You are on page 1of 23

Seminar report

On

Catalyst Deactivation

Submitted by
Akash Solanki
(U16CH082)

Under the supervision of


Dr. Mausumi Mukhopadhyay

CHEMICAL ENGINEERING DEPARTMENT

SARDAR VALLABHBHAI NATIONAL INSTITUTE OF


TECHNOLOGY, SURAT-395007
2019-20

1
Certificate

This is to certify that the B. tech IV (7th semester) seminar report entitled ‘Catalyst
Deactivation’ has been presented and submitted by Mr. Akash Solanki (U16CH082) of
Chemical Engineering Department in the partial fulfilment of the requirement for the award of
B.tech degree in Chemical Engineering at Sardar Vallabhbhai National Institute of Technology,
Surat.

They have successfully and satisfactorily finished their project in all regards. We certify that
their work is comprehensive, complete and eligible for evaluation.

Dr. Mausumi Mukhopadhyay

Professor

Department of Chemical Engineering,

SVNIT, Surat

Seminar Examiners:

Examiner 1:

Examiner 2:

Examiner 3:

2
Acknowledgment

Little achievements often require long, tortuous effort and bitter experiences including some
sacrifices. And this is only possible when the almighty GOD keeps his handful of blessings on
the head of anybody.

I would like to express my deep gratitude to my guide Dr. Mausumi Mukhopadhyay, Professor,
department of chemical Engineering, Sardar Vallabhbhai National Institute of Technology,
Surat for encouraging me as well as providing me all the necessary guidance and inspirational
support throughout this seminar work.

Furthermore, I would also like to acknowledge with much appreciation the department of
Chemical Engineering, Sardar Vallabhbhai National Institute of Technology, Surat, for
providing platform to present our report.

Thanking you,
Yours Sincerely,

Akash Solanki
(U16CH082)

3
Abstract

The objective of this report was an overview of deactivation of catalyst, their causes and
consequences as well as methods and techniques of investigation of deactivation. Any process,
chemical or physical, which decreases the activity of catalytic surface we called it deactivation
of catalyst. There are three main reasons of catalyst deactivation, i.e. poisoning, coking or
fouling and ageing. Poisoning can be reversible or irreversible, and with geometry or electronic
effect. It can also be selective, non-selective and antiselective, depending on catalyst kinetics
and affinity. The action of coke is by direct active sites coverage, or by pore plugging, and
depends on reaction,conditions, pore's type and catalyst acidity. Kinetics of coking is
determined by the mechanism of the coking reaction and its diffusion restriction. Sintering is
the main cause for catalyst ageing. It appears in two forms: chemical and thermal, depending
on prevailing reaction parameters, i.e., temperature or concentration. To cope with deactivation
two approaches are offered: either to avoid it when possible,,like in the case of feed
purification, or accept it but with an effort to minimize its effects. Accelerated deactivation
tests can be a great tools for studying catalyst deactivation in a relatively short time. By proper
selection of reaction parameters and applying deactivation compensation approach, reaction
and deactivation kinetics can be separated. Based on obtained deactivation kinetics parameters,
and by applying appropriate modeling and,simulation, the life time of a catalyst and its
performance in the commercial reactor at any time can be predicted. The status of knowledge
and needs for further work are also summarized for each type of deactivation mechanism. The
development during the past two decades of more sophisticated surface spectroscopies and
powerful computer technologies provides,opportunities for obtaining substantially better
understanding of deactivation mechanisms and building this understanding into comprehensive
mathematical models that will enable more effective design and optimization processes
involving deactivating catalysts.

4
Contents

Certificate ................................................................................................................................... 2
Acknowledgment ....................................................................................................................... 3
Abstract ...................................................................................................................................... 4
Chapter 1: Introduction .............................................................................................................. 6
Chapter 2: Deactivation Mechanisms ........................................................................................ 7
2.1 Deactivation by Poisoning ............................................................................................... 7
1.1 Deactivation by Sintering or Ageing ......................................................................... 10
2.2 Deactivation by Coking or Fouling................................................................................ 13
Chapter 3: Catalyst Deactivation in Important Industrial Processes ....................................... 15
3.1 Fluid Catalytic Cracking ................................................................................................ 15
3.2 Reduction of NOX in Exhaust Gases from Automotive Engines ................................... 16
3.3 Ethylene Epoxidation ..................................................................................................... 18
Chapter 4: Prevention of Catalyst Deactivation and Regeneration of Deactivated Catalysts . 19
Conclusion ............................................................................................................................... 22
References ................................................................................................................................ 23

5
Chapter 1: Introduction

Catalyst deactivation, the loss over time of catalytic activity and selectivity, is a problem of
great and continuing concern in the practice of industrial catalytic processes. Costs to industry
for catalyst replacement and process shutdown. Time scales for catalyst deactivation vary
considerably, for example, in the case of cracking catalysts, catalyst mortality may be in the
order of seconds, while in ammonia synthesis the iron catalyst may last for,5-10 years. But it
is inevitable that all catalysts will decay. During the lifetime catalysts undergo usually
deactivation. This happened in the time frame of seconds (fluid catalytic cracking on zeolites),
or years,(iron catalyst in ammonia synthesis). This phenomenon may significantly contribute
to the economy of catalytic process. The economic impact of catalyst deactivation expressed
as revenue loss is, in general not due to the cost of the catalyst itself, but to the lower production
rate as a result of the decrease in space time yield of the desired product with time on stream,
and intermittent shut-down in order to load the new catalyst or to,regenerate the deactivated
one. Moreover, additional costs may arise from poorer utilisation of a feedstock due to a
decrease in selectivity and from additional energy required for regeneration of catalysts.

Prevention of catalyst degradation poses substantial challenges in the design and operation of
a large-scale, catalytic process. There are many paths for catalyst decay. For example, a catalyst
may be poisoned by any one of a dozen contaminants present in the feed; its surface, pores and
voids may be fouled by carbon or coke produced by,cracking/condensation reactions of
hydrocarbon reactants, intermediates and/or products. In the treatment of a power plant flue
gas, the catalyst can be dusted or eroded by and/or plugged with fly ash. Catalytic converters
used to reduce emissions from gasoline or diesel engines may be poisoned or fouled by fuel or
lubricant additives and/or engine corrosion products. If the catalytic reaction is conducted at
high temperatures, thermal degradation may occur in the form of active phase crystallite
growth, collapse of the carrier,(support) pore structure and/or solid-state reactions of the active
phase with the carrier or pro motors. In addition, the presence of oxygen or chlorine in the feed
gas can lead to formation of volatile oxides or chlorides of the active phase followed by gas
phase transport from the reactor. Similarly,,changes in the oxidation state of the active catalytic
phase can be induced by the presence of reactive gases in the feed.

6
Chapter 2: Deactivation Mechanisms

Basically catalyst deactivation is a temporal or permanent loss of active sites, caused by


chemical and physical reasons. Deactivation occurs by: (i) chemical poisoning due to
chemisorption or reaction of a certain substrate on the surface, like H2S on Pt in hydrogenation
reactions; (ii) cooking as a result of coke deposition due to,hydrocarbon decomposition to
hydrogen-poor compounds, i.e. coke; (iii) fouling as a result of solids deposition due to dusty
materials within the feed: (iv) sintering and crystallization or segregation of the catalytic
material due to thermal effects,[1]. Although these are individual effects, it is possible that
several deactivating processes take place simultaneously, or that the dominating,deactivation
process can be initiated by another one.

2.1 Deactivation by Poisoning


Catalysts deactivation by chemical adsorption of some substance is known as poisoning. A
poison is a molecule of reactant, product or impurity in the feed with a certain affinity towards
the catalyst, the magnitude of interaction,defines the character of reversible or irreversible
poisoning. When the affinity of a poison is equal to all,active sites than this is uniform
poisoning, where all active sites have the same chance of being poisoned. In the opposite case,
when a poison preferentially chemisorbed on one category of active sites,,selective poisoning
occurs.

A poison may act by blocking the active sites - geometric effect or by changing their electronic
properties - electronic effect. In the first case the adsorption of the poison is often arranged by
some geometric rule, which is defined by both the catalyst and the poison. Sometimes for the
same combination of catalyst and poison different geometrical arrangements are possible, as it
is the case of the well-known example of Pt (100) poisoned by sulphur.[2]. Sulphur is adsorbed
on Pt in two different geometric arrangements, as,c(2x2) and p(2x2), corresponding to one half,
and one-quarter of a monolayer of S at full coverage of active sites, respectively (Figure 1). At
first glance the second geometry is preferential from the viewpoint of catalyst activity, since
one vacant adsorption site exist at the centre of,p(2x2). For bimolecular reactions, however,
being the majority of processes where Pt catalysts are used the possibility of an isolated
molecule adsorption is useless. Therefore, the c(2x2) geometric arrangement is beneficial from
a practical viewpoint,,since for a certain activity level the catalyst is capable of accepting more
sulphur.

7
(a) (b)

Figure 1: Two geometric structures of S adsorption on Pt (100): (a) c(2x2) and (b) p(2×2)

Due to the exchange of electrons for strong chemisorption, the poison can change the electronic
properties of the,catalyst. Electronic and geometric effects may be coupled, however, as is the
case of poisoning of Ni-catalysts by S, Cl, C, N and P,[3]. These poisons block the active sites
on which they are adsorbed, but they also decrease the adsorption abilities of neighboring active
sites for adsorption of reactants by electronic interaction. The poisoning effect of S and Cl is
more pronounced due to their higher electronegativity and size relative to C, N and,P.

For heterogeneous catalytic surfaces, the most common case is selective,poisoning. Due to the
different energy and geometry of active sites different mechanisms of poisoning occur. The
poisoning usually starts with preferential adsorption on sites with higher affinity towards the
poison, and once the adsorption reaches the point of saturation, it is followed by the additional
adsorption of poison over the rest of the,sites. In the case of bifunctional catalysts, poisoning
may affect their activity and/or selectivity,,depending on the reaction mechanism. For
illustration, the reaction of methyl-cyclopropane (MCP) hydrogenolysis on Pt/Al2O3, catalyst
is possible through two main routes, involving (1) metallic sites only, and (ii) both acidic and
metallic sites. In the case of unpromoted catalysts the first mechanism prevails, leading to high
selectivity to i-butane. When chlorine is added to the reaction mixture, however, only metallic
sites are selectively poisoned, resulting in a minor activity decrease, but dramatic selectivity
changes,[4]. As shown in Figure 2 the extent and profile of these changes depend on the
precursor of the poison, CCl4 or HCl, used to introduce the chlorine. A quite different result of
poisoning is obtained for a reaction with a consecutive mechanism, like isomerisation of n-
hexane. Here, poisoning of metallic sites would rather be expressed as an activity drop, than a
selectivity change; due to the absence of a dehydrogenation function, olefin production is
stopped, and straight isomerisation of paraffin on acidic sites is quite difficult when these are
of moderate acidity,[5].

8
Figure 2: Influence of addition of 0.82% Cl from different precursors, on selectivity of 0.31%

Pt/Al2O3, catalyst (47.6 % Pt dispersion) in MCP hydrogenolysis reaction [20]

The rate of catalyst deactivation may depend on the size of the catalytic sites, which means that
deactivation can be structure sensitive,[6]. The effect sometimes interferes with the structure
sensitivity of the catalytic reaction, making it difficult to decouple these two phenomena. It
was found, e.g., that sympathetic behaviour of the n-hexane reforming reaction on Pt/SiO2 is
rather due to the higher rate of coke deposition on larger particles (deactivation reaction) than
to the catalytic reaction turnover frequency,increase due to dispersivity increase. Thus, when
the catalytic reaction and the deactivation reaction belong to different types of structure
sensitivity, the effect of catalyst dispersivity on the catalytic reaction is masked by marked
structure sensitivity of,deactivation. Even a more dramatic effect occurs in the case of a
hydrogenolysis reaction, where poisoning by coke can even change the original type of
structure sensitivity of the catalytic reaction,[7]. Since this type of reaction involves destruction
of one or several C-C bonds, multi bonding of a reactant onto the catalyst surface is the most
appropriate,mechanism. Accordingly, an antipathetic behaviour of the catalyst is expected,
since the particle size increase will result in high-coordinated plain sites close to each other
being favourable for this kind of hydrocarbon,adsorption. However, the opposite was noticed
during time on stream, showing a higher rate increase with smaller particles, 1e particles having
a larger number of lower coordinated active,sites. This might be the result of preferential
poisoning of sites with high coordination number, and the reaction is forced to "switch from

9
sites on the plain (poisoned) to those which sit on edges of crystals (still unpoisoned). This
phenomenon is known as secondary structure,sensitivity.

Structure sensitivity of the deactivation reaction depends on catalyst support as,well. In the
case of two Pt-based catalysts with the same metal loading and dispersion and having the same
history of preparation and usage, the type of poisoning by CO was found to be different for
alumina and silica,supports. In both cases structure sensitive poisoning takes place, however
its patterns are different (Figure 3).,The uniform poisoning in the case of,alumina, in contrast
to the selective poisoning observed for silica [9], was explained by stronger metal-support
interaction for alumina, leading to more stable Pt particles with constant,deactivating
behaviour.

Figure 3: Structure sensitivity patterns of Pt/SiO2 (a) [25], and Pt/Al2O3 (b) [24] catalysts, having the same CO coverage of 40%

1.1 Deactivation by Sintering or Ageing


Catalyst ageing is due to slow processes of physical and chemical solid-state transformations
as a result of the catalyst being exposed to the reaction,environment. These transformations can
proceed without or with a change of catalyst overall composition,[10]. Sintering of the active
phase and/or support, segregation and transformation of phases, and all kinds of metal-support
interactions are frequent phenomena belonging to the first group, while changes of oxidation
state, formulation of carbide, and loss of catalyst component are those resulting in a change of
catalyst composition. While one of these processes may dominate under specific conditions,
more often they occur simultaneously.

Sintering is an important and frequent reason for catalyst,deactivation. In metal supported


catalysts it is usually manifested either as loss of active metal area, or decrease of catalyst
support,area. It includes, however, several other phenomena like dissociation of metal atom,

10
diffusion of metal atoms and crystals, spreading of particles, nucleation of particles,
coalescence, etc.,[12]. Depending on its cause thermal and chemical sintering are known. In
the first case a decrease of surface area occurs simply by particle enlargement due to high
reaction,temperatures. Thermal sintering is strongly influenced by the height of temperature,
and it is believed that it has to be at least as high as the Tamman temperature for the phenomena
to,occur. But it might not be a general case since the real mechanism of sintering is not known.
For example, if diffusion of atoms through the particle volume is the mechanism of sintering,
then the Tamman temperature probably must be reached for sintering to,occur. But if surface
diffusion of atoms is the mechanism,,than the Hüttig temperature (one third of the melting
temperature) is probably enough. Even assuming high temperature as the main cause of thermal
sintering other factors should not be excluded. These are the size and the shape of particles
surface roughness, metal-support interaction, the presence of promoters and the gaseous
environment. The combination of factors results sometimes in an unexpected behaviour, like
in the case of the presence of different gasses, which might influence sintering depending on
the particular metal and conditions applied [12]. Time is also an important variable since
sintering, like all other solid-state transformations, is a time consuming process. In the case of
zeolite based catalysts, sintering of the active metal depends on the preparation method, metal
loading, zeolite framework, and thermal conditions determining the position of the metal in the
zeolite structure. In Pt-exchanged Y zeolites the chance of Pt sintering depends on calcination
temperature, determining either particle migration to more stable position, or their settling in
the large,cavities. In the reduction following preparation, these particles which missed
stabilization in the previous calcination step, migrate and agglomerate on the external surface
of the zeolite,[13].

Basically, two models have been proposed to explain the growth of supported metal particles:
a) particle/crystallite migration, and atoms migration,(Figure 4). The basis of the first one is
the assumption that week metal-support interaction and a high temperature lead to migration
of crystallites which are in a quasi-liquid,state. On their way the crystallite colloid and
agglomerate. The atoms-migration model assumes that movement of atoms is important
for,sintering. The process can be assumed to occur in three stages transport of atoms from the
crystallite to the surface of the support their migration over the support and finally their
collision with bigger,crystals. The two models do not exclude each other, but most probably
occur in sequence, i.e., at lower temperatures sintering starts by migration of atoms and once
the higher temperature is reached movement of particles occur..At very high temperature, when

11
particles are of bigger size volatilization of atoms is the prevailing mechanism responsible for
transport and sintering [14]. Apart from these traditional models, an additional mechanism of
sintering has been proposed including wetting and spreading. This has been portrayed as a
consequence of the formation of films between deposited metal particles in and subsequent
rupture of this film in,H2. The explanation for this lies in the different strength of metal-
substrate interactions in different atmospheres, and the driving force for both phenomena is the
decrease of free energy of the,system.

Figure 4: Two models for crystallite growth due to sintering: Movement of atoms by migration (A), or by volatilization (A'), and
migration of particles (B)

Chemical sintering is caused by a chemical reaction between an active metal and a molecule in
the feed, resulting in a new compound forming particles of larger,size. The subsequent reverse
reaction will result in the active component again, but the size of the particle may stay
unchanged. Similarly, in alternating conditions of oxidation-reduction,reactions, active,metals
are forced to change from one compound to another usually having particles of different size.
These frequent and fast rearrangements of structure in the metal lattice cannot be followed by
all atoms, therefore they occupy thermodynamically more stable positions resulting in
enlarged,particles.

As already mentioned there are several other catalyst deactivation phenomena belonging to the
category of ageing, and most of them are caused by high temperature.,Phase transformation of
the often used active γ–Al2O3 support to its inactive c-phase is, for example, very common in
catalysts exposed to high temperature for extended periods of time.,The process is usually
promoted by an active metal sitting on the support, like in the case of a Ni/Al2O3, catalyst where
migration of Ni and formation of Ni-nucleation sites play a decisive role for phase
transformation of the,support. In addition, alumina often undergoes reactions with M2+ -type
active metals, forming spinel-like structures, like CoAl2O4. NiAl2O4,,etc. Both the
transformation of atomic to graphitic carbon, and active-carbon-rich carbides to inactive

12
carbon-poor carbides, occur often in Fe- based catalysts used in the Fischer-Tropsch process,
these are examples of catalyst composition changes due t, reaction. Sublimation of Mo from a
CoO-MoO3-Al2O3 HDS catalyst is a typical example for a loss of a catalytic active,component.
For zeolite exposed to severe hydrothermal conditions, dealumination is a frequent reason for
their acidity,change. One of the processes often accompanying sintering is a change of particle
morphology. In the case of a Pt/Al2O3, catalyst transformations of cubic to spherical particles,
and vice-versa, were found to be,dependent on both temperature and gas environment. A
similar change of particle morphology due to sintering was observed for an Ag/Al2O3
epoxidation,catalyst.

2.2 Deactivation by Coking or Fouling


Formation of hydrocarbon deposits on catalyst surface is the,common mechanism of
deactivation in processes dealing with conversion of hydrocarbons in oxygen lean conditions.
The mechanism is known as coking, or fouling, although the last expression has sometimes the
broader meaning for all kinds of deposition, such as of iron or ceramic particles displaced
originating from the reactor material,[11]. The chemical nature of coke is not easy to define
since, in principle, every hydrocarbon molecule which has a deficit in hydrogen can be
considered as a coke precursor. The nature of coke deposits and the mechanisms of their
formation depend on the feed and the nature of the catalyst employed. In the case of a bi-
functional catalyst it even depends on which part of the catalyst coking,occurs. Finally, the
mechanism of coking is a function of time since the nature of coke can change with time-on-
stream,[3]. It is believed that aromatics and/or olefins are the main precursors of coke
formation, which by reactions of dehydrogenation, condensation, and oligomerization finally
result in coke-like products,[3]. Coking by decomposition of hydrocarbons, which may result
in similar species, is in principle different, since hydrocarbons are likely to produce various H
containing species of general,formula CHX or CnHm.

To some extent the mechanism of coking is a single property of every catalyst. Thus, it was
suggested that for a bifunctional Pt-based catalyst coking starts on metallic sites by adsorption
of coke precursors like monocyclic diolefins, which than migrate to acidic sites and polymeric
to form a polycyclic molecules with several double bonds,[3]. In the case of Ni-based catalysts
for steam reforming, it seems that coke formation starts as a result of hydrocarbon
decomposition on the catalyst surface resulting in Cα. This coke species can be gasified, but
also polymerise to more stable Cβ which can form carbidic,carbon. Similarly, decomposition
of CH4 on Co-based catalysts leads to species with different quality of carbon relative to further

13
reactivity. Transformation from the most reactive to less reactive and finally to the most stable
C follows the increase in both surface coverage and temperature of,decomposition.

In principle, coke can lead to catalyst deactivation by active site coverage and by pore blockage.
The first occurs when both the catalytic and coking reactions are of a single site mechanism,
and the higher the density of active sites the more effective is coking. When catalytic and
coking reactions do not compete for the same sites, deactivation by coking is only,due to pore
blockage. The efficiency of the process of pore plugging by coke depends on catalyst texture,
and may be rather high for a favourable pore shape, e.g., ink-bottle necks. In that case quite a
small amount of adsorbed coke may bring a high activity decline. But sometimes the amount
of coke can rise to a significant fraction of the total mass of the catalyst due to an avalanche
effect, when initial coke particles act as a "glue” for new coke,particles.

Coking on zeolites is a shape selective process since both the rate of deposit build-up and its
nature depend on pore dimensions [10]. The prevailing mechanism of zeolite deactivation by
coke depends on operating conditions, zeolite acidity and its topology, Pore size and structure
are probably the most important parameters as shown in Figure 4 [10]. In systems consisting
of non-interconnecting channels, like in the case of mordenite (Fig. 4a), coking occurs through
pore blockage and is very effective in terms of deactivation. Similarly, tri-dimensional zeolites
with small apertures and large cavities like erionite, are very sensitive to deactivation by pore-

Figure 5: Different mode of deactivation by coke: pore blocking in multidimensional mordenite (a), and tridimentional erionite
(b); site coverage (c), sites accessibility hindered by channel blocking (d), and pore blocking by exterior deposits (e), all in HZSM-5

plugging,(Fig. 5b). Contrary, in systems with tri- dimensional channels without cavities, like
HZSM-5, deactivation initially starts as acid sites coverage, but with time-on-stream and
amount of coke increasing deactivation is due to blocking of the sites in the channel
intersections, and finally due to blocking of pores by exterior deposits,(Fig. 5c,d,e). The amount
of coke formed in the last case, and hence the dynamics of changes in the deactivation mode,
are proportional to the severity of operating,conditions.

14
Chapter 3: Catalyst Deactivation in Important Industrial
Processes

3.1 Fluid Catalytic Cracking


Fluid catalytic cracking (FCC) is designed to process heavy, i.e., high boiling,hydrocarbon feed
at rather high temperatures; FCC catalysts used are subject to excessive coking. Coke is a
reversible poison of,chemical structure determined by its,precursor. This can be an olefin, or
diolefin, preferentially of cyclic structure, which leads via a mechanism of cyclo dimerization
and hydrogen transfer to heavier products, like,decalin or naphthalene. Coke is the cause for
catalyst-porosity changes due to its deposition inside pores or blocking their en trance. In such
a way coke can dramatically change the,selectivity of the catalyst by preventing larger
hydrocarbon molecules to enter or leave the pore. For modem zeolite FCC catalysts the mode
of deactivation by coke depends mainly on their pore structure: one dimensional zeolites and
zeolites with large cavities are deactivated mainly by pore blocking, while for three
dimensional zeolites active sites coverage is a predominant cause for,deactivation. Different
basic molecules, e.g. nitrogen compounds are a frequent reversible poison of FCC catalysts.
Their presence is highly undesirable since they are strongly adsorbed on the catalyst surface
and can produce coke which is more difficult to remove/in the following regeneration cycle,
requiring more severe conditions,[15]. Among the irreversible poisons the most important ones
are heavy metals like V and Ni. They,damage the catalyst by destroying its active sites and
blocking the pores, but they can as well promote some undesirable reactions, like
dehydrogenation [14].

Figure 3.1: FCC Unit

15
The deactivation mechanism by V is believed to occur via its migration and formation of metal
silicates and aluminates in both the reduction and oxidation cycles finally leading to destruction
of the,zeolite matrix. Vanadium is a critical poison, its effect reaching the maximum for
this.might be related to the high mobility of this ion. Deactivation by Na is mainly due to
poisoning of acid sites, and accelerated sintering of the zeolite framework. During its life a
FCC catalyst undergoes several ten thousands alternating reaction and regeneration
cyclesk[15]. Regeneration usually occurs under severe conditions of coke burning, resulting in
catalyst ageing, i.e., surface area and porosity collapse. For zeolite catalysts the process
includes dealumination, leading to an acidity change and reduction of unit cell size.[14].

3.2 Reduction of NOX in Exhaust Gases from Automotive Engines


Since catalysts for automotive exhaust gases are very extensively used and contain expensive
precious metals, the request for their long life is a basic requirement. In platinum based three-
way catalysts for the cleaning of exhaust gases from,gasoline engines, the deactivation problem
is related to both poisoning and sintering. From a historical point of view, the first effort to
overcome the problem of poisoning of the catalytic active component by lead-based
antiknocking agent, was done by complete exclusion of Pb-compounds; this is as much a
benefit to the environment as it was to the catalyst. Besides, the usage of unleaded gasoline
made it possible to replace the very expensive Pt by Pd in Pt-Rh catalyst,[16]. However, in
order to avoid formation of bimetallic Pd-Rh particles, a special technology had to be used by
which the existence of separate layers of two active metals on a high-surface area,"wash coat"
was possible. The sintering of CeO2, which has the role for oxygen storage in noble-metal
catalyzed oxidation of CO, is an additional deactivation process of three-way catalysts.
Achievement of higher ceria dispersion and higher stability of smaller particles was
accomplished by introduction of an additional metal oxide, such as zirconia, forming a solid
solution with CeO2,[16].

Deactivation of exhaust-gas catalyst is even more demanding for fuel-lean gasoline and for
diesel engines. As a matter of fact there is no catalyst up to date which completely satisfies
activity and selectivity requirements. Deactivation is especially prominent in the presence of
H2O vapor and SO2,,which are commonly associated with exhaust gases [17], and which have
a synergistic character. The mechanism of poisoning by SO, is either by simply blocking active
sites due to poison dissociation, or by its interaction with the oxidized catalyst surface leading

16
to sulphates and,sulphites. In the case of selective catalytic reduction, accumulation of coke
from an added hydrocarbon reductant might be an additional reason for catalyst,deactivation.

Figure 3.2: NOx reduction in Exhaust gas coming from automotive engines

Figure 3.3: Catalytic converter

For the most-investigated Cu-ZSM-5 catalyst which is, however, not a suitable catalyst for
diesel-engine exhaust, the activity loss is the consequence of dealumination, especially in the
presence of water. The loss of tetrahedral Al atoms affects the population of Cu active sites
responsible for DeNOx, but it is also seen as a cause for failed hydrocarbons oxidation due to

17
the decline of the rest of,Bronsted sites. At very high temperatures, i.e., between 600 °C and
800 °C, it is believed that substantial deactivation of Cu-ZSM-5 is due to a significant,loss in
microporosity. Since there are no other effects which could be responsible for this textural
change, like dealumination or/and coking, the formation of an active-metal compound is most
likely to be the reason. The existence of Cu2O particles indeed has been,confirmed by XRD
and TEM. In contrast, recent investigation on deactivation of Cu-ZSM-5 catalyst at 400 °C in
the presence of water, showed irreversible deactivation only due to a change of Cu2+
distribution, with neither additional framework changes nor formation of Cu-compounds.

3.3 Ethylene Epoxidation


Epoxidation of ethylene to ethylene oxide is an important process in petrochemical industry.
The uniquely effective Ag-based catalyst undergoes deactivation during time-on-stream;
therefore it has been a subject of extensive research in order to improve its,stability. Under
industrial operating conditions deactivation of Ag-catalysts is compensated by increasing
reaction temperature, which in leads to more rapid deactivation, followed by a selectivity
decay. As a consequence the life expectancy of the catalyst is decreasing. Thus, the catalyst
utilization in the process can be seen as a compromise between its current activity and predicted
life, which should result in maximal yield over the total catalyst life.,Much of the work
performed on the mechanism of the reaction of ethylene oxidation on Ag catalysts has been
aimed at unveiling causes and mechanisms of catalyst deactivation. It has been clarified that
the existence of both subsurface and atomic oxygen is essential for the selective,ethylene
oxidation. Recently, the reduced ability of oxygen adsorption on catalyst ageing in industrial
condition was recognized as the possible reason for the loss of catalyst activity and,selectivity.
The reason for the lower oxygen adsorption ability may be ascribed to any one of the
phenomena which have been related to Ag-catalyst deactivation: catalyst poisoning either by
impurities in the,feed, or by chlorine and an accumulation of carbonaceous deposits on the
catalyst surface, increase in Ag particle size as well as of the support, possibly followed by
redistribution of promoters. Despite of such a wide list of possible causes, however, it is widely
accepted that silver sintering, resulting in a change of particle size distribution and decrease of
active area, is the reason for deactivation. Very often, however, these changes were not found
to be proportional to the extent of deactivation, pointing out that several other phenomena
might contribute to deactivation, like the rearrangement,of faces, changes of morphology of
Ag particles followed by changes in MSI.

18
Chapter 4: Prevention of Catalyst Deactivation and
Regeneration of Deactivated Catalysts

To overcome catalyst deactivation two approaches are possible: (1) trying to avoid or to
minimize possible causes for deactivation, or (2) to accept it, but minimizing its effect both by
both process and catalyst tuning. If the first approach is a choice, than feed purification, or the
selection of more appropriate feed is what can be done. The process tuning related to catalyst
deactivation can be approached in two different ways: i) careful selection of process parameters
in order to avoid initiation of some deactivation phenomena, and ii) careful monitoring of
parameters in the process of catalyst production in order to obtain a product with desirable and
reproducible properties,[11]. Among the process parameters reaction temperature is the most
important one when sintering is concerned. In order to minimize the chances for sintering the
temperature must be chosen in such a way that severe temperature gradients are avoided
throughout the bed of catalyst particles. The last measure can be only achieved, however, when
the catalyst itself is of uniform physical and chemical properties. By ascertaining that the
chance of appearance of hot spots is low, thermal sintering may be avoided to a certain degree.
Generally, production of catalysts of constant quality is very important from the point of view
of deactivation. The reason for this is that many catalysts, like all other solid materials, might
have a memory effect, i.e., the history of catalyst preparation affects its behaviour during its
lifetime,[11]. Quite often catalyst and process imperfections are interrelated, the first being
sometimes passive as long as there is no excessive change in desired process conditions. Once
this is provoked by some accident in the process, the result can be a prominent deactivation
due to a synergistic effect of these two imperfections.

In order to improve catalyst resistance to poisoning the distribution of the active sites in and
on a catalyst particle is of primary importance. The preferred active metal location depends on
the mechanism of poisoning, i. e., uniform distribution through the particle is favoured in the
case of a parallel mechanism, and metal deposition on the catalyst surface is preferred for the
case of consecutive,poisoning. An additional tailoring of pore structure in the first case might
be seen as a compromise between catalyst life and its activity, resulting in maximal total
catalyst,efficiency. Besides this, several other techniques in catalyst preparation are used to
increase its resistance to all kinds of deactivation phenomena. The second metal may
sometimes serve as a guard to the primary active metal, protecting it from poison. The addition

19
of Re to Pt/Al2O3 catalysts, e.g.,,which results in increased resistance to coking, was explained
by the ability of Re to remove the coke precursors, like cyclopentane, e.g., from Pt. Taking the
fact into account, however, that a physical mixture of Pt/Al2O3 and Re/Al2O3 does not result in
the same deactivation suppression as in the case of bimetallic Pt-Re/Al2O3 catalyst, the question
of the real role of Re is still open,[18]. Similarly, the addition of K to a Ni/Al2O3 catalyst for
steam reforming of hydrocarbons was found to sup press coke formation. The promotion effect
of K is explained by both a decrease of the rate of hydrocarbon decomposition on the catalyst
surface, and an increase of the rate of removing the coke deposits from the,surface. A beneficial
influence of a second metal related to deactivation is known in practice of zeolite preparation,
an inactive metal may force the active one to stay in a certain position in the zeolite matrix,
thus protecting it from sintering. This kind of stabilization, which was observed for Pt promoted
by Fe and Cr* in a Y zeolite is known as chemical anchoring. Besides previously mentioned
redispersion effects of chlorine on Pt, there is also a chlorine effect on the reduction of coke
deposition. It is believed that Cl optimizes H; chemisorption, and subsequent H-spillover,
which consequently leads to self-regeneration of the coked Pt-reforming,catalyst. Similarly,
addition of α-Sb2O4 to an active phase of MoO3 promotes spillover of O2, which reacts with a
reduced location on the surface of MoO3, hence creating an active site.

Accumulation of coke on zeolite-based catalysts doubles the negative effect: it decreases


activity due to active sites coverage, and requires subsequent oxidative treatment to burn-off
the coke. Consequently, some rules have to be followed in order to decrease chances of coking
in a hydrogenation process: a) if possible tridimensional zeolites with small apertures and large
cavities have to be avoided, b) acid strength and density have to be adjusted to the lowest values
necessary for the reaction, c) operating conditions have to be tuned in order to avoid formation
of coke precursors. In some cases, however, a beneficial effect of coke depot its has been also
demonstrated, like for n-butane skeletal isomerisation to isobutene over,ferrierite. An initial
low selectivity, followed by a selectivity increase with time-on-stream, can be explained by
preferential poisoning of the non selective strongest ive located inside of the 10-member ring
channels. At the same time shape selective Bronsted acid sites located at the pore mouth are
not too vulnerable to the medium level of coking as long as the formation of poly aromatic
compounds which may block the pore mouth is,avoided. Catalyst regeneration is the least
desirable approach to catalyst deactivation. It requires both process time and man power, and
it usually does not result in the original activity of the fresh,catalyst. Catalyst regeneration
typically consists of either removal of poison/coke, or catalyst particles redispersion, although

20
sometimes both processes occur simultaneously. In principle, the removal of coke is possible
by gasification with oxygen, steam, hydrogen, and carbon dioxide,[19]. Since catalyst
regeneration by coke burning starts with hydrogen oxidation and H2O production, the process
is performed in two steps, i. e., at low and high reaction temperature. The approach is of special
interest for zeolite catalyst regeneration, avoiding a coupling effect of high temperature,and
water.

A simulation of steam regeneration of deactivated commercial catalysts for HDS shows all the
complexity of choosing the right agent for coke,removal. Since the removal of coke by water
vapour is an endothermic reaction, the use of steam is supposed to be a more saver approach
to catalyst regeneration than using oxygen. Under laboratory conditions, however, an activity
loss of a commercial Co-Mo/Al2O3 catalyst previously regenerated in steam was found. The
results were attributed to both Mo redispersion and a molybdenum loss of the active phase
comprising molybdena; this was due to reaction of water vapor with MoO3 at temperatures
from 600 to 700 °C. This model describes the details of deactivation of a hydrotreating Ni-
Mo/Al2O3 commercial catalyst well,[20]. After regeneration in steam at a rather mild
temperature of 400 °C, the catalyst showed substantial loss of activity when afterwards exposed
to a feed stock with higher S level. The loss of active component by MoO3 segregation and
vaporization of MoO3 proved by SEM investigation of discharged catalyst samples, was
identified as the primary cause for deactivation. This might have happened as the result of hot
spots forming in the period of catalyst regeneration. In addition, an intensive sintering caused
by a liquid MoO3 phase was identified as the secondary cause for the catalyst deactivation. The
BET surface area decrease and average pore diameter increase by a factor of 5 and 10,
respectively, support this conclusion,[20].

Finally, not being directly connected to coping with catalyst deactivation, but more as a
consequence of it, the handling of deactivated catalyst is often important. Even the deactivated
catalysts are mostly still pyrophoric, and passivation by appropriate methods is required before
catalyst discharge.,In the case of CuO-ZnO/Al2O3 catalysts for methanol synthesis, wet
passivation of the spent catalyst, has been exchanged by in situ oxidation. However, due to the
extremely exothermal reaction this regeneration process is difficult to control. DSC analysis of
a deactivated commercial catalyst sample in controlled oxidizing atmosphere revealed that the
total heat released in cyclic operation is much less than for a one step catalyst oxidation in,air.
The latter procedure applied on large-scale could result in thermal reactor runaway.

21
Conclusion

Deactivation mechanisms differ by their causes and consequences. Although having separated
reasons for their appearance and different impacts on the catalyst, they are often interconnected,
even one mechanism being initiated by another. This makes situation fuzzy since the main
deactivation mechanism can be a secondary one by order of appearing. Sometimes the main
deactivation mechanism can even be masked by another,,more benign one.

While poisoning changes the character of active sites, coking or fouling mainly leave the active
sites unchanged but suppress their availability to reactants. The influence of ageing on active
sites is different, depending on which particular mechanism of ageing occurs,,i.e. with or
without change of catalyst composition.

Different factors determine the particular mechanism of deactivation: the type of substrate
responsible for deactivation (poison, coke), the substrate/catalyst affinity, main and
deactivation reaction mechanisms and kinetics, reaction parameters like,temperature,
concentration and pressure, and finally textural properties of the catalyst and its acidity.

In effort to control deactivation as much as possible, two approaches are offered: either to avoid
it when possible, like feed purification, or accept it but minimizing its effects. For the second
approach catalyst memory effects have to be acknowledged, by preventing process
irregularities as early as in the phase of catalysts,fabrication.

Accelerated deactivation tests are powerful tools for studying catalyst deactivation in a
relatively short time and at a suitable reactor scale. With properly selected reaction variables
and by applying,e.g. a deactivation compensation approach in a gradient less reactor,
deactivation kinetics can be determined. This can be the basis for further modelling, with the
final goal of predicting the catalyst life and its performance in commercial conditions at any
time of operation for a given time of reactor.

22
References

[1] M.Baerns, H. Hofmann, and A.Renken, Chemische Reaktionstechnik, Georg Thieme


Verlag, Stuttgart, 1999.
[2] T.E.Fischer and S.R.Kelemen, J.Catal. 53, 24, 1978.
[3] C.H.Bartholomew, in Catalyst Deactivation 1987, Edited by B.Delmon and G.F.Froment,
Stud.Surf.Sci.Catal. Vol. 34, 81, 1987.
[4] R.L.Ollendorf, G.Boskovic, J.B.Butt, Appl.Catal. 62, 85, 1990.
[5] B.C.Gates, J.R.Katzer, and G.C.A.Schuit, Chemistry of Catalytic Processes, McGraw-Hill
Book Company, p.282, New York 1979.
[6] P.Lankhorst, H.C.de Jongste, and V.Ponec, in Catalyst Deactivation, Edited by B.Delmon
and G.F.Froment, Stud.Surf Sci.Catal. Vol. 64, 3, 1980.
[7] M.Che and C.O.Bennett, Advan.Catal. 36, 55, 1989.
[8] D.E.Damiani and J.B.Butt, J.Catal. 94, 203, 1985.
[9] T.Onal and J.B.Butt, J.C.S.Faraday I, 781, 887, 1982.
[10] M.Guisnet and P.Magnoux, Appl.Catal. 54, 1, 1989.
[11] B.Delmon and P.Grange, in Catalyst Deactivation, Edited by B. Delmon and G.F.Froment,
Stud.Surf.Sci.Catal. Vol. 650,7, 1980.
[12] J.W.Gosselink and J.A.R.van Veen in Catalyst Deactivation 1999, Edited by B.Delmon
and G.F.Froment, Stud.Surf Sci.Catal. Vol. 12, 63, 1999.
[13] P.O'Connor and A.C.Pouwels in Catalyst Deactivation 1994, Edited by B.Delmon
and G.F.Froment, Stud. Surf.Sci.Catal. Vol. 88, 129, 1994.
[14] P.Zeuthen, J.Bartholdy, P.Wiwel, and B.H.Cooper in Catalyst Deactivation 1994. Edited
by B.Delmon and G.F.Froment, Stud.Surf Sci.Catal. Vol. 88, 199, 1994.
[15] M.Absi-Halabi and A. Stanislaus, in Deactivation and Testing of Hydrocarbon
Processing Catalysts. Edited by P.O'Connor, T.Takatsuka, and G.L.Woolery, ACS
Symposium Series 634, 229, 1996.
[16] M.Shelef and R.W.McCabe, Catal.Today 62, 35, 2000.
[17] M.lvamoto, N.Mizuno, and H. Yahiro, in New Frontiers in Catalysis, Edited by L.Guczi,
F.Solymosi, and P.Tetenyi, Elsevier Sci.Publ. and Akademiai Kiado, Buda pest, 1258, 1993.
[18] E.E.Petersen in Catalyst Deactivation 1997, Edited by C.H.Bartholomew and
G.A.Fuentes, Stud.Surf.Sci.Catal. Vol. 111, 87, 1997.
[19] R.M.Neducin, G.Boskovic, E.Kis, G.Lomic, H.Hantsche, R.Micie, and P.Pavlovic,
Appl.Catal. A: General 107, 133, 1994.
[20] R.M.Neducin, G.Boskovic, E.Kis, G.Lomic, H.Hantsche, R. Micic, and P.Pavlovic,
Appl.Catal. A: General 107, 133, 1994.

23

You might also like