Structure, texture, acidity and catalytic performance of AlPO4-caesium oxide catalysts in 2-methyl-3-butyn-2-ol conversion

Article abstract of DOI:10.1039/a807538g

A series of aluminium orthophosphate-caesium oxide systems with various caesium oxide loadings (5-30 wt%) were prepared by impregnation of ALPO₄ with a methanolic solution of caesium acetate and characterized by TG/DTA, XRD, DRIFT, Raman, SEM-EDX, XPS, ²⁷Al and ³¹P MAS NMR and nitrogen adsorption. Aluminium orthophosphate remained amorphous with caesium oxide incorporation when calcined at 873 K for 3 h. After thermal treatment at 1423 K it crystallized in the tridymite form except for materials containing 30 wt% caesium oxide. DRIFT spectroscopy showed that the P-OH stretching vibration at 3670 cm^-1 decreased in intensity with caesium oxide loading. Materials at 20-30 wt% caesium oxide did not exhibit any hydroxyl bands. Besides, Al and P atoms remained in tetrahedral coordination as in unmodified aluminium orthophosphate. Moreover, the incorporation of caesium oxide leads, simultaneously, to a progressive decrease in surface area and pore volume (larger at 30 wt% caesium oxide) as well as to an increase in the most frequently occurring pore radius. On the other hand, caesium oxide reduced both the number and strength of acid sites as the caesium content increased. Consequently, 2-methylbut-3-yn-2-ol (MBOH) underwent almost exclusively dehydration to 3-methyl-3-buten-1-yne (acid activity) on pure AlPO₄, whereas its modification with increasing amounts of caesium oxide developed AlPO₄- based materials with increased basic properties and hence high selectivities (99 mol%) to the base-catalysed cleavage of MBOH yielding acetone and acetylene.

Full text of DOI:10.1039/a807538g

J O U R N A L O F
Structure, texture, acidity and catalytic performance of
AlPO –caesium oxide catalysts in 2-methyl-3-butyn-2-ol conversion
4
Felipa M. Bautista,a Juan M. Campelo,*a Angel Garcia,a Raquel M. Leon,a Diego Luna,a
Jose M. Marinas,a Antonio A. Romero,a Jose A. Naviob and Manuel Maciasb
C H E M I S T R Y
aDepartamento de Qu ´ı mica Org a´ nica, Facultad de Ciencias, Universidad de C o´ rdoba, Avda.
S. Alberto Magno, s/n°, E-14004 C o´ rdoba, Espa n˜ a
bInstituto de Ciencia de Materiales de Sevilla, Centro de Investigaciones Cient ´ı ficas ‘Isla de la
Cartuja’, Avda. Am e´ rico Vespucio, s/n°, E-41092 Sevilla, Espa n˜ a
Received 28th September 1998, Accepted 22nd December 1998
A series of aluminium orthophosphate–caesium oxide systems with various caesium oxide loadings (5–30 wt%)
were prepared by impregnation of AlPO with a methanolic solution of caesium acetate and characterized by
4
TG/DTA, XRD, DRIFT, Raman, SEM-EDX, XPS, 27Al and 31P MAS NMR and nitrogen adsorption.
Aluminium orthophosphate remained amorphous with caesium oxide incorporation when calcined at 873 K for 3 h.
After thermal treatment at 1423 K it crystallized in the tridymite form except for materials containing 30 wt%
caesium oxide. DRIFT spectroscopy showed that the P–OH stretching vibration at 3670 cm−1 decreased in
intensity with caesium oxide loading. Materials at 20–30 wt% caesium oxide did not exhibit any hydroxyl bands.
Besides, Al and P atoms remained in tetrahedral coordination as in unmodified aluminium orthophosphate.
Moreover, the incorporation of caesium oxide leads, simultaneously, to a progressive decrease in surface area and
pore volume (larger at 30 wt% caesium oxide) as well as to an increase in the most frequently occurring pore radius.
On the other hand, caesium oxide reduced both the number and strength of acid sites as the caesium content
increased. Consequently, 2-methylbut-3-yn-2-ol (MBOH) underwent almost exclusively dehydration to 3-methyl-3-
buten-1-yne (acid activity) on pure AlPO , whereas its modification with increasing amounts of caesium oxide
4
developed AlPO -based materials with increased basic properties and hence high selectivities (99 mol%) to the base-
4
catalysed cleavage of MBOH yielding acetone and acetylene.
condensation of benzaldehyde with ethyl cyanoacetate,21,24
Michael addition of ethyl cyanoacetate to ethyl acrylate21,24
and rearrangement of v-phenylalkanals into phenylmethyl
ketones.23
In the present work a series of aluminium orthophosphate–
caesium oxide systems with various caesium oxide loadings
Introduction
Amorphous aluminium orthophosphates can be used as
catalysts and supports. As catalysts, they are active in several
organocationic chemical processes in the gas phase and in the
field of selective synthetic chemistry involving acid- and base-
catalysed liquid-phase reactions.1–3 Aluminium orthophos-
phates were also used as metal supports in the liquid-phase
hydrogenation of the double bond of alkenes bearing a variety
(
5–30 wt%) were prepared by impregnation of AlPO with a
4
methanolic solution of caesium acetate and characterized by
several physical methods. Also, we report on the effect of
caesium oxide loading on the surface acid–base character by
using pyridine chemisorption and the conversion of 2-methyl-
but-3-yn-2-ol (MBOH) as probe reaction. In this sense,
Lauron-Pernot et al.25–27 showed that MBOH is a very sensitive
molecule for the simultaneous characterization of acidic and
basic surface properties of solids. Thus, MBOH undergoes
dehydration over solid acids yielding 3-methylbut-3-en-1-yne
whereas over basic materials it gives acetone and acetylene.
Moreover, the production of 3-hydroxy-3-methylbutan-2-one
and/or 3-methylbut-3-en-2-one seems to be characteristic of
amphoteric solids possessing acid–base ion pairs. A recent
study20 showed that this test was independent of the nature of
the basic sites (Bronsted or Lewis) making it a global measure
of basic properties. In our case, the MBOH conversion was
studied in both pulse and flow reactors and, in flow experi-
ments, a comparison with MgO, ZnO and c-Al O was also
of organic functions.4 Also, AlPO -supported zinc borohydride
materials were active catalysts for the liquid-phase reductive
4
cleavage of epoxides5 and the hydration of aromatic alkenes
and alkynes.6 The modification of the acid–base properties of
AlPO by the incorporation of metal oxides (Al O , B O ,
4
2 3 2 3
TiO , ZrO , and ZnO), allowed us to control catalytic
2
2
activity.1–3
Moreover, the development of solid base catalysts (alkaline
earth metal oxides, rare earth metal oxides, modified zeolites
and clays) has recently acquired great importance in the
growing field of heterogeneous base catalysis for fine chemistry
applications.7–10 Thus, alkali metal ion-exchanged zeolites
exhibited basic properties and promoted a number of base
catalysed reactions but their basic sites were relatively weak.
Caesium ions in excess of the ion-exchange capacity or caesium
oxide impregnated zeolites generated stronger basic sites than
those of simple alkali metal ion-exchanged zeolites, thus
exhibiting high performance for reactions such as alcohol
dehydrogenation,11–13 side chain toluene alkylation,14,15
Knoevenagel condensation of benzaldehyde with active meth-
ylene compounds,16 1-butene isomerization17–19 and conver-
sion of 2-methylbut-3-yn-2-ol into acetone and acetylene.20
More recently, caesium oxide or caesium–lanthanum mixed
oxides were introduced on SiO –Al O , c-Al O and MCM-
2
3
drawn.
Experimental
Catalysts
Two AlPO supports with different pore properties were used.
4
They were obtained by precipitation, from AlCl ·6H O and
2
2 3
2 3
3
2
4
1 molecular sieve21–24 leading to catalysts with surface base
H PO (85 wt%) aqueous solutions, with ethylene (E) or
propylene (P) oxide as described elsewhere.1 The solid thus
3
4
properties showing interesting performances in Knoevenagel
J. Mater. Chem., 1999, 9, 827–835 827

obtained was washed with 2-propanol, dried at 393 K for 24 h
and calcined at 923 K for 3 h.
DRIFT Measurements
DRIFT (diffuse reflectance infrared Fourier transform) spectra
were recorded in a FTIR instrument (Bomen MB-100)
equipped with an ‘environmental chamber’ (Spectra Tech, P/N
AlPO –caesium oxide (5–30 wt% caesium oxide) catalysts
4
were prepared by impregnation until incipient wetness using a
methanolic solution of caesium acetate. The impregnated
0
030-101) placed in a diffuse reflectance attachment (Spectra
AlPO were dried at 393 K (24 h) and then calcined at 873 K
4
Tech, Collector). A resolution of 8 cm−1 was used with 256
scans averaged to obtain a spectrum from 4000 to 400 cm−1.
Single-beam spectra were ratioed against KBr at the same
temperature as that of the sample. A plot of pseudo-absorbance
was preferred. This allowed direct subtraction of spectra
obtained with the same sample under different conditions.
DRIFT spectra were recorded for all the calcined catalysts
(200 mesh size and diluted to 15% in KBr) previously dried
at 400 K for 24 h under vacuum. Afterwards, the catalyst was
placed in the environmental chamber cell with a 20 mL min−1
flow of dehydrated and deoxygenated nitrogen, heated to
(
3 h). Samples are denoted P (or E) followed by the caesium
oxide loading (in wt%; 5Cs, 20Cs, and so on) and by the
treatment temperature (P-5Cs-873, E-10Cs-873, etc.).
Unmodified AlPO (P-0 and E-0) were similarly prepared by
4
placing them in methanol containing no caesium acetate.
Thermogravimetry and differential thermal analysis
Thermogravimetric (TG) and differential thermal analyses
(
DTA) were obtained simultaneously using a high-temperature
Setaram 92, model 16.18 thermal analyzer, in the presence of
5
73 K and held at this temperature for 1 h prior to recording
static air at a heating rate of 10 K min−1 (temperature range:
the spectrum. In some cases, spectra of the O–H stretching
region were smoothed with a five-point Savitsky–Golay
algorithm.28
2
73–1423 K). Finely powdered a-alumina was used as a
reference material.
Gases evolved from TG experiments were analysed by mass
spectrometry using a VG Sensorlab quadrupole mass spec-
trometer from Fisons Instruments plc/VG quadrupoles (East
Sussex, UK) operating in the multiple-ion-monitoring (MIM)
mode.
Surface area and pore properties
Surface-area and pore-size information were obtained from
N
adsorption–desorption isotherms at 77 K, using a
2
Micromeritics ASAP 2000 analyzer. Prior to measurements,
all samples were degassed to 0.1 Pa. Assessments of possible
microporosity were made from t-plot constructions, using the
XRD Measurements
X-Ray powder diffractometry ( XRD) was carried out at 300 K
and in the 2h range between 10 and 80° (at a scanning speed
of 2° min−1) using a Siemens D500 diffractometer equipped
Harkins–Jura correlation for t as a function of p/p . Parameters
were fitted to a low-area, non-porous silica. Mesopore size
distributions were calculated using the Barret–Joyner–Halenda
(BJH) method,29 the Halsey equation30 and assuming a
cylindrical pore model.
0
˚
with Ni-filtered Cu-Ka radiation (l=1.5405 A) at 35 kV and
2
0 mA. The crystalline XRD patterns were compared with
JCPDS standards.
Surface acid properties
NMR Spectroscopy
The surface acid properties of AlPO –caesium oxide catalysts
were measured by using three procedures.
4
2
7Al (pulse; 0.6 ms; recycle delay; 0.3 s), 31P (pulse; 2.6 ms;
recycle delay: 6 s), 133Cs (pulse; 3 ms; recycle delay; 3 s) and
H (pulse: 3 ms; recycle delay: 3 s) MAS NMR spectra were
Pulse measurements. The surface acidity was measured in a
dynamic mode by means of the gas-phase adsorption of
pyridine (PY) at 373 K, using a pulse technique.2 The pulse
size was in the range of gas-chromatography linearity (pulse
size versus detector response), corresponding to 0.1–0.5 mono-
layer. The acidity measurements were repeated several times
and good reproducibility of the results was obtained (ca. 7%).
1
recorded at resonance frequencies n of 104.26, 161.98, 52.48
0
and 400.13 MHz, respectively, with a Bruker ACP-400 multi-
nuclear spectrometer (external magnetic field of 9.4 T). No
mathematical procedures of NMR signal treatment were used.
Also, the chemical shifts given for aluminium were not
corrected for second-order quadrupole effects.
Measurements were carried out at room temperature with
a standard Bruker double-bearing MAS (5.5 kHz) probe.
About 200 mg of sample material were placed in the zirconium
dioxide rotor (4 mm) with a volume of about 0.35 cm3. The
references for 27Al, 31P, 133Cs and 1H were Al(H O) 3+, 85%
Temperature programmed desorption measurements. The
TPD apparatus was described elsewhere.31 After PY adsorp-
tion at 373 K, the weakly held PY was then eluted from the
catalyst in a helium stream (50 mL min−1) for 1 h. Afterwards
the temperature was lowered to 323 K and thermal desorption
experiments were carried out from 323 to 973 K at a rate of
2
6
H PO , a 1 M aqueous solution of caesium chloride and TMS,
3
4
respectively.
1
0 K min−1 maintaining the temperature at 973 K for 10 min.
Desorption was followed by a flame ionization detector.
Repeated adsorption/TPD experiments using the same sample
did not show any change in the desorption curve.
Scanning electron microscopy (SEM)
SEM studies were carried out in a JEOL apparatus, model
JSM-5400, and the electron microprobe analysis (EPMA) was
conducted by a Link Isis Pentafet Model analyzer. A semiauto-
matic image analyzer of magnetostrictive KONTRON MOP-
DRIFT measurements. DRIFT spectroscopy with adsorbed
PY was carried out on pure catalysts previously dried at 400 K
for 24 h under vacuum. Then, the catalyst was placed in the
temperature/vacuum chamber cell and heated at 573 K (1 h)
in dehydrated and deoxygenated nitrogen (20 mL min−1).
Afterwards, the reference DRIFT spectrum of the sample was
recorded at the same temperature (573 K). Furthermore, the
cell was sequentially cooled to 473 and to 373 K under nitrogen
and the reference spectrum at each temperature was obtained.
The PY probe molecule (from a nitrogen flushed saturator)
was adsorbed at 373 K for 2 h. This procedure gives the
complete saturation of the catalyst surface as a large excess of
3
0 board was used to estimate the average size of aggregates
and particles.
XPS Measurements
The XPS spectra were recorded in a Leybold-Heraeus LHS-
1
0 spectrometer working with a constant pass energy of 50 eV
and using Al-Ka radiation (hu=1686.6 eV) as the excitation
source. Binding energy values were referred to the C(1s) peak
(
pollution carbon) at 284.6 eV.
8
28 J. Mater. Chem., 1999, 9, 827–835

the probe molecule is admitted into the DRIFT cell. Physically
adsorbed reagent was removed by flushing for 2–3 h with
nitrogen. Longer nitrogen flushing times or vacuum treatment
at 373 K did not modify the DRIFT spectra. The characteristic
spectrum of the various forms of base retained on the catalyst
was recorded. Subsequent flushings were performed at 473 and
573 K, and the corresponding spectra recorded at those tem-
peratures taking the reference spectra at the same temperature.
Catalytic activity measurements
The catalytic properties of AlPO and AlPO –caesium oxide
4
4
catalysts in the conversion of 2-methylbut-3-yn-2-ol (MBOH)
were studied by using a pulsed fixed bed reactor inserted
between the sample inlet and the analytical column [15%
TCPE/Supelcoport 80/200 (2 m ×1/8 in id)] of a HP-5890-II
GC with flame ionization detection (FID). Catalytic measure-
ments were performed under the following conditions: 2-
methylbut-3-yn-2-ol (liquid) volume/pulse size; 1 mL; reaction
temperature, 473–673 K (at 50 K intervals); catalyst weight,
1
3
0–30 mg in powder form; flow rate of nitrogen carrier gas,
0 mL min−1. A fresh catalyst was used in each run, and before
use the catalyst was pretreated by in situ heating under nitrogen
30 mL min−1) for 1 h at 473 K.
(
Some reactions were also conducted in a continuous
stainless-steel down-flow fixed-bed reactor (6 mm id) sur-
rounded by an electric heater. An iron–constantan thermo-
couple was placed in the middle of the catalyst bed. The unit
operated at atmospheric pressure. The substrate was delivered
at a set flow rate (WHSV=25.5 h−1) using a liquid syringe
pump (Harward Md. 44) and was vaporized prior to passing
it through the catalyst bed in the presence of a flow of nitrogen
carrier gas (3 L h−1) regulated by mass-flow controllers. The
catalyst charges (W) were small, usually 80 mg, retained by
quartz wool at almost the center of the reactor. The reaction
temperature was 573 K, except where otherwise specified.
Standard catalyst pretreatment was carried out in situ at 573 K
Fig. 1 (A) DTA curves (static air; 10 K min−1 scan): (a) P-0–873; (b)
P-20Cs-873. (B) TG/DTG curves for P-20Cs-873 sample (static air;
10 K min−1 scan). (C) Mass spectrometry profiles (argon flow:
50 mL min−1; 10 K min−1 scan).
that water (m/z: 18), carbon dioxide (m/z: 44) and acetic acid
(m/z: 60) were by far the major desorption products. Methanol
(m/z: 31) was not found. Therefore, these results allow us to
verify our calcination procedure.
for 1 h under a stream of high purity nitrogen (99.999%, H O
<
2
3 ppm).
Reaction products were analysed on-line by GLC with FID
(
1
Varian Md. 3400) and a column (2 m ×1/8 in id) packed with
5% TCPE/Supelcoport 80/200. Product characterization was
performed by GC-MS (HP-5890 GC coupled with a HP-
970 MSD). Retention times were compared to those of com-
Moreover, between 773 and 1423 K, a slight weight loss
(<0.02%) was observed, due mainly to the condensation of
surface hydroxyls, together with an exothermic DTA band.
The exothermic transition corresponded to the crystallization
of the amorphous AlPO into tridymite-AlPO and/or a-
5
mercially available reagents. Reaction products were 3-methyl-
but-3-en-1-yne, acetylene, acetone and minor amounts of 3-
methylbut-3-en-2-one and 3-methylcrotonaldehyde. Blank runs
showed that, under the experimental conditions used in this
work, thermal effects could be ignored.
4
4
crystobalite-AlPO crystalline forms (see XRD results below).
The position of this exothermic band also depended on the
4
caesium loading and thus, as the caesium oxide loading
increased up to 20 wt%, the exothermic peak decreased in
intensity and shifted to lower temperatures: 1365, 1195, 1145
and 1080 K for P-0, P-5Cs, P-10Cs and P-20Cs, respectively.
P-30Cs and E-30Cs samples did not exhibit any exothermic
bands at temperatures above 873 K and remained in a low
degree of crystallinity even at 1423 K (see XRD results).
Results and discussion
TG and DTA measurements
The DTA profiles for P-0 and P-20Cs samples are shown in
Fig. 1A whereas Fig. 1B shows the TG and DTG profiles for
P-20Cs. Remaining P-Cs and E-Cs samples displayed almost
the same behaviour.
XRD Measurements
XRD patterns of all AlPO –caesium oxide systems calcined
4
TG profiles were characterized by two weight losses up to
at 873 K for 3 h (results not shown) exhibited only a very
7
73 K. The first one at 293–500 K (ca. 6% weight loss),
broad band in the range 2h from 15 to 30°, a characteristic of
accompanied by a DTG peak (#440 K) and an endothermic
DTA band (#390 K), was attributed to the desorption of
adsorbed water (Fig. 1C). The second weight loss (ca. 5%) at
the amorphous AlPO . Besides, the peaks of caesium oxide
(ASTM, 9–104) were absent indicating that caesium oxide
also exhibited a low degree of crystallinity and high dispersion.
Moreover, from Fig. 2 it can be seen that the thermal
treatment at temperatures above those of the DTA peaks
4
5
00–773 K, also accompanied by a DTG peak (#687 K) and
an exothermic DTA band (#710 K), is associated with com-
bustion of acetate moieties into carbon dioxide and water as
well as the desorption of acetic acid, as indicated in Fig. 1C,
where it appeared to be continuous analysis, by mass spec-
trometry, of the gases evolved from P-20Cs sample during
thermal treatment in the range 298–923 K (10 K min−1). In
this sense, mass spectrometry analysis of the gas phase revealed
(1423 K) developed bands corresponding to crystalline AlPO .
4
The structure of P-Cs (Fig. 2) samples corresponded to that
of the pseudohexagonal tridymite identified by its characteristic
˚
reflections at d=4.34 (2h=20.4), 4.11 (2h=21.6) and 3.83 A
(2h=23.2°) [ASTM, 20–45] although the relative intensities
differed from those reported by Florke.32 In E-Cs samples
J. Mater. Chem., 1999, 9, 827–835
829

also exhibited one single component at around d −27. This
chemical shift corresponded to P atoms in tetrahedral co-
ordination with P–O–Al bonds [i.e. P(OAl) environments].
4
Therefore, the incorporation of caesium did not change the
isotropic chemical shift value of the 27Al and 31P peaks so
that the local structure of the AlPO support was not changed
4
by caesium.
On the other hand, the 133Cs MAS NMR spectra (not
shown) exhibited a non-resolved very broad gaussian band
centered at around d −30 (full width half maximum: 176 ppm)
for E-30Cs-873 and P-30Cs-873 samples. Thus, the caesium
was in chemical environments of low symmetry (highly dis-
persed caesium oxide species bonded in different ways to the
AlPO support). As the amount of caesium oxide increased in
4
4
AlPO the 133Cs resonance shifted to a lower field. This change
in peak position had been attributed to the interaction of
caesium with neighboring caesium atoms.24
Furthermore, the 1H spectra (not shown) of E-0–873 and
P-0–873 samples [amorphous AlPO ; surface composition
4
(
XPS) P/Al #0.90] contained a small line from Al–OH
groups (at around d −0.9) and an unresolved peak from P-
OH groups (at around d 3.6), as described by Mastikhin et al.36
Besides, the peak at d #3.6 decreased sharply in intensity with
caesium loading. At 20–30 wt% caesium oxide both 1H signals
disappeared indicating that chemical interaction of the sup-
ported caesium oxide species takes place with Al-OH and P-
OH groups. These results were confirmed by DRIFT
spectroscopy (see below).
SEM, EPMA and XPS Studies
Fig. 2 XRD profiles of P-Cs materials after DTA measurements (Cu-
˚
Ka, l=1.5405 A): (a) P-0–1423; (b) P-5Cs-1423; (c) P-10Cs-1423;
SEM micrographs of E-30Cs-873 and P-30Cs-873 samples are
shown in Fig. 4. SEM observation showed a very widely varied
distribution in morphology, texture and particle sizes for all
(d) P-20Cs-1423; (e) P-30Cs-1423.
(
results not shown) AlPO crystallized in a mixture of tridymite
AlPO –caesium oxide samples. Besides, no changes were
observed either in morphology or in particle sizes when the
4
4
and a-crystobalite forms. The a-crystobalite form33 was ident-
ified by its characteristic reflections at d=4.07 (2h=21.9) and
amount of caesium increased. The average size of the biggest
particles was around 40 mm, although a large population of
small particles was observed in all samples.
˚
2
.51 A (2h=35.8°) [ASTM, 11–500].
Furthermore, the XRD diagrams of P-30Cs (Fig. 3) and E-
0Cs (not shown) samples thermally treated in the range
018–1423 K, which did not exhibit any DTA exothermic
3
1
Table 1 summarizes the information about the elemental
surface composition (EPMA) and XPS binding energies (in
parentheses) of the elements present on the surface of E-Cs-
873 and P-Cs-873 samples. Thus, Table 1 indicates a surface
enrichment in caesium compared to bulk values. However,
there was a homogeneous distribution of caesium so that the
surface composition of different particles remained almost
unchanged.
peaks above 800 K, indicated partial AlPO crystallization so
4
that these systems remained at a low degree of crystallinity.
2
7Al, 31P and 133Cs MAS NMR Spectroscopy
The 27Al MAS NMR spectra (not shown) of AlPO and
4
AlPO –caesium oxide samples calcined at 873 K for 3 h exhib-
4
On the other hand, as can be seen from Table 1, the surface
concentration ratio [P] /[Al] of E-Cs-873 and P-Cs-873
ited a broad peak at around d 39, typical of tetrahedral
aluminium with phosphorus in the second coordination sphere,
i.e. sharing oxygens with four tetrahedra of phosphorus
s
s
samples decreased with caesium loading. This could be
accounted for by the fact that caesium interacted preferentially
with surface P–OH groups (see 1H NMR and DRIFT results).
Furthermore, the various core lines investigated for different
samples were very similar and they all appeared as single XPS
symmetrical lines, indicating homogeneous distribution of the
electron densities around the atoms throughout the solid.
[
Al(OP) ].34,35 Moreover, the 31P MAS spectra (not shown)
4
DRIFT Measurements
The DRIFT spectrum of pure amorphous AlPO (P-0–873
4
sample), recorded after pretreatment at 573 K (1 h) under a
nitrogen stream (20 mL min−1), is shown in Fig. 5(a) [the
DRIFT spectra of E-0–873 samples was similar]. In the OH
stretching region (4000–2500 cm−1) two isolated hydroxyl
peaks were found: a weak one at 3793 cm−1 and a strong one
at 3670 cm−1, which were due to surface Al–OH (with alu-
minium atoms in tetrahedral co-ordination) and of non-
bonded surface P–OH groups, respectively.37 There is also a
broad band around 3538 cm−1 due to bonded hydroxyl groups.
The DRIFT spectra of the hydroxyl region of P-Cs-873
samples are shown in Fig. 5(b)–(e). The incorporation of
˚
Fig. 3 XRD profiles (Cu-Ka, l=1.5405 A) of P-20Cs sample
thermally treated at increasing temperatures: (a) P-30Cs-1018; (b) P-
3
0Cs-1153; (c) P-30Cs-1423.
30 J. Mater. Chem., 1999, 9, 827–835
8

Fig. 4 SEM micrographs: (a), (b) E-30Cs-873; (c), (d) P-30Cs-873.
Table 1 Electron microprobe analysis and XPS binding energiesa of
the elements present on the surface of E-Cs-873 and P-Cs-873 systems
P (2p)
(at%)
Al (2p)
(at%)
Csb
(at%)
Cs (3d) [P] /
s
System
(at%)
[Al]
s
E-0–873
E-5Cs-873
15.9 (133.6) 17.7 (74.1)
14.9 (133.2) 18.5 (74.1) 0.7 (723.8) 1.0
—
—
0.90
0.81
0.81
0.76
0.73
0.88
0.80
0.80
0.71
0.70
E-10Cs-873 14.5 (133.0) 18.0 (74.1) 1.5 (723.8) 2.7
E-20Cs-873 14.0 (133.0) 18.4 (74.1) 3.2 (723.8) 3.0
E-30Cs-873 12.9 (132.9) 17.6 (74.1) 5.3 (723.7) 7.2
P-0–873
P-5Cs-873
P-10Cs-873 14.7 (133.1) 18.3 (74.1) 1.5 (723.7) 1.8
P-20Cs-873 13.1 (133.0) 18.4 (74.1) 3.2 (723.7) 5.5
P-30Cs-873 12.8 (132.8) 18.4 (74.1) 5.3 (723.6) 7.9
15.8 (133.6) 17.9 (74.1)
—
14.8 (133.2) 18.4 (74.1) 0.7 (723.7) 1.4
—
aIn parentheses, eV. bBulk values.
increased amounts of caesium oxide results in the disappear-
ance of Al-OH and P-OH bands, so that P-20Cs-873 and P-
3
0Cs-873 samples did not exhibit any hydroxyl bands (E-Cs
samples exhibited the same trend). This is consistent with 1H
NMR results.
Moreover, the DRIFT spectra of the skeletal region (below
2
500 cm−1) showed that, in all E-Cs and P-Cs [Fig. 5(b)–(e)]
samples, the phosphorus atom was surrounded by tetrahedra
of oxygen atoms: bands due to the triply degenerate P–O
stretching vibration (1069 cm−1), n mode of tetrahedral
3
PO 3−; triply degenerate O–P–O bending vibration
4
(
478 cm−1), n mode of PO 3− tetrahedra and stretching
4
4
Fig. 5 DRIFT spectra (at 573 K; nitrogen flow 20 mL min−1): (a) P-
vibration of Al–O bonds in combination with P-O bonds
0
8
–873; (b) P-5Cs-873; (c) P-10Cs-873; (d) P-20Cs-873; (e) P-30Cs-
73. All spectra were displaced for presentation.
(
691 cm−1). DRIFT data were in accordance with the XRD
and 27Al and 31P MAS NMR data.
J. Mater. Chem., 1999, 9, 827–835
831

Surface area and pore properties
Table 2 Textural properties, surface acidity vs. pyridine and rate
constants (k ) in MBOH conversion of AlPO and AlPO –caesium
oxide catalysts
a
4
4
The adsorption–desorption isotherms of AlPO
and
4
AlPO –caesium oxide samples calcined at 873 K for 3 h were
4
S
/
V /
D/
PYa
106b/k mol
of nearly the same type with closed and well defined hysteresis
BET
P
a
Catalyst
m2 g−1 mL g−1 nm
209
185
C
mmol g−1 s m2 atm−1
377
215
151
39
loops (Fig. 6A). From the shape of the curves it can be
assumed that samples consist mainly of mesopores, showing
type IV isotherms in the BDDT classification38 and hysteresis
loops of H1 type according to the IUPAC classification.39
Moreover, the pore volume (at p/p =0.99) decreased while
the starting point of the hysteresis loop was displaced towards
higher relative pressures on increasing caesium loading (Fig. 6A
shows the evolution of the shape of the P-Cs sample isotherm
on increasing caesium content). These facts could be accounted
for by a continuous restructuring of the porous texture as well
as an increase in mesopore diameter (Fig. 6B) in accord with
XRD and 133Cs MAS NMR results.
BET
81
P-0–873
P-5Cs-873
0.79
0.73
0.65
0.53
0.36
0.53
0.46
0.43
0.34
0.20
13.5
4.5
2.0
0.7
0.3
0.4
4.5
3.0
1.0
0.2
0.1
15.4 142
15.1 197
18.9 240
30.4 239
4.8
5.4
6.8 128
17.9 212
—c 279
P-10Cs-873 153
P-20Cs-873 107
0
P-30Cs-873
E-0–873
E-5Cs-873
48
218
185
4
80
95
421
346
163
34
E-10Cs-873 151
E-20Cs-873
E-30Cs-873
84
16
—d
aAdsorption temperature: 373 K. bAt a reaction temperature of 573 K.
cMacroporous. dThere was no adsorption of titrant.
The t plots (not shown), using the Harkins–Jura correlation40
for t as a function of p/p show two steps, the highest (an
upward deviation from the linear branch of the t plot with
ascending t values) corresponding to capillary condensation
(which starts at relative pressures similar to those of the
corresponding hysteresis loops) and the lowest (linear t plot
passing through the origin) corresponding to a mono-multi-
layer adsorption mechanism on pores in which the adsorption
desorption was reversible, thus confirming the mesoporous
texture of the solids and the absence of microporosity.
Furthermore, the surface areas calculated by the t method
from the slopes of the straight lines were found to tally with
those obtained by the BET (Brunauer–Emmett–Teller) method.
This reflects that the samples were free from micropores as
0
0
in Table 2. Here it can be seen that the incorporation of
.99), C
constant and mesopore diameter (D) are collected
BET
caesium oxide into AlPO leads, simultaneously, to a progress-
ive decrease in surface area and pore volume. This decrease
4
was larger in E-Cs samples. Moreover, the decrease in surface
area was more pronounced as the caesium oxide loading of
catalyst was increased (for example, in P-Cs samples, the
incorporation of caesium oxide at 5, 10, 20 and 30 wt%
resulted in a loss of surface area of 12, 27, 49 and 77%,
respectively) due to the surface restructuring although the
amorphous character and Al and P tetrahedral environments
remained unchanged. Fig. 6B shows that the pore size distri-
bution as well as the full width half maximum were displaced
towards higher values as the caesium amount increased. Also,
well as the high accuracy of the C
determination.
BET
The porous texture of the E-Cs-873 and P-Cs-873 samples
was analysed following the BJH method and assuming a
cylindrical pore model.29 The analysis was applied to the
adsorption branch of each isotherm, which is to be preferred
to the desorption branch for type IV isotherms.41,42
the C
value increased with caesium loading, which means
BET
that the presence of caesium enhanced the adsorbate–adsorbent
interaction.
The BET surface area (S ), pore volume (V , at p/p =
Surface acidity measurements
BET
P
0
The surface acidity (Brønsted and Lewis acid sites) of E-Cs-
8
73 and P-Cs-873 samples is given in column six of Table 2 as
the amount of pyridine (PY) chemisorbed at saturation at
73 K.2 It is evident that the incorporation of caesium oxide
3
into AlPO resulted in a remarkable decrease in surface acidity,
4
that was more pronounced as the caesium oxide loading of
catalyst increased.
The thermodesorption of PY from E-Cs-873 and P-Cs-873
samples resulted in a number of overlapping peaks, generating
complex thermodesorption spectra. Fig. 7 presents character-
istic PY thermal desorption profiles (P-Cs-873 samples)
obtained after adsorption of PY at 373 K employing a heating
rate of 10 K min−1. In order to make the comparisons easier,
Fig. 6 (A) Adsorption–desorption isotherms and (B) pore size
distributions: (a, #) P-0-873; (b, 1) P-10Cs-873; (c, %) P-20Cs-873;
Fig. 7 PY TPD profiles for P-0–873, P-5Cs-873, P-10Cs-873 and P-
20Cs-873.
(d, 6) P-30Cs-873.
8
32 J. Mater. Chem., 1999, 9, 827–835

Fig. 8 DRIFT spectra of adsorbed PY (at 573 K; nitrogen flow
2
0 mL min−1). (A) Skeletal region and (B) OH stretching region: (a)
P-0–873; (b) P-5Cs-873; (c) P-10Cs-873; (d) P-20Cs-873; (e) P-
3
0Cs-873.
areas under the experimental desorption spectra were
normalized to catalyst weight.
TPD measurements indicated that, as the caesium oxide
content was raised from 5 to 20 wt%, the total amount of PY
was remarkably reduced and the peak maximum temperature
of PY desorption shifted to lower temperatures, corroborating
acidity measurements.
Fig. 9 Overall reaction scheme for 2-methylbut-3-yn-2-ol (MBOH)
conversion.
The adsorption of PY was also studied by DRIFT
spectroscopy using a temperature controlled cell under a
such as skeletal isomerization of 3,3-dimethylbut-1-ene and
cyclohexene, cracking/isomerization of 2-methylpentane and
n-hexane or isomerization/disproportionation of o-xylene were
used to characterize the acidity and acid strength of protonic
sites. Later on, Lauron-Pernot et al.25–27 showed that the
conversion of 2-methylbut-3-yn-2-ol (MBOH) can be used as
a probe reaction for the simultaneous characterization of acidic
and basic surface properties of solid catalysts. This molecule
undergoes dehydration over solid acids yielding 3-methylbut-
nitrogen
atmosphere.
Fig. 8A
shows
the
spectra
(
1700–1400 cm−1) of the P-0–873 and P-Cs-873 samples after
adsorption and desorption of PY at 373 K. On sample P-
0
–873 [Fig. 8A(a)], PY adsorption at 373 K, followed by
desorption at the same temperature for 1 h, gives rise to the
characteristic upward absorption bands of PY interacting with
Brønsted acid sites at 1553 and 1640 cm−1 (due to the n
1
9b
and n modes of the pyridinium ions) and of species coordi-
8
a
nated to Lewis acid sites at 1451 and 1614 cm−1 (due to the
3-en-1-yne whereas over basic materials it gives acetone and
acetylene. In this latter case there was no difference in product
n
and n modes of coordinated pyridine).43,44 A band at
1
9b
8a
1
490 cm−1, which can be attributed to PY chemisorption on
selectivity between Lewis (MgO) and Brønsted [KOH/SiO
and Ca(OH) ] bases.20 Moreover, the formation of 3-hydroxy-
Brønsted and/or Lewis acid sites, was also found.
Simultaneously with the appearance of these bands, the OH
stretching vibration bands of the P–OH (3670 cm−1) and Al–
OH groups (3797 and 3747 cm−1, due to tetrahedral and
octahedral coordination, respectively) disappeared. As the ref-
erence spectrum was acquired from the fresh catalyst, in the
DRIFT spectrum after PY adsorption the Al–OH and P–OH
bands appeared as downward bands [Fig. 8B(a)]. This means
that PY indeed interacted with hydroxyl groups on the P-
2
2
3
-methylbutan-2-one and/or 3-methylbut-3-en-2-one seems to
be characteristic of amphoteric solids possessing acid–base ion
pairs.27 Furthermore, 3-methylcrotonaldehyde (prenal) can
also be obtained by acid [3,3]-rearrangement.46 The overall
reaction scheme appeared in Fig. 9.
Here we report on the results of MBOH conversion, selected
as a dynamic method, for further characterization of the surface
acid–base properties of AlPO catalysts modified by caesium
0
3
–873 and E-0–873 samples. In addition, a weak band at
087 cm−1 was due to C–H stretching vibration of the pyridine
4
oxide loading.
In the absence of boundary, inter- and intra-particle diffusion
effects, the MBOH conversion reaction data (pulse measure-
ments) for all AlPO –caesium oxide catalysts were found to
aromatic ring.
Moreover, Fig. 8A(b)–(e) shows that the incorporation of
increased amounts of caesium oxide resulted in a strong
decrease in the absorption intensity of both Brønsted-bound
4
fulfil the Bassett–Habgood rate equation47 for first-order reac-
tions where the partial reaction pressure is low and the
adsorption rate is faster than the rate of surface reaction, the
latter being the rate-determining step. In this sense, only the
data for 3–20 mol% total conversion were used for
the calculation. The Bassett–Habgood equation was in the
form
(
1553 cm−1) and Lewis-bound (1451 cm−1) PY. Besides, the
amount of Brønsted-bound PY decreased faster than Lewis-
bound PY, indicating that the Lewis acid sites were stronger
than the Brønsted ones. However, for the P-30Cs-873 sample,
PY absorption bands completely disappeared. Furthermore,
for all P-Cs-873 and E-Cs-873 samples, adsorbed PY was
easily removed by treatment at 473 K for 1 h under nitrogen.
So, the changes in band intensities with caesium oxide loading
can be associated fairly well with the changes in acid character-
istics detailed in column six of Table 2 and in Fig. 7 (TPD
measurements).
ln [1/(1−X )]=k RT(W/F)
T
where X is total conversion, k the apparent rate constant of
a
T
a
both acid and base surface processes, W the catalyst weight,
and F the flow rate of carrier gas.
The apparent rate constants, k , were obtained from the
a
slopes of the linear plots of ln [1/(1−X)] vs. W for X less
than 20 mol%, where the equilibrium reaction can be neglected.
Catalytic measurements
T
The use of model reactions was recommended45 as the best
method for characterizing industrial catalysts. Thus, reactions
The values obtained at 573 K for the different AlPO –caesium
oxide catalysts are collected in the last column of Table 2. It
4
J. Mater. Chem., 1999, 9, 827–835
833

Table 3 Reaction selectivities (in pulse reactor; at different MBOH conversions) of AlPO and AlPO –caesium oxide catalysts
4
4
3
mol%
10 mol%
15 mol%
25 mol%
Catalyst
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
ACID
BASE
AMPH
CROT
ACID
BASE
AMPH
CROT
ACID
BASE
AMPH
CROT
ACID
BASE
AMPH
CROT
P-0–873
P-5Cs-873
P-10Cs-873 86.0
P-20Cs-873 38.1
P-30Cs-873
E-0–873
E-5Cs-873
E-10Cs-873 92.4
E-20Cs-873 74.6
98.0
94.0
1.5
5.4
10.4
60.0
99.8
0.8
1.5
4.7
23.4
98.8
0.4
0.5
1.9
1.4
—
0.3
0.5
1.0
2.0
—
0.1
0.1
1.7
0.5
—
0.2
0.3
1.9
—
98.1
96.0
86.7
37.1
0.4
98.5
97.8
92.2
70.7
1.2
1.4
3.2
9.3
60.6
99.6
0.9
1.2
4.8
27.2
98.8
0.3
0.6
2.1
1.4
—
0.4
0.6
1.0
2.0
—
0.2
0.2
1.9
0.9
—
0.2
0.4
2.0
—
97.8
97.0
87.4
33.9
0.8
98.3
97.7
92.2
66.2
1.0
1.4
2.4
8.7
64.2
99.2
0.9
1.0
4.4
31.8
99.0
0.3
0.6
2.0
1.2
—
0.5
0.5
1.2
2.0
—
0.4
—
1.8
0.7
—
0.3
0.8
2.2
—
98.0
96.8
88.7
32.4
0.7
98.3
97.4
92.2
60.3
0.9
1.2
2.0
7.4
65.9
99.3
1.0
1.0
4.5
38.1
99.1
0.3
0.6
2.1
1.1
—
0.5
0.6
1.4
1.5
—
0.5
0.6
1.8
0.7
—
0.3
1.0
1.9
—
0.2
98.7
97.7
E-30Cs-873
1.2
—
—
—
—
was found that the rate constants decreased as caesium oxide
loading increased. Besides, as was to be expected there was a
fairly good relation between the catalytic activity of investi-
gated samples (except for samples with 30 wt% caesium oxide
loading) and the amount of PY chemisorbed at 573 K
Table 4 Rate constants (k , at 573 K) and reaction selectivities (S) in
a
MBOH conversion (flow measurements, WHSV=25.5 h−1) over P-
0
–873, P-30Cs-873, MgO, ZnO and c-Al O catalysts
2
3
Reaction selectivities (mol%)
(
Table 2), where dehydration to 3-methylbut-3-en-1-yne was
Catalyst
106a/k /mol g−1 s−1
31.4
7.3
S
S
S
S
a
ACID
BASE
AMPH
CROT
predominant. Nevertheless, P-20Cs-873 and P-30Cs-873
samples showed more activity (base-catalysed) than E-20Cs-
8
better textural properties in P-Cs samples, thus favoring a
better dispersion of the caesium oxide.
As far as selectivity of MBOH conversion was concerned,
P-0–873
P-30Cs-873
P-30Cs-873b 108.4
MgO
ZnO
91.2
0.5
0.3
—
2.6
12.1
1.0 1.2
6.6
0.1
0.1
—
0.1
1.5
99.4
99.6
100.0
96.8 0.5
82.0 4.4
—
—
—
73 and E-30Cs-873 ones. This fact could be explained due to
114.6
30.5
8.3
c-Al O
2
3
Table 3 compares product selectivities (S
ACID BASE AMPH
) at four similar conversion levels (3–25 mol%) from
, S
, S
and
aTime on-stream: 2 h. bAt a reaction temperature of 623 K.
S
CROT
pulse measurements.
As can be seen on pure AlPO , MBOH underwent
dehydration to 3-methylbut-3-en-1-yne (S
incorporation of increasing amounts of caesium oxide into
crotonaldehyde (S
did not show significant changes within the 3–25 mol%
conversion range.
) remained very low. Besides, selectivities
4
CROT
) exclusively. The
ACID
AlPO produced a continuous decrease in the dehydration
selectivity whereas the selectivity for the base-catalysed cleav-
Moreover, MBOH conversions in a flow-reactor (WHSV=
25.5 h−1; 573 K), fitted by a first-order rate equation, corrobor-
ated pulse data. The activity profiles, as a function of time on
stream, for MBOH conversion on MgO, ZnO, c-Al O and
4
age to acetone and acetylene (S
BASE
selectivities to 3-methylbut-3-en-2-one (S
AMPH
) increased. In all cases, the
) and 3-methyl-
2
3
Fig. 10 Reaction profiles in 2-methylbut-3-yn-2-ol (MBOH) conversion on MgO, ZnO, c-Al O and P-30Cs-873 catalysts: (#) total conversion,
2
3
X ; (%) selectivity to 3-methylbut-3-en-1-yne, S
ACID
1) selectivity to 3-methylcrotonaldehyde (prenal).
; (6) selectivity to acetone and acetylene, S
; (() selectivity to 3-methylbut-3-en-2-one;
T
BASE
(
8
34
J. Mater. Chem., 1999, 9, 827–835

P-30Cs-873 presented in Fig. 10 whereas the reaction rate
constants, after 2 h on stream, are listed in Table 4. Table 4
4
5
6
F. M. Bautista, J. M. Campelo, A. Garcia, R. Guarde n˜ o, D. Luna
and J. M. Marinas, J. Chem. Soc., Perkin Trans. 2, 1989, 493.
J. M. Campelo, R. Chakraborty and J. M. Marinas, Synth.
Commun., 1996, 26, 415.
J. M. Campelo, R. Chakraborty and J. M. Marinas, Synth.
Commun., 1996, 26, 1639.
H. Hattori, Chem. Rev., 1995, 95, 537 and ref. therein.
D. Barthomeuf, Catal. Rev. Sci. Eng., 1996, 38, 521 and
ref. therein.
compared also the product selectivities (S
ACID BASE AMPH
), at 25 mol% MBOH conversion level, for P-0–873,
, S
, S
and S
CROT
P-30Cs-873, MgO, ZnO and c-Al O catalysts.
2
3
As can be seen from Fig. 10, the MBOH conversion
decreased with the time on-stream although the selectivities
7
8
for dehydration to 3-methylbut-3-en-1-yne (S
ACID
base-catalysed cleavage (retrocondensation) to acetone and
) and for the
9
0
Y. Ono and T. Baba, Catal. Today, 1997, 38, 321 and ref. therein.
R. M. Martin-Aranda, M. A. Vicente-Rodriguez, J. M. Lopez-
1
acetylene (S ) did not show any change within 2–13 h on
Pestana, A. J. Lopez-Peinado, A. Jerez, J. D. Lopez-Gonzalez and
˜
BASE
stream. Moreover, MBOH conversion increased strongly with
temperature although the activity decay with time on-stream
was also very important.
M. A. Ba n˜ ares-Mu n˜ oz, J. Mol. Catal. A, 1997, 124, 115.
P. E. Hathaway and M.E. Davis, J. Catal., 1989, 116, 263.
P. E. Hathaway and M.E. Davis, J. Catal., 1989, 116, 279.
A. Auroux, P. Artizzu, I. Ferino, R. Monaci, E. Rombi and
V. Solinas, Microporous Mater., 1997, 11, 117.
P. E. Hathaway and M.E. Davis, J. Catal., 1989, 119, 497.
1
1
1
1
2
3
Furthermore, as can be seen from Table 4, P-30Cs-873
catalyst showed rate constants and reaction selectivities similar
to those of typical solid base catalysts, such as MgO, when
the reaction temperature was increased to 623 K. However, in
all cases, the activity decay was important.
1
4
15 N. K. Das and K. Pramanik, J. Indian Chem. Soc., 1997, 74, 701.
16 M. Lasperas, H. Cambon, D. Brunel, I. Rodriguez and P. Geneste,
Stud. Surf. Sci. Catal., 1995, 97, 319.
1
1
1
7
8
9
J. C. Kim, H. X. Li, C. Y. Chen and M. E. Davis, Microporous
Mater., 1994, 2, 413.
F. Yagi, H. Tsuji and H. Hattori, Microporous Mater., 1997, 9,
Conclusion
2
37.
F. Yagi and H. Hattori, Microporous Mater., 1997, 9, 247.
The AlPO –caesium oxide (5–30 wt% caesium oxide) materials
4
20 M. Huang and S. Kaliaguine, Catal. Lett., 1993, 18, 373.
remained amorphous when calcined at 873 K. Also, there was
2
1
K. R. Kloetstra and H. van Bekkum, J. Chem. Soc., Chem.
Commun., 1995, 1005.
a surface enrichment in caesium compared to bulk values,
although there was a homogeneous distribution of caesium
when compared to that of different particles. Moreover, caes-
ium interacted preferentially with P–OH groups. The incorpor-
ation of caesium oxide leads, simultaneously, to a progressive
decrease in surface area and pore volume (larger at 30 wt%
caesium oxide) as well as to an increase in the most frequently
occurring pore radius. Besides, caesium oxide reduced both
the number and strength of acid sites for as long as the
caesium content increased. Thus, 2-methylbut-3-yn-2-ol under-
went dehydration to 3-methylbut-3-en-1-yne (acid activity)
22 K. R. Kloetstra and H. van Bekkum, Stud. Surf. Sci. Catal., 1997,
05, 43.
1
2
3
K. R. Kloetstra, J. van den Broek and H. van Bekkum, Catal.
Lett., 1997, 47, 235.
2
4
K. R. Kloetstra, M. van Laren and H. van Bekkum, J. Chem.
Soc., Faraday Trans., 1997, 93, 1211.
25 H. Lauron-Pernot, F. Luck and J. M. Popa, Appl. Catal., 1991,
8, 213.
7
2
6
C. Lahousse, J. Bachelier, J. C. Lavalley, H. Lauron-Pernot and
A. M. Le Govic, J. Mol. Catal., 1994, 87, 329.
2
7
F. Audry, P. E. Hoggan, J. Saussey, J. C. Lavalley, H. Lauron-
Pernot and A. M. Le Govic, J. Catal., 1997, 168, 471.
almost exclusively on pure AlPO whereas its modification
28 Savitsky-Golay, Anal. Chem., 1964, 36, 1627.
4
2
9
E. P. Barrett, L. S. Joyner and P. Halenda, J. Am. Chem. Soc.,
with increasing amounts of caesium oxide developed AlPO -
4
1
951, 73, 373.
based materials with increased basic properties and hence high
3
3
0
1
G. Halsey, J. Chem. Phys., 1948, 16, 931.
selectivities to the base-catalysed cleavage of MBOH yielding
acetone and acetylene (99 mol% for E-30Cs-873 and P-30Cs-
J. M. Campelo, A. Garcia, D. Luna, J. M. Marinas and
A. A. Romero, Thermochim. Acta, 1995, 261, 175.
8
73 catalysts). So, the incorporation of caesium oxide provided
32 O. W. Florke, Z. Kristallogr., 1967, 125, 134.
33 F. d’Yboire, Bull. Soc. Chim. Fr., 1962, 1762.
a means of controlling the surface acid–base characteristics of
3
4
D. Muller, E. Jahn, G. Ladwig and U. Haubenreisser, Chem.
Phys. Lett., 1984, 109, 332.
C. S. Blackwell and R. L. Paton, J. Phys. Chem., 1984, 88, 6135.
V. M. Mastikhin, I. L. Mudrakowsky, V. P. Shmachkova and N.
S. Kotsarenko, Chem. Phys. Lett., 1987, 139, 93.
AlPO in order to achieve typical solid acid or solid base
4
catalysts.
35
36
Acknowledgements
37 J. B. Peri, Discuss. Faraday Soc., 1971, 52, 55.
3
8
S. Brunauer, L. S. Deming, W. S. Deming and E. Teller, J. Am.
Chem. Soc., 1940, 62, 1723.
The authors acknowledge subsidies from the DGESIC (Project
PB97/0446), Ministerio de Educacion y Cultura, and from the
Consejeria de Educaci o´ n y Ciencia (Junta de Andalucia),
Espa n˜ a. The authors also thank Dr. R. Ruiz (NMR Service,
Universidad de C o´ rdoba) for performing MAS NMR measure-
ments. They also would like to thank Professor M. Sullivan
for linguistic revision of the manuscript.
39 K. S. W. Sing, D. H. Everett, R. A. W. Haul, L. Moscou,
R. A. Pierotti, J. Rouquerol and T. Siemieniewska, Pure Appl.
Chem., 1985, 57, 603.
4
4
0
1
W. D. Harkins and G. Jura, J. Chem. Phys., 1943, 11, 431.
J. C. P. Broekhoff and B. G. Linsen, in Physical and Chemical
Aspects of Adsorbents and Catalysts, ed. B. G. Linsen, Academic
Press, London, 1970, pp. 1–62.
4
2
A. Mata Arjona, J. B. Parra Soto and C. Otero Arean, Stud. Surf.
Sci. Catal., 1982, 10, 175.
4
3
4
E. P. Parry, J. Catal., 1963, 2, 371.
References
4
M. R. Basila, T. B. Kantner and K. H. Rhee, J. Phys. Chem.,
1964, 68, 3197.
1
A. Blanco, J. M. Campelo, A. Garcia, D. Luna, J. M. Marinas
and A. A. Romero, J. Catal., 1992, 137, 51 and ref. therein.
F. M. Bautista, J. M. Campelo, A. Garcia, D. Luna, J. M.
Marinas, A. A. Romero, J.A. Navio and M. Macias, J. Catal.,
45 M. R. Guisnet, Acc. Chem. Res., 1990, 23, 392.
46 C. Mercier, G. Allmang and M. Aufrand, Appl. Catal. A, 1994,
2
1
14, 51.
4
7
D. Bassett and H.W. Habgood, J. Phys. Chem., 1960, 64, 769.
1
998, 173, 333 and ref. therein.
3
F. M. Bautista, J. M. Campelo, A. Garcia, J. Leon, D. Luna and
J. M. Marinas, J. Chem. Soc., Perkin Trans. 2, 1995, 815.
Paper 8/07538G
J. Mater. Chem., 1999, 9, 827–835
835