Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Volume 21
A Specialist Periodical Report
Catalysis
Volume 21
Editors
James J. Spivey, Louisiana State University, USA
Kerry M. Dooley, Louisiana State University, USA
Contributors
J. A. Anderson, University of Aberdeen, UK
Marco J. Castaldi, Columbia University, USA
Gabriele Centi, University of Messina, Italy
Claus Hviid Christensen, Technical University of Denmark, Denmark
Kresten Egeblad, Technical University of Denmark, Denmark
M. Fernandez-Garca, Instituto de Catalisis y Petoleoquimica (CSIC), Spain
Amit C. Gujar, Mississippi State University, USA
Charlotte C. Marsden, Technical University of Denmark, Denmark
Nora M. McLaughlin, Columbia University, USA
Siglinda Perathoner, University of Messina, Italy
Jeppe Rass-Hansen, Technical University of Denmark, Denmark
Esben Taarning, Technical University of Denmark, Denmark
Mark G. White, Mississippi State University, USA
Ye Xu, Oak Ridge National Laboratory, USA
If you buy this title on standing order, you will be given FREE access
to the chapters online. Please contact sales@rsc.org with proof of
purchase to arrange access to be set up.
Thank you
ISBN 978-0-85404-249-4
ISSN 0140-0568
A catalogue record for this book is available from the British Library
Apart from any fair dealing for the purpose of research or private study for
non-commercial purposes, or criticism or review as permitted under the terms
of the UK Copyright, Designs and Patents Act, 1988 and the Copyright and
Related Rights Regulations 2003, this publication may not be reproduced,
stored or transmitted, in any form or by any means, without the prior
permission in writing of The Royal Society of Chemistry, or in the case of
reprographic reproduction only in accordance with the terms of the licences
issued by the Copyright Licensing Agency in the UK, or in accordance with the
terms of the licences issued by the appropriate Reproduction Rights
Organization outside the UK. Enquiries concerning reproduction outside the
terms stated here should be sent to The Royal Society of Chemistry at the
address printed on this page.
Gordon A. and Mary Cain Dept. Chemical Engineering, Louisiana State University,
Baton Rouge LA 70803, USA. E-mail: jjspivey@lsu.edu. E-mail: dooley@lsu.edu
Cover
Preface 7
James J. Spivey* and Kerry M. Dooley*
Catalysis, 2009, 21 | 9
This journal is
c The Royal Society of Chemistry 2009
Nano-architecture and reactivity of titania catalytic materials. 82
Part 2. Bidimensional nanostructured films
Gabriele Centi and Siglinda Perathoner
Introduction 82
Outlooks for the development of catalysts based 89
on the concept of nanostructured films
Synthesis of titania nanostructured films 100
Uses, with focus on catalysis 106
Conclusions 119
10 | Catalysis, 2009, 21
This journal is
c The Royal Society of Chemistry 2009
Supported analogs of ionic liquid catalysts 181
Task specic ionic liquids (TSIL) 182
Telomerization in ILs 184
Catalysis, 2009, 21 | 11
This journal is
c The Royal Society of Chemistry 2009
Heterogeneous catalysis for production
of value-added chemicals from biomass
Kresten Egeblad, Jeppe Rass-Hansen, Charlotte C. Marsden,
Esben Taarning and Claus Hviid Christensen*
DOI: 10.1039/b712664f
1. Introduction
Almost everything around us is in some way a product of controlled
chemical processes. That is either chemical processes conducted in Nature
or chemical processes conducted in the chemical industry. In the most
developed parts of the World, it is in fact products from the chemical
industry that completely dominate our everyday lives. These products range
from fuels and fertilizers to plastics and pharmaceuticals.1 To make these
products widely available, a huge amount of resources have been invested
during the last century to develop the chemical industry to its current level
where it is the largest industry worldwide, a cornerstone of contemporary
society, and also a platform for further global economic growth.2,3 It can be
argued that the enormous success of the chemical industry can be attributed
to the almost unlimited availability of inexpensive fossil resources, and to a
continuously increasing number of catalysts and catalytic processes that
make it possible to eciently transform the fossil resources into all the
required compounds and materials. Accordingly, more than 95% of the
fuels and chemicals produced worldwide are derived from fossil resources,
and more than 60% of the processes and 90% of the products in chemical
industry somehow rely on catalysis. It has been estimated that 2030% of
the production in the industrialized world is directly dependent on catalytic
technology.4 Therefore, it is not surprising that we are continuously
expanding our already vast empirical knowledge about catalysis to further
improve the eciency of existing catalysts and processes, to discover
entirely new ways of valorizing available resources, and to lower the
environmental impact of human activities.5 Due to the overwhelming
importance of fossil resources during the 20th century, most catalysis
research eorts have, so far, concerned the conversion of these resources
into value-added fuels and chemicals. There are, however, indications that
the era of easy access to inexpensive fossil resources, especially crude oil, is
coming to an end. The resources are certainly limited and the demand from
everywhere in the world is growing rapidly. At the same time, it is becoming
increasingly clear that the emission of CO2 that follows the use of fossil
resources is threatening the climate of the Earth. Together this makes the
development of a chemical industry based on renewable resources one of the
most important challenges of the 21th century.
This challenge has two dierent facets. One is the discovery and develop-
ment of methods to use renewable resources to supply suitable energy
carriers, in sucient quantities at acceptable costs, and with minimal impact
Center for Sustainable and Green Chemistry, Department of Chemistry, Technical University
of Denmark, Building 206, Lyngby DK-2800, Denmark. E-mail: chc@kemi.dtu.dk
3.1 Introduction
This section concerns catalytic processes that transform chemicals from
renewables by CC bond breaking. Among these are thermochemical
processes, such as pyrolysis and also gasication, catalytic reactions, such
as catalytic cracking and dierent reforming reactions, and decarbonylation
and decarboxylation reactions. Many of these reactions occur simulta-
neously, particularly in the thermochemical processes. Another technically
important class of CC bond breaking reactions is the fermentation pro-
cesses, however, they will not be considered in this section since they do not
involve heterogeneous catalysis.
3.2.2 Bio-oils. In the 1970s, it was shown that bio-oils from plant extracts
such as rubber latex, corn oil, and peanut oil can be converted into a mixture
of mainly gasoline and liquid petroleum gas over a ZSM-5 catalyst, at
temperatures between 400500 1C.34 These bio-oils were investigated as
feedstocks for the reaction because they have high hydrogen to carbon
ratios and low oxygen contents and therefore a hydrocarbon-like structure.
It was suggested that such renewable plant resources, due to their siginifant
content of highly reduced photosynthetic products, would be suitable for
producing fuels or chemical raw materials.35 The high hydrogen-to-carbon
4. Catalytic hydrolysis
4.1 Introduction
Hydrolysis is the process by which a compound is broken down by reaction
with water, thus it can be thought of as the opposite reaction of dehydration,
where water is of course removed. Hydrolysis is a key reaction type in
biomass chemistry, for it is central in the depolymerisation of polysaccharides
to simpler monosaccharide building blocks, such as fructose, glucose, and
xylose.
It has been established early that acids catalyze this hydrolysis reaction,
thus liquid sulfuric acid has been used. Heterogeneous catalysis can
potentially provide simpler and environmentally more benign processes,
however, via ease of separation and recovery. Thus, solid acids, such as
acidic ion-exchange resins,65,66 zeolites and heteropolyacids, can replace the
homogeneous acids. Hydrolysis of sucrose is in fact already established on
an industrial scale using acid ion-exchangers,67 but the main route is still
via enzyme catalysis. Transfer to the heterogeneous system shows problems
regarding the microenvironment of the swollen polymer, i.e. limitation
of diusion and restricted accessibility, as well as the production of many
by-products.68
In an eort to make this switch to heterogeneous catalysts viable, various
acidic exchange resins have been tested, including those prepared by
radiation-induced grafting to produce graft copolymers capable of hosting
sulfonic groups.6971 A common problem with solid acids, including ion-
exchange resins, is that they are subject to poisoning by water. Thus,
sulfonated mesoporous silicas were investigated as a new class of solid
acids, giving glucose and fructose in 90% yield after four hours at 80 1C.72
Zeolites are also acid ion-exchangers. A conversion of sucrose of up to 90%
with close to 90% selectivity and very few by-products formed was achieved
using highly dealuminated zeolite Y at 70 1C.73 Similarly, the activity of
various dealuminated zeolites was compared, again showing high selecti-
vities and few by-products, regardless of the conversion.74 Similarly, the
hydrolysis of maltose was studied by comparing the performance of acid
zeolites, ion-exchange resins, amorphous silica-aluminas and also the
ordered mesoporous material, MCM-41.75 The best results were achieved
with zeolite beta (Si/Al = 50) at 130 1C and 10 bar where a conversion of
85% and a selectivity of 94% was reported. Most recently, the use of
organicinorganic hybrid mesoporous silica catalysts was reported for the
hydrolysis of cellubiose to yield glucose. Cellubiose was used as a model for
oligosaccharides, and it was possible to achieve 100% conversion at 175 1C
but signicant glucose degradation was observed simultaneously.76
It can be seen then that heterogeneous catalysis may nd an opportunity
for replacing the enzymatic catalysis of disaccharides to its monosaccharides,
and thereby provide industry with a more ecient and benign route.
However, it is also clear that more selective catalysts are required.
Scheme 2 Triglycerides can be hydrolyzed to give fatty acids and glycerol. This can be
catalyzed by solid acid catalysts like zeolites or acid exchanged resins.
5. Catalytic dehydrations
5.1 Introduction
There are several examples of dehydrations of chemicals derived by renew-
able resources by use of heteregeneous catalytic approaches in the literature.
These can be categorized into three types of reactions: (a) reactions in which
one (or more) molecule(s) of water is eliminated from a single substrate
molecule, (b) reactions in which one (or more) molecule(s) of water is
generated as the result of an esterication reaction between an alcohol and a
carboxylic acid or carboxylic acid derivative and (c) reactions in which one
(or more) molecule(s) of water is generated due to an etherication reaction
between two alcohol functionalities.
5.2.4 Glycerol. It has long been known that glycerol can be dehydrated
to produce acrolein by heating aqueous glycerol with a mixture of nely
powdered KHSO4 and K2SO4 (Scheme 3).92 Recently, the reaction has
received attention again, as several acidic solid oxide catalysts were tested as
catalysts for the reaction.9395 The best results were obtained with silica-
supported heteropolyacids such as silicotungstic acid with which 86%
selectivity towards acrolein was obtained at 98% conversion of glycerol at
275 1C.93 Also tungstated zirconia has been reported to be a selective
catalyst for acrolein formation in a comparative study of many dierent
catalysts; using 15 wt% WO3/ZrO2, 65% selectivity towards acrolein was
achieved at 100% conversion at 325 1C.95
Scheme 3 Acrolein can be obtained by dehydration of glycerol. The reaction was reported
many years ago using powdered KHSO4/K2SO4 as catalyst. Recently, the use of silica-
supported heteropolyacids has also been described, notably with silicotungstic acid as catalyst.
Scheme 4 Hydroxyacetone can be obtained from glycerol by dehydration. The reaction has
been reported using copperchromite as catalyst.
Scheme 5 Xylose can be dehydrated to produce furfural. The reaction has been reported using
several dierent catalysts including zeolites, sulfonic acid functionalized MCM-41 and
immobilized heteropolyacids. The best selectivity towards furfural was achieved using zeolite
H-mordenite, although at low conversion of xylose.88 Overall, the best yield of furfural was
obtained using sulfonic acid functionalized MCM-41.
5.2.6 Glucose and fructose. Sucrose is one of the largest chemicals readily
available from biomass. It is produced from sugar cane or sugar beats and
can be easily hydrolyzed into its constituent monomers, glucose and fructose.
In general, dehydration of these carbohydrates will lead to the formation of
many dierent products, however, some control of the dehydration products
obtained can be achieved using dierent acid catalysts. The target chemical
in most reports concerning solid acid catalyzed dehydration of hexoses is
5-hydroxymethyl furfural (Scheme 6), and to a lesser extent levulinic acid,
which is formed along with formic acid from HMF by a rehydration
decomposition reaction. HMF is sometimes referred to as a sleeping giant
due to its enormous potential importance as a key chemical intermediate,104
and several reviews are available concerning the production as well as
chemistry of HMF.105,106 Very recently, signicant achievements were made
in the production of HMF by homogeneous catalytic approaches.107 Typi-
cally, the starting material for HMF synthesis is fructose, however, there
exists, of course, great interest in establishing a commercially viable process
directly from glucose, since glucose is less expensive than fructose. Several
types of catalysts have been applied for dehydration of fructose to produce
Scheme 6 Fructose can be transformed into 5-hydroxymethyl furfural (HMF) via acid-
catalyzed dehydration. Solid acid catalysts applied to facilitate the reaction are zeolites, ion-
exchange resins and solid inorganic phosphates. With sporadic success, notably with inorganic
phosphates, other carbohydrate sources such as inulin can also be transformed into HMF.
With zeolites as the solid acid catalyst, the best results for HMF synthesis
were obtained by Moreau et al. who tested acidic mordenites with dierent
Si/Al ratios in batch experiments and reported that dealuminated H-Mordenite
with Si/Al ratio of 11 exhibited the highest selectivity and even so at reasonably
high fructose conversion (91% selectivity at 76% conversion after 60 min. at
165 1C using water/methyl isobutyl ketone as the reaction media).108 Other
zeolites, H-Y, H-Beta and H-ZSM-5 were also tested for the reaction, however,
none of these catalysts were as selective as H-mordenite.109
Also acidic ion-exchange resins were tested as catalysts for fructose dehydra-
tion. Using PK-216, a solution of water-DMSO-polyvinylpyrrolidone contain-
ing 10 wt% fructose was dehydrated to HMF after 816 h at 90 1C with 71%
selectivity at 80% conversion using MIBK as the extraction phase.107 Using a
more concentrated fructose solution (30 wt%), 65% selectivity was achieved at
83% conversion. Working in more dilute solution (0.5 M in DMSO), also with
PK-216 as the catalyst, an HMF yield of 90% was obtained after 5 h at
80 1C.110 Even more remarkable perhaps was the observation that the reaction
could be carried out in a continuous process with no signs of deactivation even
after 900 h, in this case using Amberlite IR-118 as the catalyst. Recently,
Amberlyst 15 was also reported as catalyst for fructose dehydration to produce
HMF at 80 1C using a solvent system comprising DMSO and either a
hydrophilic (BMIM-BF4) or hydrophobic (BMIM-PF6) ionic liquid.111 In
both cases, HMF yields of ca. 80% were achieved after 24 h, however, when
the reaction was carried out without DMSO as co-solvent, a maximum yield of
5.3.3 Levulinic acid and itaconic acid. Levulinic acid could also become an
important intermediate chemical in the future since it can be produced by acid
catalyzed dehydrationdecomposition of fructose. The synthesis of diethyl
levulinate was recently reported using sulfonated Starbon-400-SO3H. The
6. Catalytic oxidations
6.1 Introduction
Oxidation as a process to transform biomass into value-added chemicals is a
key one. Here, we focus on oxidations using molecular oxygen as the
oxidant, with the aim of illustrating selected interesting reactions that could
be important in the eorts to develop sustainable chemistry since they only
require abundant bio-resources as reactants and have water as the only, or
at least the main, byproduct.
Scheme 8 Oxidation of glycerol in basic media leads to glyceric acid, using Pd/C or Au/C as
catalyst.
During the last few years, signicant attention has been devoted to the
aerobic oxidation of glycerol using heterogeneous gold catalysts. Glycerol was
oxidized to glycerate in the presence of NaOH with 100% selectivity at 60 1C
using water as the solvent, and at an oxygen pressure of 0.3 MPa. The
catalysts used were either 1% Au/charcoal or 1% Au/graphite both giving
around 55% conversion.143 A range of dierent supports for the gold
nanoparticles catalysts were investigated, TiO2, MgO, and Al2O3, but all
showed low activity compared to Au/C. Depending on the base concentration
and the reaction time, the selectivity of the Au/C catalyzed liquid phase
glycerol oxidation could be controlled.144 Most recently, it was shown
that hydrogen peroxide is formed during oxidation of glycerol using gold
catalysts145 and that this leads to CC bond breakage and a resulting loss of
selectivity. This was independently supported by the fact the AuPd catalysts
showed higher selectivity to glycerate than the monometallic Au catalyst,
which was shown to be related to the higher eciency of Pd to catalytically
decompose the produced hydrogen peroxide in situ.146 By employing a
reaction temperature of 100 1C and an air pressure of 21 bar with methanol
as the solvent, it is possible to obtain dimethyl mesoxalate in yields as high as
89%147 and this clearly illustrates the eect of temperature on the degree of
oxidation of the glycerol feedstock.
Scheme 9 Oxidation of glucose to gluconic acid has been reported using oxidation catalysts
such as Pt/C, Pd/C, PdBi/C and Au/C.
6.3.3 Glyceric acid. The selectivity for the catalytic oxidation of glyceric
acid, and the calcium salt, can be controlled by the nature of the catalyst, and
the pH, in a similar way to that of glycerol as discussed above. In this way, it
is possible to obtain products corresponding to the oxidation of the primary
and secondary alcohol moieties, i.e., tartronic acid and hydroxypyruvic acid,
respectively. As reported by Fordham et al., oxidation of glyceric acid under
basic conditions leads to the formation of tartronic acid whereas hydro-
xypuric is aorded under acidic conditions.171 The catalyst was suspended in
glyceric acid, a calcium salt added, and oxygen gas bubbled through. NaOH
was added to keep the pH constant. Two catalytic systems were tested; 5 wt%
Pt/C at pH 10-11 yields tartronic acid with a selectivity of 60% at a
conversion of 94%; when 2% of bismuth is added, the same product is
obtained but with a selectivity of 83% at 90% conversion.171 Hydroxypyruvic
acid was obtained by aerobic oxidation of glyceric acid using a bismuth-
promoted platinum catalyst under acidic conditions (pH 34) to give a 64%
yield at 75% conversion.171 After prolonged contact with the catalyst,
tartronic acid was oxidized to oxalate whereas hydroxypyruvic acid was
oxidised even more rapidly to glycolic acid.
6.3.4 Lactic acid. Pyruvic acid and its derivatives are in increasing
demand due to their use as precursors in the synthesis of drugs and
agrochemicals.10 It has proved dicult to obtain pyruvic acid directly from
Oxidation of HMF was also attempted in situ directly from fructose, using
a membrane reactor or encapsulating PtBi/C into a polymeric silicone
matrix, and again, with air as the oxidant. However, the yield was
never more than 25%.181 A further attempt to obtain FDCA directly from
fructose involved a one pot reaction in the presence of cobalt acetyl-
acetonate encapsulated in solgel silica, at 155 1C and with 2 MPa of air
pressure giving FDCA with 99% selectivity directly from fructose at a
conversion of 72%.122
HMMF can be obtained from HMF by oxidation with Au/TiO2 (Scheme 11),
under very mild conditions25 1C, 1 bar O2 and 8% NaOMein near
quantitative yields. This is an intermediate on the route to FDMC, but the
reaction can be stopped at this stage.176
Scheme 11 Under mild conditions, oxidation of HMF in methanol can be tuned to yield
methyl (5-hydroxymethyl)-furoate via oxidation of the aldehyde moiety.
7. Catalytic hydrogenations
7.1 Introduction
Catalytic hydrogenation represents a set of reactions that will be extremely
important in the production of value-added chemicals from biomass.
Already now, they play a signicant role in todays industry, and holds
great promise for further developments. Here, selected examples of hetero-
geneously catalyzed hydrogenations of chemicals available from renewables
resources are presented.
Scheme 12 Various catalysts have been applied to facilitate the catalytic hydrogenation of
glucose to sorbitol, notably Pt supported on activated carbon cloth using which 99.5% yield of
sorbitol can be obtained.
Raney nickel, copper and platinum group metal catalysts have also been
used as catalysts for transforming fructose into mannitol via catalytic
hydrogenation (Scheme 14).197199,203,204 Ruthenium supported on carbon
is among the most studied catalysts for this reaction,205 and it is, in fact, also
eective for the combined hydrolysis-hydrogenation of inulin to mannitol
when the carbon support has been made acidic prior to the catalytic
experiments.188 The bifunctional catalyst applied in the study was made
by pre-oxidizing activated carbon (SX1G) with various oxidants such as
nitric acid and ammonium persulfate and then introducing Ru onto this
support by incipient wetness impregnation followed by reduction with
NaBH4. The oxidized carbon catalyzes the hydrolysis of inulin to a mixture
of glucose and fructose which is subsequently hydrogenated to a mixture of
glucitol and mannitol. Increasing the hydrogen pressure (up to 100 bar)
apparently also increases the rate of hydrolysis dramatically.
References
1 C. H. Christensen and J. K. Nrskov, J. Chem. Phys., 2008, 128, 182503.
2 S. Shaik, Angew. Chem., Int. Ed., 2003, 42, 3208.
3 S. Senkan, Angew. Chem., Int. Ed., 2001, 40, 312.
4 I. Maxwell, Stud. Surf. Sci. Catal., 1996, 101, 1.
5 Handbook of Heterogeneous Catalysis, eds. G. Ertl, H. Knozinger, F. Schuth
and J. Weitkamp, Wiley-VCH, Weinheim, 2008.
6 B. A. Tokay, Biomass Chemicals, in Ullmanns Encyclopedia of Industrial
Chemistry, Wiley-VCH, Weinheim, 2005.
7 C. H. Christensen, J. Rass-Hansen, C. C. Marsden, E. Taarning and
K. Egeblad, ChemSusChem., 2008, 1, 283.
8 J. Goldemberg, Science, 2007, 315, 808.
9 G. W. Huber, S. Iborra and A. Corma, Chem. Rev., 2006, 106, 4044.
10 A. Corma, S. Iborra and A. Velty, Chem. Rev., 2007, 107, 2411.
11 A. J. Ragauskas, C. K. Williams, B. H. Davison, G. Britovsek, J. Cairney,
C. A. Eckert, W. J. Frederick, Jr, J. P. Hallett, D. J. Leak, C. L. Liotta,
J. R. Mielenz, R. Murphy, R. Templer and T. Tschaplinski, Science, 2006, 311, 484.
12 D. R. Dodds and R. A. Gross, Science, 2007, 318, 1250.
1. General introduction
On the 9th of February, 2004, in its resolution A/RES/58/217, the General
Assembly of the United Nations proclaimed the period from 2005 to 2015 as
the International Decade for Action, Water for Life, which commenced on
World Water Day, 22nd March 2005. 2015 was also a deadline year set by
world leaders at the United Nations Millennium Summit where a pledge
was made to reduce by 50% the proportion of people unable to reach or to
aord safe clean drinking water. Humanitys needs are met by only 1% of
the Earths total water and so careful use of this resource is essential. The
writing of this review coincides with the release of an RSC report on
sustainable water which highlights how increasing global population,
climate change and pollution will only exacerbate issues associated with
access to fresh water. The report also highlights the key role which can be
played by chemistry in aspects of sustainable water. In this review we
address only chemical solutions based on catalytic and photocatalytic
methods. While many of the issues associated with provision of potable
water in the developing world may be resolved by the use of simple physical
methodologies such as ltration, many of the issues associated with water
purity in the developed world involve complex, stable molecules present at
low concentrations but non the less capable of producing toxic eects in
plants and animals and which require removal technologies which are more
demanding. It is with these reagents that this chapter is principally
concerned.
The purication of drinking water from various water sources generally
combines a series of physical and chemical steps to eliminate the solid
fraction, kill bacteria and reduce the level of chemical pollutants. Many
processes include an oxidation step, based on the use of a strong oxidizing
agent such as chlorine, sodium hypochlorite, ozone or hydrogen peroxide.
These treatments generally lead to an acceptable level of micro-organisms in
water, however the process also aects the chemical composition. Further-
more undesirable ions, such as nitrates, remain unaected by such oxidative
treatments.
Given that the reduction of nitrate by hydrogen to yield nitrogen must also
generate hydroxide ion (eqn (2)), this leads to loss of activity as negatively
charged hydroxide ions compete for adsorption sites with nitrate37 (and
nitrite anions), but also leads to a loss in selectivity as the pH is driven up.
The selectivity to ammonium shows a strong correlation with the solution
pH23,25,52 and the addition of a source of CO2 is often added25,42,46 to
maintain the solution at slightly acidic levels (eqn (3)).
CO2(g) + 2H2O(l) = H3O+ + HCO3(aq) (3)
The use of formic acid (or acetate)43 can be seen as an in situ buer25 where
release of CO2 takes place at the right time and place at the point where it is
required to neutralise OH ions.43 In terms of dening the right place, it
has been shown by the use of supports of a range of pore sizes, that diusion
of hydroxide ions from pores may be inuential in determining selectivity as
high localised concentrations of OH may build up due to mass transport
limitations.42,43
In addition to the use of buers such as CO2, either added directly to the
solution or through in situ generation via decomposition of an appropriate
organic reagent,25,42,43 high localised OH concentrations, and conse-
quently high levels of ammonium formation, can be avoided though
appropriate selection of support. This consists of the use of materials with
low surface areas and/or large pore sizes. This has been exploited in
particular in studies using SnO2 as support43,46 while in the case of layered
hydrotalcite materials the low ammonia formation was attributed to a
decrease of the catalyst diusion limitations where positive or neutral are
forced out of the layer spacings by ionic forces.40
These radicals may initiate oxidation reactions with organic species. The
reaction of S-containing radicals with organic moieties leading to CO2
formation is faster than the corresponding reaction with P-containing
radicals.61 However, as previously indicated, these anionic species may be
Fig. 3 Results for the photocatalytic reduction of nitrate in the presence of 0.04 M HCOOH as
hole scavenger over titania and Ag/TiO2 catalysts prepared by impregnation (IMP) or
photodeposition (Photo) and comparing use of Degussa P25 (49 m2 g1) with Hombikat UV
100 (250 m2 g1) as supports.76
6. Removal of microorganisms
The current disinfection technologies of water supplies apply either chemi-
cal or photochemical damage or physical removal of microorganisms by
ltration. Chlorination is a universally practiced water-disinfection process,
which can prevent waterborne infectious diseases, but it is not ecient in the
inactivation of spores, cysts and some viruses.197 Chlorination, combined
with ozonization, signicantly improves the disinfection eciency for all
pathogens such as bacteria, viruses, as well as cystforming protozoan
parasites.198 A key problem for these technologies (and possibly all
technologies) is related to regrowth of microorganisms.
Fig. 5 SEM images of P. aeruginosa cells sited in TiO2-coated polymer lms: (A) absence of
light; (B) after 30 min (1 kJ m2) UV-treatment. Lysed cells are in a box while cell debris are
marked with arrows.
Acknowledgements
We would like to thank researchers and colleagues who have worked in our
laboratories in the area of water treatments, including R. P. K. Wells, J. Sa,
C. Alcaraz Aguera, A. Sutherland, P. Sampedro-Tejedor, A. Kubacka, and
C. Colon and to acknowledge nancial support under project CICYT
CTQ2007-60480/BQU (Spain) and the Royal Society (London)-CSIC
(Joint international project grant). We are also grateful to the reviewers
of this article for their suggestions of additional sections and references.
References
1 http://www.mpch-mainz.mpg.de/Bsander/res/henry.html.
2 M. R. Homann, B. T. Martin, W. Choi and D. W. Bahnemann, Chem. Rev.,
1995, 95, 69.
3 A. L. Linsebigler, G. Lu and J. T. Yates, Chem. Rev., 1995, 95, 735.
4 D. W. Bahnemann, Solar Energy, 2004, 77, 445.
5 N. Serpone, J. Phys. Chem. B, 2006, 110, 24287.
6 G. Colon, C. Belver and M. Fernandez-Garc a, Nanostructured oxides in
Photocatalysis, in Synthesis Properties and Applications of Solid Oxides,
eds. J. A. Rodr guez and M. Fernandez-Garc a, Whiley, NY, 2007.
7 O. Carp, C. L. Huisman and A. Rellr, Prog. Solid State Chem., 2004, 32, 33.
8 M. A. Fox and M. T. Dulay, Chem. Rev., 1993, 93, 341.
9 J. Blanco and S. Malato, Tecnologa de Fotocatalisis Solar, Cuadernos
Monogracos, CIEMAT, 1996.
10 A. Fujishima, T. N. Rao and D. A. Tryk, J. Photochem. Photobiol. C, 2000, 1, 1.
11 U. Dietbold, N. Ruzycki, G. S. Herman and A. Selloni, Catal. Today, 2003,
85, 93.
12 J. K. Burdett, T. Hughbands, J. M. Gordon, J. W. Richarson and J. Smith,
J. Am. Chem. Soc., 1987, 109, 3639.
1. Introduction
Metal oxides are an important class of heterogeneous catalysts.13 They nd
direct application in a variety of reactions, from acid-base to redox
reactions, in photocatalytic processes, and as catalysts for environmental
protection. In addition, they are widely used as supports for other active
components (metal particles or other metal oxides), although often they act
not only as a support, but actively participate in the reaction mechanism.45
A key aspect of metal oxides is that they possess multiple functional
properties: acid-base, electron transfer and transport, chemisorption by s
and p-bonding of hydrocarbons, O-insertion and H-abstraction, etc. This
multi-functionality allows them to catalyze complex selective multistep
transformations of hydrocarbons,19 as well as other catalytic reactions
(NOx conversion, for example). The control of the catalyst multi-functionality
requires the ability to control not only the nanostructure, e.g. the nano-scale
environment around the active site,10 but also the nano-architecture, e.g. the
3D spatial organization of nano-entities. The active site is not the only relevant
aspect for catalysis. The local area around the active site orients or assists the
coordination of the reactants, and may induce sterical constrains on the
transition state, and inuences short-range transport (nano-scale level).
Therefore, it plays a critical role in determining the reactivity and selectivity
in multiple pathways of transformation. In addition, there are indications
pointing out that the dynamics of adsorbed species, e.g. their mobility during
the catalytic processes which is also an important factor determining
the catalytic performances in complex surface reaction,3,4 is inuenced by the
nanoarchitecture.
The nano-architecture is thus an important aspect to consider for the
design of novel catalysts and a critical element to consider also in analyzing
how to bridge the gap between model and real catalysts. In fact, in addition
to the issues of pressure and material gap, the complexity gap exists.1113
Goodman14 over ten years ago pointed out that despite the successes in
modelling catalysts with single crystals, there is a clear need to develop
models with higher levels of complexity and which take into account the 3D
nanoarchitecture.
Most of the studies on real nanostructured oxides are based on
materials not having a well-dened 3D structure (both on short and long-
range), being composed of irregularly shaped nano-crystals. These materials
are polycrystalline, and show several nano-interfaces, which stabilize
Fig. 1 Schematic model of the dierences between TiO2 thins lms and aligned nanorod array
of TiO2.
Titania nano-membranes
Another attractive area of development is to prepare catalytic nano-
membranes. First attempts to prepare self-organized, free-standing TiO2
nanotube membranes for ow-through photocatalytic applications was
reported by Schmuki et al.126 who showed a method to prepare by anodic
oxidation (ethylene glycol+0.2 M HF at 100 V for 10 h anodization)
robust, dense and free-standing membranes consisting of vertically oriented,
both-side-open TiO2 nanotubes. The array consists of regular tubes with a
diameter of 160 30 nm and a wall thickness of 20 5 nm. The layer is
overall 145 mm thick and has very smooth walls typical of nanotubes grown
in organic electrolytes.127 The TiO2 nanotube layers were separated from
the Ti substrates by selective metal dissolution. The TiO2/Ti specimens were
immersed into a mixture of Br2 and dry methanol for 12 h under a dry N2
atmosphere. This leads to a free standing nanotube layer oating in the
etching solution. After being rinsed with methanol and distilled water, the
layers were placed (closed tube side down) 12 cm above an open HF 48%
bottle for 30 min. This leads to HF condensation at the bottom and
preferential etching of the tube bottoms, i.e. the procedure opens the tubes.
After this treatment an opaque free-standing layer is obtained. The bottom
of the tubes is closed. Fig. 5 reports top and cross-section scanning electron
images (SEM) of this membrane prepared by Schmuki et al.126 for
size-selective, ow-through photocatalytic reactions (methylene blue
decoloration).
Similar TiO2 nanotube array membranes of uniform pore size distribution
were prepared also recently by Paulose et al.128 and tested in bioltration
applications. The size of the membranes was 12.5 cm2, a size limited by the
processing equipment. These membranes can be used in lab-scale tests, but
are fragile.
It should be mentioned that alternative possibilities to prepare similar
membranes include the use of a porous alumina membrane as matrix,
with the titania nanotubes grown in the channels.129133 Nanoporous
alumina membranes are commercial products, also synthesized by anodic
oxidation. The commercial Whatman Corporation anodic membrane has
holes of about 20-nm diameter at the top of the membrane and about
200-nm diameter at the bottom of membrane.133 Within these pores
TiO2 nanotubes fabricated by template synthesis and water vapour
hydrolysis could be grown, but non-uniform membrane characteristics
are obtained due to the non-uniform pores of the commercial alumina
Catalytic nanofactories
There is increasing interest in preparing TiO2 nanomembranes136 both for
advanced photocatalytic processes in the eld of air and water purication,137
purication of drinking water,138 novel membrane for high temperature PEM
fuel cells,139,140 Li-ion batteries,141143 advanced nanoelectrode arrays
(NEA)144 and nanoltration and pervaporation.145
The possibility of free-standing titania membranes with thickness of
about 150200 microns and straight channels of 100200 nm opens new
possibilities for catalysis. A grand-challenge for catalysis and sustainable
chemistry is to realize new materials able to perform multistep complex
reactions with ideally 100% selectivity. One approach to go in this direction
is the possibility to develop a tailored sequence of active centres inserted in a
channel (nanotube) of an ordered membrane to perform selectively the
conversion of molecules passing through this porous lm, minimizing at the
same time the possibility of side reactions. This concept was indicated as
catalytic nanofactories, because it resembles that of a production chain in a
factory, where multiple sequential operations are made in a chain-like
sequence. Fig. 6 shows schematically this concept.81 Although knowledge
Fig. 7 FESEM images of titania nanocoil produced by anodic oxidation (a). The cartoon
shows schematically the photocurrent generated by light irradiation of nanocoil containing a
catalyst particle (b) and the associated magnetic eld (c). Source: Centi and Perathoner.147
and chemical dissolution of the oxide and metal ions as soluble uoride
complexes,
TiO2 + 6F - [TiF6]2 (2)
Ti 4+
+ 6F - [TiF6] 2
(3)
Fig. 9 Rate of hydrogen generation from nanotube arrays lms of dierent lengths annealed
at 530 1C. Electrode area of 1 cm2; 100 mW/cm2 visible light. In the inset FESEM cross-
sectional image of 2.8 mm long TiO2 nanotube array prepared by anodic oxidation of a titanium
foil in an electrolyte containing potassium uoride (KF; 0.1 M), sodium hydrogen sulfate
(1 M), trisodium citrate (0.2 M) and sodium hydroxide. Elaborated from Grimes et al.270
Conclusions
Research in the eld of TiO2 nanostructured lms has progressed signicantly
over the last two years. This review integrates and extends to bidimensional
nanostructured systems (e.g. array of 1D nanostructures) the analysis
reported in part 160 on quasi-1D TiO2 nanostructures, e.g. nanorods,
nanowires and nanobres, nanotubes and nanopillars. However, it must be
commented that in the last two years, e.g. approximately the time spam from
the preparation of part 1,60 the progress in this eld has been so rapid that
several of the doubts presented in the earlier review have been in large part
solved. The aim of these two reviews is to highlight the potential and issues in
using these materials for catalysis, in particular the opportunities due to the
nanostructure. Still the number of examples of application of these materials
as solid catalysts in gas or liquid phase reactions is limited, but it was shown
that several of the barriers hindering their larger use have been solved or
nearly solved. We thus expect a large growth of examples and probably
applications in the near future.
Better established is their use as photocatalysts, in the photoelectro-
catalytic production of H2 or the elimination of pollutants, and in devel-
oping advanced electrodes for fuel cells, particularly for direct methanol or
ethanol oxidation. Nevertheless, also in this case the eld can be still
considered to be at an earlier stage. It has been shown how several of the
results have to be further demonstrated, and issues and limits better dened.
However, there are clear indications that this will be a major area of
research not only for this specic eld, but in general for all catalysis.
The recent US DoE report Catalysis for Energy also indicates that the
development of better tailored nanostructures for photo- and electro-
catalytic applications, particularly for better use of renewable resources, is
one of the priority areas of research in catalysis and in general of science.
We have limited the discussion here to TiO2 materials, but various
examples have been shown to indicate that both TiO2 nanotubes are a well
suited support for other catalytic elements, including bio-catalysts, and that
Acknowledgements
This work was realized in the frame of the EU Network of Excellence
IDECAT (Integrated Design of Catalytic Nanomaterials for a Sustainable
Production; NMP3-CT-2005-011730) and of the EU project NATAMA
(NMP3-CT-2006-032583) which are gratefully acknowledged.
References
1 Metal Oxide Catalysis, eds. S. D. Jackson and J. S. J. Hargreaves, Wiley VCH,
Weinheim, Germany, 2008, vols. 1 and 2.
2 Metal Oxides. Chemistry and Applications, ed. J. L. G. Fierro, CRC Taylor &
Francis, Boca Raton, US, 2006.
3 G. Centi, F. Cavani and F. Triro`, Selective Oxidation by Heterogeneous
Catalysis. Recent Developments, Kluwer/Plenum Publishing Corporation,
New York & London, 2001.
4 G. Centi and S. Perathoner, Curr. Opin. Solid State Mater. Sci., 1999, 4, 74.
5 Y. Ono, Catal. Today, 2003, 81, 3.
6 M. Misono, Catal. Today, 2005, 100, 95.
7 R. Schlogl, A. Knop-Gericke, M. Havecker, U. Wild, D. Frickel, T. Ressler,
R. E. Jentoft, J. Wienold, G. Mestl, A. Blume, O. Timpe and Y. Uchida, Top.
Catal., 2001, 15, 219.
8 B. K. Hodnett, Heterogeneous Catalytic Oxidation, J. Wiley & Sons,
New York, US, 2000.
9 P. Arpentinier, F. Cavani and F. Triro`, The Technology of Catalytic
Oxidations, Editions TECHNIP, Paris, France, 2001.
10 G. Centi and S. Perathoner, Catal. Today, 2003, 7980, 3.
11 H.-J. Freund, H. Kuhlenbeck, J. Libuda, G. Rupprechter, M. Baumer and
H. Hamann, Topics in Catal., 2001, 15, 201.
12 T. Ressler, B. L. Kniep, I. Kasatkin and Robert Schlogl, Angew. Chem., Int.
Ed., 2005, 44, 4704.
13 W. Weiss. R. Schlogl, Top. Catal., 2000, 13, 75.
14 D. W. Goodman, Chem. Rev., 1995, 95, 523.
15 C. R. Henry, in Catalysis and Electrocatalysis at Nanoparticle Surfaces,
eds. A. Wieckowski, E. R. Savinova and C. G. Vayenas, 2003, p. 239.
16 M. A. Barteau, J. E. Lyons and I. K. Song, J. Catal., 2003, 216, 236.
17 A. K. Datye, Top. Catal., 2000, 13, 131.
1. Introduction
The development of the catalytic science since its advent in the early 19th
century1 laid the foundation for the tremendous growth of the petroleum and
chemical industries in the 20th century, one of the factors directly responsible
for the substantial increase in the standards of living in the industrialized world.
Today, catalysts are responsible for over $3.3 trillion in global GDP (2002
gures2), much of which due to heterogeneous catalysis. Traditionally hetero-
geneous catalysis has been an empirical science. Commercially successful
catalyst formulations have emerged through thousands of experiments, many
of which involving complex mixtures of metals, metal compounds, promoters
and inhibitors, and functional supports.3,4 The development process is often
time- and resource-intensive and discoveries often serendipitous. Therefore,
researchers have always sought detailed understanding of how their catalysts
work, in order to approach catalyst development in a rational and scientically
well-grounded manner. The wedding of surface science methods, including
various spectroscopic, microscopic, and temperature-programmed techniques,
to catalysis in the latter half of the 20th century, an approach pioneered by such
prominent scientists as Gerhardt Ertl and Gabor Somorjai, is a huge step
forward for the catalytic science.
The beginning of the new millennium nds the cost of raw material and
energy rapidly rising, traditional feedstock showing signs of depletion, and
pollution taking a toll on our planets eco-system. Researchers are once
again being called upon not only to improve existing catalytic processes but
to develop entirely new ones in order to meet the continuously changing
demands for energy and chemicals production, which now also need to
reduce energy consumption, reduce or eliminate pollutants and toxic
wastes, and use alternative feedstock. The mainstream approach to catalyst
synthesis, testing, and characterization, however, remains largely in a top-
down mode and is limited in its ability to meet these new demands
eciently.
A new wave of research tools have emerged at the end of the 20th century that
hold the promise of ushering in a new area in heterogeneous catalysis research.
New imaging techniques such as STM and AFM oer a direct, spatially
resolved view of catalytic surfaces, important clues for the nature of individual
adsorbates and adsorbate-surface interaction, and sometime glimpses of mole-
cular transformation events, for a growing number of systems. These and many
spectroscopic techniques are also being continuously enhanced to provide better
time/spatial resolutions and to operate in increasingly realistic conditions (e.g.
high pressure or liquid phase instead of the traditional UHV).5,6 At the outer
Fig. 2 2006 price of 70 binary intermetallic compounds plotted against the calculated methyl
binding energies. The left region represents low reactivity and high selectivity, and the right
region represents high reactivity and low selectivity, for acetylene hydrogenation. Adapted
from Studt et al., Science, 2008, 320, 1320.41 r 2008, Reprinted with permission from AAAS.
Fig. 3 Measured kinetic current densities at 0.80 V as a function of the calculated repulsion
energy between two OHs on (Pt3M)ML/Pd(111) (M = Au, Pt, Ir, Pd, Rh, and Ru) or between
OH and O (M = Re and Os). All energies are in eV. With kind permission from Springer
Science+Business Media: A. U. Nilekar et al., Topics in Catalysis, 2007, 46, 276, Figure 7.47
The exchange of material with the environment, e.g., via the adsorption of
gas-phase species or the segregation of a component into/from the bulk of a
mixture, is accounted for by including the chemical potentials (Nimi) of
the independent chemical species involved. The chemical potential of
e.g. adsorbed oxygen, can be related to that of an O2 gas, which can
be calculated from tabulated thermodynamic properties,18,67 through the
assumptions of chemical equilibrium between surface O and gas-phase O2.
The resulting information can be graphically represented by phase diagrams
like those shown in Fig. 4.
For its simplicity and exibility FPT has been applied to many systems
relevant to heterogeneous catalysis, including adsorbates ranging from simple
gasses such as hydrogen and oxygen, to water, H2S, and larger compounds
such as ethylene and acrolein, on metals,6975 metal compounds,67,7680 and
nite particles.8183 For adsorption systems, the FPT framework is used to
compare the thermodynamic stability of disparate but related structures
(e.g., chemisorption, surface reconstructions, surface and bulk compounds)
on the same basis, thereby predicting surface coverage and possible phase
change. FPT has also been used to provide a measure of the acidity/basicity
of metal oxides through the evaluation of the concentration of Lewis
(unsaturated metal centers) and Brnsted (surface hydroxyls) sites in thermal
and chemical equilibrium with water.84,85
Fig. 4 Normalized surface energy as a function of temperature at (a) 1012 atm and (b) 15 atm,
for various phases of O on Ag(111). (c) The most stable phases plotted in the (T, pO2) space.
Reprinted with permission from Michaelides et al., Chemical Physics Letters, 2003, 367, 344.68
r Elsevier.
Fig. 5 Montage image combining an STM image of the Ag oxide structure (from bottom left)
superimposed over the proposed oxide structure (from top right). The numbers, n = 15,
correspond to the symmetrically dierent positions within the middle silver layer sandwiched
between two O layers. Ag1 and Ag2 have metallic character, as they are exclusively bonded to
silver atoms in the substrate below, whereas Ag3, Ag4, and Ag5 are directly bonded to oxygen
inside the oxide rings and are ionic in nature. Both Ag4 and Ag5 sites sit above threefold sites of
the underlying (111) lattice atoms, whereas Ag3 occupies a top site. Reprinted with permission
from Bocquet et al., Journal of the American Chemical Society, 2003, 125, 3119.65 r 2003,
American Chemical Society.
Fig. 6 STM images and partial structural models of the Ag-oxygen overlayers. (a) A 200 A2 patch
of Ag(111) covered by the p(4 4), c(3 5O3)rect, and p(4 5O3)rect overlayers, measured at
0.51 nA and 131.5 mV. (b) The Ag1.83O model for the p(4 4) proposed by Carlisle et al.106
Solid (open) large gray balls represent the overlayer (substrate) Ag atoms, and small balls the
oxygen atoms. (c, e) STM images of 35 A2 patches of the p(4 4) and c(3 5O3)rect over-
layers, respectively. (c) is at 0.42 nA and 21.7 mV, and (e) is at 0.40 nA and 34.2 mV.
The inset in (e) displays the six-atom structural element of the c(3 5O3)rect phase measured
with a tip state that allowed the resolution of the central parts of the proposed Ag6 triangles
(0.42 nA, 21.7 mV). (d, f) The proposed Ag6-based models for the p(4 4) and c(3 5O3)rect
phases, respectively. Figure adapted with permission from Schnadt et al., Physical Review Letters,
2006, 96, 146101.107 r American Physical Society.
Fig. 7 Surface phase diagram indicating the most stable surface structures for CuAg
bimetallic surface as a function of the Cu surface content and oxygen chemical potential.
Reprinted gure with permission from Piccinin et al., Physical Review B, 2008, 77, 075426.128
r American Physical Society.
Zq
E EDFT
hVidQ qU:
0
Fig. 8 Phase diagram showing the free energy of dierent surface structures of water at
pH = 0 in contact with Au(111), Pt(111), and Ni(111). The lowest line represents the
thermochemically most stable phase. The crossing of the two bottom lines indicates a phase
change. Reprinted with permission from Rossmeisl et al., Journal of Physical Chemistry B,
2006, 110, 21833.140 r 2006, American Chemical Society.
4. Outlook
Impressive progress in heterogeneous catalysis has been made with the help
of theory and rst-principles methods. The PdAu alloy catalyst created by
Goodman and co-workers for vinyl acetate synthesis represents an encoura-
ging trend in which mechanistic insights from experiment and theory merge
to become a general way of approaching catalysis research. Tremendous
opportunities as well as challenges lie ahead. Take the ab initio design of
catalysts as an example. The successes to date have largely focused on
optimizing a single step, and have been made possible in part by the
predictability of the geometric and electronic structures of metals and their
alloys. More complex reactions and applications will demand the simulta-
neous consideration of multiple descriptors. For instance, Janik et al. have
identied a set of criteria for designing improved anode catalysts for direct
methanol fuel cells.130 These include preferential breaking of CH vs. OH
bonds, weakened binding of CO vs. Pt, ability to activate water and oxidize
CO, and comparable rate of methanol oxidation to CO to Pt. Their
calculations suggest the ternary surface alloy PtRuAu/Ru(0001) as a
promising candidate, which remains to be tested. The complexity in design
variables and the amount of data that theory will generate may one day rise
to a level where they pose a challenging optimization and organization
problem.
A comparable paradigm of success is still to be established for more
complex catalytic materials (e.g., nite metal/compound clusters, mixed
oxides, zeolites), many of which are highly important technological catalysts
and often contain multiple active centers that are required to achieve a series
of specic transformations. The local structures of these catalysts and
Fig. 9 STM image (48 53 A2; Vt = 5.3 mV; It = 1.28 nA) showing a MoS2 nanocluster
synthesized on a reconstructed Au(111) surface. The bright brim extending all the way around
the edge is due to localized metallic edge states. Reprinted gure with permission from Bollinger
et al., Physical Review Letters, 2001, 87, 196803.147 r American Physical Society.
Acknowledgements
The author acknowledges support by Oak Ridge National Laboratorys
Center for Nanophase Materials Sciences, which is sponsored by the
Scientic User Facilities Division, Oce of Basic Energy Sciences, US
Department of Energy.
References
1 M. W. Roberts, Catalysis Letters, 2000, 67, 1.
2 D. Filmore, Todays Chemist At Work, 11/2002, 33.
3 J. H. Sinfelt, Bimetallic Catalysts: Discoveries, Concepts, and Applications,
Wiley, New York, 1983.
4 C. N. Sattereld, Heterogeneous Catalysis in Industrial Practice, McGraw-Hill,
1991.
5 J.-D. Grunwaldt and A. Baiker, Physical Chemistry Chemical Physics, 2005, 7,
3526.
6 MRS Bulletin, 2007, 32.
7 S. Baker, J. S. Robinson, C. A. Haworth, H. Teng, R. A. Smith, C. C. Chirila,
M. Lein, J. W. G. Tisch and J. P. Marangos, Science, 2006, 312, 424.
8 M. B. J. Roeaers, B. F. Sels, H. Uji-i, F. C. De Schryver, P. A. Jacobs,
D. E. De Vos and J. Hofkens, Nature, 2006, 439, 572.
9 J. P. Perdew, J. A. Chevary, S. H. Vosko, K. A. Jackson, M. R. Pederson,
D. J. Singh and C. Fiolhais, Physical Review B, 1992, 46, 6671.
10 J. P. Perdew, K. Burke and M. Ernzerhof, Physical Review Letters, 1996, 77,
3865.
11 D. R. Hamann, M. Schluter and C. Chiang, Physical Review Letters, 1979, 43,
1494.
12 D. H. Vanderbilt, Physical Review B, 1990, 41, 7892.
13 P. E. Blochl, Physical Review B, 1994, 50, 17953.
14 G. Kresse and D. Joubert, Physical Review B, 1999, 59, 1758.
Introduction
The development of ionic liquids dates to 1914. The rst research eorts
involved the synthesis of ethylammonium nitrate.1 Hurley and Wier at the
Rice Institute in Texas, 1948, developed the rst ionic liquids with chloro-
aluminate ions as bath solutions for electroplating aluminum.2 These liquids
were studied primarily for their applications as electrolytes in electro-
chemistry technologies such as electroplating, batteries and alloy
preparations.
An ionic liquid (IL) is a substance that is composed entirely of ions, and is a
liquid at room temperature. Frequently the ionic liquid consists of organic
cations and inorganic anions, although it is not limited to these combinations.
While some people have said that the ionic liquid can have a high melting
temperature such as in the case of the molten salt form of NaCl, the most
commonly held understanding of this term is one that has a melting point of
less than 100 1C, more preferably less than 50 1C. For example, many
preferred ionic liquids are liquid at room temperature, or less.
The cations of the ionic liquid include organic and inorganic cations.
Examples of cations include dialkylimidazolium ion, tetra-alkylphosphonium
ion, etc. and these cations can be associated with a number of dierent
anions (Fig. 1). The anion includes organic and inorganic anions such as
PF6, CF3SO3, CF3COO, etc. and Lewis acids such as AlCl4, GaCl4,
etc.
Many ionic liquids have been widely investigated with regard to applica-
tions other than as liquid solvents: such as electrolytes, phase-transfer
reagents,3 surfactants,4 and fungicides and biocides.5,6 The physical and
chemical properties of ionic liquids can be varied over a wide range by the
selection of suitable cations and anions. Some of the properties that
depend on the cation and anion selection includes: melting point, viscosity,
density, acidity and coordination ability, solvation strength and solubility
characteristics.7
Changes in ion types, substitution, and composition produce new ionic
liquid systems, each with a unique set of properties that can be explored.
With the potential large matrix of both anions and cations, it becomes clear
that it will be impossible to screen any particular reaction in all possible
ionic liquids. Work is clearly needed to determine how the properties of
ionic liquids vary as functions of anion/cation and establish which, if any,
properties change in a systematic way.
Scheme 1
Review articles
We begin with a brief summary of some of the review articles that have been
written on the subject of ionic liquids. Wilkes13 wrote a short history of ionic
liquids describing the chronological development of ionic liquids with an
emphasis on listing the names and pictures of those involved in the research.
Holbrey and Seddon14 and Earle and Seddon15 reviewed the literature of
ionic liquids composed entirely of ions which were mainly of interest to
electrochemists. Recently, however, it has become apparent that, inter alia,
their lack of measurable vapor pressure characterizes them as green solvents,
and that a wide range of chemical reactions (reviewed here) can be performed
in them. Wassercheid and Keim16 reviewed the literature of ionic liquids, not
only the synthesis and physical properties of the ILs, but also their use as
Alkoxy carbonylation
Much less attention has been focused on carbonylation reactions in ionic
liquids. The biphasic palladium-catalyzed alkoxycarbonylation of styrene,
Scheme 2, in [bmim][BF4]cyclohexane has been reported.28
Scheme 2
Arene carbonylation
Carbonylation of the arene system can be achieved using highly acidic media
such as triic acid or sulfated zirconia. Such highly acidic conditions can also
be created in certain ionic liquids derived from the chloroaluminate anions
and the appropriate organic cations, such as the dialkylimidazolinum
cations or pyridinium cations. Ionic liquids containing chloroaluminate
(AlCl4, Al2Cl7) anions are strong Lewis acids and if protons are present
they are superacidic. Coupled with the fact that they are relatively easy to
handle makes these materials attractive non-volatile alternatives for standard
Scheme 3
Catalytic oxidations
Considering the commercial importance of catalytic oxidations, and the fact
that ionic liquids are expected to be relatively inert towards autoxidation
with O2, surprisingly little attention has been devoted to performing such
reactions in ionic liquids. The Ni(acac)2-catalyzed aerobic oxidation of
aromatic aldehydes, to the corresponding carboxylic acids, in [bmim][PF6]
has been described.40 However, rather high (3 mol%) catalyst loadings
were used and this can hardly be considered a challenging oxidation. The
methyltrioxorhenium (MTO)-catalyzed epoxidation of olens with the
ureaH2O2 adduct (UHP) in [emim][BF4] has been reported.41 Both
the UHP and the MTO are soluble in [emim][BF4] and the medium remains
homogeneous throughout the reaction. It should be noted, however, that
the substrates were generally highly reactive olens and when the more
challenging dec-1-ene was used, a long reaction time (72 h) was needed for
moderate conversion (46%), using 2 equivalents of oxidant. When 30% aq.
H2O2 was used as the oxidant, there was ring opening of sensitive epoxides.
The asymmetric Jacobsen-Katsuki epoxidation with NaOCl, catalyzed by a
chiral Mn Schi-base complex, has been conducted in [bmim][PF6].42
However, dichloromethane was required as a cosolvent, as the ionic liquid
solidies at the reaction temperature (0 1C), which nullies a primary
incentive for using an ionic liquid. The ionic liquid containing the catalyst
could be recovered and recycled four times, albeit with a signicant loss in
yield. A more recent and very exciting development is the electro-assisted
biomimetic activation of molecular oxygen by a chiral Mn Schi-base
complex in [bmim][PF6], described by Gaillon and Bedioui.43 Evidence
was provided for the formation of the highly reactive oxomanganese(V)
intermediate that could transfer its oxygen to an olen. This would appear
to oer potential for clean, electro-catalytic oxidations with molecular
oxygen in ionic liquid media.
Scheme 5
The potential activation of dierent Lewis acid catalysts and their load
eect when used in combination with this solvent were explored, in order to
determine the improvement of rates and selectivity to the endo and exo
isomers. The list of Lewis acid catalysts included: Li(OTf), Li(NTf2), ZnI2,
AlCl3, BF3, HOTf, HNTf2, Ce(OTf)4*5H2O, Y(OTf)3, Sc(OTf)3, Sc(NTf2)
and a blank without any Lewis acid. The reaction conditions were as follows:
2.2 mmol of cyclopentadiene + 2.0 mmol of dienophile + 0.2 mol% of
catalyst in 2 mL [hmim][BF4]. When no catalyst was added, the two ketones
(R1QMeCQO; R2 = R3 = H; and R1QEtCQO; R2 = R3 = H) showed
modest activity (B50% in 1 h) with endo/exo selectivity = 85/15. Whereas
acrolein showed modest activity (59% conversion in 2 h), methacrolein
and crotonaldehyde were inert without a Lewis acid catalyst. Acrylonitrile
and methyl acrylate underwent low conversions in 1 h (1617%) whereas,
N-phenylmaleimide, maleic anhydride and 2-methyl-1,4-benzoquinone
showed complete reaction in 5 min with high endo isomer yields.
In subsequent studies, methyl vinyl ketone (2.0 mmole) was chosen as the
dienophile so as to determine the combined eect of the ionic liquid (2 mL)
and the Lewis acids (0.2 and 0.5 wt%) upon the yield and selectivity. Without
the Lewis acid catalyst, this system demonstrated a 52% conversion of the
cyclopentadiene (2.2 mmol) in 1 h with the endo/exo selectivity being 85/15.
The cerium triate-catalyzed reaction was quantitative in 5 min and the
endo:exo selectivity was very good for this experiment as well (94:6, endo:exo).
Also with the scandium or yttrium salts tested, reactions came to completion
in a short time with high stereo-selection. Cerium, scandium and yttrium
triates are strong Lewis acids known to be quite eective catalysts in the
cycloadditions of cyclopentadiene with acyclic aldehydes, ketones, quinones
and cycloalkenones. These compounds are expected to act as strong Lewis
acids because of their hard character and the electron-withdrawing triate
group. On the other hand, reaction times of 1 hour were required for
Scheme 6
The reaction conditions were room temperature and pressure and the IL
was developed from 1-butyl-3-methylimidazolium hexauorophosphate
using 1-4 mol% Pd(OAc)2, 1 mol % Lewis acid, and 0.5 mL of IL. The
reaction times were from 2.5 to 4.5 h. The Lewis acids were chosen from the
following: Cu(OTf)2, Cu(CF3CO2)2, Zn(OTf)2, and In(OTf)3.
Complete conversion of the substrate was observed with reaction times as
short as 2.5 h when copper triate was the Lewis acid, and the selectivity
was usually greater than 98% towards the desired compound. Upon reuse
of the same catalyst system with fresh substrate, it was reported that the
conversion was 80% and 76% for the second and third reuse of the catalyst.
Scheme 7
By this scheme, the Pd and Cu complexes were consumed with repeated use
and thus the conversion should decrease accordingly. More evidence must
be presented to conrm this mechanism.
Wasserscheid and Eichmann47 have investigated the dimerization of
1-butene in chloroaluminate ILs using (cod)Ni(hfacac) catalyst. This
biphasic mode of operation had several advantages over running the
reaction in toluene as the solvent, such as high activity even at low
temperatures, high turnover frequency and no detectable leaching of
the catalyst. Thiele and de Souza48 have studied the eect of adding a
co-catalyst (AlEtCl2) when [Ni(MeCN)6][BF4] was used as a catalyst in
chloroaluminate IL. They concluded that the neutral IL was the best
medium for the dimerization reaction.
Enzyme-catalyzed reactions
Kim and Lee49 disclosed an ionic liquid-coated enzyme that remarkably
improves enzyme functions, such as enantioselectivity and stability, when
the enzyme, which may be lipase, is coated with an ionic liquid. Further,
even in the case where the ionic liquid-coated enzyme is reused, the
Scheme 8
Scheme 9
Scheme 10
Scheme 11
Scheme 12
Scheme 13
Heck reaction
The Heck reaction is a CC coupling reaction where an unsaturated
hydrocarbon or arene halide/triate/sulfonate reacts with an alkene in
presence of a base and Pd(0) catalyst so as to form a substituted alkene.
Kaufmann et al.59 showed that the Heck reaction carried out in presence
of ILs such as tetra-alkyl ammonium and phosphonium salts without
the phosphine ligands, resulted in high yields of product. They attributed
the activity to the stabilizing eect of ammonium and phosphonium salts on
Pd(0) species. Carmichael et al.60 used ionic liquids containing either
N,N 0 -dialkylimidazolium and N-alkylpyridinium cations with anions
such as halide, hexauorophosphate or tetrauoroborate to carry out
reactions of aryl halide and benzoic anhydride with ethyl and butyl
acrylates in presence of Pd catalyst. An example of iodobenzene
reacting with ethyl acrylate to give trans-ethyl cinnamate is shown in
Scheme 14.
The authors showed that the above reaction carried out in N-hexyl-
pyridinium salts gave higher yields than imidazolium salts, that the addition
of a phosphine ligand to Pd reduced the yield in pyridinium salts, and that
higher reaction temperatures were required to obtain high yields. Howarth
and Dallas61 studied the use of [bmim][PF6] as an IL solvent for the reaction
of aryl halides and methyl acrylate. They found that the yields of the
products were comparable with that obtained with DMF, and that the
solvent and catalyst could be reused several times.
Calo`s group62,63 have reacted both a-substituted and b-substituted
acrylates with haloaromatics in presence of a Pd-benzothiazole carbene
complex using tetrabutyl ammonium bromide as the solvent. The reactions
were found to be fast and ecient in the ILs when compared to
conventional solvents.
Scheme 15
The authors oered no explanation for the apparent order of reactivity with
these ILs, and they could be reused up to ve times with only small losses of
apparent reactivity.
Hydrogenation in ILs
Researchers65 performed the biphasic hydrogenation of cyclohexene with
Rh(cod)2 BF4 (cod = cycloocta-1,5-diene) in ILs. They observed roughly
equal reaction rates, reported as turnover frequencies of ca. 50 h1, in either
[bmim][BF4] or [bmim][PF6]. The presumption here was that the [bmim][BF4]
was free from chloride. In a separate report, the same group showed that
RuCl2(Ph3P)3 in [bmim][BF4] was an eective catalyst for the biphasic
hydrogenation of olens, with turnover frequencies up to 540 h1.66
Similarly, (bmim)3-Co(CN)5 dissolved in [bmim][BF4] catalyzed the hydro-
genation of butadiene to but-1-ene, with 100% selectivity at complete
conversion.
More recently, the ruthenium-catalyzed hydrogenation of sorbic acid to cis-
hex-3-enoic acid, Scheme 16, was achieved in a biphasic bmim-PF6methyl
tert-butyl ether (MTBE) system.67 The ruthenium cluster [H4Ru(Z6-C6H6)4]
[BF4]4, in [bmim][BF4], was shown to be an eective catalyst for the
hydrogenation of arenes to the corresponding cycloalkanes at 90 1C and
60 bar.68 The cycloalkane product formed a separate phase, which was
decanted and the IL phase, containing the catalyst, could be repeatedly
recycled.
Scheme 16
Scheme 18
Scheme 20
Hydroformylation
Hydroformylation of propene in an aqueous biphasic system, using a water-
soluble rhodium complex of the sodium salt of trisulfonated triphenyl-
phosphine (tppts), forms the basis of the Ruhr Chemie-Rhone Poulenc
process for the manufacture of butanal.75 Unfortunately this process is
limited to C2 to C5 olens owing to the very low solubility of higher olens
in water. Hence, one can envisage that the use of an appropriate IL could
provide the basis for biphasic hydroformylation of higher olens. As noted
earlier, Parshall showed in 1972 that platinum-catalyzed hydroformylations
could be performed in tetraethylammonium trichlorostannate melts.76
More recently, Waenschmidt and Wasserscheid77 studied the platinum-
catalyzed hydroformylation of oct-1-ene in [bmim][SnCl3], Scheme 21,
which is a liquid at room temperature. Despite the limited solubility of
oct-1-ene in the IL, high activities (TOF = 126 h1) were observed together
with a remarkably high regioselectivity (n/iso = 19). The product was
recovered by phase separation and no leaching of platinum was observed.
Scheme 21
Scheme 22
Scheme 23
Isomerization
Herbst et al.,95 report a process for the conversion of linear and/or branched
parans based on the use of an IL catalyst in combination with a metal salt
additive, which provides a catalytic composition of increased activity
compared with the IL alone. Under suitable reaction conditions this
conversion leads to paran hydrocarbon fractions having higher octane
numbers.
The hydrocarbon feed used for the isomerization experiments was a mixture
of 19 wt% n-heptane, 21 wt% 2-methylhexane, 21 wt% 3-methylhexane,
36 wt% methyl-cyclohexane, 1 wt% 2,4-dimethylpentane, 1.7 wt%
2,3-dimethylpentane and 0.3 wt% of other C7 isomer compounds. The ILs
described here were developed from excess AlCl3 and trimethylamine chloride.
Anhydrous metal chlorides, such as CuCl2, CuOHCl, CuSO4, CoCl2, FeCl3,
MnCl2, MoCl5, NbCl5, NiCl2, TiCl3, or ZnCl4, were then added to the IL
before the linear alkane was added. Data for the eect of adding these Lewis
acids to the reaction mixture showed that incremental yields per mole of
additive are dierent. Adding a small amount of AlCl3 (B2 g) to the mixture
results in the highest incremental yield increase; whereas, adding more AlCl3
(5 g) results in lower incremental yields, i.e., diminishing returns are observed.
Co, Zr and Nb chlorides were also shown to have positive incremental yields
upon their addition; however, smaller amounts of AlCl3 showed the highest
incremental yields. On the other hand, negative incremental yields were
observed for the addition of Ni, Mn, and Ti chlorides (Fig. 4).
AlCl3 4 ZrCl4 4 TiCl3 B NbCl5 B MnCl2 B CoCl2 B NiCl2
Apparently, these metal chlorides that are softer acids than AlCl3 are
therefore not as eective as the harder acids in catalyzing the paran
isomerization.
Metathesis of olens
Hembre et al.96 disclosed a process for which one olen selected from certain
a,b-dihydroxyalkenes and 4-(alkenyl)ethylenecarbonates was reacted with a
second olen to produce a metathesis product. When the rst olen reactant
was an optically enriched or enantiomerically pure a,b-dihydroxyalkene,
cross metathesis reactions result in products having the same optical purity.
The a,b-dihydroxyalkenes and the 4-(alkenyl)ethylene carbonates may be
converted to hydrogenated products, and the 4-(alkenyl)ethylenecarbonates
may be decarboxylated to provide the corresponding epoxides. The
decarboxylation process may be carried out in a solvent in the presence of
a decarboxylation catalyst. Halide anions are eective decarboxylation
catalysts and thus a catalyst may be selected from group IA halide salts.
The use of a solvent to favor the solubility of these salts or the addition of
agents such as crown ethers to enhance the solvation of a salt in a given
solvent may be used to increase their catalytic activity. For similar reasons,
one or more ammonium or phosphonium salts of the halides may be used as
decarboxylation catalysts. In this case, the ammonium and/or phosphonium
salts are known as ionic liquids and can serve both as the reaction solvent
and the decarboxylation catalyst. Because ILs have very low vapor pressures,
they can be a very eective catalyst/solvent medium for the production of
relatively low-boiling epoxides, which can be removed by fractionation.
Michael reaction
The Michael reaction is the nucleophilic addition of a carbanion to
a,b-unsaturated carbonyl compounds. It is a useful way to make CC
and Chetero atom bonds. Karodias group97 studied the use of the ionic
liquid ethyltri-n-butylphosphonium tosylate (n-Bu3PEtOTs) as a solvent for
Scheme 24
Scheme 25
Scheme 26
Sonogashira reaction
The Sonogashira reaction is a CC coupling reaction of terminal alkynes
with aryl or vinyl halides in presence of Pd(0) metal and/or Cu(I) catalyst.
These compounds are useful in synthesizing species having pharmaceutical
Scheme 27
Scheme 28
Nucleophilic additions were studied using the same TSIL with pyrrolidine
and thiophenol as models. As with the Diels-Alder reaction above, the reaction
gave the required adducts which were then transesteried to give the nal
products. Heck coupling catalyzed by a transition metal and the Stetter
reaction, Scheme 30, to prepare 1,4-dicarbonyl compounds were also studied
by the same group using similar TSILs.
Scheme 30
Telomerization in ILs
Telomerization is the reaction of olens having conjugated double bonds
(conjugated dienes) in the presence of a nucleophile (telogen). The main
Acknowledgements
The authors acknowledge helpful conversations with Dr E. J. Angueira
(Ciba Specialty Chemicals, McIntosh, AL) in selecting the literature to be
reviewed. We also acknowledge support from the Earnest W. Deavenport, Jr
endowed chair.
References
1 S. Sugden and H. Wilkins, J. Chem. Soc., 1929, 12911298.
2 (a) F. H. Hurley, US Patent 2,446,331, 1948; (b) F. H. Hurley and T. P. Wier,
J. Electrochem. Soc., 1951, 98, 207212.
3 D. Albanese, D. Landini, A. Maia and M. Penso, J. Mol. Catal. A, 1999, 150,
113.
4 E. Blackmore and G. Tiddy, J. Chem. Soc., Faraday Trans. 2, 1988, 84, 1115.
5 J. Pernak, J. Krysinski and A. Skrzypczak, Pharmazie, 1985, 40, 570.
6 J. Pernak, A. Czepukowicz and R. Pozniak, Ind. Eng. Chem. Res., 2001, 40,
2379.
7 P. Wasserscheid and W. Keim, Angew. Chem. Int. Ed., 2000, 39, 37723789.
1. Introduction
The initial observations regarding a catalytic eect are attributed to Sir
Humphrey Davy and Russian chemist Gottlieb Kirchho whom both
observed catalytic eects by 1812. Davy was evidently the rst to observe
that platinum induced the oxidation of alcohol vapor in air1 and Kirchho
was the rst to observe the liquid phase catalysis of acid breaking down
starch.2 That was the likely beginning of the experimentally based study of
catalysis that continues today. Over the years, catalysis research developed
broad categories that still govern most current research eorts. The
categories can be generally classied as experimental and theoretical and
more specically, the experimental consists of preparation techniques,
activity-selectivity testing and kinetic and mechanistic elucidation. The
specic theoretical techniques are recently more capable of predicting
electronic and mechanical bulk properties of more complex systems based
on electronic wave function or electron density calculations.
There has been signicant integration of these categories that has enabled
a tremendous understanding of catalytic reactions, cycles and eect. One of
the many signicant contributions to the catalysis eld was the observation
that the catalyst structure changes during the reaction that it is aecting.
Gabor Somorjai conducted the rst experiments that compared surface
diraction patterns of a crystal both before and after it was exposed to
reactive gases.3 That work set the stage for the molecular investigation of
catalysis.
The grand challenge to catalysis has become the a-priori design of a
catalyst for a given reaction. That is, select the specic chemicals (i.e. Pt, Re,
Ni, La, CeO) in the correct ratios, prepare the catalyst, and deposit it onto a
support (i.e. g-Al2O3) and substrate (e.g. monolith), put the substrate into
the reactor and achieve a predicted result. This is not possible currently.
Thus we must resort to sets of preparation-activity correlations, structure-
activity correlations and idealized modeling eorts. The combination of
these sets has and will continue to provide a very detailed understanding of
catalytic reaction engineering and development. Not until we can accurately
and completely describe the electronic nature of irregular surface structures
containing several dierent atoms will we be able to come close to this
a priori ability. Therefore, the frontier of catalysis today is the development
of analysis tools to help us see the reactions occurring during the catalytic
cycle along with the modeling tools to match that data generated. This
2. Structure to kinetics
The relationship between the morphology of the catalytic surface and the
performance of a system is referred to as the structure-activity relationship
(SAR). When a catalytic reaction takes place, there exists intermediate
chemical species, which are dierent from both the reactants and products.
Intermediate state species are the result of inter- and intra-molecular
processes and bond changes when the reacting substance and the catalyst
interact. This change enables faster reaction rates and higher activity.
Compounds are often classied together if they have similar structural
characteristics, including shape, size, stereochemical arrangement, and
distribution of functional groups. Factors contributing to SAR include
chemical reactivity, electronic eects, resonance, and inductive eects. SAR
was observed and postulated in 19074,5 and has since been extended and
formalized by Boudart.68 Currently there are two main categories of
SAR: primary SAR, caused by geometric eects, and secondary SAR,
caused by preferential poisoning of sites of a given geometry and coordina-
tion. Moreover, structure sensitive reactions typically occur on large multi-
ple atom sites, which are very sensitive to alloy or poison eects and involve
the activation of bonds (i.e. CC or NH). A signicant number of studies
have been done on various catalytic metals ranging from base metals such as
Ni, Cu and Fe to precious metals such as Au, Pt, Pd and Ir.
The use of two SAR categories is a very powerful methodology to
organize catalyst performance. It should be recognized, however, that
SAR performance is not just a two category phenomena; rather, it is a
spectrum where the catalysts have varying degrees of behavior.9 Informa-
tion concerning the mechanism and kinetics of a reaction can be determined
by knowing the intermediate chemical species formed on the surface by
activating bonds. The connection to kinetics is the impact of structure on
activity, manifested in the evaluation of the Turnover Frequency (TOF).7
The TOF is the intrinsic value of the rate of reaction on a given active site
aorded by its structure. The active site is only expected to stay constant
when it is comprised of a single metal atom. The number of surface metal
atoms (counted via titration or other techniques) is equated to the number
of points at which the reaction can proceed, varying the capacity of a
catalyst to adsorb gas and promote reaction.10 Researchers have quantied
Fig. 1 Raw and smoothed data of conversion (x) versus clock time (t) for dierent runs
showing varying degrees of catalytic activity.34
Fig. 2 Reactor system used to quench the catalyst surface after testing and that allows for
in situ measurements during reaction.37
Infrared spectroscopy
Since 1905, when Coblentz obtained the rst IR spectrum, vibrational
spectroscopy has become an important analytical research tool. This
technique was then applied to the analysis of adsorbates on well-dened
surfaces, subsequently moving towards heterogeneous reaction studies.
Terenin and Kasparov (1940) made the rst attempt to employ IR in
adsorption studies using ammonia adsorbed on a silica aerogel containing
dispersed iron. This led to a prediction by Eischens et al. from Beacon
Laboratories in 195653 that the IR technique would prove to be extremely
important in the study of adsorption and catalysis. For an excellent review
article in IR spectroscopy, see Ryczkowski54 and references therein; and for
a more recent review with applications, see Topsoe.55
IR spectroscopy is likely the most frequently used technique in studying
heterogeneous catalysis. Its conguration and use is fairly straight forward,
making it relatively simple to carry out in situ studies. Transmission
spectroscopy involves passing infrared radiation through a sample and then
measuring the extent of absorption. The information obtained provides
insight into important parameters, ranging from characterization of
supports to the nature of adsorbed molecules and reaction intermediates.
IR studies have helped establish correlations between observed IR para-
meters and the catalyst activity, allowing researchers to move towards a
more rational design of industrial catalysts. Fig. 6 shows a typical data set
that is obtained. The data shows the dierence in the spectra going from an
as-prepared, non-reacting sample to the sample in the working state. TOS is
Fig. 6 FT-IR spectra of catalytic surface during n-butane isomerization at 323 K taken every
16 minutes. This shows the dierence in IR spectra from before activation to during reaction.
The black lines show the spectra rst at the start of the reaction, and the next is at the maximum
reaction rate.56
Fig. 7 Schematic IR reactor system enables time-resolved kinetic and IR spectroscopic data
measurements: (a) IR cell; (b) total set up.5759
Fig. 8 (a) DRIFTS spectra of catalyst surface in NO ow. Absorbance increases with time.
(b) In situ Raman spectra measured while methanol owing over catalyst.61
Raman spectroscopy
The previous discussion focused on IR spectroscopy, a method that involves
probing molecules with photons that resonate its vibrational frequencies.
Raman spectroscopy, however, is based on detection of scattered photons
from absorbed species which interfere weakly with signals from the gas phase.
This makes Raman analysis complementary to IR analyses. It is an excellent
in situ technique for catalyst structure, but is not very surface sensitive.
Moreover, the common alumina supports of most catalysts are weak Raman
scatterers. This positions Raman to be well suited for in situ analysis because
of its wide frequency range, from 50 to about 5000 cm 1 and high resolution
of nearly 1 cm 1. Since the gas phase yields almost no signal, Raman is
practical for studying surface reactions. The information obtained from a
single measurement gives insight into the molecular reactants and the solid
catalyst surface. Raman has been most valuable in analysis of oxide catalysts,
since the metal-oxygen vibration frequency is dicult to probe with IR
because it is typically between B400 and 1000 cm 1.
Not all vibrations that get excited by the incident light are observable,
such as the weak vibrations of a highly polar moiety (i.e. the OH bond). An
external electric eld is not capable of inducing large changes in the dipole
moment, and stretching or bending the bond does not change this. A
vibration will be Raman active if the molecule changes shape, thus changing
the polarizability of the molecule. Typical strong Raman scatterers have
distributed electron clouds, such as carboncarbon double bonds. The
pi-electron cloud of the double bond is easily distorted in an external
electric eld. Bending or stretching the bond changes the distribution of
electron density substantially, and causes a large change in induced dipole
moment.
Raman typically provides information on a catalysts crystallinity,
coordination of site structure and spatial distribution of phases through
can answer all questions about the catalyst structure activity and structure
selectivity relationship.
As noted above, it should be realized that understanding the activity on a
surface is not the only issue needed to fully characterize the catalytic
system. There must be an understanding of how the surface reactions
produce intermediates that can desorb and initiate gas phase reactions close
to the surface. To that extent, the author is developing a catalytic shock
tube technique to probe the interface of the catalyst surface and the
immediate gas phase layer, bridging the gap between surface mechanisms
and gas phase mechanisms.
Fig. 13 Schematic of proposed apparatus for catalyst surface reaction studies in the absence of
transport eects.
Fig. 15 Analysis showing experimental conditions do not have diusion in the washcoat.
Fig. 17 Representative data from catalytic shock tube: pressure trace of a Mach 1.3 shock.
un-catalyzed screen with combustible gases, and catalyzed screen with inert
gases. The pressure traces in Fig. 16 are examples of the test data, showing
the pressure at a point just downstream of the SCT. The data shows the
shocks arrival and followed by rapid return to non-reacting steady state
conditions. The three tests enable comparison between non-reacting and
reacting conditions. The resulting pressure traces show that both the
catalyzed screen with inert gas and the un-catalyzed screen with combustible
gases behave the same. The reactive mix of the catalyzed screen with
combustible gases exhibits dierent pressure behaviors, holding the pressure
longer after an observed ignition delay. As evidenced from Fig. 17 the
tests show that after the ignition delay time, there was nearly 1% conversion
of the methane, giving o heat and manifesting as a temperature increase
(20 K) and pressure increase (1.5 psi).
Computational techniques
To be sure all experimental methods need to be complemented by theoretical
techniques. The calculational techniques started with ab-initio and quantum
calculational methods, such as MOPAC and GAMESS. These methods focus
on the solution for the wave functions of the system being modeled. Those
computations enabled calculations to be done in a sequence of frozen
congurations of the catalyst and the gas phase molecules approaching the
surface. The calculations produced thermodynamic energetic and entropic
eects as the reaction coordinate changed, bring a reactant closer to the
5. Future directions
The review just presented has provided an overview of the more well used
techniques for catalyst mechanism investigation. As faster computer power
becomes available and more precise engineering, i.e. fabrication of nano-
structures, becomes more developed catalyst science will progress to the next
level. The next 30 years will include coupling of multiple investigation
techniques with sophisticated computational analysis to further understand
the role of catalyst surfaces for many industrially relevant reaction sequences.
It is likely the next signicant area of progress will be modication of
existing catalysts to achieve better selectivity toward desired reactions.
Activity improvements will continue but at a slower pace. The advent of
nano-engineering with unprecedented visualization of surfaces is leading in
the direction of a priori designed catalysts. As our understanding of multiple
classes of reactions grows, it is very likely within the next 50 years a catalyst
will be designed from rst principles to achieve a desired selectivity. Once
that occurs, the next frontier will be to design a catalyst for a desired
durability with predetermined activity and selectivity.
References
1 The Collected Works of Sir Humphry Davy, ed. J. Davy, Smith, Elder, and Co.,
Cornhill, London, 1840.
2 L. Gmelin, Hand-book of Chemistry, Harrison and Sons, London, 1862.
3 S. Westerberg, C. Wang and G. A. Somorjai, Surf. Sci., 2005, 582, 137144.
4 J. Boeseken, Recl. Trav. Chim. Pays-Bas Belg., 1907, 26, 285288.
5 J. Boeseken, Recl. Trav. Chim. Pays-Bas Belg., 1911, 29, 330339.
6 M. Boudart, Actas Simp. Iberoam. Catal., 1984, 1, 318.
7 M. Boudart and G. Djega-Mariadasson, Kinetics of Heterogeneous Catalytic
Reactions, 1984.
8 M. Boudart and D. G. Loer, J. Phys. Chem., 1984, 88, 5763.
9 G. C. Bond, Acc. Chem. Res., 1993, 26, 490495.
10 H. S. Taylor, Proc. R. Soc. London. Ser. A, 1925, 108, 105111.
11 D. Farin and D. Avnir, J. Am. Chem. Soc., 1988, 110, 20392045.
12 J. J. Carberry, J. Catal., 1987, 107, 248253.
13 R. D. Cortright and J. A. Dumesic, Adv. Catal., 2001, 46, 161264.
14 M. J. Castaldi, A. M. Vincitore and S. M. Senkan, Combust. Sci. Technol.,
1995, 107, 119.
15 M. J. Castaldi and S. M. Senkan, Combust. Sci. Technol., 1996, 107, 141150.
16 M. J. Castaldi and S. M. Senkan, Ullmans Encyclopedia, 6th edn., 2002.
17 M. J. Castaldi, N. M. Marinov, C. F. Melius, J. Huang, S. M. Senkan,
W. J. Pitz and C. K. Westbrook, Symposium (International) on Combustion,
[Proceedings], 1996, 26th, 693702.
18 M. J. Castaldi and S. M. Senkan, Combust. Sci. Technol., 1996, 116117,
167181.
19 M. J. Castaldi and S. M. Senkan, J. Air Waste Manage. Assoc., 1998, 48, 7781.
20 N. M. Marinov, M. J. Castaldi, C. F. Melius and W. Tsang, Combust. Sci.
Technol., 1998, 131, 295342.