (2002)
Busca G M A Larrubia L Arrighi and G Ramis Catalytic abatement of NOx Chemical and mechanistic aspects Catalysis Today 107-108 139-148 (2005)
Busca G L Lietti G Ramis and F Berti Chemical and mechanistic aspects of the selelctive catalytic reduction of NOx by ammonia over oxide catalysts A Review Appl Catal B Environmental 18 (1-2) 1-36 (1998)
Butler A and R Nicholson Life Death and Nitric Oxide Royal Society of Chemistry Cambridge UK (2003)
Centeno M A I Carrizosa and J A Odriozola NO-NH3 coadsorption on vanadiatitania catalysts determination of the reduction degree of vanadium Applied Catalysis B-Environmental 29 (4) 307-314 (2001)
142
Chen J P M A Buzanowski R T Yang and J E Cichanowicz Deactivation of the vanadia catalyst in the selective catalytic reduction process Journal of the Air amp Waste Management Association 40 (10 Oct) 1403-1409 (1990)
Chen J P and R T Yang Mechanism of Poisoning of the V2O5TiO2 Catalyst for the Reduction of NO by NH3 Journal of Catalysis 125 (2) 411-420 (1990)
Chen J P and R T Yang Role of WO3 in Mixed V2O5-WO3TiO2 Catalysts for Selective Catalytic Reduction of Nitric-Oxide with Ammonia Applied Catalysis a-General 80 (1) 135-148 (1992)
Chen J P and R T Yang Selective catalytic reduction of NO with NH3 on SO4-
2TiO2 superacid catalyst Journal of Catalysis 139 277-288 (1993)
Choo S T Y G Lee I-S Nam S-W Ham and J-B Lee Characteristics of V2O5 supported on sulfated TiO2 for selective catalytic reduction of NO by NH3 Applied Catalysis AGeneral 200 (1 Aug 28) 177-188 (2000)
Choung J W I-S Nam and S-W Ham Effect of promoters including tungsten and barium on the thermal stability of V2O5sulfated TiO2 catalyst for NO reduction by NH3 Catalysis Today 111 (3-4) 242-247 (2006)
Coates J Interpretation of infrared spectra a practical approach Encyclopedia of analytical chemistry 10815-10837 (2000)
Djerad S L Tifouti M Crocoll and W Weisweiler Effect of vanadia and tungsten loadings on the physical and chemical characteristics of V2O5-WO3TiO2 catalysts Journal of Molecular Catalysis A Chemical 208 (1-2) 257-265 (2004)
DOE Technologies for the combined control of sulfur dioxide and nitrogen oxides emissions from coal-fired boilers Clean Coal Technology Topical report number 13 (1999)
DOE N Selective Catalytic Reduction (SCR) Technology for the Control of Nitrogen Oxide Emissions from Coal-Fired Boilers (2005)
Dunn J P J-M Jehng D S Kim L E Briand H G Stenger and I E Wachs Interactions between surface vanadate and surface sulfate species on metal oxide catalysts Journal of Physical Chemistry B 102 (32 Aug 6) 6212-6218 (1998a)
Dunn J P P R Koppula H G Stenger and I E Wachs Oxidation of sulfur dioxide to sulfur trioxide over supported vanadia catalysts Applied Catalysis B-Environmental 19 (2) 103-117 (1998b)
EPA Clean Air Markets-Progress and Results Presentation from Internet (2004)
Fenimore C P Formation of nitric oxide in premixed hydrocarbon flames Symp (Int) Combust [Proc] 13th 373-80 (1971)
143
Fogler H S Elements of Chemical Reaction Engineering Upper Saddle River New Jersey 07458 Prentice Hall PTR (1999)
Forzatii P and L Lietti Recent Advances in De-NOxing Catalysis for Stationary Applications Heterogeneous Chemistry Reviews 3 33-51 (1996)
Forzatti P Present status and perspective in de-NOx SCR analysis Appl Catal A General 222 221-236 (2001)
Forzatti P and L Lietti Catalyst deactivation Catalysis Today 52 165-181 (1999)
Forzatti P L Luca and T Enrico Nitrogen Oxides Removal-Industrial Encyclopedia of Catalysis John Wiley amp Sons Inc (2002)
Franklin H N The effect of fuel properties and characteristics on selective catalytic reduction systems ASME International Joint Power Generation Conference (1996)
Ghodsi M and J P Prossnitz Differential thermal analysis of the anatase to rutile transformation in boron-silicon glasses Proc Eur Symp Therm Anal 2nd 572-5 (1981)
Glarborg P A D Jensen and J E Johnsson Fuel nitrogen conversion in solid fuel fired systems Progress in Energy and Combustion Science 29 (2) 89-113 (2003)
Hadjiivanov K Indentification of neutral and charged NxOy surface species by IR spectroscopy Catal Rev-SCIENG 42 (1amp2) 71-144 (2000)
Janssen F J J G and K F M G V Den Mechanism of the reaction of Nitric oxide ammonia and oxygen over vanadia catalysts 2 Isotopic transient studies with Oxygen-18 and nitrogen -15 J Phys Chem 91 6633-6638 (1987)
Jung S M and P Grange Characterization and reactivity of pure TiO2-SO42- SCR
catalyst influence of SO42- content Catalysis Today 59 (3-4) 305-312
(2000)
Kamata H K Takahashi and C U I Odenbrand Role of K2O in the selective reduction of NO with NH3 over a V2O5(WO3)TiO2 commercial selective catalytic reduction catalyst Journal of Molecular Catalysis A Chemical 139 189-198 (1999)
Khodayari R and C U I Odenbrand Deactivating effects of lead on the selective catalytic reduction of nitric oxide with ammonia over a V2O5WO3TiO2 catalyst for waste incineration applications Industrial amp Engineering Chemistry Research 37 (4) 1196-1202 (1998)
Khodayari R and C U I Odenbrand Selective catalytic reduction of NOx A mathematical model for poison accumulation and conversion performance
144
Chemical Engineering Science Elsevier Science LtdExeterEngl 54 1775-1785 (1999)
Khodayari R and C U I Odenbrand Regeneration of commercial SCR catalysts by washing and sulphation effect of sulphate groups on the activity Applied Catalysis B-Environmental 33 (4) 277-291 (2001a)
Khodayari R and C U I Odenbrand Regeneration of commercial TiO2-V2O5-WO3 SCR catalysts used in bio fuel plants Applied Catalysis B-Environmental 30 (1-2) 87-99 (2001b)
Lani B W I Thomas J Feeley J Murphy and L Green A review of DOENETLs Advanced NOx control Technology RampD Program for Coal-Fired Power Plants DOENETL NOx RampD Program Review March 2005 (2005)
Li F B-s Jin T Liu and H-m Fan Study of DeNOx behavior of V2O5-based catalyst with different catalyst promoters supported on commercial anatase TiO2 Huanjing Huaxue 24 (4) 374-378 (2005)
Lietti L J L Alemany P Forzatti G Busca G Ramis E Giamello and F Bregani Reactivity of V2O5-WO3TiO2 catalysts in the selective catalytic reduction of nitric oxide by ammonia Catalysis Today 29 (1-4) 143-148 (1996)
Lietti L and P Forzatti Temperature-Programmed Desorption - Reaction of Ammonia over V2O5TiO2 De-NO(X)Ing Catalysts Journal of Catalysis 147 (1) 241-249 (1994)
Lietti L G Ramis F Berti G Toledo D Robba G Busca and P Forzatti Chemical structural and mechanistic aspects on NOx SCR over commercial and model oxide catalysts Catalysis Today 42 (1-2) 101-116 (1998)
Lin W Y and P Biswas Metallic Particle Formation and Growth Dynamics During Incineration Combustion Science and Technology 101 (1-6) 29-43 (1994)
Lisi L G Lasorella S Malloggi and G Russo Single and combined deactivating effect of alkali metals and HCl on commercial SCR catalysts Applied Catalysis B-Environmental 50 (4) 251-258 (2004)
Liu J Z Zhao C Xu A Duan L Zhu and X Wang Diesel soot oxidation over supported vanadium oxide and K-promoted vanadium oxide catalysts Applied Catalysis B Environmental 61 (1-2) 36-46 (2005)
Moradi F J G M Brandin and M Sohrabi Deactivation of oxidation and SCR catalysts used in flue gas cleaning by exposure to aerosols of high-and low melting point salts potassium salts and zinc chloride Applied Catalysis 46 65-76 (2003)
Nova I L dallAcqua L Lietti E Giamello and F Pio Study of thermal deactivation of a de-NOx commercial catalyst Appl Catal B Environmental 35 31-42 (2001)
145
Orsenigo C L Lietti E Tronconi P Forzatti and F Bregani Dynamic Investigation of the Role of the Surface Sulfates in NOx Reduction and SO2 Oxidation over V2O5-WO3TiO2 Catalysts Ind Eng Chem Res 37 2350-2359 (1998)
Ozkan U S Y Cai and M W Kumthekar Mechanistic studies of selective catalytic reduction of nitric oxide with ammonia over V2O5TiO2 (anatase) catalysts through transient isotopic labeling at steady state Journal of Physical Chemistry ACSWashingtonDCUSA 99 2363-2371 (1995)
Ozkan U S M W Kumthekar and Y P P Cai Selective Catalytic Reduction of Nitric-Oxide over VanadiaTitania Catalysts - Temperature-Programmed Desorption and Isotopically Labeled Oxygen-Exchange Studies Industrial amp Engineering Chemistry Research 33 (12) 2924-2929 (1994)
Pena D A Development and characterization of mixed oxide catalysts for the selective catalytic reduction of nitric oxide from stationary sources using ammonia Chemical engineering University of Cincinnati (2003)
Pritchard S S Kaneki and K Suyama Opitmizing SCR catalyst design and performance for coal-fired boilers
Ramis G and G Busca Fourier Transform Infrared Study of the Adsorption and Coadsorption of Nitric oxide Nitrogen Dioxide and Ammonia on TiO2 anatase Appl Catal 64 243-257 (1990)
Ramis G G Busca F Bregani and P Forzatti Fourier transform-infrared study of the adsorption and coadsorption of nitric oxide nitrogen dioxide and ammonia on vanadia-titania and mechanism of selective catalytic reduction Applied Catalysis 64 (1-2 Sep) 259-278 (1990)
Reddy B M I Ganesh and B Chowdhury Design of stable and reactive vanadium oxide catalyst supported on binary oxides Catalysis Today 49 115-121 (1999)
Samaniego J M B Labegorre F N Egolfopoulos M Ditaranto J C Sautet and O Charon Mechanism of nitric oxide formation in oxygen-natural gas combustion Symposium (International) on Combustion [Proceedings] 27th (Vol 1) 1385-1392 (1998)
Saur O M Bensitel A B M Saad J C Lavalley C P Tripp and B A Morrow The Structure and Stability of Sulfated Alumina and Titania Journal of Catalysis 99 (1) 104-110 (1986)
Siemens A Powder River Basin (PRB) Coal and High Arsenic Concentrations catalyst specific issues and operating experience Workshop on selective catalytic reduction (2000)
Srnak T Z J A Dumesic B S Clausen E Tornqvist and N Y Topsoslashe Temperature-Programmed Desorption Reaction and Insitu Spectroscopic
146
Studies of Vanadia Titania for Catalytic Reduction of Nitric-Oxide Journal of Catalysis 135 (1) 246-262 (1992)
Stuart B J and D S Kosson Characterization of Municipal Waste Combustion Air-Pollution Control Residues as a Function of Particle-Size Combustion Science and Technology 101 (1-6) 527-548 (1994)
Tokarz M J S Persson B Eka Nobel AB Surte Swed Poisoning of de-NOx SCR catalyst by flue gases from a waste incineration plant Studies in Surface Science and Catalysis 68 (Catal Deact 1991) 523-30 (1991)
Topsoslashe N Y Characterization of the Nature of Surface Sites on Vanadia Titania Catalysts by FTIR Journal of Catalysis 128 (2) 499-511 (1991)
Topsoslashe N-Y Mechanism of the selective catalytic reduction of nitric oxide by ammonia elucidated by in situ on-line Fourier transform infrared spectroscopy Science 265 (5176 Aug 26) 1217-1219 (1994)
Topsoslashe N Y J A Dumesic and H Topsoslashe Vanadia-Titania Catalysts for Selective Catalytic Reduction of Nitric-Oxide by Ammonia 2 Studies of Active-Sites and Formulation of Catalytic Cycles Journal of Catalysis 151 (1) 241-252 (1995)
Topsoslashe N Y J A Dumesic and H Topsoslashe The Nature of the Active-Sites in Vanadia-Titania Catalysts for Selective Catalytic Reduction of Nitric-Oxide by Ammonia Abstracts of Papers of the American Chemical Society 207 10-PETR (1994a)
Topsoslashe N-Y J A Dumesic and H Topsoslashe Nature of the active sites in vanadiatitania catalysts for selective catalytic reduction of nitric oxide by ammonia NOx Reduction Proceedings of the Symposia on Distillate Fuel Auto-Oxidation ChemistryMar 13-18 1994 San DiegoCAUSA Publ by ACSWashingtonDCUSA 39 129 (1994b)
Topsoslashe N Y and J A D TZ Srnak BS Clausen E Tornqvist Temperature-Programmed DesorptionReaction and in Situ Spectruscopic Studies of VanadiaTitania for Catalytic Reduction of Nitric Oxide J Catal 135 246-262 (1992)
Topsoslashe N Y H Topsoslashe and J A Dumesic Vanadia-Titania Catalysts for Selective Catalytic Reduction (SCR) of Nitric-Oxide by Ammonia 1 Combined Temperature-Programmed in-Situ FTIR and Online Mass-Spectroscopy Studies Journal of Catalysis 151 (1) 226-240 (1995)
Wachs I E Molecular engineering of supported metal oxide catalysts oxidation reactions over supported vanadia catalysts Catalysis 13 37-54 (1997)
Wachs I E Recent conceptual advances in the catalysis science of mixed metal oxide catalytic materials Catalysis Today 100 (1-2) 79-94 (2005)
Wachs I E L E Briand J-M Jehng L Burcham and X Gao Molecular structure and reactivity of the group V metal oxides Catalysis Today The 3rd
147
International Symposium on Group Five ElementsMar 22-Mar 23 1999 Rio de JaneiroBraz Elsevier Science Publishers BVAmsterdamNetherlands 57 3223-330 (2000)
Wachs I E Y Chen J-M Jehng L E Briand and T Tanaka Molecular structure and reactivity of the Group V metal oxides Catalysis Today 78 (1-4) 13-24 (2003)
Wachs I E G Deo B M Weckhuysen A Andreini M A Vuurman M deBoer and M D Amiridis Selective catalytic reduction of NO with NH3 over supported vanadia catalysts Journal of Catalysis 161 (1) 211-221 (1996)
Wachs I E J-M Jehng and W Ueda Determination of the Chemical Nature of Active Surface Sites Present on Bulk Mixed Metal Oxide Catalysts Journal of Physical Chemistry B 109 (6) 2275-2284 (2005)
Wachs I E and B M Weckhuysen Structure and reactivity of surface vanadium oxide species on oxide supports Applied Catalysis A General 157 (1-2) 67-90 (1997)
Wang C-B C Y and I E Wachs Langmuir 15 1223 (1999)
Wang X and I E Wachs Designing the activityselectivity of surface acidic basic and redox active sites in the supported K2O-V2O5Al2O3 catalytic system Catalysis Today 96 (4) 211-222 (2004)
Went G T L J Leu and A T Bell Quantitative Structural-Analysis of Dispersed Vanadia Species in TiO2(Anatase)-Supported V2O5 Journal of Catalysis 134 (2) 479-491 (1992a)
Went G T L J Leu S J Lombardo and A T Bell Raman-Spectroscopy and Thermal-Desorption of NH3 Adsorbed on TiO2 (Anatase)-Supported V2O5 Journal of Physical Chemistry 96 (5) 2235-2241 (1992b)
Went G T L J Leu R R Rosin and A T Bell The Effects of Structure on the Catalytic Activity and Selectivity of V2O5TiO2 for the Reduction of NO by NH3 Journal of Catalysis 134 (2) 492-505 (1992c)
Yang R T W B Li and N Chen Reversible chemisorption of nitric oxide in the presence of oxygen on titania and titania modified with surface sulfate Applied Catalysis a-General 169 (2) 215-225 (1998)
Zheng Y A D Jensen and J E Johnsson Deactivation of V2O5-WO3-TiO2 SCR catalyst at a biomass-fired combined heat and power plant Appl Catal B Environmental 60 253-264 (2005)
Zheng Y J A D Jensen and J E Johnsson Laboratory investigation of selective catalytic reduction catalysts Deactivation by potassium compounds and catalyst regeneration Industrial amp Engineering Chemistry Research 43 (4) 941-947 (2004)
148
149
APPENDICES
150
151
Appendix A REI Slipstream reactor
REI and University of Utah designed and built the slipstream reactor where one
BYU monolith catalyst and 5 commercial catalysts experienced flue gases from coal
fired boilers and biomass-coal co-firing boilers up to 3800-hour of exposure The
following cited information of this slipstream reactor comes from REI report
The SCR slipstream reactor was installed in the flue gas duct downstream of the
economizer and upstream of the air preheater Figure A1 shows a schematic of the
slipstream SCR reactor A sampling probe was inserted through an existing port in the
duct wall The probe extended approximately three feet into the duct and had a 2-ft
long slot oriented 90 ordm from the direction of flow in the duct An isolation valve was
placed on the inlet line just outside the duct wall This valve was coupled to the
control system and closed automatically if the flue gas became too cold in order to
prevent condensation in the catalyst units The reactor exhaust line was connected to
the horizontal duct downstream of the air preheater Anhydrous ammonia was injected
into the flue gas stream near the entrance to the reactor and blended with a static
mixer
A schematic drawing of the SCR slip-stream reactor appears in Figure A2
There were six identical chambers The overall flow through the system was
controlled by a single educator just upstream of the system exhaust Flow rate through
each catalyst chamber were ensure equally by achieving same pressure drop across
the six catalyst sections (divided by catalyst type) which were adjusted with butterfly
152
valves so that the flow velocity per catalyst chamber should be same throughout the
duration of the tests
Anhydrous ammonia provides the ammonia feed maintained by a mass flow
controller according to the NOx flow rate with a molar ratio of NH3NOx typically set
at approximately 11
Figure A1 SCR slipstream reactor
Flue Gas
Flue Gas
Duct Wall
Duct Wall
Pneumatic Isolation Valve One SCR Inlet Heated Sample Line
Six SCR Outlet Heated Sample Lines to Sequencer
Ammonia Injector
SCR Reactor
153
Figure A2 Schematic of SCR slipstream reactor
The six catalysts four monolith and two plate were configured as shown in
Figure A3
The four monolith catalysts were installed in four sections each Each section
had a cross section of 225 by 225 inches and was housed in a 48-inch long
aluminum square tube with outer dimensions of 25 by 25 inches and 18rdquo wall
thickness The overall cross section of each monolith catalyst was 45 times 45 inches
The four tubes were bunched together as a square with outer dimensions 50 times 50
inches
154
Figure A3 Arrangement of catalysts (plain view)
The plate catalysts were housed in square aluminum tubes with an inside
dimension of 475 inches (50 inches outside 18rdquo wall thickness) Roughly 20 plates
were placed in the tube resting in slots along opposite walls This configuration had
the same outer dimensions as the groupings of monolith catalysts
The system had seven sampling ports one before the catalyst chambers and one
after each of the six catalyst sections The ports themselves consisted of thin tubes
that entered the channel and bent downwards in line with the gas flow Each
sampling port was fitted with a stainless steel frit for removal of particles laden in the
sample
The reactor flue gas intake was attached to boiler at the economizer outlet where
up to 250 scfm of gases could be withdrawn through a probe inserted in an existing
port The probe was also fitted with a thermocouple for monitoring the temperature of
the flue gas as it exited the economizer After passing through the reactor the gases
were exhausted through an existing port at the air hearter exit The reactor itself was
155
approximately 8 feet long with a 25times30 inch footprint and weighs approximately
1000lbs The reactor was insulated and securely fastened
The reactor operated as follows when the pneumatic gat valve on the inlet of
the reactor opened flue gas from the power plant flue gas duct was allowed to enter
the reactor Upon entering the nitrogen oxide concentration in the gas stream was
measured using the gas analyzer Using this information a mass flow controller
connected to ammonia storage tanks injected a stoichiometrically appropriate amount
of reagent into the stream From here the gas entered six catalyst chambers each with
a different type of catalyst The flow rate through each of the chambers was controlled
using educators (which create low pressure in the chambers) and associated pressure
control valves which drove the educators Feedback for this control came from six
venture flow meters connected to differential pressure transducers This system
allowed the flow rate through each of the chambers to be controlled independently as
required for different catalyst types and conditions Since the flow rates through the
catalysts were not independent of one another an iterative approach was used where
PID was applied to each chamber several times in sequence This sequence through
the chambers was then repeated to convergence After leaving the chambers the gas
was returned to the flue duct
The temperature of the reactor was closely monitored and controlled using
electric heaters and thermocouple measurements taken at various locations within the
reactor
Flow gas flowing through the rector was sample to measure oxygen nitrogen
oxide and carbon dioxide both before and after the catalysts These data defined the
effectiveness of the chemical reactions in the catalysts a sootblowing system
minimized ash buildup and maintained catalyst activity
156
157
Appendix B Commercial monolith catalyst
Five vendor-supplied (Cormetech Haldor Topsoslashe Hitachi and Siemens)
commercial catalysts three of which are monoliths and two of which are plates and a
BYU prepared monolith (M4) have been exposed to the flue gas in a slipstream
reactor Those catalysts were analyzed to help characterize the deactivation that
occurs in coal-boiler flue gas over time Of each catalyst type a fresh unused sample
is available a sample that has been exposed for about 2063-hour and a sample that
has been exposed for 3800-hour are available for examination
The flow rate capacity of mass flow controllers at BYU laboratory requires
small size of samples therefore each monolith and plate catalyst were cut into small
pieces for testing
A common scroll saw was used to cut sections out of the monolith (M1 and M2)
catalysts These sections were then sanded down around the sides and on the ends to
yield samples of four channels in a two by two arrangement (Figure B1) When
preparing the exposed catalyst samples some ash was dislodged due to movement
and vibrations caused by the scroll saw and sanding Care was taken to dislodge as
minimal amount of ash as possible
All samples taken from catalysts that had been exposed in the slipstream reactor
were taken from the upstream end of the catalyst M1and M2 pictures appear in
Figure B2-Figure B3
158
Figure B1 Sampling of M1 and M2 catalysts
Figure B2 M1 monolith
159
Figure B3 M2 monolith
160
161
Appendix C CCS Overview
The catalyst characterization system (CCS) provides capabilities for long-term
catalyst exposure tests required for ascertaining deactivation rates and mechanisms
and a characterization facility for samples from the slipstream reactor An overview of
the system is illustrated in Figure C1 A series of experiments designed to clarify the
kinetics and deactivation mechanisms of commercial (vendor-supplied) and BYU-
manufactured SCR catalysts after exposure in coal and biomass effluent provide the
basis of much of our work This system simulates industrial flows by providing a test
gas with the following nominal composition NO 01 NH3 01 SO2 01 O2
2 H2O 10 and He 877 Both custom and commercial catalysts are tested as
fresh samples and after a variety of laboratory and field exposures under steady
conditions Undergraduate assistants were largely responsible for the construction of
this laboratory most notably Aaron Nackos Kenneth Crowther Seth Herring Ben
Severson Aaron Nackos began the analysis of samples in this laboratory as part of his
MS thesis but later switched topics to a different area The work was completed under
the direction of the author with the assistance of additional undergraduate assistants
Figure C2 illustrates the essential features of this system up to the analytical
train Helium functions as the carrier gas in this system There is no indication in any
literature of which we are aware that substitution of helium for nitrogen in any way
alters rates or mechanisms of selective catalytic reduction of NOx on vanadium-based
catalysts
162
Figure C1 CCS overview
Figure C2 Summary Schematic Diagram of the CCS up to (but excluding) the Analytical Train
163
The CCS is composed of gas cylinders manifolds mass flow controllers water
bubblers tubing reaction chambers a ten-way selector valve a NH3SO2 analyzer a
water condeser and a NOx analyzer The components are connected by copper brass
stainless steel polyethylene and teflon tubing with SwagelockTM fittings These
components are described in order from upstream to downstream
The monolith test reactor (MTR) (also used to test plate catalyst samples)
consists of an 8rdquox 1rdquox 1rdquo hollow square tube that has flanges on either end Machined
aluminum inserts that hold pieces of monolith or plate catalysts are inserted in this
square tube and are placed near the center (see Figure C3) Aluminum inserts are held
into place in the tube by a stainless steel spring The tube is heated with four plate-
type heaters on the outside The feed gas is routed through a frac14rdquo tube that is placed
lengthwise against the outside of one of the heaters in order to preheat the gas
Product gases exit on the opposite end The MTR is insulated and placed on a welded
stand which stands on a bench top Only one MTR is used at a time and this uses the
same CCS feed and outlet tubes in which one of the PTRrsquos can connect
Plate heater (x4)
thermocouple
8rdquo
In
frac14rdquo gap between round tube and square tube
FlowTC
075rdquo (middle of tube to end of flange)
075rdquo (middle of tube to end of flange)
6rdquo
Ceramic plugs for insulation and to reduce dead volume
catalystMonolith piece25rdquo
Aluminum catalyst holder 2rdquo long (side view)
Plate heater (x4)
thermocouple
8rdquo8rdquo
In
frac14rdquo gap between round tube and square tube
FlowTC
075rdquo (middle of tube to end of flange)
075rdquo (middle of tube to end of flange)
6rdquo6rdquo
Ceramic plugs for insulation and to reduce dead volume
catalystMonolith piece25rdquo
Aluminum catalyst holder 2rdquo long (side view)
Figure C3 Schematic of Monolith Test Reactor
164
Figure C4 Front View of Aluminum Catalyst Holders (displaying a square
honeycomb monolith a corrugated monolith and plate catalysts)
Temperature is controlled manually by an external variable AC transformer
which sends a variable amount of power to the plate heaters Although an exact
temperature set-point cannot be reached by using this manual-control setup it was
used because the manufacturer of the plate heaters said that if more than fifty percent
of the outlet power reaches the heaters they would burn out The variable AC
transformer allows us to keep outlet power under fifty percent Temperature is
measured by two thermocouplesmdashone extending inside the MTR chamber just
downstream of the catalyst exit and the other fastened to the outside of the MTR body
between two of the heater plates on one of the corners
This MTR design is advantageous because it allows for minimal changes to the
existing reactor system does not require a bulky and high-energy-consuming furnace
allows for economical gas usage and provides versatility for this and future
applications since various shapes of aluminum inserts may be used for different
monolithic catalysts
165
Appendix D Experimental Design
This investigation requires substantial mechanistic and kinetic experimentation
The intention is to supplement the existing literature by investigation of sulfur-laden
gases using equipments and techniques described below The effort to understand
vanadia catalyst reaction and deactivation mechanisms in typical coal and coal-
biomass co-combustion involves several different types of analytical systems
Conceptual mechanistic details regarding the active sites mechanistic role of
substrates and catalyst and impacts of catalyst contaminants (sulfur and alkali metals)
on reaction pathways are postulated and supported by spectroscopic and activity data
1 In situ FTIR surface spectroscopic investigations of lab-prepared powder
vanadia catalysts (fresh sulfated and contaminated) provide mechanistic
information such as definite indications of surface-adsorbed species
distinguishing between Broslashnsted and Lewis acid sites and among adsorbed
species active sites impacts of sulfate and contaminants on catalyst surface
chemistry
2 MS reactivity investigations on laboratory-prepared powder catalysts within
intrinsic kinetic regime provide global kinetic parameters such as activity and
activation energy for NOx reduction of fresh sulfated and poisoned
laboratory-prepared powder catalysts Comparisons of results from 1 and 2
determine the extent to which laboratory experiments simulate field behavior
3 Other surface characterizations provide information such as the effects of
catalyst ingredients sulfate species and poisons on BET surface area pore-
166
size distribution surface elemental compositions and oxidation states by XPS
(x-ray photon spectroscopy) and standard (bulk) analyses that supplement the
reactor data
4 Activity and surface characterization data results from industrial samples are
compared with laboratory data using both contaminated and fresh catalysts
There are no literature reports documenting such comparisons
D1 Samples
D11 Fresh Samples
a Four vanadia-based catalysts with total vanadia concentrations of 0 1 2 and 5
(by mass)
b 1 vanadia ndash 9 tungsten titania catalyst
D12 Contaminated Samples
a Each of three contaminants (K Na and Ca) doped into 1 V minus 9 W TiO2
vanadia catalyst Table D4 lists the detailed information
b K doped 1 vanadiatitania catalyst
D13 Sulfated Sample
One sample of each of the fresh (0 2 and 5 V2O5 TiO2) and deliberately
contaminated samples (K Na and Ca doped 1 V-9 W TiO2) after complete
sulfation of surface
Laboratory-prepared catalysts were prepared by impregnating titania with
various amounts of vanadia tungsten and contaminants (K Na and Ca) followed
with drying and calcination The procedure results in intimate association of catalyst
167
and contaminant All contaminants dissolve in solution in nitrate form but eventually
form oxides
Details of the experimental equipment and procedure appear in the task
statements below
D2 Vanadia Catalyst In Situ Surface Chemistry
Investigation
The purpose of this task is to gain knowledge of surface chemistry of vanadia-
based SCR catalysts with the FTIR in situ spectroscopy reactor (ISSR) The ISSR
provides in situ transmission FTIR spectra of adsorbed SO2 NH3 and NOx among
other species a definitive indication of surface-active species through in situ
monitoring of infrared spectra from catalytic surfaces exposed to a variety of
laboratory and field conditions Adsorption and desorption behaviors of these and
other species change with temperature catalyst formulation extent of sulfation and
gas composition as quantitatively indicated by changes in the spectral features of the
sample provides identification of acid sites interaction pattern between reactant
gases (NH3 NO and SO2) and surface sites before and after contamination and the
extent of sulfation on fresh and poisoned SCR catalyst surfaces These investigations
indicate how catalyst ingredients sulfation and poisons impact vanadia catalyst
surface chemistry
This task includes the following specific activities
7D21 Transient Adsorption
NH3 transient adsorption and NO transient adsorption (each 1000 ppm in
helium) at temperatures from 25-380 ˚C proceed by monitoring in situ FTIR spectra
168
of adsorbed species on laboratory-prepared catalyst surfaces with various vanadia
tungsten sulfate species and poisons contents This investigation provides qualitative
and quantitative critical parameters including Broslashnsted and Lewis acid site
identification their relative acidities and changes of acidities induced by surface
sulfation and poisoning
D22 Mechanism Investigation
Similar techniques identify the surface active sites (surface titania or vanadia or
sulfated species) for each reactant gas (NH3 NO SO2) and interacting surface species
(vanadia and sulfate species) to help elucidate SCR reaction mechanisms and
specifically impacts of sulfur and poisons on such mechanisms Hypotheses are
established and tested with different experiments as shown in follow schemes
NO Adsorption Site Identification
Possibilities A NO adsorbs on titania sites (A)
B NO adsorbs on vanadia sites Absorption frequencies observable
in the infrared do not distinguish between the various sites (1 2 3
and 4) on which NO absorbs
TiO
O
O
OTi Ti
O
O
O
OTi
O
OTi
O
OTi Ti
O
O
O
OTi
O
O
V
O
V VO
O O OO
NO 13
4
2
A
B
Figure D1 Possible NO adsorption sites
Expected experimental outcomes
169
Table D1 Possible NO adsorption trends
Hypothesis Observations A NO adsorption intensity darr as V uarr B NO adsorption intensity uarr as V uarr
Sulfate Adsorption Site Identification
Possibilities A Sulfate interacts with titania surface
B Sulfate interacts with vanadia surface
TiO
O
O
O
Ti TiO
O
O
OTi
O
OTi
O
OTi Ti
O
O
O
OTi
O
O
V VO
O O OO
S
O
OA
B
O
O
O
S
Figure D2 Possible SO2 interacting sites
Expected experimental outcomes
Table D2 Possible SO2 interactionadsorption trends
Hypothesis Observation A Sulfate IR adsorption intensity or Sulfur darr as V uarr B Sulfate IR adsorption intensity or Sulfur uarr as V uarr
NH3 Adsorption Sites Identification
Possibilities A NH3 adsorbs on titania site
B NH3 adsorbs on vanadia site
170
C NH3 adsorbs on sulfate site (which could be attached to
a vanadia or a titania atom)
TiO
O
O
O
Ti TiO
O
O
OTi
O
OTi
O
OTi Ti
O
O
O
OTi
O
O
V VO
O O OO
S
O
O
A
B
O
O
O
SNH3
NH3
NH3
C
Figure D3 NH3 possible adsorption sties
Expected experimental outcome
Table D3 NH3 possible adsorption trends
Hypothesis Peak (cm-1) Observation
A 1170 NH3 IR adsorption most intense on pure TiO2 B 1430 NH3 IR adsorption intensity increases as V increases C 1430 NH3 IR adsorption intensity increases as S increases
Identification of active adsorption sites for NO NH3 and SO2 provides
additional information to SCR reaction and poisoning mechanism
D23 Surface Sulfation
Surface sulfation represents a critical issue in this investigation since the
practical applications of low-rank coal combustion and coal-biomass co-firing involve
SO2-laden gases As discussed earlier literature opinion regarding the impacts of SO2
on SCR surface sulfation differ and the majority of the literature comes to conclusions
different from those indicated by our results This test uses in situ FTIR spectra
obtained during 24-hour sulfation of each fresh laboratory-prepared catalyst IR
171
spectra of fresh sulfated vanadia catalyst and vanadyl sulfate indicate with which site
sulfate interacts and where it forms Subsequent XPS surface chemistry analyses of
both fresh and sulfated vanadia catalysts provide evidence for identifying sulfate
species oxidation state and concentration In addition the extent to which each field-
exposed catalyst sulfates is determined
D3 NOx Reduction Kinetic Investigation
An NO reduction kinetics investigation conducted in the in situ spectroscopy
reactor (ISSR) provides first-of-their-kind data detailing mechanisms and rates
FTIR-MS in situ spectroscopy reactor based kinetics (activity at steady state)
investigations compare reactivity of the various SCR catalysts under overall nominal
gas-phase conditions of 700 ppm NH3 700ppm NO 5 O2 and helium with
Helium rather than nitrogen forms the bulk flow in all experiments for several
reasons mostly related to attempts to measure N2 as a product of the reactions All
reactivity measurements in this investigation are based on relatively simple reaction
mechanisms such as mechanisms assumed to be first order in NO and zero order in
ammonia water oxygen and all other reactants The details of the assumed
mechanism vary but in any case the detailed mechanisms exclude elementary or
completely fundamental descriptions as these unrealistically expand the scope of this
work These tests involve temperatures and catalyst composition relevant to
commercial operation but involve intrinsic kinetic regimes (unlike commercial
operation)
172
D31 Statistical Experiment Design
Three aspects of the statistical analyses performed in this investigation are
summarized here number of replications experimental design and determining
parameters from data
This investigation attempted to eliminate experimental precision and minimize
random errors Calibration reduces systematic errors in this investigation to below
instrumental detection limits All analysis gases are NIST-traceable standards and all
flow rates and flow controllers are calibrated using a Gilibratortrade Model TD5 air flow
calibration system a NIST-certified primary standard Standard thermocouples
measure temperatures within published limits (typically plusmn 2 K) These traceable
sources provide gas streams that calibrate gas analyzers in this investigation All
equipment measurements can be traced to calibrated sources which should essentially
eliminate systematic errors from these data
Random errors are minimized largely through data replication and minimization
of measurement uncertainty Figure D4 illustrates the logic used to choose sample
replications The decrease in 95 and 99 confidence intervals normalized by
standard deviations indicates that data become increasingly precise with increased
replication approximately inversely proportional to the square root of the sample size
However benefit of additional data points becomes increasingly small as sample size
increases Assuming the effort involved in collecting data scales proportional to the
number of replications a cost-to-benefit ratio (product of the number of data points
and the size of the confidence interval) behaves as illustrated again for 95 and 99
confidence levels The minimum in this curve appears somewhere between 5 and 7
data points depending on the confidence level chosen This indicates that precision
most efficiently balances effort with this sample size Additional considerations
173
sometimes determine the sample size such as resource availability and precision
required to provide statistically meaningful results but these numbers generally set
target sample sizes in this work in the absence of other compelling considerations
Precision is further reduced by decreasing the variation in the measurements
through careful materials preparation and experimental execution Considerable effort
went into developing procedures that result in repeatable and consistent data
0
2
4
6
8
10
12
14
16
18
0 5 10 15 20data points
cost
ben
efit
ratio
or c
onf
ints
td d
ev
99 CostBenefit
95 CostBenefit
99 Confidence IntervalStandard Deviation
95 Confidence IntervalStandard Deviation
Figure D4 Dependence of cost (effort)benefit ratio and confidence
intervalstandard deviation ratio on number of data points used to calculate an average value
Most of the laboratory portion of this work involved seven factors (amounts of
V W Na K Ca and S in the catalyst and temperature) and three responses (NO
adsorption NH3 adsorption and reaction kinetics) A full factorial design for all these
factors even if considered at only two levels and only accounting for linear and
correlated impacts would involve 384 different experimental conditions each
174
requiring typically 7 replications for a total of 2688 experiments In practice both
composition and kinetics must be determined at more than two levels to develop
reliable results since essentially none of these factors has linear impacts on the
responses Such a set of experiments exceeds substantially the resources available for
this investigation The approach here uses single factor variation in an exploratory
mode to illustrate overall trends in the mechanistic responses In addition a formal
statistically designed experiment based on a fixed catalyst composition typical of
commercial systems helps determine the impacts of poisons and sulfation
This statistically designed systematic investigation (Table D4) determines
effects of poisons and sulfates on catalyst activity as well as interactions among
sulfates and poisons No previously published investigation clarifies whether
interactions among poisons and sulfates exist and how important interactions are if
they do exist Table D4 summarizes factors and factor levels for the full factorial
design for these four factors at two levels Three poisons (K Na and Ca) and
sulfation represent the four factors in this experimental design All experiments
involve a 1 vanadia 9 tungsten on titania catalyst ndash the most common
commercial formulation Each factor appears at two levels either no poison or a
poison-to-vanadium elemental ratio of 05 in the case of the poisons and either not
sulfated or fully sulfated catalyst surfaces in the case of sulfation NOx reduction
activity as expressed by kinetic rate constant represents the response
Table D4 indicates the full factorial design for this experiment Theoretical
considerations suggest that there should be little interaction among poisons but
possibly substantial interaction between any given poison and sulfation A fractional
factorial design derived as a subset of this full design includes all single factor
impacts (Na K Ca and S) and all interactions with sulfur (Na-S K-S and Ca-S) but
175
not other binary interactions and no higher-order interactions This reduces the
number of experimental conditions within this design to by half with numbers 1-4
and 9-12 in Table D4 used in this experiment although these are done at several
temperatures each with at least two common temperatures among all experimental
conditions Seven replicate measurements provide statistically qualified data at each
of the experimental conditions Additional experiments demonstrate the lack of a
binary interaction between some poisons justifying the fractional factorial design
Table D4 Statistical experimental design of this investigation
Runs Composition Factor Runs Composition Factor
K Na Ca SO4 K Na SO4
1 0 0 0 0 9 0 0 0 1
2 05 0 0 0 10 05 0 0 1
3 0 05 0 0 11 0 05 0 1
4 0 0 05 0 12 0 0 05 1
5 05 05 0 0 13 05 05 0 1
6 05 0 05 0 14 05 0 05 1
7 0 05 05 0 15 0 05 05 1
8 05 05 05 0 16 05 05 05 1
D4 Other Surface Characterization Investigations
BET surface area and pore size distribution analyses for all samples provide
physical and structural information about the catalysts A Micromeritics Tri-star
Instrument (Model 3000) using the N2 surface area method provides all data for these
176
measurements The test matrix includes all samples that is fresh and exposed
commercial samples sulfated and non-sulfated laboratory samples and contaminated
and uncontaminated laboratory samples Several other surface-sensitive laboratory
diagnostics such as XPS and ESEM (Environmental Scanning Electron Microscopy)
supplement the reactor data collected in our laboratory
The above experiments involve comparisons of sulfated and non-sulfated
samples of uncontaminated and contaminated laboratory-prepared catalysts with
known amounts and forms of contaminants and catalyst Uncontaminated SCR
material and at least one sample of the same material contaminated with each poison
provide a database with which to compare commercially exposed materials (discussed
next) This investigation results in a database of FTIR-MS results describing surface
spectra reactor effluent compositions and transient concentration profiles for
contaminated and uncontaminated catalysts is completed for both sulfated and non-
sulfated catalyst surfaces The experimental design appears in Table D5
Table D5 Experimental design for sulfation and poison tests of catalyst samples Samples Characterization
Catalyst Poisons Sulfaton NH3 NO adsorption
(FTIR)
Activity (MS)
BET XPS
TiO2 times times times TiO2 times times times times
2 V2O5TiO2 times times times 2 V2O5TiO2 times times times times 5 V2O5TiO2 times times times times 5 V2O5TiO2 times times times times
1 V2O5-9WO3TiO2 times times 1 V2O5-9WO3TiO2 times times times 1 V2O5-9WO3TiO2 K times times times 1 V2O5-9WO3TiO2 K times times times 1 V2O5-9WO3TiO2 Na times times times 1 V2O5-9WO3TiO2 Na times times times 1 V2O5-9WO3TiO2 Ca times times times 1 V2O5-9WO3TiO2 Ca times times times
177
Based on the above designed experiments the effect of sulfur and poison
addition on vanadia catalyst surface chemistry and kinetics should be acquired to
supplement the existing literature to help elucidate the mechanism of SCR catalysts
deactivation and to support developing deactivation modeling in Task 2
178
179
Appendix E Pore and Film Diffusion Calculation
E1 Pore Diffusion Limitation
SCR tests on 5 and 2 V2O5TiO2 catalysts at 350 ˚C resulted in NO
conversions of about 80 and 72 respectively which indicates that increasing
vanadia content enhances NO reduction activity However at conversions as high as
80 and 72 pore diffusion resistance could be dominant For example the calculated
Thiele modulus (MT) for 5 V2O5TiO2 is 106 substantially exceeding the 04 upper
limit for MT customarily accepted for negligible pore diffusion resistance In addition
the Weisz modulus (MW) for 5 V2O5TiO2 is 113 exceeding the 013 upper limit
for MW customarily accepted for negligible pore diffusion resistance
( )Aseff
Asn
T CDCknLM
21+
= (E1)
Aseff
obsAw CD
rLM )( 2 minus= (E2)
=L Z2 for flat plate Z = thickness
=L r2 for cylinders r = radius
=L r3 for spheres
n = reaction order
Deff = effective diffusivity
CAs = reactant concentration on catalyst surface
180
Pore diffusion may influence observed kinetic reaction rates Operation in
regimes without such influence provides more accurate intrinsic kinetic data in
addition to simplifying the analysis Therefore NO reduction experiments used 1
V2O5TiO2 at nominally 18 conversion with temperatures of about 250 ˚C and
nominally 30 conversion at temperatures of 300 ˚C These conditions correspond to
Thiele moduli (MT) of about 025 (250 ˚C) and 034 (300 ˚C) respectively
corresponding to effectiveness factors of 096 and 093 respectively Therefore pore
diffusion does not appreciably influence the results at these temperatures with the 1
vanadia catalyst Commercial catalysts typically contain about 1 vanadia
The remaining kinetic tests on vanadia catalysts (fresh contaminated and
sulfated) maintained a Thiele modulus (MT) below 04 by flow rate and temperature
adjustment prior to kinetic investigations
E2 Film Diffusion Limitation
Both theoretical and experimental results show that film diffusion represents a
trivial consideration during SCR tests on 1 V2O5TiO2 at temperatures up to 350 ˚C
and under the conditions of these experiments
Film diffusion resistance is determined according to
1 Shd
DKp
ABc ⎟
⎟⎠
⎞⎜⎜⎝
⎛minus=
φφ (Fogler 1999) (E3)
oslash = void fraction of packed bed
DAB = gas-phase diffusivity m2s
dp = particle diameter m
Shrsquo =Sherwood number
181
The calculated result indicates that the film resistance accounts for about 03
of the total resistance (combined resistance of film diffusion and kinetic resistance)
Experimentally film diffusion investigations on SCR catalysts (1 V2O5TiO2)
involved three different flow rates (93 121 187 mlmin) corresponding space
velocities are100000 130000 and 200000 hour-1 This range of space velocities in
the catalyst provides significant variation in the boundary layer thickness along the
catalyst surface and therefore should result in different conversions if film resistance
plays a significant role in NO reduction Similar NO conversions (176 at 933
mlmin 18 at 121 mlmin and 174 at 187mlmin at 250 ˚C) resulted from each
experiment consistent with the mathematical expectation of negligible film transport
resistance The following SCR reactions involved 1 V2O5TiO2 catalysts with 700
ppm NH3 and NO 5 O2 helium (balance) and a total flow rate of 187 mlmin at
temperatures ranging from 250 to 300 ˚C where both film diffusion and pore
diffusion resistance can be neglected This investigation involved catalyst reacting in
the intrinsic kinetic range Similar film diffusion resistance determination procedures
were followed for the rest catalyst samples which were also investigated under
intrinsic kinetic regime
182
183
Appendix F Derivation of the Chen model1
Figure F1 schematically illustrates a two-dimensional reactor in which a
reactant from the bulk flow is transported to a porous wall containing catalyst The
dimension in the direction of flow is z and the dimension perpendicular to the low is
x The origin is taken from the reactor entrance at the center of the porous catalyst If
we assume Fickian diffusion that the catalyst is isothermal and homogeneous and that
the surface reaction is first order in reactant that the flux in the flow direction is
negligible compared to the flux in the direction perpendicular to the flow and that
bulk diffusion does not influence the conversion rate then the flux at any point in the
catalyst can be equated to the rate of reaction in the catalyst as follows where the
dependence of the mole fraction of reactant on both coordinate directions is
emphasized
)(22
2
zxycakhdx
ydcDe = (F1)
where c represents gas concentration De represents the diffusivity of the reactant in
the porous media and a represents a time-dependent and dimensionless activity
factor defined as the ratio of the chemical activity in the catalyst at arbitrary time
divided by its initial value The value of a generally decreases from unity with
chemical deactivation but could exceed unity because of catalyst activity increases
caused for example by catalyst sulfation Extensions of this model to accommodate
184
surface fouling bulk diffusion and similar impacts will be discussed in the final
report
Figure F1 Schematic diagram of a two-dimensional reactor
This equation can be written in dimensionless form as follows
eDkayh
dxyd 2
2
2
= (F2)
where
hxx = (F3)
and
h
YNO
catalyst bulk flow
x
z
185
infin=NO
yyy NO (F4)
are based on the half-thickness of the wall (h) and the bulk mole fraction in the cell
(yinfinNO) The boundary conditions are
1 1
0 dxdyBiy
xminus
=+= (F5)
0
1
==xdx
dy (F6)
The solution gives the concentration profile within the wall
( )
( )11
22
2
minusminusminus
+=
minusminus
minusminus
φφ
φφ
φ eBi
e
eeyxx
(F7)
where
eDkah2
2 =φ (F8)
and
e
m
DhkBi = (F9)
This equation describes the relative impacts of film mass transfer pore diffusion
and surface reaction on conversion
Considering the reactor the mass balance along the axial direction of the reactor z
is
( ) 0=minus+ infininfin
sNONO
mNO yyuAk
dzdy σ (F10)
where u is the linear gas velocity in the cell which is assumed to be constant σ is the
perimeter length of a cell in the monolith and A is the cross-sectional area of a cell
The boundary condition is
186
0
0
infin
=
infin = NOzNO yy (F11)
and the bulk and surface NO concentrations are related by
⎟⎟⎠
⎞⎜⎜⎝
⎛+minus
minus= minus
minusinfin
111 2
2
φ
φφee
Biyy s
NONO (F12)
The overall conversion X of NO in the reactor at axial position L is given by
0
0
infin
infin minus=
NO
LNONONO
yyyX (F13
Combining these results the NO conversion is given by
⎥⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢⎢
⎣
⎡
⎟⎟⎠
⎞⎜⎜⎝
⎛
minus+
minus
minusminus=
minus
minus
1111
exp1
2
2
φ
φ
σ
ee
kaDkuA
LX
em
(F14)
200 220 240 260 280 300 320 340 360 380 40010
20
30
40
50
60
70
80
90
100
Chen model M1 freshChen model M1 exposed 2063 hrChen model M1 exposed 3800 hrData M1 freshData M1 exposed 2063 hrData M1 exposed 3800 hr
Temperature (degC)
NO
Con
vers
ion
()
Figure F2 Comparison of M1 data to Chenrsquos model prediction
187
200 220 240 260 280 300 320 340 360 380 4000
20
40
60
80
100
Chen model M2 freshChen model M2 exposed 2063 hrChen model M2 exposed 3800 hrData M2 freshData M2 exposed 2063 hrData M2 exposed 3800 hr
Temperature (degC)
NO
Con
vers
ion
()
Figure F3 Comparison of M2 data to Chenrsquos model prediction
188
189
Appendix G In situ IR spectra of 24-hour sulfation
Dry Sulfation of 2 V2O5TiO2 and TiO2
In situ IR spectra of the dry sulfated 2 V2O5TiO2 catalysts include a minor
doublet with the major sulfate-related feature located near 1378 cm-1 shown in Figure
G1
20
15
10
05
00
Abs
orba
nce
1440 1420 1400 1380 1360 1340 1320 1300Wavenumber cm-1
137789
137789 24 hours 23 hours 6 hours 5 hours 4 hours 3 hours 2 hours 1 hour
Peak area is 28975
Figure G1 Time-dependent in situ FTIR spectra of a 2 vanadium catalyst
exposed to a typical vitiated gas (see VTOF in Table 516for details of experimental conditions) Ordinate is offset for each time
190
Spectra are offset upward along the ordinate with time for clear comparison
Sulfate peaks display much larger areas and intensities than those from 5
V2O5TiO2 and varied in wave number 1378 cm-1 during the dry sulfation
Consequently sulfation occurs fast on 2 V2O5TiO2 catalyst as well The sulfate
peak area on 24 hour dry sulfated 2 V2O5TiO2 is 29 about 70 larger than the
sulfate peak area (17) of 24 hour dry sulfated 2 V2O5TiO2
25
20
15
10
05
00
Abs
orba
nce
1440 1420 1400 1380 1360 1340 1320Wavenumber cm-1
24 hours 23 hours 6 hours 5 hours 4 hours 3 hours 2 hours 1 hour
137307140296
138078140778
area is 463708
Figure G2 Time-dependent in situ FTIR spectra of a titania exposed to a typical
vitiated gas (see TiO7 in Table 516 for details of experimental conditions) Ordinate is offset for each time
Sulfation of pure TiO2 with SO2-laden gas under conditions typical of SCR
operation provides a baseline comparison for experiments involving sulfation of
vanadia catalysts Different from 2 and 5 V2O5TiO2 Figure G2 clearly shows
doublet sulfate peak formation on dry sulfated TiO2 with individual peaks located at
about 1405 cm-1 and about 1375 cm-1 Spectra are offset upward along the ordinate
191
with time for clear comparison The doublet peak intensities increased noticeably with
time and the peak positions shifted to higher frequencies with increasing time during
the sulfation test indicating increasing sulfate acidity with timesurface coverage
Therefore sulfate species gradually built up on titania surfaces unlike the fast
saturation of sulfate species on 2 and 5 V2O5TiO2 After 24 hours sulfate peak
area reached 46 larger than the corresponding peak areas of both 5 (17) and 2
V2O5TiO2 (29) The peak intensity increases and peak position shifts show no
apparent shifts after 6 hours of sulfation However the sulfation continued to 24 hours
to ensure a consistent surface
Wet Sulfation of 5 and 2 V2O5TiO2 and TiO2
Similar to dry sulfation results from a series of twenty-four hour sulfation
experiments on titania a 2 vanadia on titania catalyst and a 5 vanadia on titania
catalyst in an SO2-laden moist environment appear in Figure G3 through Figure G5
Figure G3 illustrates in situ IR spectra collected during wet sulfation of 5
V2O5TiO2 Spectra are offset upward along the ordinate with time for clear
comparison Similar to the IR spectra from dry sulfated 5 V2O5TiO2 the sulfate
peak intensities and positions remain the same between the first (1 hour) and the last
(24 hour) data indicating sulfation rapidly reached saturation on 5 V2O5TiO2 In
addition the sulfate peaks appear weaker than the corresponding peaks from dry
sulfated 5 V2O5TiO2 the area of sulfate peak after 24 hours exposure on the 5
V2O5TiO2 is 4 which is much less than the peak area from 24 hour wet sulfated 5
V2O5TiO2 (17)
192
04
03
02
01
00
Abs
orba
nce
1440 1420 1400 1380 1360 1340 1320 1300Wavenumber cm-1
24 hours 6 hours 5 hours 4 hours 3 hours 2 hours 1 hour
Peak area is 39
Figure G3 Time-dependent in situ FTIR spectra of a 5 vanadium catalyst
exposed to a typical vitiated gas (see VTHF in Table 516 for details of experimental conditions) Ordinate is offset for each time
12
10
08
06
04
02
00
Abs
orba
nce
1440 1420 1400 1380 1360 1340 1320 1300Wavenumber cm-1
24 hours 5 hours 4 hours 3 hours 2 hours 1 hour
137885
138078Peak area is 188
Figure G4 Time-dependent in situ FTIR spectra of a 2 vanadium catalyst
exposed to a typical vitiated gas (see VTHG in Table 516 for details of experimental conditions) Ordinate is offset for each time
193
In situ IR spectra of the wet sulfated 2 V2O5TiO2 catalysts include an obscure
doublet and one major sulfate-related feature located near 1378 cm-1 shown in Figure
G4 Spectra are offset upward along the ordinate with time for clear comparison
Sulfate peaks display much larger areas and intensities than those from wet sulfated
5 V2O5TiO2 The peak positions shifted from 1378 cm-1 at the first hour to 1380cm-
1 at 24 hours of wet sulfation indicating minor increase in sulfate acidity with
timesurface coverage No obvious peak area increase occurred during the wet
sulfation on 2 V2O5TiO2 after the first hour thus rapid sulfation also occurs on 2
V2O5TiO2 but probably slower than 5 V2O5TiO2 because of a slight peak position
shift to higher frequencies The sulfate peak area on 24 hour dry sulfated 2
V2O5TiO2 is 19 smaller than the corresponding peak area from 24- hour dry sulfated
2 V2O5TiO2 but larger than sulfate peak area of 24- hour wet sulfated 5
V2O5TiO2
Sulfation of pure TiO2 with SO2-laden gas under conditions typical of SCR
operation provides a baseline comparison for experiments involving sulfation of
vanadia catalysts Different from 2 and 5 V2O5TiO2 Figure G5 shows obvious
doublet sulfate peaks formed on sulfated TiO2 locate at ~1405 cm-1 and ~1375 cm-1
Spectra are offset upward along the ordinate with time for comparison The doublet
peak intensities increased noticeably with time and the peak positions shifted to
higher frequencies with increasing time during the sulfation test indicating increasing
sulfate acidity with timesurface coverage Therefore similar to dry sulfated TiO2
sulfate species gradually built up on titania surface unlike the fast saturation of
sulfate species on 2 and 5 V2O5TiO2 After 24 hours sulfate peak area reached to
41 slightly smaller than the corresponding peak area from dry sulfated TiO2 (46) and
larger than the sulfate peak areas of 24 hour sulfated 5 (4) and 2 V2O5TiO2 (19)
194
25
20
15
10
05
00
Abs
orba
nce
1440 1420 1400 1380 1360 1340 1320 1300Wavenumber cm-1
13672813856
137307
140392
25 hours 6 hours 5 hours 4 hours 3 hours 2 hours 1 hour
Peak area is 41487
Figure G5 Time-dependent in situ FTIR spectra of a TiO2 exposed to a typical
vitiated gas (see TiO5 in Table 516 for details of experimental conditions) Ordinate is offset for each time
Doublet Sulfate Peak from TiO2
The doublet sulfate IR peak signals appear on sulfated titania samples under
both dry and wet sulfation conditions and this sulfate peak increases gradually with
time A doublet peak is also reported in Yang et alrsquos work although their IR peak
around 1380 cm-1 is more intense than the one around 1401 cm-1 (Chen and Yang
1993) Our results on the other hand showed the IR peak around 1401 cm-1 to be
more intense Moreover the peak positions shifted to higher frequencies with
increasing time during the sulfation test indicating increasing sulfate acidity with
time
- Poisoning and Sulfation on Vanadia SCR Catalyst
-
- BYU ScholarsArchive Citation
-
- Title
- ABSTRACT
- ACKNOWLEDGMENTS
- Table of Contents
- List of Figures
- List of Tables
- Chapter 1 Introduction
-
- 11 NOx Definition and Properties
- 12 NOx Formation
- 13 NOx Regulations
- 14 NOx Control Technologies
- 15 SCR Technology
- Chapter 2 Literature Review
-
- 21 Background of SCR and SCR catalysts
-
- 2141 Surface Structure of VanadiaTitania
-
- 2142 Active Site Identification
-
- 21421 Active site investigation based on structures
- 21422 Active sites investigation based acid sites
-
- 2143 NH3 Adsorption
-
- FTIR and Raman investigations
-
- 2144 NO Adsorption
- 2145 NH3 and NO Coadsorption
-
- 22 Interactions with Sulfur Dioxide
- 23 Deactivation of Vanadia Catalysts
-
- 2321 Pore Plugging
-
- 2322 Channel Plugging
- 2331 Arsenic
- 2332 Lead
- 2333 HCl
- 2334 Alkali and Alkaline-earth Metals
-
- 24 Summary of Previous Work and Current State of Knowledge
-
- Chapter 3 Objectives
- Chapter 4 Experimental Apparatus and Procedures
-
- 41 Sample Information
- 42 Catalyst Preparation
- 43 BET Analyses
- 44 X-ray Photoelectron Spectroscopy (XPS)
- 45 Environmental Scanning Electron Microscopy (ESEM)
- 46 In Situ Surface Reactor (ISSR)
- 47 NH3 and NO Adsorption
- 48 Sulfation
- 49 Activity Measurement
- 410 Catalyst Activity Characterization System (CCS)
-