You are on page 1of 21

ELSEVIER Catalysis Today 34 (1997) 307-327

Chapter 4

Catalyst characterization: characterization techniques

G. Leofanti a, G. Tozzola a, M. Padovan a, G. Petrini a, S. Bordiga b, A. Zecchina b
a ENICHEM S.p.A., Centro Ricerche di Nouara, Via Fauser, 4, 28100 Nouara, Italy
b Department oflnorganic, Physical and Chemistry, Universitir degli Studi di Torino, Via P. Giura, 7, 10125 Torino, Italy

1. Characterization techniques able techniques it is impossible to review all of

them in few pages. So we have to decide what
techniques to describe and what to treat deeply.
1. I. Introduction The choice has been done on the basis of:
(i) applicability to real catalysts (so spectro-
The most remarkable difference between scopies on single crystals like LEED, UPS, etc.
catalysis during the 50s and 60s and nowadays were rejected);
is the remarkable development of characteriza- (ii) easy access, either in cost and/or in
tion techniques. necessary facilities (however also techniques
Various factors contributed to that growth: both expensive like XPS, NMR, TEM and of
(i) need of having a better control on the difficult access like EXAFS, requiring syn-
catalyst in order to meet the requirements for chrotron radiation, will be reported because of
further optimization of processes already in op- their importance);
eration and/or for realization of new processes; (iii) applicability to many different systems
(ii) appearance of new classes of solids with (then little space will be devoted to Miissbauer
catalytic properties and peculiar characteristics and PAS, much more to N, adsorption, IR
(zeolites, zeolite-like materials, oxides with con- spectroscopy and temperature programmed
trolled porosity, very high pore volume silica, techniques, etc.);
superacids, etc.); (iv) informative content, or capacity to give a
(iii) technology and computer development manifold information (multipurpose techniques
have placed on the market more powerful and like IR spectroscopy are preferred).
friendly to use versions of known techniques On the whole 35 techniques, i.e. a number
(FT-IR, FT-Raman, calorimeter, etc.) or new enough to face adequately nearly all the prob-
techniques, often initially developed for other lems of daily practice, are described; on a few
sectors use (XPS, EXAFS, SIMS, ISS, EPMA, other (about 20) a short description is given. For
etc.). all of them an adequate bibliography useful for
To realize how large is the growth of the a first deepening will be given.
characterization influence in the development of Techniques have been grouped on the basis
catalysts it is enough to look at scientific maga- O f:

zines and catalysis congresses proceedings. 1. morphology and physical characteristics,

Because of the tremendous variety of dispos- 2. surface characteristics,

0920-5861/97/$17.00 Copyright 0 1997 Published by Elsevier Science B.V. All rights reserved.
PZI SO920-5861(96)00056-9
308 G. Leofanti et al. / Catalysis Today 34 (1997) 307-327

3. bulk characteristics, The pores are classified in different classes

4. technological properties (granulometric dis- depending on their size:
tribution and mechanical properties). 1. micropores (+ < 2 nm),
Specific arguments are discussed by choosing 2. mesopores (2 < 4 < 50 nm),
information difficult to find in literature, conse- 3. macropores (4 > 50 nm).
quently: The porous structure enables the solid to
(i) the space dedicated to theory has been have a total surface much higher than that corre-
reduced to the minimum indispensable level; sponding to the external one. Most common
(ii) instrumentation is described only in case catalysts have a specific surface area between 1
of laboratory-made apparatus; and 1000 m2 g-l, while their external specific
(iii) the problems usually encountered in the surface area is in the range 0.01-10 m2 g-l.
practice, the obtainable information and the lim- The knowledge of the morphological parameters
its of the different methods are the most closely is very important to understand the catalyst
examined topics; evolution during the preparation procedure and
(iv) when different approaches to determine to give a feedback useful for modifying the
the same properties can be used, the preferable method and to obtain the desired results. It is a
techniques, on the basis of our experience, are matter of fact that a better comprehension of the
indicated. catalytic behaviour requires the knowledge of
This choice, aiming to underline the practical the morphological characteristics. In fact the
application of characterization, is strengthened catalytic process takes place at the surface.
in the second part of this work, where the To reach the surface sites the molecules of
approach to specific problems will be examined. the reagents must diffuse through the porous
system, while the reaction products follow the
1.2. Morphology and physical properties opposite path. Mass transfer process inside the
granules depends on pore size (bulk diffusion in
Most heterogeneous catalysts are porous macropores, Knudsen diffusion in mesopores
solids. The porosity arises from the preparation and molecular diffusion in micropores). Also
methods of the solids. For example the precipi- deactivation phenomena (i.e. organic material
tation from a solution originates precursor parti- deposition) are greatly influenced by pore size;
cles that agglomerate and form a porous struc- deposition of carbonaceous material with block-
ture. Crystallization in hydrothermal conditions age of micropore mouth and coverage of the
can produce zeolites, in which the peculiar dis- whole wall in meso and macropores are typical
position of the building units generates in- examples. The morphological characteristics of
tracrystalline cavities of molecular size. As interest are the specific surface area, the specific
mentioned above, during the thermal treatment pore volume, and the distribution of the area
the elimination of the volatile material (burning, and of the pore size.
evaporation) produces cavities that represent The methods for pore determination are nu-
both the result of the solid rearrangement and merous; the right choice depends on type of
the exit way of the removed material. Different pore and, as will be clarified later, on the use
shaping procedures, employed to obtain the cat- we want to do of such data. The most utilized
alysts suitable for industrial reactors (tableting, techniques are schematically reported in Table
extrusion, spray drying, etc.), give rise to stable 1.
aggregates of particles that contain a porous
structure. Then a typical catalyst contains one or 1.2.1. Vapour adsorption at low temperature
more group of pores whose size and volume Nitrogen adsorption at boiling temperature
depend on preparation method. (77 K) represents the most used technique to
G. Leofanti et al. / Calalysis Today 34 (1997)307-327 309

Table 1
Techniques and methods for the determination of morphological characteristics

Techniques Evaluation method Information Micropores Mesopores Macropores

N, adsorption at 77 K BET a Surface yes yes yes

t-plot yes yes
cu-plot yes yes
N, adsorption at 77 K BJH Surface area = (pore size) yes
modelless yes
N, adsorption at 77 K r-plot Pore volume yes
a-plot yes
DRK yes
MP yes
Horvath-Kavazoe yes
Oliver-Conklin yes yes
BJH yes
Gurvitsch yes yes
Hg porosimetry yes yes
Incipient wetness yes yes yes
From chemical and particles density yes yes yes
N, adsorption at 77 K Horvath-Kavazoe Pore volume = (pore size) yes
Oliver-Conklin yes yes
Hg porosimetry yes yes
All techniques Calculated from + = KVp/S Average pore size yes yes yes
Permeametry yes yes Yes
Counterdiffusion yes yes yes

a The value of the surface area in the presence of micropores is questionable (see [l]).

determine catalyst morphology. The starting Some valid alternatives to nitrogen adsorp-
point of that technique is the determination of tion at 77 K exist in special cases. We remem-
the adsorption isotherm, i.e. of the nitrogen ber argon and krypton adsorption at 77 K (to
volume adsorbed vs. its relative pressure. determine surface area < 1 m2 g- > and argon
By analyzing the data with suitable models adsorption at 87 K to study micropores in which
(reported in Table 1) the nitrogen adsorption at it is less interactive than nitrogen.
77 K allows the determination of: All these techniques utilize automatic com-
(i) total surface of the solid (BET method); mercial apparatuses giving good performances.
(ii) total surface, external to micropores (t- For further details on the arguments see [l-7].
plot or (Y, plot method);
(iii) mesopore surface distribution vs. their 1.2.2. Mercury porosimetry
size (BJH method); Mercury enters the pores whose size is in-
(iv) micropore volume (t-plot or OL, plot versely proportional to the applied pressure.
method); Usually mercury porosimetry based on that
(v) mesopore volume and volume distribution principle allows to study pores in the range
vs. their size (Gurvitsch and BJH methods). 7.5-1.5 X lo4 nm (2000 atm). The technique
On the basis of our experience the methods utilizes only commercial apparatuses. Working
that give the best results are reported inside the under pressure requires some care because it
brackets. Additionally we put remark that the can cause compression of the solid or breakage
H-K and the Oliver-Conklin methods (although of the pore walls (this occurs in very highly
needing further verification) represent the only porous silica with VP > 3 cm3 g- obtained by
tools to obtain data on the size and size distribu- sol-gel method).
tion of micropores. Mercury porosimetry, carefully utilized, is
310 G. Leofanti et al./ Catalysis Today 34 (1997) 307-327

the sole technique capable to determine the size alyst particles. Suitable models allow to calcu-
distribution of macropores. In the mesopore late the desired parameters. All apparatus are
range both porosimetry and nitrogen adsorption laboratory-made.
techniques can be used and the results are,
generally, in good agreement. If different results 1.25 A4icroscopy
are obtained, the following parameters can be By microscopy we intend both the conven-
wrong: tional optical and the modem electron micro-
(i) the geometric model of pores; scopies: i.e. SEM (scanning electron mi-
(ii) the value of the contact angle between croscopy) and TEM (transmission electron mi-
mercury and the solid surface; croscopy), with increasing resolution power up
(iii) the equation that defines the thickness of to few Angstroms.
adsorbed nitrogen vs applied pressure. They are substantially qualitative techniques.
More details can be found in [8,9]. However they have, in respect of the above
described quantitative techniques, the advantage
1.2.3. Incipient wetness method to give a direct view of the solid under study.
Following this method [lo] the solid is im- So microscopy and quantitative morphological
pregnated with a non solvent liquid, usually techniques can be considered as complemen-
water or hydrocarbons, just to fill the pores, tary*
without any meaningful excess. Addition of liq- SEM and TEM are quite expensive and they
uid must be stopped when the solid starts to be usually are not directly available in characteriza-
sticky or, if a large excess has been used, the tion laboratories. Nevertheless this electron mi-
liquid can be removed by centrifugation. In that croscopy is frequently used to study catalysts,
condition the pore volume is equal to the vol- particularly to obtain data about shape, size,
ume of the liquid adsorbed by the solid. crystalline habit, homogeneity, contemporary
That is a good, precise method, very useful to presence of amorphous and crystalline (or dif-
control values obtained with other ones and it is ferent crystalline) compounds and their distribu-
the only one that can be applied on high porous tion, relative surface of different crystalline
silica where the other methods fail. faces.
Most interesting results have been obtained
1.2.4. Permeametry and counterdifision on zeolites or zeolite-like solids, on which even
Up to now all the efforts to utilize pore size the pore mouth can be visible, and on supported
distribution to interpret mass transfer processes metals, on which data about distribution and
inside the pores failed. preferentially exposed faces (affecting the struc-
Permeametry and counterdiffusion [ 11,121 al- ture sensitive reactions) can be obtained.
low to obtain the values of the average diffu- Electron microscopy is reviewed in [13-151.
sive diameter and the tortuosity factor that
are of direct use in the equations describing 1.3. Sur$ace properties
mass transfer in catalyst pores.
Both techniques require that the gas can flow The importance of the total surface area mea-
only through the pores of a shaped catalyst surement has been pointed out in the previous
fixed to a non porous support. In permeametry a paragraph. However, although the surface is the
slight pressure difference between the two faces place of catalytic activity, only a part is utilized
of catalyst particles is established and the equili- in the catalytic reaction. In catalysts based on
bration pressure rate is recorded. Counterdiffu- supported metals only a little fraction of the
sion measures the diffusional counterflux of two solid total surface is occupied by the active
gases of different molecular weight through cat- centre (that is the metal particle). In acid cata-
Table 2
Main techniques for surface characterization of catalyst using probe molecules
Techniques Information Quantitative analysis
Volumetric adsorption Amount of adsorbate as a function of pressure Yes (high precision)
Gravimetric adsorption Amount of adsorbate as a function of pressure Yes (high precision)
Dynamic adsorption Amount of irreversible adsorbate Yes
Calorimetry Heat of adsorption as a function of coverage Yes (high precision)
IR-spectroscopy Surface functional groups Possible
Functional groups of adsorbates
Raman spectroscopy Surface functional groups Possible
Functional groups of adsorbates
UV-Vis spectroscopy Local environment of surface groups perturbed by adsorption Possible
TPD (temperature programmed desorption) Amount of desorbed species as a function of temperature Yes
TPD-MS (TPD coupled with Mass spectrometry) Amount and composition of desorbed species as a function of temperature Yes
TPD-IR (TPD coupled with IR spectroscopy) Composition of desorbed species as a function of temperature No
TPSR (MS) Competitive interaction of different probe molecules-Reactivity of surfaces Possible
Table 3
Main spectroscopies and related techniques for surface characterization of catalysts. Other techniques: LEED (low energy electron diffraction), HREELS (high resolution electron %
energy loss spectroscopy), ellypsometry, SEXAPS (surface extended X-ray absorption tine structure) R
Techniques Excitations Response Information Monolayers Quantitative Gaseous
sampled analysis atmosphere s
XPS (X-ray photoelectron spectroscopy Photons Electrons Surface atomic composition 2-20 Yes Difficult 2.
Nature and binding of the atoms
AES (Auger electron spectroscopy) Electrons Electrons Surface atomic composition 2-20 Possible Difficult {
SIMS (secondary ions mass spectrometry) Ions Ions Surface atomic composition l-3 Possible No ti
Short distance order 2
ISS (ion scattering spectroscopy) Ions Ions Surface atomic composition 1 No No %
G. Leofanti et al. / Catalysis Today 34 (1997) 307-327 313

A 8

Irreverslbla (on the metal) I

Fig. 1, Schematic diagram of a static-volumetric apparatus for adsorption measurements (A) and typical representation of the results (B): 1,
vacuum pump; 2, cold trap; 3, liquid reservoir; 4, gas reservoir; 5, high precision pressure transducer: 6, calibrated volume; 7, sample
holder: 8, furnace.

lysts the acid sites not only occupy a little group (Table 2). They are very versatile tech-
fraction of the surface, but also differ in acid niques allowing to operate practically on all the
strength and sometimes in nature. To under- type of catalysts and in conditions (temperature,
stand how a catalyst works, why a sample gives pressure) not too far from the reaction ones.
better performance than others or why activity Moreover they can be used to study many dif-
decays by time, the knowledge on the number ferent characteristics and to choose from them
and on the nature of active sites is indispens- those relevant for the case under investigation.
able. So it is easy to understand why researchers These advantages, joined to the low and medium
are so interested in surface methods giving data cost of the instruments required, explain the
on atomic and molecular levels. wide diffusion of these techniques. So nowa-
Techniques actually available can be divided days they are part of the standard equipment of
in two groups: catalysis laboratories.
1. techniques using probe molecules; The techniques belonging to the second group
2. techniques enabling the direct study of the (Table 3), being more expensive and less avail-
catalyst surface. able, are usually administrated by specialists
Volumetric and gravimetric static chemisorp- which operate in big research centres or in
tion, adsorption calorimetry, spectroscopies of universities. Moreover they are techniques which
adsorbed molecules and temperature pro- can be used only under vacuum or under limited
grammed desorption methods belong to the first gas pressure. For these reasons their use is more



Fig. 2. Schematic diagram of a static-gravimetric apparatus for adsorption measurements (A) and typical representation of the results (B): 1,
vacuum balance; for other components see Fig. 1.
314 G. L.eofanti et al. 34 (1997) 307-327


Fig. 3. Schematic diagram of a dynamic pulse apparatus for adsorption measurements (A) and typical representation of the results (B): 1, gas
reservoir; 2, liquid reservoir; 3, four way valve; 4, sampling valve; 5, sample holder; 6, furnace; 7, katharometer.

limited than those of first group. XPS is an devices with good performance are now avail-
exception because of its surface sensitivity to able.
atomic chemical environment. Static techniques give more precise data than
dynamic ones, being the last preferentially used
1.3.1. Volumetric, gravimetric and dynamic ad- for routine measurements (for example for pro-
sorption methods duction control). The static-volumetric tech-
All the techniques for the quantitative deter- niques are preferred in gas adsorption; while the
mination of the adsorbed molecules are grouped gravimetric ones give better performances in
into: vapour adsorption and particularly in the case of
(i) static-volumetric techniques in which the organic vapours.
outgassed solid is contacted with a known vol- Static methods are slow and delicate because
ume of gas and the adsorbed quantity is mea- of the difficulty of recognizing the reaching of
sured through the pressure decrease (Fig. 1); the equilibrium conditions and of distinguishing
(ii) static-gravimetric techniques similar to between the chemisorbed and the physisorbed
the previous ones, except the quantity of ad- fraction (a further measure after an intermediate
sorbed gas is determined by the weight increase outgassing is necessary). However the results
of the solid (Fig. 2); can at last greatly depend on the measurement
(iii) dynamic techniques in which the probe of the temperature: a preliminary investigation
gas is fed continuously or by pulse on the solid with TPD is suggested.
and the update is followed by gascromatography Typical applications are the determination of
(Fig. 3). exposed surface of dispersed metals and the
The laboratory-made instruments predomi- measurement of acidity as discussed in the sec-
nate in this field, even if some commercial ond part of this paper. It is also possible to

Fig. 4. Schematic diagram of an apparatus for calorimetric measurements (A) and typical representation of the results (B): 1, sample holder;
2, reference holder; 3, isothermal calorimeter. For other components see Fig. 1.
G. Leofanti et al. /Catalysis Today 34 (1997) 307-327 315

follow the adsorption of reagents and products for the sample holder which is inserted in the
of reaction, obviously in non-reactive condi- spectrophotometer to allow the acquisition of
tions. spectra upon dosage of probe molecules (Fig.
More details are available in [16-181. 5). IR spectroscopy. IR spectroscopy is the

1.3.2. Adsorption calorimetry
most diffuse technique. It is possible to operate
Equipment and methodology are in large part
both in transmittance (sample is a thin tablet
identical to those used in volumetric methods
crossed by the beam) and in reflectance (sample
(Fig. 4).
is placed in a horizontal holder on which a
The sample holder is positioned in a
mirrors system conveys the beam: a parabolic
calorimeter that enables the measure of the de-
mirror collects all the scattered and reflected
veloped heat and of the adsorbed volume. So it
light and send it to the detector). The first
is possible to obtain data on the nature of the
method is preferred, because it allows to work
interactions between adsorbate and adsorbent
with cheaper laboratory-made cells and to ob-
and on the heterogeneity of the surface sites.
tain spectra without artefacts due to reflected
For example, by using a base as probe molecule,
this technique allows to determine the number
The most used probe molecules are CO, NO,
of surface acid sites and their distribution vs.
NH,, C,H,, CH,OH, H,O and pyridine be-
acid strength.
The limits of this technique are the length of
(i) they are simple, and have spectra easy to
the measurements and a great fan of execution
difficulties. Anyway this is the only technique
(ii) they have vibrations significatively af-
giving both quantitative and qualitative data on
fected by adsorption located in the spectral range
the energy distribution of sites.
1200-4000 cm- (MID-IR) where skeleton vi-
Some examples of its employment are re-
brations of solid do not interfere.
ported in the second part of this paper and in
There are no limitations to the use of more
[191 .
complex compounds: of course the spectro-
scopic response becomes increasingly more
I . 3 . 3Spectroscopy
. of adsorbed molecules complex. It is possible in this way to study
The best techniques to obtain a qualitative interactions of molecules present in the reaction
information on the structure and reactivity of environment and to obtain more direct pieces of
adsorbed molecules are the spectroscopic ones. information on surface reactivity and reaction
Equipment and methodology are in large part mechanism (see for example the paragraph in
identical to those of volumetric methods except the second part of this paper).

Fig. 5. Schematic diagram of an apparatus for spectroscopic measurements of adsorption of probe molecules on catalysts (A) and typical
representation of the results (B): 1, sample holder; 2, spectrophotometer; 3, furnace for thermal treatment of the sample. For other
components see Fig. 1.
316 G. L.eofanti et al. / Catalysis Today 34 (1997) 307-327

IR spectroscopy can give also a direct infor- formation on the adsorption centre. The best
mation on the surface; particularly in the case of results are reached when all the optically active
oxides: in fact the surface atoms tend to com- compounds under investigation are on the sur-
plete their coordination sphere using mainly OH face (because there is no need to separate sur-
groups, whose stretching vibrations fall in the face and bulk contributions). The study of sup-
3000-4000 cm- range. Position of bands gives ported metal precursors and of zeolites contain-
pieces of information on the nature (i.e. the acid ing optically active elements like transitions
strength) of hydroxyl group that are often the elements, Ti, Zn, etc. have received a great help
catalytic active sites. Moreover it is possible to by this technique.
verify the influence on surface sites of probe Among the numerous papers on the argument
molecules (see Fig. 5). we recommend [20-301.
IR spectroscopy has two main limitations:
(i) the low frequency region can not be used
because of skeleton vibrations of the solid; 1.3.4. Temperature programmed desorption
(ii) most IR transparent materials (NaCI, KBr, The sample is placed in a tube fluxed by an
CaCl,, and so on) are not resistant to water or inert gas. The substance to be adsorbed is fed
other liquids. by pulses or continuously till the equilibrium is
reached. After outgassing the physisorbed frac- Raman spectroscopy. Raman spec- tion, the temperature is raised at constant rate:
troscopy allows to overcome these limitations so the adsorbate undergoes a progressive release
because: at temperature as higher as stronger is its inter-
(i) structural bands are often sharp and weak; action with the solid surface. A detector system
(ii) even water (which is strongly absorbing placed beyond the sample-holder allows the
in IR) gives rise to weak bands; monitoring of the process (Fig. 6). The most
(iii) inert and resistant materials, for instance used analytical systems are:
quartz, can be used. (i) a katharometer when the adsorbate is re-
Weak Raman bands usually correspond to leased unchanged;
strong IR bands and vice versa: in most cases (ii) an IR spectrophotometer equipped with a
the same bands are detectable in both spectra. gas cell or a mass spectrometer when more than
This spectroscopy, whose use was once lim- one probe molecules are used together or when
ited to expert spectroscopist, is in the last years the adsorbate released is modified by interaction
become available for industrial applications due with the solid surface. The techniques are named
to significative improvements in the commercial respectively TPD-IR and TPD-MS.
Traditional application fields are the study of
heteropolyacids and of molibdates. Very inter-
esting results have been recently obtained on
zeolites, in both structural and adsorption stud-
Most limitations arise from luminescence
phenomena due to organic contaminants of the
sample. I


1.3.3,3. UV- Vis spectroscopy. UV-Vis spec- Fig. 6. Schematic diagram of a TFD apparatus (A> and typical
representation of the results (B): 1, katharometer (TPD) or mass
troscopy is less used in surface studies, some- spectrometer (TFD-MS) or gas cell for IR spectroscopy. For other
times wrongly, because it supplies a direct in- components see Fig. 3.
G. Leofanti et al. / Catalysis Today 34 11997) 307-327 317

The technique, especially in the simplest XPS is exciting core electrons by means of a
form, is widely diffuse because: soft X-ray beam (i.e. Mg K,); while Auger
(i) it is possible to build up a cheap labora- spectroscopy utilizes electrons to bombard the
tory-made instrument even if good commercial surface. The emitted electrons bring to us an
devices are available; information coming from a depth of l-4 nm, so
(ii) the use is simple; providing a good surface picture.
(iii) the qualitative interpretation of results is XPS provides more useful data than AES
also simple. since the exciting beam does not usually dam-
For a deeper understanding of the results we age the surface. On the contrary Auger spec-
have to consider that: troscopy possesses a higher sensitivity and can
(i) the release of adsorbed substances is a discover contaminants better than XPS (at least
complex process, with contributions due to dif- in traditional commercial instruments). Never-
fusion and readsorption: consequently without theless the use of XPS in catalyst investigation
an adequate model the interpretation can be is increasing and nowadays the availability of
misleading; synchrotron light instrumentation (super ESCA)
(ii) during the experiment, the exposure to is favouring the XPS investigations because they
increasing temperatures can modify the solid as are of great help in surmounting also the sensi-
well as the adsorbate-adsorbent interaction. tivity problem.
An advantageous variation of TPD is often More details on XPS and other surface tech-
the so called TPSR (temperature programmed niques can be found in [37-451.
surface reaction) in which mixed probe
molecules are fed during all the experiment.
I .4. Bulk properties
Important data can be obtained on the catalyst
functionality and on reaction mechanism, as
pointed out in the second part of this paper, Even if the main effort in catalysts character-
using molecules present in the reaction environ- ization is devoted to the surface knowledge,
ment. bulk properties maintain their importance being
Among the numerous papers on the argument the nature of surface sites determined by those
we recommend [3 l-361 properties.
There is a large number of techniques avail-
able for analyzing bulk characteristics but most
1.3.5. Techniques for direct characterization of of them are expensive and not specific being
su$ace their main application in material science. So
All the techniques presented in the previous they are usually placed in general analytical
paragraphs use probe molecule to investigate services or in specialistic groups, while catalysis
the solid surface. The techniques allowing the laboratories are more frequently equipped with
direct characterization of surface are here briefly less expensive traditional spectroscopic and
discussed. Among those reported in Table 1 thermal analysis.
only XPS and, to a little extent, AES are widely In spite of the large number of those tech-
used in catalyst characterization. niques, only a few are frequently used in study-
Both the techniques reported give pieces of ing catalysts.
information on the composition and chemical (i) elemental analysis techniques;
status of the elements on the solid surface be- (ii) spectroscopic and diffraction (XRD, IR,
fore and after reaction test. Both the techniques Raman, UV-Vis, NMR) techniques;
provide data, via expulsion and analysis of the (iii) thermal analysis (TG, DTA, DSC, TPR,
related energies, of electrons from the solid. TPO) techniques.
Table 4
Main spectroscopies and related techniques for bulk characterization: + = molecular spectroscopy; ??= X-ray spectroscopy; 0 = resonance
Techniques Excitation Response Information Quantitative Gaseous Applicability
analysis atmosphere

+ IR (infrared spectroscopy) Photons Photons Local environment Possible Yes Quite all solids except C and few others
Functional groups
+ Raman Photons Photons Local environment Possible Yes Quite all solids, fluorescence can give problems
Functional groups
+ PAS (photoacustic spectroscopy) Photons Photons Local environment Possible Yes Like fR (in particular for solids opaque to usual fR)
Functional groups
+ UV-Vis (ultraviolet-visible spectroscopy) Photons Photons Chemical bonds Possible Yes Solids containing transition metal ions and few others
Oxidation degree
XRLr (X-ray diffraction) Photons Photons Crystalline species Yes Yes All crystalline solids
Crystalline degree
Crystal size
?? EXAFS (extended X-ray absorption Photons Photons Local structure Yes Yes Quite all metal compounds
fine structure) (coordination
number, inter-
atomic distances)
??XANES (X-ray absorption near Photons Photons Local structure Possible Yes Like XAFS
edge structure)
0 NMR (nuclear magnetic resonance) Photons Photons Local environment Yes Yes Quite all solids
Functional groups
Molecular diffusion
?? EPR (ESR) (electron paramagnetic Photons Photons Oxidation degree Yes Yes Paramagnetic species (transition metal ions).
spin resonance
Nature of ligands
0 MGssbauer Photons Photons Local environment Yes Yes Few elements, in particular Fe, Sn
Oxidation degree
(few elements)
TEM (transmission electron microscopy) Electrons Electrons Structure No No Quite all, in particular zeolites, supported metals
Crystal shape
EPMA (electron probe microanalysis) Electrons Electrons Atomic composition Yes No
(resolution 1 Pm)
XRF (X-ray fluorescence) Photons Photons Atomic composition Yes Yes Elements with Z > 5
NS (neutron scattering) Neutrons Crystal structure Yes Yes Quite all solids
Molecular diffusion
G. Leofanti et al. /Catalysis Today 34 (1997) 307-327 319

1.4.1. Elemental analysis techniques

The description of elemental analysis tech-
niques is out of the scope of this paper: usually
researchers working in catalysts characterization
are discussing problems with analysts which
choose the most suitable method.
The knowledge of solid chemical composi-
tion must be the starting point of every investi-
gation. This statement is not so obvious as it r I I I I -r
could seem. A lot of literature results cannot be 1200 loo0 0 600 cm -1

evaluated because of lack of composition data. Fig. 7. Example of information obtainable by IR and Raman
It is noteworthy to remember that: spectroscopy.

(i) catalyst composition at the end of prepara-

tion can be different from that calculated from
the weight of reagents; this is frequently ob- potential of the most diffuse techniques is
served in coprecipitation, hydrothermal syntbe- schematically reported.
sis, calcination of solids containing volatile
compounds; Molecular spectroscopies. IR spec-
(ii) composition reproducibility is a very troscopy allows us to obtain spectra of solids,
common problem in catalyst preparation, so fre- working both in transmission and reflectance;
quent experimental checks are necessary; sample dilution must be generally used because
(iii) impurities can always be present; their of the high intensity of the IR bands of the
origin is extremely variable as they can come skeleton vibration. In reflectance mode phenom-
from polluted reagents, previous preparation ena such as specular reflection can deeply affect
residues, release of substances from vessels, the spectra.
etc.; IR spectroscopy does not find much applica-
(iv) exposure to the reaction atmosphere tions in structural characterization of solids be-
changes almost every time the catalyst composi- cause most of the inorganic supports, particu-
tion: because of loss of volatile compounds, larly oxides, give rise to broad bands of difficult
deposition of organic substances, reactions be- interpretation, with the exception of few com-
tween solid and environment. pounds, like molibdates, vanadates and zeolites.
Further details are available in [46] and refer- The last give rise to well defined bands due to
ences therein. secondary building units vibrations then charac-
teristic of each framework (Fig. 7).
1.4.2. Spectroscopic techniques Raman spectroscopy suffers by different
The spectroscopic techniques are widely used problems: it is difficult to obtain good spectra,
for bulk characterization and this is not a sur- particularly of some oxide species, due to lumi-
prise. On the whole they allow to determine nescence phenomena. Alike IR techniques the
both local (oxidation state, ligands nature and application of this spectroscopy is limited to
number, symmetry) and structural (framework, MO, V, W, Ti compounds and few other sub-
crystalline degree, crystal size) characteristics. stances. Anyway the possibilities of this tech-
Unfortunately a single spectroscopic method nique has not already been investigated deeply:
able to supply all these pieces of information recently we have obtained important informa-
does not exist. Many different techniques must tion on zeolites (Fig. 7).
be chosen from time to time to obtain the UV-Vis spectroscopy is certainly the most
desired information. In Table 4 the informative interesting molecular spectroscopic technique to
320 G. Leofanti et al. / Catalysis Today 34 (1997) 307-327

Most problems derive from difficulty in in-

15000 30000 15000 30000 15000 30000 cm-
terpretation of spectra, particularly for charge
transfer transitions, characterized by broad bands
covering large spectra region. From a practical
point of view it must be pointed out that:
(i) it is necessary to use reflectance tech-
Fig. 8. Example of information obtainable by UV-Vis spec- (ii) easy to use materials, like quartz, are UV
troscopy: (A) change in oxidation state; (B) change in ligands
transparent, so there are not limitations as in IR
type; (C) change in coordination number (0, d-d transition; ??,
charge transfer transition). spectroscopy;
(iii) commercial instruments are reliable and
relatively cheap.
study bulk characteristics. In fact in the spectral On the whole the UV-Vis spectroscopy is a
range usually available (10000-50000 cm- ) it technique relatively easy to use and reliable,
is possible to investigate d-d transitions, which also if non exhaustively used.
are characteristic of transition elements, and It is not possible to summarize all the UV-Vis
other transitions due to charge transfer from spectroscopy applications in the study of solids.
ligand to metal or vice versa (Ti, Zn, Ce, Zr, A few examples are reported in the second part
MO, Sn, rare earths, alkaline earths metal ox- of this paper.
ides, and many transition metal oxides). With To go deeper into molecular spectroscopy
UV-Vis spectroscopy a direct information on techniques see [22,27-30,47-551.
electronic structure and on the first coordination
sphere of the examined ions can be obtained. In X-ray difSraction and X-ray absorption
particular oxidation state, type of ligands, lig- spectroscopies. Most catalysts are crystalline
ands number and coordination can be studied. solids such as zeolites, many oxides, supported
Changes in oxidation state involve a variation metals, salts and so on. So X-ray powder
of electronic configuration and this in turn usu- diffraction becomes a fundamental technique
ally greatly influences the spectra. See for ex- enabling us to evaluate:
ample in Fig. 8A Cr,O, (d,) and CrO, (da) 1. nature of crystalline phases;
spectra. 2. their concentration in the solid;
Electronic transitions are also influenced by 3. crystallite size.
ligand nature, as foreseen by ligand field theory, The last two characteristics can be deter-
and the series of optical electronegative was mined only in simple spectra. In the case of
proposed, parallel to Paulings series to qualify solids with low symmetry (i.e. zeolites), the
this influence: calculation of crystallinity degree is very diffi-
I - < Br-< S*-< SCN-< Cl-< NO, <F- cult.
During the last ten years these difficulties
< OH-< H,O < NH, <NO, < CN-< CO have been partially overcome by the help of
Fig. 8B presents as an example Cr,O, and new methods of structural refinement, like the
CrF, . Rietveld one, which permits the reproduction of
Likewise the influence of coordination type the whole diffractogram, through the optimiza-
on both d-d transition and charge transfer is of tion of both structural (peak position and inten-
interest. Fig. 8 shows spectra of titanium having sity) and non-structural (peaks shape) data. For
oxygen in the coordination sphere respectively instance this method enables an accurate deter-
in tetrahedral (Ti-silicalite) and octahedral mination of cell parameters of MFYlzeolites and
(amorphous TiO,-SiO,) coordination. then makes possible to evaluate the volume
G. Leofanti et al. / Catalysis Today 34 (1997) 307-327 321

11 22 23 2. 2% Is 27 rB zb

Fig. 9. Example of information obtainable by XRLJ spectroscopy on solid-solid transition induced by thermal treatments in silicalite; the
transition from monoclinic to orthorhombic symmetry is revealed by the substitution of the two doublets at about 24 and 29 (20) with two

increase due to the insertion of guest atoms in XANES (X-ray absorption near edge structure)
the framework. gives indications on the coordination symmetry.
An interesting possibility offered by XRD is These techniques can be used even at high
to follow solids evolution during thermal treat- temperature and under controlled atmosphere,
ment or exposure to gases or vapours. An exam- so it is possible to work in reaction conditions.
ple is presented in Fig. 9, where spectra of Some studies on XAFS reactors are known. An
silicalite at different temperatures, showing the example is reported in the second part of the
transition from orthorhombic to monoclinic paper, where the technique has been used to
symmetry, are reported. demonstrate the tetrahedral coordination of Ti in
XRD technique can be applied without par- Ti-silicalite, (and then its insertion in the frame-
ticular limitations, although it requires a not work), and also the change in coordination fol-
negligible quantity of samples. However the lowing the exposure to different molecules (Ti
instrumentation is expensive: so usually it is is a reactive centre).
managed in specialized laboratories rather then The techniques are illustrated in detail in
in characterization groups. [46,56-631.
A great number of applications are possible.
Some examples are reported in the second part Nuclear magnetic resonance (NMR).
of this paper. The application of NMR on solids is now possi-
EXAFS (extended X-ray absorption fine ble because of improvement in instrumentation
structure) technique deserves a mention because and of introduction of new techniques, such as
it is extremely interesting in catalysis even if MAS (magic angle spinning).
not easily available, as it is requiring the avail- The technique is useful in zeolite study of Al
ability of synchrotron radiation. distribution c2Si, 27A1>,silanols number ( 29Si),
The technique allows the determination of: acidity (HI, porosity (129Xe). Surely its impor-
(i) average coordination number of investi- tance is destined to increase in spite of the
gated element; required investment as shown in Refs. [64-671.
(ii) interatomic distances of the first (some-
times also of the second) shell. Other spectroscopies. The number of
In the region before the adsorption edge available spectroscopic techniques is larger than
322 G. Leofanti et al. / Catalysis Today 34 (1997) 307-327

that described before. Most of them have a has a great influence: low rate guarantees a
limited use because they require expensive in- better peak resolution, but gives broad signals
vestment or must be used under vacuum condi- and vice-versa. Experiments at different scan-
tions or they are applicable to a limited number ning rate are used for kinetic analysis with
of elements. Some of them would deserve a alternate results.
mention (EPR, Mijssbauer) but they are not As regard the study of the solid, each tech-
described for sake of brevity. See Refs. [46,68- nique can reveal:
781 to examine closely the argument. (i) loss of volatile components and reactions
involving mass change such as decomposition
1.4.3. Temperature programmed techniques (TG);
Under this denomination are grouped all the (ii) reduction (TPR) or oxidation reactions
techniques where the variation of a physical (TPO);
parameter as a function of temperature (when it (iii) solid-solid, solid-gas or liquid-gas
is varied with constant rate) is measured. transitions (DSC and DTA).
We can distinguish: Examples concerning TPR, TG, TPO and
1. TG (thermal gravimetry) ; DSC are reported in the second part of this
2. DTA (differential thermal analysis); paper.
3. DSC (differential scanning calorimetry); The ability of TPR to identify different
4. TPR (temperature programmed reduction); species of the same element is particularly no-
5. TPO (temperature programmed oxidation); ticeable. For instance in the Fe-silicalite case Fe
The techniques are generally reliable and well in the zeolite framework and Fe in extraframe-
tested. TG, DTA, and DSC are executed only work position are reduced at different tempera-
on commercial instruments, while the labora- ture, so allowing the quantitative determination
tory-made instruments for TPO and TPR are of the two species. But not always two signals
progressively substituted by commercial de- mean two different species. That is the case of
vices. Usually the selected parameter variation chlorinated Cu compounds that are reduced in
vs. temperature is reported, obtaining a plot like two steps with the formation of Cur compound
that showed in Fig. 10. as intermediate. Another application of TPR is
Analyzing the results it must be considered the quantitative determination of oxidized
that peaks are representative of several kinetic species, for example Crvr, in the presence of the
phenomena happening on solids changes altered species in lower oxidation state.
by readsorption, diffusion into the sample, heat TG, TPO and DSC find their main applica-
transport to the detector, etc. Even scanning rate tions in studies of activation of the fresh cata-
lysts and regeneration of exhausted ones or to
evaluate the thermal stability (see paragraph 6
la1 (bl (cl (dl lel (fl in the second part of the paper).
Useful data are reported in [19,33,35,79-811.

WC-DTA 1.5. Particle size distribution and mechanical

This part is devoted to properties of appar-
ently low scientific interest, but surely of great
importance in practical use: granulometric dis-
Fig. 10. Typical representation of temperature programmed tech-
nique results: (a) loss of volatile compounds; (b) adsorption; (c)
tribution and mechanical properties.
crystallization; (d) reduction; (e) oxidation; (f) melting. As regard the importance of particle size
G. Leofanti et al. /Catalysis Today 34 (1997) 307-327 323

distribution it must be taken into account its Table 5

Main techniques for particle size analysis
effect on:
Techniques Detectable size
1. mass transport processes (i.e. interparticle
Sieving 20-104
2. heat transport processes; Optical microscopy 0.1-104
3. pressure drop through catalytic beds; SEM (scanning electron microscopy) 0.003-104
TEM (transmission electron microscopy) o.OOO-lo4
4. catalyst deactivation by fouling.
Gravitational sedimentation l-100
As for as the mechanical strength, we have to Resistive pulsed (Coulter effect) 0.5-200
distinguish between fixed bed and fluid bed Light obscuration (1) 10-9000
catalysts. Fraunhofer diffraction (1) 10-2000
Centrifugal sedimentation 0.05-5
In fixed bed catalysts problems arise from the PCS (photon correlation spectroscopy) 0.01-l
presence of powder or fragments generated dur- HCD (hydrodynamic chromatography) 0.01-l
ing the transport, from the reactor charge and FFF (field flow fractionation) 0.01-100

the industrial run. The results are:

(i) change in particle size distribution with
(ii) crushing and abrasion of the most parti-
the consequence above illustrated;
(ii) irregular distribution of the pressure drop
(iii) agglomeration of electrostatically charged
(then of the fluxes, in tubular reactors);
(iii) local accumulation of powder (then pos-
With an adequate procedure it is possible to
sibility of too high activity and heat transfer
overcome points (ii) and (iii) in most cases.
Besides sieving it is possible to use other
In fluid bed catalysts, low resistance of the
instruments, based on different principles, whose
solid, produces a great amount of small particles
(on the whole) are applicable for particle ranges
which are eliminated from the reactor. So it
between 10 nm and 1 cm. The best use of them
becomes necessary to reintegrate frequently the
is in the range where sieving is not applicable (1
catalyst: consequently problems of catalyst con-
and 50 pm). The most used instruments are
sumption and disposal arise.
those based on sedimentation (the size is deter-
In conclusion the working capacity of a cata-
mined via XRD), on resistive pulse, on the light
lyst depends on particle size distribution and
obscuration, and on the Fraunhofer diffraction.
mechanical resistance as well as on its chemi-
Techniques working even under 1 p,m are
cal, structural and morphological characteristics.
used only in particular cases and are not re-
ported here for sake of brevity (see references).
Techniques based on microscopy usually are
1.5. I. Particles size analysis
used as reference in catalysis, however it is
A large number of techniques, based on many
difficult to obtain quantitative results (even if
different principles, are nowadays available,
more sophisticated instruments of automatic im-
most of them were developed and offered by
age counting are now available).
commercial producers. In Table 5 a choice of
It is noteworthy that this method gives a size
suitable techniques, divided on the basis of size
distribution based on particles number and not
of analyzing particles, are reported.
on their mass (like sieving).
The most diffuse technique for particle size
Refs. [82-841 allow deepening in particle
analysis is sieving. Most catalysts can be ana-
size analysis.
lyzed with this technique. The limits of the
technique are: 1.5.2. Crushing strength measurements
(i) difficulty to complete the sieving with They are applied to catalysts shaped for fixed
particles smaller than 40 Frn; bed (tablets, extrudates, granulates).
324 G. Leofanti et al. /Catalysis Today 34 (1997) 307-327

Two different type of measurements have The bulk crushing strength technique mea-
been developed (Table 6): sures the mean value of the mechanical resis-
1. crushing strength of the single granule, tance subjecting to a predetermined pressure a
2. bulk crushing strength. catalyst volume enclosed in a cylinder and by
The measurement of crushing strength is determining, through granulometric analysis, the
made by gradually compressing a single catalyst powder formed by granule crushing.
granule and by continuously registering the Since the granules in contact with walls are
value of the applied force until the crush hap- preferentially crushed we must work with a
pens. The measurement is repeated on a signi- large cylinder (+cylinder/+granule> 10) to mini-
ficative number of granules (at least 50): the mize this effect. For the same reason the force
mean value of crushing strength and its distribu- applied must be enough to induce the formation
tion is so obtained. Generally, the distribution is of a significant amount of powder. On the other
narrow for tablets, broad for pellets, and very hand too high pressure must be avoided other-
broad for extrudates. wise the difference among samples will not be
Crushing strength depends not only on the appreciable. Meaningful measurements are pos-
material intrinsic resistance, but also on several sible with powder fraction between 1 and 10%.
factors, such as shape, size, wetness degree: Really it would be more correct to determine
consequently the comparison is realistic only the pressure necessary to obtain a given powder
among homogeneous series of catalysts (i.e. fraction, but as in this way several measures
different supplies of the same batch). For the would be indispensable, this method is seldom
same reasons it is difficult to predict the be- used.
haviour of a single sample only on the bases of Based on our experience, the determination
the results of these techniques. of bulk crushing strength supplies results able to
The main utility of the method consists on predict the catalyst resistance during its practi-
the capacity of underlining the presence of frac- cal application. Anyway when problems in cata-
tions with particularly low resistance. lyst charging of industrial reactors are found,

Table 6
Main techniques for mechanical properties determination
G. Leofanti et al. / Catalysis Today 34 (1997) 307-327 325

the best test is a simulation of that operation, for [4] K.S.W. Sing, in J.M. Thomas and R.M. Lambert (Editors),
Characterisation of Catalysts, Wiley, Chichester, 1980, p. 12.
example by filling a tube having the same height [5] K.K. Unger, J. Roquerol, K.S.W. Sing and H. Kral (Editors),
and diameter of one of the tubular reactors, and Characterization of Porous Solids, Elsevier, Amsterdam,
by measuring the amount of formed powders. 1988.
[6] F. Rodriguez-Reinoso, J. Roquerol, K.S.W. Sing and K.K.
See for more details [85-891.
Unger (Editors), Characterization of Porous Solids II, Else-
vier, Amsterdam, 1991.
1.5.3. Attrition resistance measurements [7] R.C. Bansal, J.B. Donnet and F. Stoeckli, Active Carbon,
They are applied, in different versions, either Marcel Dekker, New York, 1988.
[8] H.L. Ritter and L.C. Drake, Ind. Eng. Chem., 17 (194.5) 782.
on fixed bed or on fluid bed catalysts (Table 6).
[9] S. Lowell, Powder Technol., 37 (1980) 37.
For fixed bed catalysts the granules are placed [lo] M.P. McDaniel and T.D. Hottovy, J. Colloid Interf. Sci., 78
in a turning drum or in a horizontal cylinder (1980) 31.
swinging along its axis and, after a prefixed 1111C.N. Satterfield, Mass Transfer in Heterogeneous Catalysis,
MIT Press, Cambridge, Massachusetts, 1970.
time, the formed powder is determined by siev- 1121J. Valus and P. Schneider, App. Catal., (19851 329.
ing. [I31 F. Delannay, in F. Delannay (Editor), Characterization of
In most cases the measurement gives results Heterogeneous Catalysts, Marcel1 Dekker, New York, 1984,
p. 71.
comparable to those obtained by bulk crushing
1141H. Dexpert, P. Gallezot and C. Leclercq, in B. Imelik and
strength test, so meaning that a catalyst gives J.C. Vedrine (Editors), Les techniques physiques detude des
good or bad results with both measurements. catalyseurs, Editions Tchnip, Paris, 1988, 655.

For fluid bed catalysts the powder is flu- 1151A. T. Bell and K. Klier (Editors), Special Issue on Electron
Microscopy, Catal. Rev. Sci. Eng., 34 (1992).
idized under controlled conditions in a tube [I61V. Rudzinski and D.H. Everett, Adsorption of Gases on
collecting the fines formed in a fixed time. The Heterogeneous Surfaces, Academy Press, London, 1992.
collection begins after a time long enough to 1171P. Hippie, Chemisorption and Catalysis, Elsevier, New York,
warrant the elimination of the fines previously [181 D. A. Cadenhead and N. J. Wagner, in R.B. Anderson, P.T.
present in the catalyst. Dawson (Editors), Experimental. Methods Catalysis. Re-
The measurement is delicate and often pre- search, Academy Press, New York, 1976.
1191A. Auroux, in B. Imelik and J.C. Vedrine (Editors), Les
sents reproducibility problems due, for example, techniques physiques detude des catalyseurs, Editions Tch-
to adhesion of a part of the sample to the tube nip, Paris, 1988, p. 823.
surface and wear of the nozzle from which the 1201M.L. Hair, Infrared Spectroscopy in Surface Chemistry, Mar
ccl Dekker, New York, 1967.
flux exits. It is noteworthy that, as the fluid bed
El1 J.T. Yates, Jr. and T.E. Madey, Vibrational Spectroscopy of
catalyst is subjected to mechanic solicitations Molecules on Surfaces, Plenum Press, New York, 1987.
during all the run, it can happen that its mechan- La G. T. Haller, Catal. Rev. Sci. Eng., 23 (1981) 477.
ical properties are modified by exposure to the 1231M. C. Kung, H. H. Kung, Catal. Rev. Sci. Eng., 27 (1985)
reaction environment. b41 A. Terenin, Adv. Catal., 30 (19811, 227.
Notwithstanding these limitations, the attri- WI A. Zecchina, E. Garrone and E. Guglielminotti, Catalysis, 6
tion resistance measurements are in some case (1983) 90.
bl J.L.G. Fierro, in J.L.G. Fierro (Editor), Spectroscopic Char-
the only available technique to control the re- acterization of Heterogeneous Catalysts-Part B: Chemisorp-
sults of catalyst shaping.Recommended refer- tion of Probe Molecules, Elsevier, Amsterdam, 1990, p. 67.
ences are [87,90]. D71 G. Coudurier and F. Lefebvre, in B. Imelik and J.C. Vedrine
(Editors), Les techniques physiques detude des catalyseurs,
Editions Technip, Paris, 1988, p. 29.
ml G. Coudurier and F. Lefbvre, in B. Imelik and J.C. Vedrine
References (Editors), Les techniques physiques detude des catalyseurs,
Editions Technip, Paris, 1988, p. 39.
[ll S.J. Gregg and K.S.W. Sing, Adsorption, Surface Area and b91 Coudurier and E. Garbowski, Les techniques physiques
Porosity, 2nd edn., Academy Press, New York, 1982. detude des catalyseurs, Editions Technip, Paris, 1988, p. 85.
[2] M. Jaroniec and R. Madey, Physical Adsorption on Hetero- [301 E. Garbowski, Les techniques physiques detude des cataly-
geneous Solids, Elsevier, Amsterdam, 1988. seurs, Editions Technip, Paris, 1988, p. 103.
[3] G. Hot-vat and K. Kawazoe, J. Chem. Eng. Jpn., 16 (1983) [311 J.L. Falconer and J.A. Schwarz, Catal. Rev. Sci. Eng., 25
470. (1983) 141.
326 G. L.eofmti et al. /Catalysis Today 34 (1997) 307-327

[32] R.J. Cvetanovic and Y. Amenomiya, Adv. Catal., 30 (1981) [56] H.P. Klug and L.E. Alexander, X-ray Diffraction Procedures,
103. Wiley, New York, 1974.
[33] S. Bhatia, J. Beltramini and D.D. Do, Catal. Today, 7 (1990) 1571 H.M. Rietveld, J. Appl. Crystallogr., 2 (1969) 65.
309. [58] G. Bergeret and P. Gallezot, in B. Imelik and J.C. Vedrine
[34] P. Forzati, E. Tronconi and L. Lietti, Temperature Pro- (Editors), Les techniques physiques detude des catalyseurs,
grammed Desorption Technique in Catalyst Technology, Editions Technip, Paris, 1988, p. 599.
1988, p. 299. [59] A.J. Renouprez, in B. Imelik and J.C. Vedrine (Editors), Les
[35] J.L. Lemaitre, in F. Delamay (Editor), Characterization of techniques physiques detude des catalyseurs, Editions Tech-
Heterogeneous Catalysts, Marcel Dekker, New York, 1984, nip, Paris, 1988, p. 587.
p. 29. [60] B. Moraweck, in B. Imelik and J.C. Vedrine (Editors), Les
[36] P. Malet, in J.L.G. Fierro (Editor), Spectroscopic Characteri- techniques physiques detude des catalyseurs, Editions Tech-
zation of Heterogeneous Catalysts-Part B: Chemisorption of nip, Paris, 1988, p. 633.
Probe Molecules, Elsevier, Amsterdam, 1990, p. 333. [61] J. C. Bart, Adv. Catal., 34 (1986) 203.
[37] J. Alvarez and MC. Asensio, in J.L.G. Fierro (Editor), [62] J. C. Bart and G. Vlaic, Adv. Catal., 35 (1987) 1.
Spectroscopic Characterization of Heterogeneous Catalysts [63] J.C. Conesa, P. Esteban, H. Dexpert and D. Bazin, in J.L.G.
Par-A: Methods of Surface Analysis, 1990, p. 79. Fierro (Editor), Spectroscopic Characterization of Heteroge-
[38] M.W. Roberts, Adv. Catal., 29 (1980) 55. neous Catalysts Part-A: Methods of Surface Analysis, Else-
[39] C. DefossC, in F. Delamray (Editor), Characterization of vier, Amsterdam, 1990, p. 225.
Heterogeneous Catalysts, Marcel1 Dekker, New York, 1984, [64] Y.B. Taarit and J. Fraissard, in B. Imelik and J.C. Vedrine
p. 225. (Editors), Les techniques physiques detude des catalyseurs,
[40] J.C. Vedrine and Y. Jugnet, in B. Imelik and J.C. Vedrine Editions Technip, Paris, 1988, p. 161.
(Editors), Les techniques physiques detude des catalyseurs, [65] R.E. Wasylishen and CA. Fyfe, in G.A. Webb (Editor),
Editions Technip, Paris, 1988, p. 365. Annual Reports on NMR Spectroscopy Academic Press,
[41] Y. Jugnet and J.C. Vedrine, Les techniques physiques detude New York, 1982, p. 1.
des catalyseurs, Editions Technip, Paris, 1988, p. 407. [66] U. Haeberlen, High Resolution NMR in Solids-Selective
[42] J.-C. Bertolini, in B. Imelik and J.C. Vedrine (Editors), Les Averaging, Accademy Press, New York 1976.
techniques physiques detude des catalyseurs, Editions Tech- [67] W.W. Pudler, Nuclear Magnetic Resonance: General Con-
nip, Paris, 1988, p. 433. cepts and Applications, Wiley, New York, 1987.
[43] M. Abon, in B. Imelik and J.C. Vedrine (Editors), Les [68] J. A. Dumesic and H. Topsoe, Adv. Catal., 26 (1977) 122.
techniques physiques detude des catalyseurs, Editions Tech- [69] J.C. Vedrine, in F. Delannay (Editor), Characterization of
nip, Paris, 1988, p. 475. Heterogeneous Catalysts, Marcel1 Dekker, New York, 1984,
[44] M. Abon, in B. Imelik and J.C. Vedrine (Editors), Les p. 161.
techniques physiques detude des catalyseurs, Editions Tech- [70] M. Che and E. Giamello, in J.L.G. Fierro (Editor), Spectro-
nip, Paris, 1988, p. 513. scopic Characterization of Heterogeneous Catalysts Part-B:
[45] J.L.G. Fierro, in J.L.G. Fierro (Editor), Spectroscopic Char- Chemisorption of Probe Molecules, Elsevier, Amsterdam,
acterization of Heterogeneous Catalysts Part-A: Methods of 1990, p. 265.
Surface Analysis, 1990, p. 1. [71] M. Che, D. Olivier, L. Bonueviot and P. Meriaudeau, in B.
[46] F. Dehumay and B. Delmon, in F. Delannay (Editor), Char- Imelik and J.C. Vedrine (Editors), Les techniques physiques
acterization of Heterogeneous Catalysts, Marcel Dekker, New detude des catalyseurs, Editions Technip, Paris, 1988, p.
York, 1984, p. 1. 231.
[47] P. He&a, C. Jones and G. Warners, Fourier Transform [72] F.J. Berry, in J.L.G. Fierro (Editor), Spectroscopic Character-
Raman Spectroscopy-Instrumentation and Chemical Appli- ization of Heterogeneous Catalysts Part-A: Methods of Sur-
cations, Ellis Horwood, New York, 1991. face Analysis, Elsevier, Amsterdam, 1990, p. 299.
[48] N.B. Colthup, L.H. Daly and SE. Wiberley, Introduction to [73] P. Bussiere, in B. Imelik and J.C. Vedrine (Editors), Les
Infrared and Raman Spectroscopy, 3rd edn., Academic Press, techniques physiques detude des catalyseurs, Editions Tech-
Boston, 1990. nip, Paris, 1988, p. 323.
[49] K. Nakamoto, Infrared Spectra of Inorganic and Coordina- [74] W. Jones, in J.M. Thomas and R.M. Lambert (Editors),
tion Compounds, Wiley, New York, 1963. Characterisation of Catalysts, Wiley, Chichester, 1980, p.
[50] J.M. Stencel, Raman Spectroscopy for Catalysis, Van Nos- 114.
trand Reinhold, New York, 1990. [75] D. Olivier and L. Bonneviot, in B. Imelik and J.C. Vedrine,
[51] R.A. Schoonheydt, in F. Delamtay (Editor), Characterization Ediyors, Les techniques physiques detude des catalyseurs,
of Heterogeneous Catalysts, Marcel Dekker, New York, 1984, Editions Technip, Paris, 1988, p. 293.
p. 125. [76] S.J.B. Reed, Electron Microprobe Analysis, Cambridge, Uni-
[52] D. Nicholls, Complexes and First-Row Transition Elements, versity Press, Cambridge, 1975.
Macmillan, London, 1974. [77] H. Jobic and A. Renouprez, Electron Microprobe Analysis,
[53] R.P. Cooney, G. Curthoys and N. The Tam, Adv. Catal., 30 Cambridge, University Press, Cambridge, 1975, p. 557.
(1975) 293. [78] C.J. Wright, in J.M. Thomas and R.M. Lambert (Editors),
[54] J. Ferguson, Prog. Inorg. Chem., 12 (1970) 159. Characterisation of Catalysts, Wiley, Chichester, 1980, p.
[55] C.K. Jorgensen, Prog. Inorg. Chem., 12 (1970) 101. 169.
G. Leofanti et al. / Catalysis Today 34 (1997) 307-327 321

[79] W. Hemminger and G. Hohne, Calorimetry-Fundamentals dards, Vol. 5.03, American Society for the Testing of Materi-
and Practice, Verlag Chemie, Weinheim, 1984. als, New York, 1984.
[80] B. D. McNicol, in J.M. Thomas and R.M. Lambert (Editors), [87] R.J. Bertolacini, in A. Bradley, J. Gattuso and R.J. Berto-
Characterisation of Catalysts, Wiley, Chichester, 1980, p. lacini (Editors), Characterization and Catalyst Development,
135. ACS, Washington, 1989, p. 380.
[81] N.W. Hurst, S.J. Gentry, A. Jones and B.D. McNicol, Catal. [88] S.A. Bradley, E. Pitzer and W.J. Koves, in A. Bradley, J.
Rev.-Sci., 24 (1982) 233. Gattuso and R.J. Bertolacini (Editors), Characterization and
[82] H.G. Barth, Modem Methods of Particles Size Analysis, Catalyst Development, ACS, Washington, 1989, p. 398.
Wiley, New York, 1984. [89] P.K. Brienza, in A. Bradley, J. Gattuso and R.J. Bertolacini
[83] J.D. Stokham and E.G. Fochtman, Particle Size Analysis, (Editors), Characterization and Catalyst Development, ACS,
Ann Arbor Science, AM Arbor, 1978. Washington, 1989, p. 407.
[84] P.J. Lloyd, Particle Size Analysis 1985, Wiley, New York, [90] L.A. Pedersen, J.A. Lowe and C.K. Matocha Sr., in A.
1987. Bradley, J. Gattuso and R.J. Bertolacini (Editors), Characteri-
[85] J.T. Richardson, Principle of Catalyst Development, Plenum zation and Catalyst Development, ACS, Washington, 1989,
Press, New York, 1989. p. 414.
[86] ASTM Committee D-32 1985 Annual Book of ASTM Stan-