Study of Catalyst Deactivation in Three Different Industrial Processes

Full text

Study of Catalyst Deactivation
in Three Different Industrial Processes

Acta Wexionensia
No 106/2007
Chemical Engineering/Bioenergy Technology
Study of Catalyst Deactivation
in Three Different Industrial Processes
Ann-Charlotte Larsson
Växjö University Press

Study of Catalyst Deactivation in Three Different Industrial Processes. The-
sis for the degree of Doctor of Technology, Växjö University, Sweden 2007.
Series editors: Tommy Book and Kerstin Brodén
ISSN: 1404-4307
ISBN: 978-91-7636-533-5
Printed by: Intellecta Docusys, Göteborg 2007

Abstract
Larsson, Ann-Charlotte (2007). Study of Catalyst Deactivation in Three Different
Industrial Processes, Acta Wexionensia No 106/2007. ISSN: 1404-4307,
ISBN: 978-91-7636-533-5. Written in English.
Deactivation of catalysts were investigated focusing on three industrial proc-
esses: 1) Selective Catalytic Reduction (SCR) for abatement of NOx from bio-
mass combustion using V2O5-WO3/TiO2 catalysts; 2) Catalytic oxidation of vola-
tile organic compounds (VOC) from printing industries using a Pt/γ-Al2O3 cata-
lyst; and 3) Ni and Pt/Rh catalysts used in steam reforming reaction of bio-
syngas obtained from biomass gasification.
The aim has been to simulate industrial conditions in laboratory experiments
in order to comprehend influence of compounds affecting catalysts performance.
Typical catalyst lifetimes in industrial processes are several years, which are a
challenge when accelerating deactivation in laboratory scale experiments where
possible exposure times are few hours or days. Catalysts can be introduced to
deactivating compounds through different routes. The first method examined
was gaseous exposure, which was applied to deactivate VOC oxidation catalyst
through exposure of gaseous hexamethyldisiloxane. The second method involved
wet impregnation and was used for impregnation of SCR catalyst with salt solu-
tions. The third method was based on exposure and deposition of size selected
particles of deactivating substances on the catalyst. The latter device was devel-
oped during this work. It was applied to monolithic SCR catalysts as well as to
pellet catalysts intended for steam reforming of biomass gasification syngas. De-
activated SCR catalyst samples by size selected exposure method were verified
and compared with SCR catalysts used in a commercial biomass boiler for 6 500 h.
Evaluations of fresh and deactivated samples were investigated using BET
surface area; chemisorption and temperature programmed desorption (TPD); sur-
face morphology using Scanning Electron Microscopy (SEM) and poison pene-
tration profile through SEM with an Electron Micro Probe Analyser (EMPA)
also equipped with a energy dispersive spectrometer (EDS); chemical analysis of
accumulation of exposed compounds by Inductively Coupled Plasma - Atomic
Emission Spectroscopy (ICP-AES); and influence on catalyst performance. The
size selected generated particles of deactivating substances were characterized
with respect to mean diameter and number size distribution through Scanning
Mobility Particle Sizer (SMPS) and mass size distribution applying an Electric
Low Pressure Impactor (ELPI). Results from catalyst characterization methods
were useful tools in evaluation of catalyst deactivation routes.
Understanding deactivation processes and impact on catalyst performance is
vital for further optimization of catalysts with respect to performance and life-
time. Further research in this field can provide more resistant catalysts for appli-
cation in industry leading to higher commercial benefits and further application
of environmental catalysts in thermo-chemical conversion of biomass.
Key words: deactivation; catalyst; SCR; VOC; steam reforming; aerosol particle;
potassium; zinc; ash salts; biomass; organosilicon;

qç=lä~I=gç~âáã=~åÇ=^äÉñ~åÇÉê=

=

1
List of Papers
This thesis is based on the following papers appended in the thesis. They are re-
ferred to in the text by their Roman numerals:
I Deactivation of SCR Catalysts by Exposure to Aerosol Parti-
cles of Potassium and Zinc Salts
Ann-Charlotte Larsson, Jessica Einvall and Mehri Sanati
Accepted for publication in Aerosol Science and Technology
(Dec 2006)
II Targeting by Comparison with Laboratory Experiments the
SCR Catalyst Deactivation Process by Potassium and Zinc
Salts in a Large Scale Biomass Combustion Boiler
Ann-Charlotte Larsson, Jessica Einvall, Arne Andersson and
Mehri Sanati
Energy & Fuels (2006) 20, 1398-1405
III Physical and Chemical Characterisation of Potassium Deac-
tivation of a SCR Catalyst for Biomass Combustion
Ann-Charlotte Larsson, Jessica Einvall, Arne Andersson and
Mehri Sanati
in Proceedings 12th Nordic Symposium on Catalysis Confer-
ence, Trondheim, Norway, May 28-30, 2006, (P51, p 198)
Accepted for publication in a special issue of Topics of Catalysis
(Dec 2006)
IV Pilot-Scale Investigation of Pt/Alumina Catalysts Deactiva-
tion by Organosilicon in the Total Oxidation of Hydrocar-
bons
Ann-Charlotte Larsson, Mohammad Rahmani, Karl Arnby,
Morteza Sohrabi, Magnus Skoglundh, Neil Cruise, and Mehri
Sanati
in Proceedings 12th Nordic Symposium on Catalysis Confer-
ence, Trondheim, Norway, May 28-30, 2006, (O24, p 82)
Accepted for publication a special issue of in Topics of Catalysis
(Dec 2006)

2
V Investigation of Reforming Catalyst Deactivation by Expo-
sure to Fly Ash from Biomass Gasification in Laboratory
Scale
Jessica Einvall, Simone Albertazzi, Christian Hulteberg, Fran-
cesco Basile, Ann-Charlotte Larsson, Jan Brandin and Mehri
Sanati
Submitted for publication in Energy & Fuels
VI Comparison of Deactivated Catalysts in Laboratory and
Large Scale Biomass Combustion Focussing on Potassium
and Zinc Salts
Ann-Charlotte Larsson, Mikael Strand, Simone Albertazzi,
Francesco Basile and Mehri Sanati
Submitted for publication in Applied Catalysis A: General
Results related to this thesis are also presented in:
Developing a Method to Deactivate a V2O5-WO3/TiO2 Mono-
lithic Catalyst by Aerosol Salts
Ann-Charlotte Larsson, Jessica Einvall, Mehri Sanati,
Poster presentation at European Aerosol Conference, Ghent,
Belgium, August 30-September 1, 2005, p 665, ISBN 90-8091-
5939
Environmental Catalyst for Abatement of NOx from Bio-
mass Boilers
Ann-Charlotte Larsson, Jessica Einvall, Mehri Sanati,
Oral presentation at Swedish Finnish Flame Days, October 18-
19, 2005, 176-184, ISBN 91-7178-185-4
Deactivation of Catalysts for Reforming of Gas from Biomass
Gasification by Exposure to Aerosol Particles
Jessica Einvall, Simone Albertazzi, Christian Hulteberg, Fran-
cesco Basile, Ann-Charlotte Larsson, Eva Gustafsson, Jan
Brandin, Ferruccio Trifirò and Mehri Sanati
Poster presentation at 12th Nordic Symposium on Catalysis Con-
ference, Trondheim, Norway, May 28-30, 2006, (P18, p 132)
ISBN 82-9955-69-1-0

3
List of Contents
Introduction........................................................................................................... 5
Catalyst Deactivation................................................................................ 5
Deactivation in Industrial Processes......................................................... 6
Selective Catalytic Reduction (SCR) in Biomass Combustion.................6
Deactivation of VOC Oxidation Catalyst ................................................. 7
Steam Reforming Catalyst Applied to Product Gas of Biomass
Gasification............................................................................................... 8
Simulation of Catalyst Deactivation in Laboratory Scale....................... 10
Scope of Work........................................................................................ 10
Experimental ....................................................................................................... 13
Exposure to Generated Poisons ..............................................................13
Catalyst Reactor Configuration .............................................................. 13
Particle Generation and Deposition Process........................................... 14
Gaseous Exposure Process .....................................................................17
Wet Impregnation of Catalyst Samples by Poisons................................ 18
Commercial Deactivation of Catalyst Samples ......................................19
Characterisation .................................................................................................. 21
Catalyst Characterisation ........................................................................ 21
Physical Characterisation of Catalyst Samples (BET)............................21
Chemisorption ........................................................................................ 21
Surface Analysis of Catalyst Samples .................................................... 22
Poison Penetration Profile ......................................................................23
Chemical Analysis .................................................................................. 23
Physical Characterisation of Generated Particles ................................... 24
Scanning Mobility Particle Sizer (SMPS) .............................................. 24
Tandem Differential Mobility Analyser (TDMA).................................. 25
Electric Low Pressure Impactor (ELPI)..................................................26
Catalyst Performance.............................................................................. 28
SCR Catalyst Activity Measurement...................................................... 28
VOC Catalyst Activity Measurement..................................................... 29
Steam Reforming Catalyst Activity Measurement .................................29
Results................................................................................................................. 31
Deposition of Generated Poisons............................................................ 31
Impact of Poisoning on Catalyst Activity............................................... 35
Influence of Poisoning on Physical Properties of Catalysts....................36
Penetration Profile of Poisoning............................................................. 38
Chemical Characterisation by Chemisorption ........................................ 40
Discussion ........................................................................................................... 43
Future recommended work .................................................................................47

4
Acknowledgment ................................................................................................ 49
Publications......................................................................................................... 51
Paper I..................................................................................................... 51
Paper II ...................................................................................................51
Paper III.................................................................................................. 52
Paper IV.................................................................................................. 52
Paper V ................................................................................................... 53
Paper VI.................................................................................................. 53
Authors Contribution to Presented Papers.............................................. 53
References........................................................................................................... 55

5
Introduction
Catalyst Deactivation
Catalyst deactivation, the loss over time of catalytic activity or selectivity, is a
problem of great economical concern in application of commercial catalytic
processes. Catalyst deactivation is attributed to interaction between the catalyst
and the impurities present in process effluent in which the catalyst is used. Any
chemical or physical interaction that reduces catalyst activity or selectivity is
classified as catalyst deactivation phenomena. In general, deactivation leads to a
shortened catalyst lifetime, and the replacement of an aged catalyst to a new one
is determined by the industrial processes for which the catalyst is used. Industrial
catalytic deactivation can range from short term to several years. Given that re-
duced catalyst lifetime has a strong negative impact on the process economics
improved catalyst lifetime is of great commercial value.
The causes of catalyst deactivation can be grouped into: chemical deactivation
through reversible or irreversible poisoning; physical deactivation through foul-
ing; thermal deactivation through sintering; loss of active material by vaporiza-
tion; and mechanical deactivation through attrition or erosion (Petersen et al.,
1987, Bartholomew, 2001, Forzatti et al., 1999, Chen et al., 1992, Moulijn et al.,
2001).
Deactivation of catalysts have been investigated focusing on three industrial
processes: 1) Selective Catalytic Reduction (SCR) for abatement of NOx from
biomass combustion using V2O5-WO3/TiO2 catalysts; 2) Catalytic oxidation of
volatile organic compounds (VOC) in printing industries which use a Pt/γ-Al2O3
catalyst; and 3) Ni and Pt/Rh catalysts used in steam reforming reaction of bio-
syngas obtained from biomass gasification.
Catalyst deactivation in the processes studied are mainly poisoning of the cata-
lysts by impurities in the effluent gas. Poisonous compounds related to the inves-
tigated processes are arsenic, phosphorous, alkali metals, heavy metals, iron, sul-
phur, and chlorides present in gaseous form or as submicrometer size particles
(Ertl et al., 1997, Satterfield, 1996).
With an aim to simulate industrial conditions in laboratory experiments, the in-
fluence of poisoning compounds on catalysts performance has been evaluated.
Catalysts were introduced to deactivating compounds through the following dif-
ferent routes: gaseous exposure, which was applied to deactivate VOC oxidation
catalyst through exposure to gaseous organic silicon compounds; wet impregna-

tion used for impregnation of SCR catalyst with potassium and zinc salt solu-
tions; and exposure and deposition of size selected particles of potassium and
zinc salts applied to the catalyst SCR as well as to steam reforming catalysts.
Understanding of the catalyst deactivation mechanism and impact on catalyst
performance is vital for further optimization of catalyst structures as well as
physical and chemical properties with respect to tailoring crystal structures to re-
sist deactivation.
Deactivation in Industrial Processes
Selective Catalytic Reduction (SCR) in Biomass Combustion
SCR is commonly used for removal of NOx emission for combustion processes
of different fuels. The catalyst used is a V2O5-WO3/TiO2 monolithic structure. In
the SCR reaction NO is reduced by NH3 producing N2 and water in accordance
with:
4 NO + 4 NH3 + O2 4 N2 + 6 H2O
The chemical reaction is assumed to consist of adsorption of ammonia on vana-
dium active sites followed by an adsorption of NO to form a complex with am-
monia reacting to form nitrogen and water. After desorption of nitrogen and wa-
ter the vanadium active sites are regenerated by oxygen present in the flue gas
(Topsoe et al., 1995, Forzatti, 2001).
A monolithic catalyst is extruded from a homogeneous material as elements of
150 mm by 150 mm with a length of up to 1300 mm. Flue gas is passed through
the catalyst channels, having a cell opening of 4 to 10 mm, under laminar flow
conditions. One or multiple layers of catalyst are applied in order to achieve the
desired conversion of NO. The catalyst is generally placed in the exit duct of the
boiler economizer at temperatures of 300 to 400 °C having high dust loads in the
gas, up to 20 g/m3n.
Deactivation mechanisms related to combustion processes are poisoning of the
catalyst by exposure to flue gas containing different impurities from fuels like
biomass (such as arsenic, phosphorous, zinc and alkali metals (Ertl et al., 1997,
Hums, 1998, Beck et al., 2004 and 2005, Herrlander, 1990, Chen et al., 1990,
Chen and Yang, 1990, Kamata et al., 1999, Lisi et al., 2004), fouling by ash,
plugging, erosion, precipitation of ammonia salts in the catalyst pore structure or
sintering of the titanium oxide support structure at elevated temperatures above
450 °C (Forzatti, 2001).
Mechanical deactivation phenomena can be minimised through proper catalyst
system design, i.e. fouling or plugging can be avoided by installation of cleaning
6

systems such as soot-blowers or sonic horn cleaning. Deactivation by thermal
sintering or salt precipitation can also be avoided by the choice of preferred op-
erating temperature.
Poisoning by fuel compounds often involves a reaction between poisonous com-
pounds and catalyst active sites producing a permanent blocking of the active
site. The catalyst poisons penetrate the catalyst either by diffusion of the gaseous
poison, by capillary condensation of poisons in the pore structure or by penetra-
tion of submicrometer sized particles of poisonous material. Typical degradation
of SCR catalysts involves a process taking several years with lifetimes of
roughly 2 to 3 years.
In woody biomass combustion the catalyst deactivation is attributed to potassium
poisoning (Zheng et al., 2004 and 2005, Khodayari et al., 2000 and 2001, Kling
et al., 2006). Wood contains high levels of potassium and in the ash the potas-
sium concentration amounts to 11 weight %. Ash from coal contains in compari-
son 1 weight % of potassium. The potassium in biomass combustion is present as
submicrometer particles. Ash from biomass combustion consists of a fine mode
and a coarse mode of particles. The fine mode (< 0,1 µm) consists mainly of po-
tassium chlorides and sulphates (Lighty et al., 2000, Pagels et al., 2003 and 2005,
Strand et al., 2002, Wierzbicka et al., 2005). A fraction of the submicrometer
size potassium particles can penetrate the catalyst pore structure and deactivate
the catalyst (Moradi et al., 2003).
In commercial operation the catalyst is simultaneously exposed to deactivating
substances of different composition and mechanisms providing little opportunity
to investigate possible catalyst modification in order to improve the catalyst life-
time.
Deactivation of VOC Oxidation Catalyst
Total catalytic oxidation is a common method to reduce emissions of volatile or-
ganic compounds (VOC) from industrial processes. The total oxidation of the
VOC is energy advantageous with carbon dioxide and water as reaction products
in accordance with:
CxHy + (x+y/4) O2 x CO2 + (y/2) H2O
Typical catalysts used consist of supported platinum catalyst, for example sup-
ported Pt/γ-Al2O3 in pellet form with the active phase distributed slightly below
the external surface with monolithic catalyst shapes also being utilized. The reac-
tion mechanism is assumed to follow a Langmuir-Hinshelwood model (Mor-
bidelli et al., 2001).
Catalytic oxidation of hydrocarbons can be used to reduce VOC in industrial
processes such as printing, coating and painting. The catalyst is placed in the exit
gas at temperatures of 300 to 500 °C (Ertl et al., 1997, Satterfield, 1996). The
7

catalyst pellets or monoliths are placed in a tubular reactor with sufficient resi-
dence time to achieve the requested degree of VOC oxidation.
Catalyst deactivation of alumina supported platinum catalysts is generally related
to gaseous poisons. Typical poisons are lead, phosphorous, zinc, iron, silicon,
chlorides and SO2 (Spivey et al., 1992, Neyestanaki et al., 2004, Hegedus et al.,
1984). Thermal degradation through sintering and redispersion of the platinum
can also reduce catalyst activity at elevated temperatures.
In the printing industry, organosilicon compounds are often present in the flue
gas originating from the printing ink, which is evaporated during the drying. Gas
concentrations of the silicon compounds are very low, in the range of ppb. Nev-
ertheless the organosilicon residues are foreseen to deposit on the catalysts active
sites and cause deactivation. It has been reported that the silicon compounds can
form a SiO2 layer, which covers the Pt surface (Gentry et al., 1978, Cullis et al.,
1984, Matsumiya et al., 2003, Libanati et al., 1998, Rahmani et al., 2004).
It can be possible to regenerate the catalyst if the SiO2 film can either be reorgan-
ised to open up the Pt surface or it can be released the silicon from the surface.
The degradation of platinum catalyst in the printing industry is a process that
takes several years, with approximate lifetimes of three to five years.
Deeper understanding of the deactivation mechanism can lead to development of
more silicon tolerant catalysts to be used even in the presence of silicon com-
pounds. Modifications of the catalyst could be through introduction of promoters
or inhibitors of transient metal oxides. The protection of the Pt surfaces by intro-
duction of a non-active eggshell coating or tailoring of the pore structure is also
possible.
Steam Reforming Catalyst Applied to Product Gas of Biomass
Gasification
The use of biomass for production of synthetic liquid fuels can involve gasifica-
tion of biomass producing mainly a gas consisting of lower hydrocarbons (CH4
and C2), H2, H2O, CO and CO2 as well as ash and tars (Satterfield, 1996). Steam
reforming can be used to convert the hydrocarbons and tar to CO, CO2 and H2 in
accordance with:
CnHm + n H2O n CO + (n+m/2) H2
Further processing of the obtained gas mixture can produce liquid fuels such as
methanol, ethanol, synthetic diesel or dimethylether through catalytic processes.
Commercial reforming catalysts usually consist of supported nickel catalysts
with supporting materials alumina and alkali metals as promoters. The platinum
group metals are also highly catalytically active but have so far been considered
expensive (Rostrup-Nielsen, 1975, Satterfield, 1996).
8

9
Catalysts are supplied as rings, pellets with multiple holes or wagon wheels to
obtain high surface area and good mass and heat transfer. Reactor configuration
is of tubular form in either thin heat exchanger tubes or packed beds. Application
of catalysts of monolithic type is also possible.
Deactivation related to steam reforming involves the unwanted formation of car-
bon deposition, which can lead to blocking of catalyst pores as well as to catalyst
deterioration. Design of process operating conditions is important to avoid car-
bon formation as the nature of deposition varies with operating conditions.
Nickel catalysts as well as catalysts of the platinum group metals are very sensi-
tive to sulphur poisoning and sulphur levels of less than 0.5 ppm are required in
commercial operation. High temperature effects on the support material are also
of importance for both nickel and platinum group catalysts.
Important commercial applications of reforming are conversion of nafta to town
gas, conversion of high methane contents, synthesis of ammonia, formation of
CO-H2 mixtures for Fischer-Tropsch processes, or for manufacturing of hydro-
gen. Typical steam reforming conditions depend on the processes but tempera-
tures may range from 400 to 900 °C and pressures of 1.5 to 3 MPa or higher with
catalyst lifetimes of one to three years.
In biomass gasification process steam reforming is to be applied at 800 to 900 °C
after hot gas particle filtration with steam addition to the catalytic reactor (Alber-
tazzi et al., 2005). Several catalysts are to be evaluated with respect to perform-
ance and lifetime with focus on commercial nickel catalysts and novel platinum-
rhodium catalysts.
Chemical composition of biomass is different in comparison to refinery feed-
stock, which may influence catalyst deactivation mechanisms. Biomass contains
very little sulphur but still the amounts of sulphur may be sufficient to induce
sulphur poisoning of the applied catalysts. The level of alkali metals such as po-
tassium is much larger in biomass feedstock as compared to refinery products
(Lighty et al., 2000, Pagels et al., 2003, Strand et al., 2002, Wierzbicka et al.,
2005). Depending on temperature, the alkali metals can be present either in gase-
ous form or as submicrometer aerosol particles that may be enriched in the fine
fraction of ash particles penetrating the hot gas filter. Catalyst deactivation can
be induced from alkali both through gaseous exposure and particle deposition
(Moradi et al., 2003). Other trace metals such as for example zinc can also influ-
ence catalyst activity. The trace metals may also be present in both particulate
and gaseous form.
Understanding catalyst lifetime is crucial for the future commercial success of
producing liquid fuels from gasified biomass as lifetime influences both initial
investment costs as well as fuel production costs.

10
Simulation of Catalyst Deactivation in
Laboratory Scale
A model of catalyst deactivation was simulated in process conditions that were
carried out in laboratory scale experiment. Deactivation processes in commercial
operation are complex involving a number of simultaneous mechanisms. To
study a simplified mechanism it should be duplicated as closely as possible un-
der controlled laboratory conditions. The catalyst exposure mostly takes place
with impurities in product gases, which interact between the catalyst structure
and the gaseous or particulate compounds. Impact of mechanical or thermal in-
fluence can also be simulated in laboratory experiment.
A common method for investigation of deactivation impact of different com-
pounds has been wet impregnation of catalyst samples (Chen et al., 1990, Ka-
mata et al., 1999, Zheng et al., 2004). Catalysts used for abatement of gaseous
emissions in industrial processes are generally not exposed to poisons through
wet conditions. The SCR and VOC catalyst as well as the steam reforming cata-
lysts investigated are exposed at elevated temperatures assuming dry gas condi-
tions. Wet impregnation may not, in these applications, be a suitable recognition
of the commercial deactivation process. The wet impregnation method has been
widely used to distinguish between different catalyst poisons and as a screening
method.
Exposure time of the catalysts under laboratory conditions can also be a chal-
lenge as the deactivation processes can take between hours and up to several
years before a recognisable impact is considerable on the catalyst sample. A suit-
able process for accelerating the poison impact may be necessary to achieve rea-
sonable laboratory deactivation time frames.
Laboratory investigation allows for evaluation of specific poisonous compounds
as well as evaluation of different deactivation procedures. The design of a labo-
ratory deactivation process must be carefully considered taking into account the
commercial process conditions input as well as the deactivation time aspects.
Scope of Work
The aim of this work has been to compare deactivation processes for environ-
mental catalysts and the focus has been to simulate in laboratory or pilot scale
the deactivation characteristics of commercial processes related to biomass com-
bustion and gasification as well as SCR and VOC oxidation.
Simulation of deactivation induced by biomass combustion and gasification re-
quired the development of laboratory deactivation processes involving genera-
tion and characterisation of submicrometer particles followed by subsequent ex-
posure and deposition of the particles on monolithic catalyst samples as well as

11
on catalysts of pellet or grain types. For the SCR and steam reforming catalyst
deactivation considered poisons were alkali and heavy metal compounds, found
in product gases from biomass combustion and gasification, while the corre-
sponding poison in the VOC oxidation was a silicon compound involved in
printing processes.
For all catalysts, both fresh and exposed samples were chemically and physically
characterised and catalyst deactivation was compared with respect to influence of
process condition as well as deactivation processes parameters.

12

13
Experimental
Exposure to Generated Poisons
Generation of deactivating substances under laboratory conditions are related to
the production of poison in gas mixtures of suitable poison concentration, tem-
perature and gas mass flow to be able to obtain a considerable amount of catalyst
exposure to be investigated. Depending on the deactivation mechanism, the re-
sulting experimental set up needs to be modified in order to simulate the degree
of catalyst decay present in the industrial processes.
The challenge is to obtain an exposure of poisoning agent as close to the indus-
trial exposure as possible with respect to mass transfer and chemical interaction
as well as reaction between poison and catalyst structure. It is necessary to accel-
erate the deactivation process under laboratory conditions in order to obtain suf-
ficient poisoning. Commercial deactivation processes occur in the long-term and
need to be slow to allow a catalyst lifetime of multiple years, whereas laboratory
deactivation processes should be targeted for exposure times of hours or days.
Acceleration of deactivation can be facilitated by increased poison concentration,
through choosing a model compound as deactivation agent that is a strong poison
or by designing the laboratory deactivation experiment to increase poison depo-
sition rates. In the following sections investigated deactivation processes will be
discussed.
Catalyst Reactor Configuration
Catalysts are of different forms and shapes, ranging from small grains to pellets
of various sizes and shapes to monolithic elements of various channel configura-
tion. Depending on catalyst shape and configuration as well as process applica-
tion, for the proposed performance of the catalysts different appropriate chemical
reactors will be designed. A common type is tubular reactors of fix bed type that
can be used for both pellet and monolithic catalysts (Sanati et al., 1990, Fogler,
1992). Fluidized bed reactors can also be applied for catalysts in the form of
small catalyst grains (Satterfield, 1996, Butt, 1999).
The reactor configuration chosen for an application take into consideration proc-
ess conditions and impact on catalyst system such as endothermic or exothermic
reactions, internal and external heat and mass transfer, chemical reaction rates,

14
operation temperature, gas compositions, dust or ash in the gas and pressure drop
constrictions (Ertl et al., 1997, Murzin et al., 2005). The catalyst reactor systems
industrially applied for the investigated processes are all fix bed tube reactors,
the VOC oxidation and steam reforming with pellets of different sizes and SCR,
NOx reduction with monolithic catalyst elements.
A fix bed tube reactor utilizing a pellet catalyst has turbulent gas flow through
the catalyst bed allowing for good mass transfer outside the catalyst pellet result-
ing in a relatively high pressure drop (Fogler, 1992, Treybal, 1980 Cussler,
1984). A monolithic catalyst reactor on the other hand is operated with laminar
gas flow through the catalyst channels decreasing the mass transfer but achieving
a lower pressure drop (Fogler, 1992, Hayes et al., 1994, Khodayari et al., 1999).
In laboratory deactivation experiments the catalysts should be exposed to deacti-
vation compounds under conditions as close to industrial as possible including
the applied catalyst reactor type and configuration.
Particle Generation and Deposition Process
Catalyst exposure by particle deposition is a deactivation process previously not
commonly utilized in laboratory deactivation of catalysts. The process is flexible
with respect to different poison agents but it involves the generation of submi-
crometer particles of the poisoning substances concerned. The particle mass con-
centration generated is restricted by the particle generation methods. Generation
of the particulate poisons need attention with respect to particle size and concen-
tration in order to recreate the commercial deactivation process adequately. In-
fluence of gas compositions on oxidation or reduction of generated particle spe-
cies must also be understood and taken into account.
A laboratory set up involving exposure by particle induced poisoning consists of
the sections: feeding system, catalyst reactor, heating system, on line characteri-
zation techniques and control system.
The feeding system for the particle exposure includes the generation of the sub-
micrometer particles through different methods. Most commonly used is atomi-
zation of liquids and evaporation of liquids or solids followed by condensation
and drying.
Particles can be generated through an atomizer producing small droplets of a so-
lution followed by drying in an oven or heated gas stream. The size and concen-
tration of particles are dependent on the atomizer equipment but particles formed
through atomisation have particle sizes from 30 to 500 nm and concentrations in
the order of 106 particles per cm3n. The sizes and concentrations can be influ-
enced by changing the solution concentration and the atomization pressure
(Hinds, 1999).

15
The generated droplet flow is diluted with dry pressurised particle free air or ni-
trogen and dried in an oven or heated gas stream. Dilution ratios as well as dry-
ing time and temperature are determined to avoid moist conditions that could
impact the particle drying or condensation of water on the particles in combina-
tion with particle measurements. Hygroscopic particles can take up water if the
relative moisture content is above the salts deliquescent and efflorescence points
resulting in moist particles. Potassium salts are reported to have efflorescence
points in the range of 80 to 95 % relative moisture allowing for generation of dry
particles through liquid atomisation (Pagels et al., 2003).
Another particle generation method is evaporation of a liquid or a solid precursor
followed by forming of particles from the saturated gas phase by cooling of the
gas mixture. Particles are formed through nucleation and subsequent agglomera-
tion. Particles produced through nucleation can have sizes from 10 to 10 000 nm
and concentrations of 109 particles per cm3n (Hinds, 1999).
Deposition of generated aerosol particles on catalyst samples take place in a cata-
lyst tube reactor of either pellet or monolithic type placed in an oven to allow the
deposition to occur at elevated temperatures normally present in commercial
processes. Deposition of submicrometer particles in a packed fix bed reactor
filled with pellets or catalyst grains show a high level of deposition due to the
high mass transfer between the particle flow and catalyst surface (Murzin et al.,
2005). A packed bed has good filtration properties resulting in a high degree of
particle capture. In a monolithic catalyst channel on the other hand, the submi-
crometer particles will mostly pass through the channel unaffected due to the
laminar flow conditions (Hinds, 1999, Fogler, 1992). In commercial operation
this limits catalyst deactivation due to particle deposition and it has been found
that only a small fraction of a poisonous compound is found on the catalyst sur-
face as compared to the amount present in the gas. (Khodayary, 1999 and 2001).
For laboratory induced particle deactivation, an acceleration of the particle depo-
sition can be achieved by applying turbulent flow conditions or by utilization of
an electrostatic field that facilitate deposition of the particles on the catalyst sur-
face.
A control system can be used for temperature and flow control as well as particle
measurement techniques.
The deactivation process for the investigation of the SCR monolithic catalyst un-
der biomass combustion conditions was designed using potassium and zinc salt
particles of a mean diameter of 100 nm as model compounds to simulate submi-
crometer particle poisoning. Particle concentrations in the range of 106 particles
per cm3n were applied (Strand et al., 2004).
To enhance the particle deposition rate and accelerate the deactivation process in
the monolithic catalyst channel, an electrostatic field was applied across the cata-
lyst channel. The application of the electric field is shown in Figure 1. The gen-
erated salt particles are positively and negatively charged leaving the atomiser.

An electrical field was created inside the catalyst channel by placing a steel wire
inside the catalyst channel and surrounding the catalyst channel walls with alu-
minium foil.
+
Catalyst sample with an
aluminum foil coating
connected to ground
potential
Positvely charged steel electrode
Gas flow
direction
Glass connections
+
Catalyst sample with an
aluminum foil coating
connected to ground
potential
Positvely charged steel electrode
Gas flow
direction
Glass connections
+
Catalyst sample with an
aluminum foil coating
connected to ground
potential
Positvely charged steel electrode
Gas flow
direction
Glass connections
Figure 1. Application of electrical field in the monolithic catalyst channel.
The overall laboratory set up is shown in Figure 2. The particles were generated
through a pneumatic atomizer fed with a liquid solution of deionised water and
the respective salt with deactivation conditions presented in Paper I. The cata-
lyst sample consisted of 1 monolithic channel of commercial V2O5–WO3/TiO2
catalyst with a cell opening of 6.4 mm and a length of 160 mm.
Precursor
atomizer
Inlet gas
Particle measurement
SMPS and ELPI
Catalyst
sample
Evacuation
Electric field
source
0-3,5 kV Vacuum
pump
DMA
ELPI
SMPS
Filter
Critical
orifice
Dilution air
CPC
Dilution air
1/10
Dilution air
1/10
Precursor
atomizer
Inlet gas
Particle measurement
SMPS and ELPI
Catalyst
sample
Evacuation
Electric field
source
0-3,5 kV Vacuum
pump
DMA
ELPI
SMPS
Filter
Critical
orifice
Dilution air
CPC
Dilution air
1/10
Dilution air
1/10
Figure 2. Experimental set up of the aerosol particle deposition for monolithic catalysts.
Investigations of deactivation of steam reforming catalysts of Ni or Pt/Rh were
carried out also using interaction with potassium salts as well as biomass com-
bustion ash salts. The experiments were carried out in a set up shown in Figure 3.
The particles are also generated in an atomizer using diluted salt solutions and
subsequent drying of the salt droplets in an oven. The catalyst was crushed to
powder before it was placed in a fix bed packed reactor. The deposition of parti-
cles was carried out at 800 °C. Further deactivation conditions are presented in
Paper V.
16

Due to the elevated temperature and the thermodynamic properties of the salts
the deactivating species will be present as either solid particulate matter, liquid
droplets as well as gaseous form. A detailed analysis is presented in Paper V.
N2
P=1 bar
Atomizer
containing
salt solution
Oven
Reactor
Impaction
vessel
Aerosol flow 4 l/min
Cooling flow
T= 800°C
Catalyst holder
Filter
Filter
Pump
N2
P=1 bar
Atomizer
containing
salt solution
Oven
Reactor
Impaction
vessel
Aerosol flow 4 l/min
Cooling flow
T= 800°C
Catalyst holder
Filter
Filter
Pump
Figure 3. Experimental set up of the aerosol particle deposition.
Gaseous Exposure Process
The gaseous poison exposure process is flexible and can use different poison
agents available as liquids, gases or in solid form. Deactivation rates can be ac-
celerated by the preference of poison agent as well as suitable concentration lev-
els. Exposure times can be alternated from hours to months.
A laboratory set up involving exposure by gaseous poisons comprise of four sec-
tions: a feeding system, a catalyst reactor, a heating system and a control system.
The feeding system can consist of a dosing pump for liquid poison, a vaporiser
or a mass flow controller for gaseous poison or a controlled evaporation of a
solid poison. The concentrated poison flow is diluted with a gas consisting of air
or nitrogen or a simulated gas mixture to the desired poison concentration.
A catalyst reactor of either tubular fixed bed type with pellet catalyst or mono-
lithic catalyst reactor can be applied.
Depending on the process, the heating system can be a heat exchanger or an
oven. The control system includes temperature control as well as measurement
equipment for gaseous process and exposure compounds, for example Flame
Ionization Detectors (FID) or Gas Chromatographs (GC).
The deactivation process for the investigation of the supported alumina platinum
catalyst was designed using Hexamethyldisiloxane (HMDS) as a model com-
pound to simulate gaseous silicon poisoning (Gentry et al., 1978, Cullis et al.,
17

1984, Colin et al., 1996, Ehrhardt et al., 1997). Ethyl acetate was used as hydro-
carbons model compound in order to investigate the catalysts oxidation ability
(Sawyer et al., 1994).
The pilot set up used is shown in Figure 4. It consisted of a feeding system, a tu-
bular fixed bed reactor, a heat exchanger and a control unit. The feeding system
consisted of a dosage pump, a vaporizer, and a small air fan. A solution of ethyl
acetate and HMDS was injected into the hot air inside the vaporizer, fed by an air
fan. This flow was diluted with air and then fed to the heat exchanger, and fi-
nally, passed through the fixed bed reactor. An electrical heating element was
placed at the reactor inlet for the start up and to control the feed temperature to
350° C.
H-1
F-2
F
-
1
P
-
1
V
-
1
H-2
Vent
R-1
FI
TIC
TI
TIS
TIC
FI
Figure 4. Schematic flow diagram of pilot unit for gaseous silicon exposure.
A fixed bed reactor made of stainless steel, consisting of 12 vertical tubes, was
integrated with the heat exchanger in an insulated compartment.
A control unit was used to adjust the reactor inlet temperature, vaporizer tem-
perature, and feed flow rate.
Deactivation conditions are presented in Paper IV. Deactivation was performed
in stages for 350, 650 and 1000 hours with tubes filled with Pt/γ-Al2O3 catalyst
and γ-Al2O3 support.
Wet Impregnation of Catalyst Samples by
Poisons
Wet impregnation is a commonly used laboratory deactivation process. The cata-
lyst sample is impregnated with a solution of catalyst poisons. The solution con-
18

19
centration as well as the impregnation time can be varied in order to study the in-
fluence of poison concentration in the catalyst matrix.
After impregnation the catalyst samples are dried in ambient temperature or dur-
ing heating in an oven. The catalyst samples can also be calcined at elevated
temperatures to allow formation of poisonous species on the catalyst surface
(Chen et al., 1990, Kamata et al., 1999, Zheng et al., 2004).
The impregnation process not does often resemble an industrial application, but
allows for a comparison of different catalyst poisons and evaluation of impact of
poison concentration on catalyst deactivation.
In the study on deactivation of commercial V2O5-WO3/TiO2 SCR catalysts were
impregnated with solutions of KCl, K2SO4 and ZnCl2 at conditions described in
Paper II.
Commercial Deactivation of Catalyst Samples
Exposure of catalyst samples in industrial processes can allow for an understand-
ing of the overall deactivation mechanisms. The catalyst samples can be placed
in the gas flow either as part of the original catalyst installation or as designated
test samples. It is of importance to ensure that the samples are exposed to the gas
under the same conditions as for full-scale installations with respect to tempera-
ture and gas flow. If the catalyst is installed as a separate sample the pressure
drop across the sample may require a fan or a pump to make sure the gas flow
through the sample is correct. Another alternative is to use a catalyst sample con-
figuration that has a very low pressure drop. A common exposure form is to re-
move for example a monolith element of SCR catalyst or samples of pellet cata-
lyst from commercial installation after specific operating intervals. A SCR
monolith catalyst sample placement is shown in Figure 5.
It is also possible to expose catalyst samples in slipstream pilot plants extracting
a fraction of the gas flow from the commercial installation. The pilot plant is
then designed to expose the catalyst samples to the condition of the industrial
process. In the operation of the pilot plant it is important to ensure that the frac-
tion of gas taken from the full-scale process is sampled isokinetically from the
gas duct allowing both gas as well as particulate matter to come in contact with
the catalyst samples. Otherwise concentration of gas components and particles
may differ from the full-scale process and the evaluation of the impact on the
catalyst samples may not be correct. A pilot plant has the advantage that tem-
perature and gas flow conditions can be controlled and monitored. A slipstream
pilot installation is shown in Figure 6.

Figure 5. Catalyst sample placement in a full scale SCR reactor.
Industrial exposure of catalyst samples allows all existing deactivation mecha-
nisms to influence the catalyst simultaneously. The distinction between different
deactivation mechanisms may be difficult depending on the application. The
overall activity decrease can be determined while the dominating mechanism
may not be revealed.
Figure 6. Industrial slipstream pilot installation.
In the study of SCR catalysts, samples were exposed to commercial biomass
combustion conditions in a commercial biomass fired boiler during an operating
season. The catalyst sample was exposed to flue gas without any prior particle
removal for 6 500 h in the placement shown in Figure 5. Operating conditions
are described in Paper II. The exposed catalyst sample was used for verification
of the laboratory deactivation processes related to potassium poisoning of SCR
catalysts used in biomass combustion processes.
20

21
Characterisation
Catalyst Characterisation
Catalysts can be characterised chemically and physically with respect to catalyst
compositions and internal structures. Characterisations of both fresh and exposed
catalyst samples can indicate interaction between the catalyst matrix and poisons
and give input to deactivation mechanism studies. The methods used in this work
are described in this section.
Physical Characterisation of Catalyst Samples (BET)
The surface area of a catalyst sample can be evaluated through gas adsorption on
the catalyst surface under controlled pressure conditions. Brunauer, Emmett and
Teller developed the most common method, BET, in 1938 and it is routinely
used for catalyst studies (Satterfield, 1996). The surface area is determined by
adsorption of a monolayer of nitrogen molecules on the catalyst surface at the
temperature of liquid nitrogen.
As BET is a common analysis method several instruments are available. For the
investigation of the V2O5-WO3/TiO2 catalyst (Paper II and III) and Pt/γ-Al2O3
catalyst (Paper IV) a Micromeritics Tristar 3000 instrument was used. The
V2O5-WO3/TiO2 catalyst was also analysed using a Micrometerics ASAP 2400
(Paper VI). The steam reforming catalysts of nickel and platinum-rhodium was
analysed using a BOL Sorpty 1750 (Paper V).
Pore volume and pore size distributions in the range of 17-3000 Å were calcu-
lated using the method based on nitrogen desorption developed by Barrett, Joy-
ner and Halenda (BJH) (Satterfield, 1996). All catalyst samples were degassed at
vacuum before the analysis at elevated temperatures of at least 200 °C for a pe-
riod of 36 h.
Chemisorption
To investigate the effect of poisons on the catalyst active sites chemisorption of
gas molecules on the catalyst surface can be employed. A suitable gas molecule
is chosen for the chemisorption study. The gas molecule should not dissociate on

22
adsorption nor change the catalyst structure irreversibly by chemical reaction
with the catalyst material. CO is a commonly used gas molecule (Satterfield,
1996).
Chemisorption investigations may indicate changes in the rate-limiting step of
the catalysts. It is also possible to evaluate the heat of chemisorption as it is a
measure of the bonds formed between the catalyst and the adsorbing gas mole-
cule. Chemisorption also indicates the number of active sites available for reac-
tion as well as the metallic surface area and dispersion of a metallic catalyst.
The influence of potassium and zinc salts on the SCR V2O5-WO3/TiO2 catalyst
was evaluated by ammonia chemisorption of the fresh and exposed SCR catalyst
samples using Temperature Programmed Desorption (TPD) with ammonia as the
adsorbing gas. The reaction mechanism of the SCR reaction includes the adsorp-
tion of ammonia on the acidic vanadium sites (Topsoe et al., 1995) preceding the
reaction of ammonia with NO. Conditions are presented in Paper VI.
CO chemisorption was used for the evaluation of the dispersion and surface area
of the platinum based alumina catalysts for VOC oxidation (Holmgren et al.,
1998, Skoglundh et al., 1996, Arnby et al., 2004) with conditions described in
Paper IV. The steam reforming catalysts of nickel and platinum-rhodium were
evaluated using H2 chemisorption (Paper V).
Surface Analysis of Catalyst Samples
Evaluations of catalyst surfaces with respect to morphology, structure, as well as
chemical composition can involve different techniques some of which are pre-
sented here.
Scanning Electron Microscopy (SEM)
SEM uses a scanning of a finely focused electron beam to give a magnified im-
age of a surface. The image is produced from electrons backscattered from the
sample surface. Magnifications make it possible to detect surface markings in the
size of less than 50 nm (Satterfield, 1996).
SEM is particularly used for examination of catalyst surface texture and mor-
phology. It can also be used to give information on size and shape of deposited
particles. SEM can also be equipped with EDS (Energy Dispersive Spectros-
copy) in order to evaluate the chemical composition on a designated point or part
of the investigated area. A beam of electrons are focused on the catalyst surface
and atoms are ionized to produce characteristic X-rays that are representative of
the elements and intensity that is proportional to the concentration of the ele-
ment. The electron beam penetrates 100 nm into the material so the composition
is an average of the sample.
In the current investigations SEM was used to evaluate catalyst morphology us-
ing a LEO, Gemini 1550 microscope. The surface of the monolithic V2O5-

23
WO3/TiO2 catalyst was evaluated with SEM to investigate particle deposition
morphology related to the electrostatic deposition of salt particles in the catalyst
channel (Paper I). SEM-EDS was also used to evaluate the chemical
composition of surface deposits on a commercially exposed SCR catalyst sample
(Paper I).
X-Ray Powder Diffraction (XRPD)
X-ray powder diffraction (XRPD) is a non-destructive analytical technique al-
lowing for identification and quantification of crystalline bulk phases in powder
samples of crystalline materials (Arnby, 2004). XRPD uses the elastic scattering
of X-ray photons by atoms in the periodic lattice. A diffraction pattern shows
phases present, concentrations, amorphous content and crystalline size by ex-
pressing peak position, peak height, broad features and peak widths. A limitation
is that particles that are too small cannot be detected and it is thus not possible to
exclude the presence of phases not detected by XRPD.
XRPD was used in the structural evaluation of the steam reformer catalysts, Ni
and Pt/Rh, where the metal crystallite size was also determined by using the peak
broadening as presented in Paper V.
Poison Penetration Profile
SEM-EDS can also be used to evaluate concentrations of deposited substance
penetrating into the catalyst matrix. The method is very useful for obtaining a
distribution profile of an element through a sample of catalyst, for example the
penetration of a poison in a catalyst pellet or through a catalyst wall.
SEM with an Electron Micro Probe Analyser (EMPA) also equipped with an en-
ergy dispersive spectrometer (EDS) was used for penetration analysis. The cata-
lyst samples were impregnated with epoxy resin as pre-treatment.
The Pt/γ-Al2O3 catalyst was analysed for the radial distribution of silicon in indi-
vidual pellets. The deactivated cylindrical pellets were cut radially and their
cross-sections were scanned along their diagonal (Paper IV).
The monolithic V2O5- WO3/TiO2 catalyst was analysed for potassium and zinc
penetration across the wall thickness. The catalyst samples were cut across the
wall thickness and were scanned through the wall (Paper II). The wall of the
monolithic catalyst was evaluated perpendicular to the gas flow direction in the
catalyst channel.
Chemical Analysis
Chemical analysis of the catalyst is used to determine possible accumulation of
compounds on the catalyst caused by the exposure to deactivating substances.
Chemical analysis can also be used to determine initial concentrations of active
components. A variety of methods can be employed depending on the catalyst

24
composition and elements to be analysed. The methods used in the studies are
ICP-AES (Inductively Coupled Plasma - Atomic Emission Spectroscopy) and
AAS (Atomic Absorption Spectrophotometery). The bulk compounds analysis of
fresh and deactivated samples was carried out with an ARL 3560 ICP-AES (In-
ductively Coupled Plasma - Atomic Emission Spectroscopy) instrument. For
each analysis, the sample was ignited at 1000°C for one hour. The sample was
smelted with LiBO2 and dissolved in HNO3. This sample was used to run ICP-
AES. The method allows for simultaneous analysis of the elements in the cata-
lyst material. For an exposed catalyst the analysis will present all elements pre-
sent both in the catalyst material as well as any elements accumulated on the
catalyst during exposure. This technique was used to analyse the bulk silicon
concentration of the Pt/γ-Al2O3 catalyst, as well as the elemental analysis of the
V2O5-WO3/TiO2 catalyst (Paper II and IV).
The AAS analysis involves dissolving the catalyst sample in hydrochloric acid
followed by heating. An analysis is then performed on an absorption spectropho-
tometer at a designated wavelength.
Physical Characterisation of Generated
Particles
The aerosol salt particles generated for the investigation of particle-induced de-
activation were characterised with respect to particle size, mass concentration,
particle number size distribution and mass size distribution as well as particle
charge distribution. The instrument techniques used are described in this section.
Scanning Mobility Particle Sizer (SMPS)
A Scanning Mobility Particle Sizer (SMPS) classifies aerosol particles according
to their electrical mobility and it is used to measure particle size distribution in
the range of 20 to 248 nm.
A SMPS is an electrical mobility spectrometer consisting of a bipolar diffusion
charger, a differential mobility analyser (DMA) and a condensation particle
counter (CPC) (Willeke et al., 1993, Hinds, 1999, Strand, 2004), as shown in
Figure 7. Computer software is used to convert particle penetration characteris-
tics to a particle size distribution.
The particles are charged by a well-defined charge distribution when passing
thorough the bipolar diffusion charger before entering the DMA. A DMA con-
sists of a centre electrode surrounded by a grounded outer concentric cylinder. A
“sheet airflow” of particle free gas flows downwards and axially between the two
cylinders. The particle sample flow is introduced to the sheet flow at the top of
the cylinders. By applying an electrical field between the two cylinders, a scan-

ning of the electrical potential, the poly-dispersed particles can be separated
based on their mobility in the electrical field. The particle size distribution is thus
determined based on the electric-mobility-equivalent particle size.
Figure 7. The electric mobility spectrometer (From TSI Inc.).
The CPC is used to establish the concentration of particles at the outlet of the
DMA. A CPC is a particle counter, growing the particles by condensation of
buthanol onto the particle before counting the particles as they are passed across
a laser beam.
In the evaluation of generated salt particles in the submicrometer mode the num-
ber size distribution was measured with a SMPS (Model 3936, TSI Inc.) incorpo-
rating a CPC 3010 (TSI Inc.). The measurements are presented in in Papers I,
II, and V.
Tandem Differential Mobility Analyser (TDMA)
Determination of the particle charge distribution requires a Tandem Differential
Mobility Analyser (TDMA) set up employing two separate DMA units. The
method is illustrated in Figure 8. The first DMA separates the aerosols based on
their electrical mobility. The aerosols are then charged in a diffusion charger
with a determined charged distribution (Boltzmann equilibrium charge distribu-
25

tion) and the SMPS (DMA2 and CPC) measures the particle number size distri-
bution (Kim et al., 2005, Willeke et al., 1993).
Applying both positive and negative electrical potentials to the first DMA allows
for measurement of both positively and negatively charged aerosol particles.
Neutral aerosol particles can be analysed by removing the air sheet flow of the
first DMA and applying maximum electrical potential.
A charge distribution measurement was applied to the generated salt particles. A
DMA Model 3071(TSI Inc.) was used in combination with a SMPS (Model
3936, TSI Inc.) The charge distribution measurements are discussed in Paper I.
Figure 8. The Tandem Differential Mobility Analyser set up (From Kim et al. (2005)).
Electric Low Pressure Impactor (ELPI)
The electrical low pressure impactor (ELPI), designed at Tampere University of
Technology, Finland, is a real time particle size spectrometer. The measurement
principle is to charge the particles before an inertial classification in 12 steps and
electrical detection of the particles. The instrument consists of a corona charger,
a low pressure cascade impactor and a multi channel electrometer shown in Fig-
ure 9.
26

The particle sample flow passes a unipolar positive polarity charger where the
particles are charged electrically by the production of small ions in the cororna
discharger. The charged particles are size classified in the low pressure impactor.
The impactor stages are insulated electrically and each stage is individually con-
nected to an electrometer current amplifier. The charged particles collected in a
specific impactor stage produce an electrical current proportional to the number
of particles collected. The current is recorded by the respective electrometer
channel. The current values are converted to a size distribution taking into ac-
count particle dependent relationships describing the properties of the charger
and the impactor stages as well as the effective particle densities (Strand, 2004).
The ELPI can be used to classify the aerodynamic diameter of the aerosol parti-
cles in the range 30–10 000 nm, and by the effective particle density the mass
size distribution can be obtained.
The ELPI (Dekati Ltd.) was used to measure the mass size distribution of the
generated salt particles, Paper I, II and V. To avoid possibilities of overestima-
tions of mass for large particles sizes larger than 3000 nm were neglected in the
evaluations (Pagels et al., 2005),. The sample flow to the ELPI was diluted with
dry particle free pressurized air with a total dilution of 1/10.
Figure 9. The electrical low pressure impactor (From Dekati Oy).
27

Catalyst Performance
The impact of catalyst poisons on catalyst activity can be evaluated using a
model reaction under controlled reaction conditions. The reaction, the test condi-
tions and reactor configuration are designed, based on the catalyst and the proc-
ess application. The activity measurements are performed on fresh and exposed
catalyst samples allowing for comparison. The reaction and conditions applied
must be carefully evaluated in order not to change the catalyst composition or
regenerate the catalyst by removal of the deposited material. If inappropriate
conditions are applied the catalyst could for example be oxidised or reduced with
loss of activity as a result.
SCR Catalyst Activity Measurement
The investigation of catalyst activity for the monolithic V2O5- WO3/TiO2 catalyst
was based on the SCR reaction, where NO is reduced by reaction with NH3 ac-
cording to (Topsoe et al., 1995, Forzatti, 1999):
4 NO + 4 NH3 + O2 4 N2 + 6 H2O
The catalyst activity measurements were carried out in a laboratory fixed bed re-
actor set up including synthetic gas generation, an electrical heater, catalyst reac-
tor and gas measurement equipment, temperature and pressure control system
and gas concentration analyzers.
28
The synthetic gas is produced by injection of gas components, NO and NH3, into
a preheated ambient air stream. The flow of the NO and ammonia was measured
by mass flow controllers (Brooks MFC 5850).
The reactor was a square vertical tube of stainless steel with an inside dimension
of 40 mm and height of 420 mm. The analysis was carried out with catalyst sam-
ples of 6 monolithic channels with a length of 160 mm. The reactor was placed
in a heating compartment to minimize influences of temperature gradients. The
inlet and outlet concentration of NO and NO2 were analysed using IR/UV on line
analysers (Binos UV/IR and Unor 6N).
The SCR catalyst activity test conditions are presented in Table 1. Further details
are given in Papers II, III and VI.
Table 1. Reference conditions used for the activity measurementsof the SCR catalyst.
Catalyst volume 52.5 cm3
Temperature levels 250, 275, 300, 325, 350 °C
Pressure 1 atm (1.01×105 Pa)
Gas feed flow rate 13.5 liters/min, NTP condition
Gas hourly space velocity 15 500 h -1, NTP condition
NO concentration 500 ppmv
NH3 concentration 550 ppmv

VOC Catalyst Activity Measurement
The activity measurements of the Pt/γ-Al2O3 catalyst utilized pure ethyl acetate
as model compound for the oxidation to CO2 and water according to (Sawyer et
al., 1994):
C4H8O2 + 5 O2 4 CO2 + 5 H2O
The activity tests were performed in a cylindrical adiabatic reactor of stainless
steel. The reactor was 36 mm in diameter and was surrounded with an insulation
section containing a temperature-controlled heat barrier. The insulation was
heated to the same temperature as the inlet gas (Hinz et al., 2001).
The gas flow of dry industrial air free from CO2 was fed with a mass flow con-
troller (Brooks), and a liquid pump (ALiTEA) was used to dose the ethyl acetate
to the gas flow. The inlet and outlet gas compositions were analysed with respect
to CO2 and CO using an IR instrument (Fuji Electronic, ZRF IR-analyser). The
total hydrocarbon concentration was analysed using FID-detectors (Bernath
Atomic, model 3006).
The reference operating conditions used for the activity measurements are shown
in Table 2. More details are available in Paper IV.
Table 2. Reference conditions used for the activity measurements of the Pt/
γ
-Al2O3 cata-
lyst.
Catalyst volume 60 cm3
Temperature levels 250, 300, 325, 350 °C
Pressure 1 atm (1.01×105 Pa)
Gas feed flow rate 11.5 l/min, NTP condition
Gas hourly space velocity 11500 h -1, NTP condition
Ethyl acetate in the feed 425 ppmv
Steam Reforming Catalyst Activity Measurement
The activity of the Pt/Rh catalyst used the steam reforming reactions of methane
according to:
CH4 + H2O ↔ CO + 3 H2
CH4 + 2 H2O ↔ CO2 + 4 H2
The main parts of the activity measurement set-up are an oven delivering the re-
action temperatures and gas mixing equipment including mass flow controllers
to generate the gas mixture. The reaction gases are mixed and preheated in a
heated vessel and then led into a tubular fix bed reactor containing 10 ml of cata-
lyst.
29

30
Pressure and temperature as well as steam to carbon feed ratio were varied in the
experiments. Activity test conditions are summarized in Table 3 and further de-
tails are presented in Paper V.
Table 3. Reference conditions used for the activity measurements of the Pt/Rh catalyst us-
ing stream reforming reactions.
Catalyst volume 10 ml
Temperature levels 600 to 800 °C
Pressure 1 to 9 bar
Gas feed flow rate 500, 1000 l/h
Gas hourly space velocity 50 000 and 100 000 h -1
Steam to Carbon ratio 1.7 and 2.5

31
Results
Evaluation of the laboratory methods applied in order to expose catalysts to gen-
erated poisons involves chemical and physical characterization of the induced ef-
fects. For a complete evaluation, the laboratory results should be compared to
catalysts exposed at intended operation conditions in commercial plants. Com-
mercial operation results are not always accessible due to lack of commercial ex-
perience of the catalyst investigated. The three catalytic processes, NOx abate-
ment by SCR for biomass combustion, oxidation of VOC’s from printing proc-
esses and steam reforming after biomass gasification have been evaluated based
on: actual deposition and accumulation of poisons in the exposed catalysts, in-
fluence on catalyst reaction activity, deposition of poison with respect to pore
structures and poison penetration and influence of blocking of active sites as well
as influence of active phase dispersion. Comparison with commercially exposed
catalysts was available only for one of the processes, NOx reduction by SCR.
Deposition of Generated Poisons
The methods applied for catalyst exposure to poisonous compounds all aimed to
give maximum accumulation of poison in the catalyst after minimum exposure
times.
Investigations regarding particle induced catalyst deactivation, for SCR catalyst
and steam reformer catalysts, both involved generation of salt particles of a des-
ignated size distribution with maximum mass concentration. Both investigations
utilized the same atomizer set up for particle generation although different salts
and exposure temperatures were employed. The evaluated aerosol particles are
presented in Table 4 and number size distribution for three of the salts used are
presented in Figure 10. A difference in mass concentration was detected when
comparing the generated K2SO4 aerosols. A higher mass concentration is avail-
able at the conditions applied for the SCR catalyst, as a fraction of particles lar-
ger than 1000 nm was removed for the steam reforming measurements. Number
concentration and the mean particle sizes are comparable. Further details of the
generated aerosols are available in Papers I and V. The particles generated are
comparable in size, concentration and chemical composition to measurements of
particles measured in commercial biomass boilers, Paper I.

Table 4. Characterisation of generated aerosol particle.
Deactivation spe-
cies
KCl K2SO4 ZnCl2 K
2SO4
Ash
salt
Temperature 200 200 200 150 150 °C
Bulk density 1.98 2.66 2.91 2.66 2* g/cm3
Number concentra-
tion
5.7x106 6.0x106 8.0x106 7.4x106 6.5×106 #/cm3
Mean particle size 104 103 91 119 98 Nm
Mass concentration 121 131 130 37 13 mg/m3
*Estimated bulk density
0.0E+00
4.0E+06
8.0E+06
1.2E+07
1 10 100 1000
Mobility Equivalent Diameter, dp [nm]
dN/dlg(dp) [cmn
-3]
KCl
K2SO4
ZnCl2
Figure 10. The number size distribution of generated salts particles measured by SMPS.
The steam reformer catalyst was exposed to the aerosol particle flow using a
packed bed of small grains of catalyst. Deposition of aerosol particles can be as-
sumed to be high due to good mass transfer as the gas and particles pass through
the packed bed of catalyst. The amount of salts deposited on the steam reforming
catalysts analysed by ICP-AES were 0.3 wt % of K2SO4 for both the Ni and
Pt/Rh catalysts and 0.07 and 0.1 wt % of ash salt for the Ni and Pt/Rh catalysts
respectively (Paper V). As the application of catalytic steam reforming for gasi-
fication of biomass is not yet a commercial process no comparison with full scale
commercial operation is yet possible.
For the SCR monolithic catalyst, the flow in the catalyst channel is laminar un-
der exposure conditions due to restrictions of the experimental set up. The parti-
cle deposition required an enhancement of the deposition rate, which was
achieved by applying an electrical field. The influence of the electrical field was
evaluated by varying voltage and by measurement of the fraction of charged par-
ticles generated by the atomizer. An optimum voltage of 2.5 kV was chosen as
32

increased voltage did not significantly enhance the particle deposition but created
instability to the electrical field and increased spark-over between the electrode
and the catalyst wall. The fraction of charged particles was shown to be small for
particle sizes below 100 nm leading to negligible deposition of smaller particles.
The electrostatic deposition was evaluated with respect to number concentration
and the relative differences are shown in Figure 11. A relative difference below 1
indicates that particles have been deposited on the catalyst surface. It can be seen
that the deposition is predominantly acting on the charged particles with a di-
ameter above 100 nm. Further discussions are presented in Paper I.
0,0
0,5
1,0
1,5
2,0
1 10 100 1000
Mobility Equivalent Diameter, dp [nm]
Relative difference in
number concentration
b
c
a
Figure 11. Relative differences of number concentration of generated salts aerosol parti-
cles measured for the honeycomb SCR catalyst with and without applying a voltage of 2.5
kV using SMPS, (a) KCl, (b) K2SO4, (c) ZnCl2.
Chemical bulk analysis with ICP-AES was employed for the SCR catalyst and
indicated concentrations of 0.3 wt % KCl, 0.2 wt % K2SO4 and 0.3 wt % ZnCl2
deposited in the catalyst matrix after 10 h of exposure to particles.
Wet impregnation deactivation was investigated in order to compare with parti-
cle-induced deactivation by potassium and zinc salt. Two impregnation solutions
with different salt concentrations were used, 1 and 10 g/L. Chemical bulk con-
centrations of potassium and zinc were analysed by ICP-AES and are presented
in Table 4. Results show levels of 0.1 to 0.8 wt % of potassium salts and 0.1 to
0.6 wt % of zinc.
A commercial catalyst was exposed to biomass combustion conditions for 6 500
h for verification of the laboratory investigations. The commercially exposed
catalyst sample showed a bulk concentration of 0.5 wt % potassium. No zinc ac-
33

cumulation was detected in the commercial catalyst and no comparison of the
zinc concentration was possible.
Both the electrostatic particle exposure method as well as the wet impregnation
method was shown to be able to accelerate potassium and zinc exposure under
laboratory conditions producing comparable amounts of potassium when com-
pared to commercially exposed samples, Paper II.
Table 6. Bulk concentrations of potassium and zinc for wet impregnated SCR catalysts.
Catalyst samples K; Zn concentration
(ICP-AES)
(wt %)
KCl Low conc. 0.2
High conc. 0.8
K2SO4 Low conc. 0.1
High conc. 0.4
ZnCl2 Low conc. 0.1
High conc. 0.6
Deactivation of alumina-supported platinum VOC catalysts by gaseous hexame-
thyldisiloxane (HDMS) was studied in a pilot scale fix bed of catalyst pellets for
up to 1000 h, Paper IV. The bulk silicon loading for the Pt/γ-Al2O3 catalysts and
blank γ-Al2O3 support is presented in Figure 12.
0
2000
4000
6000
8000
10000
12000
14000
300 550 800 1050
Time on Stream, h
Si Loading, mg/kg
Pt-Catalyst
Carrier
Figure 12. Silica loading for the inlet fraction of
γ
-Al2O3 carrier and Pt/
γ
-Al2O3 catalyst.
Silicon deposition shows a linear relationship with exposure time with a slightly
faster deposition for the platinum catalyst compared to the blank support. Silicon
deposition was also compared with respect to axial bed position and a linear rela-
tionship was also found. Maximum silicon concentrations are found at the reac-
34

tor inlet increasing with exposure time. The silicon concentration in the inlet of
the reactor can be slightly overestimated at accelerated exposure, as the inlet
concentration of HDMS is typically 25 times higher in the pilot study. Silicon
bulk concentrations after 1000 h of exposure was 4 000 to 13 600 mg/kg.
No comparison with commercially applied VOC catalysts was available in this
study. The HDMS concentration was increased from typically 20 ppb in com-
mercial operation to 0.5 ppm in the pilot study implying a 25 fold increase of the
concentration of the poisonous substance. The 1000 h exposure could be compa-
rable to about 3 years of commercial operation.
Impact of Poisoning on Catalyst Activity
The impact of deactivating substances on catalyst activity was evaluated in the
respective reactions, reduction of NO with ammonia for the V2O5- WO3/TiO2,
SCR catalyst, oxidation of ethylacetate for the Pt/γ-Al2O3 catalyst and stream re-
forming of methane and water for the Pt/Rh catalyst.
Studies of the deactivation processes for the SCR catalyst were carried out by
comparing the catalyst activity at 300 °C taking into account the accumulation of
potassium and zinc in the catalyst, Papers II and VI. The results are presented
in Figure 13. Catalysts exposed to particles of potassium salts show a slightly
higher NO reduction than the sample supplied from a commercial boiler,
whereas catalyst samples prepared by wet impregnation of the potassium salts
show a stronger deactivation with increasing potassium salt concentration. The
wet impregnation results are comparable with results presented by other investi-
gations (Paper II).
0
25
50
75
0,0 0,5 1,0 1,5
KCl/K2SO4/ZnCl2 [w%]
NO reduction [%]
K Particle Exposure
K Impregnated
Zn Particle Exposure
Zn Impregnated
Commercial Exposure
Figure 13. Comparison of NOx reduction as function of bulk salt concentration in the ca-
talyst matrix.
35

The activity of Pt/γ-Al2O3 catalyst as evaluated with respect to the total oxidation
of ethyl acetate and a decrease in CO2 yield was found with increased exposure
time, Figure 14. Significant differences in catalyst activity was found at 300 °C
giving the highest deactivation level after 1000 h of exposure time (Paper IV).
Influences of the HDMS on the γ-Al2O3 support were also investigated. These
showed a substantial decrease of the support activity after 1000 h exposure.
225 250 275 300 325 350 375
Inlet Temperature, °C
CO2 Yield, (%)
Fresh Pt-Catalyst
350h, T op Frac .
650h, T op Frac
1000h, Top Frac.
295 320 345 370
Inlet Temperature, °C
CO2 Yield, (%)
0
10
20
30
40
50
60
70
80
90
100
60
65
70
75
80
85
90
95
100
Figure 14. Effect of HDMS exposure time on the total oxidation of ethyl acetate over the
aged inlet fraction of Pt/
γ
-Al2O3 catalyst.
The Pt/Rh steam reforming catalyst was evaluated for activity impact in the
steam reforming reaction involving methane and water as presented in Paper V.
It was shown that exposing the Pt/Rh catalyst to aerosol particles of K2SO4 de-
creases the catalyst activity. The conversion of methane decreased from 8 % to 5
% after particle exposure, indicating a negative effect on the catalyst activity by
particle deposition.
Influence of Poisoning on Physical Properties
of Catalysts
Catalyst samples were evaluated with respect to BET surface area, pore volume
and pore diameters as presented in Table 6. BET results show differences be-
tween the fresh and aged VOC catalysts as well as for Ni and Pt/Rh catalysts
used for steam reforming. No significant changes were detected for the fresh and
aged SCR catalysts.
Operation temperature may also influence the internal structures. The highest in-
ternal surface area is seen for the alumina supported Pt catalyst for VOC oxida-
tion with internal surface areas of 130 m2/g catalyst. The SCR catalyst shows an
internal surface area of 80 m2/g catalyst while both catalysts evaluated for steam
reforming only have internal surface areas of about 16 to 28 m2/g catalyst. The
36

37
supported alumina Pt catalyst also has larger pore volumes, 0.7 cm3/g catalyst
compared to 0.03 cm3/g catalyst for the SCR catalyst.
Table 6. BET surface area, pore volume and pore diameter of fresh and exposed catalysts.
Catalyst samples Surface area Pore volume Pore diameter
(m2/g) (cm3/g) (nm)
SCR catalyst
Fresh 81 0.03 2.14
Particle exposed
KCl 73 0.03 2.17
K2SO4 71 0.03 2.15
ZnCl2 78 0.03 2.02
Salt impregnated
KCl 90 0.04 2.07
K2SO4 82 0.03 2.08
ZnCl2 82 0.05 2.10
Commercial SCR samples
Fresh catalyst 64 0.04 2.09
Deactivated catalyst 67 0.03 2.14
VOC oxidation catalyst
Pt/
γ
-Al2O3 Catalyst
Fresh catalyst 124 0.77
350 h 118 0.63
650 h 119 0.61
1000 h 118 0.73
γ
-Al2O3 Support
Fresh 130 0.68
350 h 119 0.40
650 h 117 0.48
1000 h 113 0.78
Steam reforming catalyst
Ni Catalyst
Fresh 15
K2SO4 11
Ash-salt 10
Pt/Rh Catalyst
Fresh 16
K2SO4 12
Ash-salt 18
BET results can indicate differences in catalyst influences between different de-
activation routes by distinguishing between pore poison penetration. For the
comparison of the deactivation of SCR catalysts the catalyst samples prepared by
particle deposition showed slight decreases in BET surface area while the surface
area of the samples prepared by impregnation remained almost unchanged. Pore

38
sizes of the catalyst samples prepared by impregnation are also generally re-
duced. The commercially exposed samples show no impact on BET surface area
or pore diameters but a decrease in pore volume. BET evaluation for the SCR
samples was also carried out at different degassing conditions, which to some ex-
tent influenced the results (Papers II, III and VI).
Specific surface area and pore volume were evaluated for the Pt/γ-Al2O3 catalyst
and the γ-Al2O3 carrier (Paper IV). The surface area and pore volume for the
inlet fraction of the blank support and the platinum catalyst decreases somewhat
after long-time exposure to HMDS. The decrease is slightly more pronounced
for the blank support than for the platinum catalyst. It also seems that the surface
area of the blank support is diminished with increasing exposure time.
BET investigation of the catalysts for steam reforming was carried out on fresh
catalysts, catalysts deactivated by K2SO4 and biomass ash salts (Paper V). The
Ni catalyst showed loss of internal surface area after deactivation, possibly due
to particles covering the surface. The supported Pt/Rh catalyst also showed a
slight loss of area due to particle deposition.
Penetration Profile of Poisoning
SEM with an Electron Micro Probe Analyser (EMPA) also equipped with a en-
ergy dispersive spectrometer (EDS) was applied to investigate the penetration of
a compound into a catalyst structure. This technique was applied to the evalua-
tion of potassium and zinc penetration for the SCR catalyst as well as for the sili-
con penetration for the VOC oxidation catalyst. Both catalysts were analyzed for
a cross section of a catalyst wall and a catalyst pellet respectively.
Comparison of the penetration profiles of potassium and zinc for the SCR cata-
lyst samples prepared through particle deposition and impregnation is shown in
Figure 15, a and b. For verification, the commercially exposed SCR catalyst was
also analyzed for potassium penetration. No comparison was possible for zinc
due to lack of zinc accumulation during commercial exposure (Paper II).
The profiles of potassium and zinc show a difference in the behaviour between
the catalyst samples prepared by particle exposure and impregnation. With the
salt solution, catalyst pores might by filled with solution through capillary forces
distributing the solution into the pore structure. An evaluation of the samples
prepared by impregnation show a penetration profile of both potassium and zinc,
evenly distributed through the catalyst structure.
Particle exposure profiles indicate particles penetrating into the catalyst matrix
forming local high concentrations. For zinc the concentration is at a maximum
around 0.10 mm with an exponentially decreasing concentration profile.

0
0,5
1
1,5
2
2,5
3
0,0 0,1 0,2 0,3 0,4
Penetration Depth [mm]
Concentration
KCl particle depostion
K2SO4 particle deposition
KCl Impregnation
K2SO4 Impregnation
Commercial Exposure
0
0,1
0,2
0,3
0,4
0,5
0 0,1 0,2 0,3 0,4
Penetration Depth [mm]
Concentration
ZnCl2 Particle Deposition
ZnCl2 Impregnation
Figure 15.Penetration profiles of (a) potassium and (b) zinc across the catalyst wall.
For the VOC oxidation catalyst the silicon deposition was evaluated radially
across an individual pellet taken from the inlet fraction exposed to HDMS for
650 h (Paper IV). The concentration of silicon along the pellet diagonal as
shown in Figure 16 indicates a maximum concentration at the outer surface of
the pellet, which decreases toward the centre. After about 500 µm inside the
catalyst pellet the silicon concentration is almost zero showing that the silicon
has penetrated well beyond the active catalyst phase located approximately some
hundreds of micrometers into the catalyst pellet. The silicon forms an eggshell
coating of the outer surface of the pellets.
39

0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
0
270
540
811
1081
1351
1621
1892
2162
2432
2702
2972
µ m
c/s/nA
Figure 16. Penetration profile of silicon in an inlet fraction pellet deactivated for 650 h.
Chemical Characterisation by Chemisorption
Chemisorption was applied to the three catalysts in order to study: partition of
active sites blocked by poisoning substances for the V2O5-WO3/TiO2 catalyst;
metallic surface area, dispersion and average metal particle size for the steam re-
forming Pt/Rd and Ni catalysts; and available surface area for the Pt/γ-Al2O3.
Ammonia, H2 and CO was used as adsorption gases. X-ray powder diffraction
(XPRD) was also applied to the Pt/Rd and Ni catalysts for the evaluation of
metal crystallite sizes.
Ammonia chemisorption of the V2O5-WO3/TiO2 catalyst is discussed in Paper
VI. Ammonia TPD of particle exposed and salt impregnated samples showed
two peaks, one very small at about 350 °C and the other, much larger, at about
500 °C. The first may be due to physically adsorbed ammonia not completely
removed by previous treatment. The main peak at 500°C is related with the ac-
tive site of the catalyst, present as V(IV). Indeed, it shows a reverse correlation
with the amount deposited, thus indicating that the deposition affects the active
site of the catalysts.
The commercial exposed catalyst with higher vanadium content shows a peak at
650 °C. This peak could be attributed to crystalline vanadium oxide phase that
forms at high vanadium contents, above monolayer concentration, V(V), (Sanati
et al., 1990 and 1993).
40

Comparison of the catalyst activity vs. the available active sites (Figure 17)
showed that a substantial NO reduction can still be obtained for the catalyst sam-
ples deactivated by potassium particle deposition as compared to the samples de-
activated by impregnation. The amounts of available active sites are comparable
while the NO reduction is substantially lower for the impregnated samples com-
pared to the samples deactivated by particle deposition.
Evaluation of the poison penetration profiles (Figure 15) showed the salt parti-
cles to penetrate into the material while impregnation also acts on the outer cata-
lyst surface. The NO reaction takes place in the surface region (Tronconi et al.,
1992), which can explain the maintained catalyst activity for the catalyst samples
exposed to particle deposition. Further discussion is presented in Paper VI.
0
25
50
75
0 100 200 300 400
Desorbed ammonia [micromol/g cat]
NO reduction [%]
K Particle Exposed
K Impregnated
Zn Particle Exposed
Zn Impregnated
Commercial Exposure
Figure 17. Comparison of NO reduction as function of available active sites for the salt
exposed SCR catalyst.
CO chemisorption was applied in order to study the impact of silicon deposition
on platinum surfaces for the Pt/γ-Al2O3 catalyst (Paper IV). The evaluation
showed that nearly all of the platinum sites were blocked after exposure to
HDMS for 1000 h. The initial available platinum surface area of 14.4 decreased
to 0.12 dm2/g for the inlet catalyst fraction, leaving only 1 % of the initial plati-
num surface area available (Arnby et al., 2004). It was noted that the low number
of active sites still possessed high activity for ethyl acetate oxidation (Figure 14).
However, it may be that the active sites for total oxidation of VOC are not the
same as the sites taking part in CO chemisorption.
For the steam reforming catalyst evaluation H2 chemisorption was applied for
investigation of metallic surface area, average metal particle size and metal dis-
41

42
persion. X-Ray Powder Diffraction (XRPD) was used to determine the size of
the metal crystallites. The results are discussed in Paper V.
Commercial Ni catalyst exposed by K2SO4 and ash salt particles showed a
decrease in the metallic surface area measured by H2-chemisorption possibly due
to salt particles covering or fouling the surface. The crystallite size measured by
XRPD was not affected by the particle exposure.Supported Pt/Rh catalyst
average metal particle size and dispersion of the metal also showed an increase
in metal particle size and loss in dispersion when exposed to salt particles. Pt/Rh
catalyst crystallite size measured by XRPD remained unchanged. Changes in
both Ni and Pt/Rh catalyst parameters may be attributed to surface coverage of
salts particles.

43
Discussion
Deactivation of three catalysts used in different applications has in this work
been investigated. Deactivation has been carried out by using particle exposure,
gas exposure as well as impregnation by salt solutions of poisons. The basis for
the experiments has been to mimic commercial operation conditions under accel-
erated laboratory conditions. The phenomena of poison deposition and penetra-
tion is linked to the catalyst matrix and the condition and behaviour of the poison
with respect to different physical state (solid, liquid or gaseous) and apparent dif-
fusion characteristics in the investigated process flue gas containing the respec-
tive poison. The catalysts investigated have consisted of a homogeneous catalyst
material in the monolithic V2O5-WO3/TiO2 catalyst for abatement of NOx, a pel-
let eggshell platinum catalyst on an alumina support for the Pt/γ-Al2O3 catalyst
for total oxidation of VOC and supported nickel and platinum/rhodium catalysts
for the steam reforming process.
Interaction between poisons and catalyst matrix can act selectively or non-
selectively on the catalysts active sites. At selective deactivation the poison re-
acts preferably with the active sites whilst with a non-selective deactivation a
blocking or coating acts similarly on the active sites and support structure.
The penetration of a liquid poison into a catalyst pore structure and its deposition
and impact on catalyst performance may be explained by the following hypothe-
sis. At first the liquid fills the pore and after drying either a coating or scattered
layer forms blocking the catalyst internal surface area. Liquid penetration is not
foreseen in the commercial processes studied as the catalysts are applied at ele-
vated temperatures from 300 to 800 °C.
A gaseous catalyst poison will enter the catalyst pores by diffusion similarly to
the diffusion by reactant molecules into the catalyst matrix. Gaseous poison
penetration is determined by concentration gradients and diffusion coefficients,
which are influenced by the molecule size as well as temperature and pressure.
The gaseous poison is likely to gradually attach to the catalyst surface. Depend-
ing on the poison molecule size and the exposure residence time the poison
deposition may take place close to the surface or throughout the material. For
large poison molecules the penetration can be limited due to diffusion restric-
tions in the catalyst pores and deposition can be accumulated at the pore open-
ings. Smaller poison molecules that can diffuse unhindered into the pore struc-
ture can give depositions extending into the catalyst matrix probably similar to
impregnation with poisonous salt.

44
Particle deposition on the catalyst surface and penetration into the pore structure
may take place through diffusion of the micrometer-sized particle into the cata-
lyst matrix. The diffusion coefficients for particles are smaller than for gaseous
molecules assuming slower interaction between the particulate poison and the
catalyst. For the accelerated laboratory deposition method electrostatic forces can
be applied to induce enhanced particle deposition in the catalyst matrix. Particle
agglomerates may be formed at the pore mouth or further into the pore resulting
in blocking of the pores. The nature of the blocking can be physical and possibly
also chemical due to interaction with the catalyst material.
Evaluation of catalyst deactivation routes involves comparison of the physical
and chemical characteristics for fresh and exposed catalyst samples. Applying
different chemical and physical evaluations are necessary in order to distinguish
between deactivation routes related to commercial and laboratory catalyst sam-
ples.
It is of primary interest to validate the amount of poisonous material deposited in
the catalyst and it's influence on catalyst activity. A laboratory deactivation ex-
periment that does not create accumulation of poison or impact on catalyst activ-
ity cannot be used for evaluation of poison impact. Other researchers have exten-
sively used both impregnation and gaseous poisons exposure while particle
deposition exposure under laboratory conditions has not been studied. The exter-
nal mass transfer of the catalyst reactor configuration determines particle deposi-
tion as it is limited by diffusion forces. For a fix bed reactor with small catalyst
grains the particle deposition is sufficient, while for a monolithic reactor the par-
ticle deposition is a very slow process with only marginal deposition of gener-
ated particles. Utilizing an electrical field to enhance the deposition rate enabled
larger deposition amounts on the catalyst in laboratory scale experiments. The
investigated exposure methods in this work all resulted in acceptable poison ac-
cumulation as well as detectable impact on catalyst activity.
Investigation of BET surface area, pore volume and pore sizes can indicate poi-
soning mechanisms, for example coating of the pores or pore mouth blocking or
loss of internal surface area by sintering. The catalysts and deactivation routes
investigated showed some influence on BET results although dependencies were
not conclusive.
Evaluation of poison penetration profiles was valuable in detecting differences
between deactivation routes both for the SCR catalyst and the VOC catalyst.
Comparing impregnation and particle deposition for SCR with a commercially
exposed catalyst confirmed the resemblance between the particle deposited cata-
lyst and the commercially exposed catalyst showing a penetration of the submi-
crometer particles into the catalyst wall rather than deposition on the surface.
The eggshell silicon penetration profile obtained for the VOC catalyst was also
valuable for understanding the poisoning phenomena.

45
Chemisorption studies can also give information on the nature of catalyst poison-
ing. Evaluation of the nature of the active sites, the size and dispersion of metal
particles as well as blocking or loss of active sites is necessary in understanding
deactivation routes. Chemisorption, as applied to the catalyst evaluated, assisted
in the evaluation of the interactions.
Understanding the poisoning mechanisms and the impact on catalyst perform-
ance is the key to further optimize catalyst structures with respect to catalyst per-
formance and apparent lifetime. This will lead to more economical industrial
catalyst processes and further application of catalysts for small-scale units espe-
cially in biomass combustion.

46

47
Future recommended work
Understanding the influence of industrial process conditions on catalyst behav-
iour and performance is an important factor in successful application of a cata-
lytic system. The evaluation of possible interactions between the catalysts and
impurities present in the process flue gases can improve both investment and op-
erating costs by prolongation of the lifetimes of catalysts. A long-term effect of
lower investment and operating costs can result in more commercial installations
reducing overall emissions of NOx and VOC.
Neither of the investigations give a full picture of the mechanisms involved in
catalyst deactivation either by silicon gaseous exposure or by salt particle deposi-
tion and further investigation is needed to fully understand the phenomena in-
volved.
Particle deposition investigations regarding monolithic SCR catalysts showed
that it was possible to accelerate laboratory exposure to generated particles by
the application of electric field. The deposition method can be further improved
by controlled charging of the generated particles prior to deposition. Further
evaluation of the deposition method should also include investigations of deposi-
tion conditions such as temperature, gas velocity and residence time.
Investigation of the deactivation mechanism related to particle deposition in SCR
catalyst is of further interest. In our investigations only three salts of one particle
size distribution have been assessed. Further evaluation of particle sizes and
other inorganic salts can give insight into the deactivation mechanism. Under-
standing the nature of the interaction between the catalyst structure and the parti-
cles with respect to size and chemical nature can provide suggestions on how to
improve catalyst structure to avoid or detain catalyst deactivation.
Biomass combustion is also a process involving not only wood and peat combus-
tion but also other vegetable fuel fractions such as nuts, olive stones, garden
waste, straw, sludge or animal based fuel fractions such as bone and meat meal
or manure. These new fuel fractions contain other compounds that can be poi-
sonous to the SCR catalyst for example phosphor related compounds. Further
understanding of the fuel influence on catalyst deactivation is needed for future
application of environmental catalysts.
Other industrial processes are facing tightened emission regulations involving
NOx emissions. These processes are of combustion nature also involving high
concentrations of submicrometer particles of different sizes and compositions.

48
Such industrial processes can be kraft recovery boilers used to recover chemicals
in the pulp industry and sinter processes used in metallurgical industry.
An increasing usage of wood as raw material for energy production can extend
the influence of inorganic salts on catalytic processes. Gasification of biomass
with the intention to produce a synthetic gas for further processing into liquid fu-
els will involve different catalytic processes such as steam reforming, water gas
shift and fuel generation processes for dimethyl ether or synthetic diesel. The
catalysts may be exposed to aerosols of the alkali salts present after the gasifica-
tion process depending on process conditions and gas pre-treatment such as gas
filtration. Investigation of particle-induced deactivation has started with com-
mercial steam reforming catalysts and will be extended to other catalysts devel-
oped specially for the biomass gasification process. Further evaluation of the
particle influence on other catalyst processes will also be needed. The influence
of salt particle deposition will be of interest for the future commercial adaptation
of the overall manufacturing process of biomass generated liquid fuels.
Improvement of the VOC catalytic oxidation process also involves extension of
catalyst lifetime as this also reduces related process costs. Process development
can focus on either improving silicon resistance of the alumina supported plati-
num catalyst or by changing the catalyst composition for a more resistant active
component.
A broader application for VOC oxidation catalysts is also foreseen in the future.
In the application for new industrial processes other poisonous compounds may
be encountered which may lead to other improvements of the catalyst and the
support structure.
The above recommended work is important for the future as the use of environ-
mental catalysts will continue to grow in line with more stringent emission legis-
lation leading to improved air quality for us all.

49
Acknowledgment
I would like to express my gratitude to my supervisor, Prof. Mehri Sanati at the
School of Technology and Design/Department of Bioenergy for her continuous
support and guidance in the academic world and for never giving up in trying to
raise funds for our project. Also to my co-supervisors Prof. Erik Swietlicki at
Lund University and Dr. Jan Brandin of Catator AB for sharing all their experi-
ence and knowledge with me.
I greatly appreciate the support I have received from ALSTOM Power Sweden
AB and all the patience of my fellow colleagues there too. I would also like to
thank Växjö Energy AB, Perstorp AB and Cormetech Inc. for providing all the
catalyst samples.
I would like to thank all my colleagues at the Department of Bioenergy, espe-
cially Dr. Mikael Strand and Jessica Einvall for their full assistance and co-
operation. It has been a pleasure working with you all and I will always remem-
ber the good spirit within our department.
My time at the School of Technology and Design at Växjö University has been
very inspirational for me and I am grateful to all my colleagues at the University
for the interesting discussions and friendly atmosphere. My thanks go to all the
students from the department and from the external universities who have par-
ticipated with excellent performances in the laboratory.
I would also like to extend my thanks to Prof. Roland Wimmerstedt, my former
supervisor at Lund University, for all his encouragement over the decades and to
Lena Lillieblad for all our inspiring discussions.
And most of all my love and thanks go to my family. To my loving husband Ola,
with whom I have shared all thoughts and had so many discussions with on all
kinds of topics for more than 25 years. Thank you for inspiring me to pursue my
dreams and always believing in me! And to my sons, Joakim and Alexander -
you bring love and joy to my life and I hope I can make you proud. From now
on, no more late nights or work at weekends. I’m all yours!
The Swedish Research Council is gratefully acknowledged for the financial sup-
port of this project. The European Commission is also acknowledged for co-
financing the Framework-6 project related to this research: Integrated Project
"CHRISGAS" (contract SES6-CT-2004-502587).

50

51
Publications
Paper I
Deactivation of SCR Catalysts by Exposure to Aerosol Particles of Potassium
and Zinc Salts,
Ann-Charlotte Larsson, Jessica Einvall and Mehri Sanati,
Accepted for publication in Aerosol Science and Technology (Dec 2006)
The focus of the work was to generate aerosol particles for deposition in a mono-
lithic catalyst channel under laminar flow conditions. An electrostatic field was
applied to enhance deposition rates. Three model compounds of potassium and
zinc salts, KCl, K2SO4 and ZnCl2, were used to induce catalyst deactivation.
Comparison was made between the laboratory exposed catalyst samples and
catalyst samples obtained from a commercial biomass combustion boiler. It was
shown that the laboratory deposition method could produce accumulation of po-
tassium of the same orders of magnitude after 10 h of deposition compared to 6
500 h of commercial operation. The impact on catalyst activity was comparable
between the samples and the particle deposition also showed resembling mor-
phology. The laboratory deposition method utilizing an electrical field to en-
hance particle deposition can thus be further used in order to study catalyst deac-
tivation mechanisms induced by deposition of inorganic salts particles.
Paper II
Targeting by Comparison with Laboratory Experiments the SCR Catalyst Deac-
tivation Process by Potassium and Zinc Salts in a Large Scale Biomass Combus-
tion Boiler,
Ann-Charlotte Larsson, Jessica Einvall, Arne Andersson and Mehri Sanati,
Energy & Fuels (2006) 20, 1398-1405
Catalyst deactivation by potassium and zinc salts was investigated comparing
exposure by particle deposition; impregnation by salts solution and exposure by
commercial biomass combustion. Potassium and zinc salts, KCl, K2SO4 and
ZnCl2 were used as deactivating substances. The catalysts samples were com-
pared with respect to increased concentrations of potassium and zinc, penetration
behaviour of potassium and zinc into the catalyst wall, BET surface area, pore
volume and pore sizes as well as catalyst activity reductions. Both wet impregna-

52
tion and electrically enhanced particle deposition showed increases in potassium
concentrations comparable with commercial operation during 6 500 h. Catalyst
samples exposed to laboratory generated particles showed penetration profiles as
well as impact on physical structures of a resembling nature indicating pore
mouth blocking. Catalysts deactivated through solution impregnation showed
penetration profiles and physical structures pointing towards coating of the pore
structure with the poisoning compound. Wet impregnated samples also showed a
stronger influence of increased potassium concentration in the catalyst material
on catalyst activity as compared to particle deposition exposure and commercial
operation.
Paper III
Physical and Chemical Characterisation of Potassium Deactivation of an SCR
Catalyst for Biomass Combustion
Ann-Charlotte Larsson, Jessica Einvall, Arne Andersson and Mehri Sanati
Accepted for publication in a special issue of Topics of Catalysis (Dec 2006)
In order to investigate the catalyst deactivation mechanisms induced by exposure
to generated particles, wet impregnation and commercial biomass operation of
SCR catalysts the catalyst samples were characterised physically and chemically.
Good resemblance was found between characteristics of commercially exposed
catalysts and laboratory particle deposited ones, while wet impregnation samples
characteristics deviated in comparison.
Paper IV
Pilot-Scale Investigation of Pt/Alumina Catalysts Deactivation by Organosilicon
in the Total Oxidation of Hydrocarbons
Ann-Charlotte Larsson, Mohammad Rahmani, Karl Arnby, Morteza Sohrabi,
Magnus Skoglundh, Neil Cruise, and Mehri Sanati
Accepted for publication in a special issue of Topics of Catalysis (Dec 2006)
The work focused on deactivation of VOC oxidation catalysts by Si compound.
The influence of the long-term deactivation test by silica compounds on Pt/γ-
Al2O3 catalyst in total oxidation of ethyl acetate was studied. Experiments were
carried out in a pilot-scale tubular fixed bed reactor, with exposure times of 350,
650, and 1000 h. The surface area of active phase and radial silicon distribution
in individual catalysts sample was also investigated by using CO chemisorption
and microscopic studies. The performance of the fresh, aged catalyst samples for
both support and Pt-supported samples were measured in the total oxidation of
ethyl acetate. The most excessive deactivated catalyst was observed for 1000 h
of exposure to HMDS. The microscopy results for radial distribution of silica in
a 650 h exposed Pt/γ-Al2O3 samples indicate a maximum silicon concentration at
the outer surface of the pellet, which decreases toward the centre.

53
Paper V
Investigation of Reforming Catalyst Deactiviation by Exposure to Fly Ash from
Biomass Gasification in Laboratory Scale
Jessica Einvall, Simone Albertazzi, Christian Hulteberg, Francesco Basile, Ann-
Charlotte Larsson, Jan Brandin, Mehri Sanati
Submitted for publication in Energy & Fuels
Work has focused on deactivation of reforming catalysts of Ni and Pt/Rh type
during exposure to aerosol particles of potassium salts and biomass ash salt in
laboratory scale. The catalysts were characterised with respect to salt accumula-
tion, induced catalyst deactivation, physical characterisation and chemisorption.
Paper VI
Comparison of Deactivated Catalysts in Laboratory and Large Scale Biomass
Combustion Focussing on Potassium and Zinc Salts
Ann-Charlotte Larsson, Michael Strand, Simone Albertazzi, Francesco Basile
and Mehri Sanati
Submitted for publication in Applied Catalysis A: General
In order to investigate the catalyst deactivation mechanisms induced by exposure
to generated particles, wet impregnation and commercial biomass operation of
SCR catalysts; ammonia chemisorption by Temperature Programmed Desorption
(TPD) of ammonia was employed. The influence on deactivation was evaluated
with respect to partition of active vanadium sites (V(IV) and V(V)). The cata-
lysts were also characterised with respect to activity, poison penetration, BET
area, and accumulation of the exposed poison salts.
Authors Contribution to Presented Papers
Paper I
Author has been responsible for planning and execution of the experimental
work regarding aerosol deposition and catalyst activity testing and has been re-
sponsible for the evaluation of the data. Author has written the manuscript.
Paper II
Author has been responsible for planning and execution of the experimental
work regarding catalyst exposure and activity testing and has been responsible
for the evaluation of the data. Author has written the manuscript.

54
Paper III
Author has been responsible for planning and execution of the experimental
work and has been responsible for the evaluation of the results. Author has writ-
ten the manuscript.
Paper IV
Author has taken part in the evaluation of the experimental data. Author has been
responsible for writing major parts of the manuscript.
Paper V
Author has participated in planning and development of the deactivation experi-
ments as well as the evaluation of obtained results.
Paper VI
Author has been responsible for planning and execution of the experimental
work and has been responsible for the evaluation of the results. Author has writ-
ten the manuscript.

55
References
Albertazzi, S., Basile, F., Brandin, J., Einvall, J., Hulteberg, C., Fornasari, G., Rosetti, V.,
Sanati, M., Trifirò, F., Vaccari, A., (2005). The technical feasibility of biomass gasifi-
cation for hydrogen production, Cat. Today 106; 297-300
Arnby, K., (2004). Low-temperature oxidation of CO and volatile organic compounds
over supported platinum catalysts, Doctorial Dissertation. Chalmers University of
Technology, Göteborg, Sweden, ISBN 91-7291-541-2
Arnby, K., Rahmani, M., Sanati, M., Cruise, N., Ambertsson Carlsson, A., Skoglundh, M.,
(2004). Characterisation of Pt/γ-Al2O3 Catalysts Deactivated by Hexamethyldisilox-
ane, Appl. Catal. B Env. 54: 1-7
Arnby, K., Törncrona, A., Andersson, B., Skoglundh, M., (2004). Investigation of Pt/γ-
Al2O3 Catalyst with Locally High Pt Concentrations of Oxidation of CO at Low Tem-
peratures, J. Catal. 221: 252-261
Bartholomew, C. H., (2001). Mechanisms of Catalyst Deactivation, Appl. Catal. A Gen.
212: 17-60
Beck, J., Brandenstein, J., Unterberger, S., Hein, K. R. G., (2004) Effects of sewage
sludge and meat and bone meal Co-combustion on SCR catalysts, Appl. Catal. B:
Env. 49: 15-25
Beck, J., Müller, R., Brandenstein, J., Matscheko, B., Matschke, J., Unterberger, S., Hein,
K. R. G., (2005) The behaviour of phosphorus in flue gases from coal and secondary
fuel co-combustion, Fuel 84: 1911-1919
Butt, J. B., (1999). Reaction Kinetics and Reactor Design, 2nd ed., New York: Marcel
Dekker Inc.
Chen, J. P., Buzanowski, M. A., Yang, R. T., Cichanowicz, J. E., (1990). Deactivation of
the Vanadia Catalyst in the Selective Catalytic Reduction Process, J. Air Waste Man-
age. Assoc. 40: 1403-1409
Chen, J., Heck, R.M., Farrauto, R.J., (1992). Deactivation regeneration and poison - resis-
tant catalysts: commercial experience in stationary pollution abatement, Catal. Today
11: 517-545
Chen, J. P., Yang, R. T., (1990). Mechanism of Poisoning of the V2O5/TiO2 catalyst for
the reduction of NO by NH3, J. Catal. 125: 411-420

56
Colin, L., Cassuto, A., Ehrhardt, J.J., Ruiz-Lopez, M.F., Jamois, D., (1996). Adsorption
and decomposition of hexamethyldisiloxane on platinum: an XPS, UPS and TDS
study, Appl. Surf. Sci. 99: 245-254
Cullis, C. F., Willatt, B. M., (1984). The inhibition of hydrocarbon oxidation over su-
ported precious metal catalysts, J. Catal. 86: 187-200
Cussler, E. L., (1984). Diffusion, Mass transfer in fluid systems, Cambridge University
Press
Ehrhardt, J.J., Colin, L., Jamois, D., (1997). Poisoning of platinum surfaces by hexame-
thyldisiloxane (HMDS): application to catalytic methane sensors, Sens. Actuators B
40: 117-124
Ertl, G., Knözinger, H., Weitkamp, J., (1997). Handbook of Heterogeneous Catalysis, Vol.
4,Weinheim, VCH
Fogler, H. S.,(1992) Elements of Chemical Reaction Engineering, New Jersey: Prentice-
Hall Inc.
Forzatti, P., (2001). Present Status and Perspectives in de-NOx SCR Catalysis, Appl.
Catal. A: Gen. 222: 221-236
Forzatti, P., Lietti, L., (1999). Catalyst Deactivation, Catal. Today 52: 165-181
Gentry, S.J., Jones, A., (1978). Poisoning and Inhibition of Catalytic Oxidations: The Ef-
fect of Silicon Vapour on the Gas-Phase Oxidations of Methane, Propene, Carbon
Monoxide and Hydrogen over Platinum and Palladium Catalysts, J. Appl. Chem. Bio-
technol. 28; 727-732
Hayes, R. E., Kolaczkowski, S. T., (1994). Mass and heat transfer effects in catalytic
monolith reactors, Chem. Eng. Sci. 21: 3587-3599
Hegedus, L. L., McCabe, R.W., (1984). Catalyst Poisoning, New York: Marcel Dekker
Inc.
Herrlander, B., (1990). SCR DeNox at the Munich South Waste Incinerator, Air Waste
Manage. Assoc., 83 rd Annual Meeting & Exhibition Pittsburg, June 24-29, Pennsyl-
vania, USA.
Hinds, W. C., (1999). Aerosol Technology: Properties, Behavior, and Measurements of
Airborne Particles, New York: John Wiley & Sons Inc.
Hinz, A., Larsson, P-O., Skårman, B., Andersson, A., (2001). Platinum of Alumina, Tita-
nia, and Magnesium Supports for the Combustion of Methanol in a Waste Gas with
Trace Amounts of Ammonia, Appl. Catal. B Env. 34: 161-178
Holmgren, A., Andersson, B., (1998). Oxygen Storage Dynamics in Pt/CeO2/Al2O3 Cata-
lysts, J. Catal., 178: 14-25

57
Hums, E., (1998). Is advanced SCR technology at a standstill? A provocation for the aca-
demic community and catalyst manufacturers, Cat. Today 42; 25-35
Kamata, H., Takahashi, K., Odenbrand C. U. I., (1999). The role of K2O in the selective
reduction of NO with NH3 over a V2O5 (WO3)/TiO2 commercial selective catalytic
reduction catalyst, J. Mol. Catal. 139: 189-198
Kim, S. H., Woo, K. S., Liu, B. Y. H., Zachariah, M. R., (2005). Method of measuring
charge distribution of nanosized aerosols, J. of Coll. Interf. Sci., 282: 46-57
Khodayari, R., (2001). Selective Catalytic reduction of NOx: Deactivation and regenera-
tion Studies and Kinetic Modelling of Deactivation, Doctorial Dissertation. Lund In-
stitute of Technology, Lund, Sweden, ISBN 91-7874-122-X
Khodayari, R., Andersson, C., Odenbrand, C. U. I., Andersson, L. H., (2000). Proceedings
of the 5th European Conference on Industrial Furnaces and Boilers, vol II, A.Reis,
J.Ward, W. Leuckel, R.Collin, 11-14 April 2000, Espinho-Porto-Portugal.
Khodayari, R., Odenbrand, C. U. I., (1999). Selective Catalytic Reduction of NOx: A
Mathematical Model for Poison Accumulation and Conversion Performance, Chem.
Eng. Sci. 54: 1775-1785
Khodayari, R., Odenbrand, C. U. I., (2001). Regeneration of Commercial SCR Catalysts
by Washing and Sulphation: Effect of Sulphate Groups on Activity, Appl. Cat. B Env.
3: 277-291
Kling, Å., Andersson, C., Myringer, Å., Eskilsson, D., Järås, S. G., (2006). Alkali deacti-
vation of high-dust SCR catalysts used for NOx reduction exposed to flue gas from
100 MW-scale biofuel and peat fired boilers: Influence of flue gas composition, Appl.
Cat. B Env. 69; 240-251
Libanati, C., Ullenius, D.A., Pereira C.J., (1998). Silica deactivation of bead VOC cata-
lysts, Appl. Catal. B Env. 15: 21-28
Lighty, J. S., Veranth, J. M., Sarofim, A. F., (2000). Combustion Aerosols: Factors Gov-
erning Their Size and Composition and Implications to Human Health, J. Air Waste
Manage. Assoc.50: 1565-1618
Lisi, L., Lasorella, G., Malloggi, S., Russo, G., (2004). Single and combined deactivation
effect of alkali metals and HCl on commercial SCR catalysts, Appl. Catal. 50: 251-
258
Matsumiya, M., Shin, W., Qiu, F., Izu, N., Matsubara, I., Murayama, N., (2003). Poison-
ing of platinum thin film catalyst by hexamethyldisiloxane (HMDS) for thermoelec-
tric hydrogen gas sensor, Sens. Actuators B 96: 516-522
Moradi, F., Brandin, J., Sohrabi, M., Faghihi, M., Sanati, M., (2003). 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, Appl. Catal. 46: 65-76

58
Morbidelli, M., Ravriilidis, A., Varma, A., (2001). Catalyst Design: Optimal Distribution
of Catalyst in Pellets, Reactors and Membranes, Cambridge University Press
Moulijn, J. A., van Diepen, A. E., Kapteijn, F., (2001). Catalyst deactivation: is it predict-
able? What to do?, Appl. Cat. A Gen. 212; 3-16
Murzin, D., Salmi, T., (2005). Catalytic Kinetics, Elsevier, Amsterdam, The Netherlands
Neyestanaki, A.K., Klingstedt, F., Salmi, T., Murzin, D.Y., (2004). Deactivation of post-
combustion catalysts, a review, Fuel 83: 395-408
Pagels, J., Strand, M., Rissler, J., Szpila, A., Gudmundsson, A., Bohgard, M., Lillieblad,
L., Sanati, M., Swietlicki, E., (2003). Characteristics of aerosol particles formed dur-
ing grate combustion of moist forest residue, J. Aerosol Sci. 34: 1043-1059
Pagels, J., Gudmundsson, A., Gustavsson, E., Asking, L., Bohgard, M., (2005). Evaluation
of Aerodynamic Particle Sizer and Electrical Low-Pressure Impactor for Unimodal
and Bimodal Mass-weighted Size Distributions, Aerosol Sci. Technol. 39: 871-887
Petersen, E. E., Bell, A. T., (1987). Catalyst Deactivation, New York, Marcel Dekker, Inc.
Rahmani, M., Badii, K., Faghihi, M., Sanati, M., Cruise, N., Augustsson, O., Spivey, J.J.,
(2004). Deactivation of Oxidation Catalysts for VOC Abatement by Si and P Com-
pounds, in J.J. Spivey, G.W. Roberts (Ed.), Catalysis, Specialist Periodical Reports,
Vol. 17. The Royal Society of Chemistry, Cambridge, 210
Rostrup-Nielsen, J. R., (1975). Steam reforming catalysts, Teknisk Forlag A/S, Copenha-
gen
Sanati, M., Andersson, A., (1990). Ammoxidation of Toluene over TiO2(B)-supported
Vanadium Oxide Catalysts, J. Mol. Cat. 59: 233-255
Sanati, M., Andersson, A., (1993). DRIFT study of the oxidation and ammoxidation of
toluene over TiO2(B)-supported vanadia catalysts, J. Mol. Cat. 81: 51-62
Satterfield, C. N., (1996). Heterogeneous Catalysis in Industrial Practice, 2 ed., New
York, Mc Graw–Hill Inc.
Sawyer, J.E., Abraham, M.A., (1994). Reaction Pathways during the Oxidation of Ethyl
Acetate on a Platinum/Alumina Catalyst, Ind. Eng. Chem. Res. 33: 2084-2089
Skoglundh, M., Johansson, H., Löwendahl, L., Jansson, K., Dahl, L., Hirschaouer, B.,
(1996). Cobolt-promoted Palladium as a Three-way Catalyst, Appl. Catal. B Env. 7:
299-319
Spivey, J. J., Butt, J. B., (1992). Literature review: deactivation of catalysts in the oxida-
tion of volatile organic compounds, Catal. Today 11: 465-500
Strand, M., (2004). Particle Formation and Emission in Moving Grate Boilers Operating
on Woddy Biofuels, Doctorial Dissertation. Växjö University, Växjö, Sweden, ISBN
91-7636-430-5

59
Strand, M., Pagels, J., Szpila, A., Gudmundsson, A., Swietlicki, E., Bohgard, M., Sanati,
M., (2002). Fly Ash Penetration through Electrostatic Precipitator and Flue Gas Con-
densor in a 6 MW Biomass Fired Boiler, Energy Fuels, 16: 1499-1506
Strand, M., Bohgard, M., Swietlicki, E., Gharibi, A., Sanati, M., (2004). Laboratory and
Field Test of a Sampling Method for Characterization of Combustion Aerosols at
High Temperatures, Aerosol Sci. Technol. 38: 757-765
Treybal, R. E., (1980). Mass-transfer operations, Mc Graw-Hill Inc.
Tronconi, E., Forzatti, P., Gomez Martin, J. P., Malloggi, S., (1992). Selective catalytic
removal of NOx: A mathematical model for design of catalyst and reactor, Chem.
Eng. Sci. 47: 2401
Topsoe, N. Y., Dumesic, J. A., Topsoe, H., (1995). Vanadia/Titania Catalysts for Selective
Catalytic Reduction of Nitric Oxide by Ammonia, II. Studies of Active Sites and
Formulation of Catalytic Cycles, J. Catal. 151; 241-252
Zheng, Y., Jensen, A. D., Johnsson, J. E., (2004). Laboratory Investigation of Selective
Catalytic Reduction Catalysts: Deactivation by Potassium Compounds and Catalyst
Regeneration, Ind. Eng. Chem. Res. 43: 941-947
Zheng, Y., Jensen, A. D., Johnsson, J. E., (2005). Deactivation of V2O5-WO3-TiO2 SCR
Catalyst at a Biomass-Fired Combined Heat and Power Plant, Appl. Catal. B Env. 60:
253-264
Wierzbicka, A., Lillieblad, L., Pagels, J., Strand, M., Gudmundsson, A., Gharibi, A.,
Swietlicki, E., Sanati, M., Bohgard, M., (2005). Particle Emissions from District Heat-
ing Units Operating on Three Commonly Used Biofuels, Atmospheric Environment
39: 139-150
Willeke, K., Baron, P. A., (1993). Aerosol Measurement; Principles, Techniques, and Ap-
plications, New York: John Wiley & Sons Inc.