Você está na página 1de 11

Journal of Chemical Technology and Biotechnology J Chem Technol Biotechnol 82:513–523 (2007)

Perspective
Catalytic performance of Brønsted acid
sites during esterification of acetic acid with
ethyl alcohol over phosphotungestic acid
supported on silica

Abd El-Aziz A Said,∗ Mohamed M M Abd El-Wahab and Alian Mohamed Alian
Chemistry Department, Faculty of Science, Assiut University, Assiut, Egypt

Abstract: Different ratios of phosphotungestic acid supported on silica gel were prepared by an impregnation
method with PWA loadings ranging from 1 to 30% w/w and calcined at 350 and 500 ◦ C for 4 h in a static air
atmosphere. The catalysts were characterized by thermogravimety (TG), differential thermal analysis (DTA),
X-ray diffraction, FT-IR spectroscopy and N2 adsorption measurements. The surface acidity and basicity of the
catalyst were investigated by the dehydration–dehydrogenation of isopropanol and the adsorption of pyridine (PY)
and 2,6-dimethyl pyridine (DMPY). The gas-phase estrification of acetic acid with ethanol was carried out at
185 ◦ C in a conventional fixed-bed reactor at 1 atm using air as carrier gas. The results clearly revealed that the
catalyst containing 10% w/w PWA/SiO2 is the most active and delivers reaction selectively to ester with 85% yield.
The Brønsted acid site resulting from hydroxylation of tungsten oxide plays the main role in the formation of ester.
 2007 Society of Chemical Industry

Keywords: phosphotungestic acid; silica; structure; surface area; acidity; esterification; acetic acid; ethanol

INTRODUCTION and carbon are good supports, and among them silica
Keggin type heteropoly acids (HPAs), as represented is the most often used.
by phosphotungestic acid, are unique and versatile According to Ohlman et al.8 the thermal stability
catalysts since they possess dual catalytic functions of silica-supported heteropolyanions decreases with
of strong acidity and high oxidizing ability.1 – 5 This respect to that of the bulk acid, particularly at low
allows them to be used as catalysts in various liquid concentrations. The chemical interactions between
phase reactions at moderate temperature6 with yields the HPA and the support are a matter of great
higher than those of other currently used acids such interest because a strong interaction could fix the
as sulfuric and perchloric acids. It is important to HPA to the carrier, avoiding the leaching of HPA
emphasize that the acidity of HPA can be controlled in liquid phase reactions9,10 or maintaining a high
by selecting suitable HPA compounds, the partial HPA dispersion in gas-phase reactions. However,
neutralization of the acid and the dispersion degree of the continuous dissolution of HPA in polar medium
different supports.7 was the most severe problem for the supported
Studies of the HPA catalytic behavior have focused samples.11,12 To overcome this problem, gas-phase
mainly on the acids and their salts. However, reactions may be one alternative method. Recently, we
supported systems are of great practical interest have reported that the gas-phase reaction of acetic acid
because the catalytic activity is directly related to with ethyl alcohol on supported metal oxides13 – 16 and
the surface area in some HPA-catalyzed reactions. phosphomolybdic acid17 showed good results towards
One of the main drawbacks of the Keggin type HPA is ester formation. The main advantages of this method
their low specific surface areas (1–10 m2 g−1 ),4,6 which are the possibility of regeneration of the catalysts and
limit application in gas-solid phase reactions. These reduction of the environmental impact.
disadvantages can be overcome by dispersing the HPA On the other hand, the catalytic gas-phase esterifica-
on high surface area supports. The use of a support tion of acetic with ethyl alcohol on phosphotungestic
allowing the HPA to be dispersed over a large surface acid supported on silica, to our knowledge, has not
must result in an increase in its catalytic activity. It has been reported. In this paper, the acid catalysis of the
been found that silica, mesoporous molecular sieves, silica including H3 PW12 O40 is examined for gas-phase


Correspondence to: Abd El-Aziz A Said, Chemistry Department, Faculty of Science, Assiut University, Assiut, Egypt
E-mail: a.a.said@acc.aun.ed.eg
(Received 17 December 2006; revised version received 20 March 2007; accepted 26 March 2007)
Published online 15 May 2007; DOI: 10.1002/jctb.1704

 2007 Society of Chemical Industry. J Chem Technol Biotechnol 0268–2575/2007/$30.00


Perspective

esterification of acetic acid with ethyl alcohol and their desorption curves using Nova enhanced data reduction
characterization by physical and chemical methods. software (Version 2.13). A highly pure nitrogen gas
In particular, the results of the esterification reaction was used as adsorbate.
are presented and their correlation with Brønsted acid
sites on the catalyst surfaces is discussed. Acidity and basicity determination
The acidity and basicity of the catalysts under inves-
tigation were determined by studying the adsorption
EXPERIMENTAL of pyridine (PY), 2,6-dimethyl pyridine (DMPY) and
Materials the dehydration–dehydrogenation of isopropyl alco-
Acetic acid, ethyl alcohol, isopropyl alcohol, pyridine hol (IPA). The decomposition of IPA was carried
and 2,6-dimethyl pyridine were obtained as pure out in a conventional fixed-bed flow Pyrex glass tube
reagents and were used without further purification. reactor, at atmospheric pressure using nitrogen as a
Silica gel (grade, 60–200 mesh), the support material, carrier gas. The reaction conditions were: 0.5g cat-
and phosphotungestic acid were supplied by Merck alyst weight, 2% IPA in the gas feed, 130 ml min−1
(Darmstadt, Germany). The catalysts were prepared total flow rate and 185 ◦ C reaction temperature. The
by impregnation of silica gel with various loadings measurement of propene yield (%) was made after
of phosphotungestic acid (PWA) dissolved in doubly 2 h to achieve steady-state conditions. The adsorption
distilled water. The samples produced were dried in an of PY and DMPY were carried out by injection of
oven at 100 ◦ C for 24 h before being calcined at 300, different volumes at steady state conditions. The exit
350, 400 and 500 ◦ C for 4 h in a static air atmosphere. feed was analysed by direct sampling of the gaseous
The content of PWA added was varied between 1 and products into a Unicam ProGC gas chromatograph
30% w/w (silica supported PWA denoted SPWA). using a flame ionization detector (FID) with a 10%
PEG 400 glass column (2 m).
Apparatus and techniques
Thermal analysis Catalytic activity measurements
Thermogravimetry (TG) and differential thermal anal- The catalytic activity of the catalysts for the vapor-
ysis (DTA) were performed on heating (at 10 ◦ C phase esterification of acetic acid with ethyl alcohol
min−1 ) test samples to 600 ◦ C in a dynamic atmo- was carried out at 185 ◦ C in a conventional fixed-
sphere of air (40 mL min−1 ) using a computerized Shi- bed flow type reactor at atmospheric pressure using
madzu Thermal Analyzer TA6O Apparatus (Japan). dry air as carrier gas. The system comprised two
reactors. One was used without any catalyst and
filled with glass beads (control reactor), to enable
Fourier transform infrared (FT-IR) spectroscopy
a measurement of the ‘control’ conversion (if any),
FT-IR spectra of the prepared catalysts calcined
which was subtracted from that measured in the flow
at 350 and 500 ◦ C for 4 h were recorded in the
reactor. 0.5 g of catalyst was placed in the middle of the
4000–200 cm−1 region with a Nicolet (Madison, WI,
second reactor with quartz wool. Space in the reactor
USA) 710 FT-IR equipped with data station. Dried
pre- and post-heating zone was filled with glass beads
samples of about 10 mg were mixed with 100 mg
to reduce the effect of auto-oxidation of the substrate
of spectral grade KBr and pressed into discs, under
and products in the gas phase. A mixture of ethanol,
hydraulic pressure.
acetic acid and air was introduced into the reactor after
air was bubbled through ethyl alcohol and acetic acid
X-ray diffraction (XRD) saturators. The total flow rate was fixed at 130 mL
XRD analysis of the test samples was performed with min−1 and used 1.6 and 2.1% ethanol and acetic acid
a Philips (The Netherlands) diffractometer (Model in the gas feed, respectively. The gases after reaction
PW 2103, λ = 1.5418 Å, 35 kV and 20 mA) with were chromatographically analysed by FID with a
a source of CuKα radiation (Ni filtered). An on- Unicam ProGC using a 2 m DNP glass column for
line data acquisition and handling system facilitated analysis of the reaction products of ethanol and acetic
automatic JCPDS library search and match for phase acid on the tested catalysts. The connection between
identification purposes. the reactor and the gas chromatograph (GC) was
heated by resistance wire to prevent any condensation.
Nitrogen gas adsorption Measurements of the conversion and yield (%) were
Nitrogen gas adsorption–desorption isotherms were recorded after 2h from the initial introduction of the
measured at −196 ◦ C using a Nova 3200 instrument reactants into the reactor to ensure the attainment of
(Quantachrom Instrument Corporation, USA). Test the reaction equilibrium, (steady state conditions).
samples were thoroughly outgassed for 2 h at 200 ◦ C to
a residual pressure of 10−4 torr, and the weight of the
outgassed sample was that used in calculations. The RESULTS AND DISCUSSION
specific surface area, SBET was calculated by applying Thermal analysis
the Brunauer-Emmett-Teller (BET) equation. The Figure 1 shows TG and DTA curves of the pure silica,
porosity of the catalysts was determined from the pure PWA and PWA supported on silica. It is seen

514 J Chem Technol Biotechnol 82:513–523 (2007)


DOI: 10.1002/jctb
Perspective

that the silica precursor16,17 loses weight with heating


from ambient temperature to 600 ◦ C. This weight
loss, which was accompanied by a broad band on
the DTA curve peaking at 80 ◦ C, can be attributed
to the physically adsorbed water. Above 400 ◦ C and
up to 600 ◦ C a little weight loss was observed. The
weight loss in this region can be considered as a result
of dehydroxylation of hydroxyl groups. Unsupported
PWA shows that the weight loss exhibits three stages
on heating to 600 ◦ C, maximized at 85, 185 and 485 ◦ C
as indicated by the DTA curve. This behavior reflects
the fact that PWA loses its water of crystallization18
near 200 ◦ C to give the Keggin structure, which is
stable on heating to 350 ◦ C. The endothermic peak
observed on the DTA curve at 485 ◦ C may be
attributed to the removal of PWA protons, which
may react with anionic oxygen atoms19 (from PWA
itself) to form water between 400 and 500 ◦ C. Curves
for 10, 20 and 30% w/w PWA supported on silica
show that the first two broad peaks maximized at 85
and 185 ◦ C correspond to the physically adsorbed
water and removal of structural water molecules,
respectively. Further heating to 600 ◦ C is reflected
in one loss step maximized at 485 ◦ C, which can be
attributed to the decomposition of keggin structure
due to deprotonation of PWA into simple oxide.20,21
It is important to note that our findings match the
results of Hodjati et al.22

Infrared analysis of the calcinated products of


Figure 1. TG and DTA curves for pure SiO2 , pure PWA and SiO2
pure PWA and supported on silica supported PWA.
FTIR spectra were measured in the range
4000–200 cm−1 using a KBr disc technique.
Figure 2(a) and (b) show the FTIR spectra obtained Nowinska et al.27 neither a systematic shift nor a
for pure silica, pure PWA and PWA supported on weak band at 925 cm−1 , which may result from the
SiO2 calcined at 350 and 500 ◦ C, respectively, for higher calcination temperature, could be found for
4 h in air atmosphere. Figure 2(a) shows the FTIR silica supported PWA samples with different loadings.
spectra of pure SiO2 and supported PWA catalysts On increasing the calcination temperature to
calcined at 350 ◦ C for 4 h. It shows a typical infrared 500 ◦ C, Fig. 2(b), the results indicate that the spectra
spectrum of silica23,24 with bands assigned at 1700, are similar to those observed on calcination at
1640, 980 and 800 cm−1 together with two broad 350 ◦ C. On the other hand, the spectrum of pure
bands at 3430 and 1100 cm−1 . The spectrum of pure PWA indicates the disappearance of some bands
PWA shows six bands in the range 1250–500 cm−1 . in the range 1250–500 cm−1 . This reflects the fact
These bands are assigned at 1089, 984, 889, 789, that the Keggin structure decomposes into WO3 on
594 and 520 cm−1 , which are the characteristic bands increasing the calcination temperature from 350 to
of the Keggin structure. These were assigned25 to 500 ◦ C. In the context of supported catalysts, FTIR
Vas (W = Ot ), Vas (W – Ob – W), Vas (W – Oc – W), spectroscopy28 – 30 has shown that water is eliminated
δ (O – P – O) and Vs (W – O – W), respectively. from heteropoly acids at two different temperatures,
In addition, the band observed at 594 cm−1 may be one approximately 200 ◦ C, and the other in the range
attributed to the δ (O–P–O) vibration. Moreover, the 400–500 ◦ C. The former is due to the desorption
weak bands assigned at 1629, 1700 and 3374 cm−1 of hydrogen-bonded water; the latter apparently
may be characteristic H2 O and OH stretching results from water formed by the acidic protons
vibrations. The IR spectra of the different loadings and anionic oxygen. These findings are supported
of PWA on silica calcined at 350 ◦ C indicates that using temperature programmed desorption studies18
most of the characteristic bands of the parent Keggin of the bulk PWA, which shows that upon increasing
structure can be found in the PWA fingerprint region temperature, a number of peaks due to evolution
(1250–500 cm−1 ), not shown, or appearing in the of water appeared. Therefore, on calcinations above
same assignable position of the bands corresponding 350 ◦ C, we expect that the heteropoly acid has
to the SiO2 host material. In comparison with the been decomposed forming (W12 O36 .1/2P2 O5 ) mixed
IR results reported by Mohana Rao et al.26 and oxides.

J Chem Technol Biotechnol 82:513–523 (2007) 515


DOI: 10.1002/jctb
Perspective

to an addition of 20% w/w. With further increase


in the loading percentage to 30% w/w, a small peak
corresponding to PWA starts to appear. On increasing
the calcination temperature to 500 ◦ C, Fig 3(b), small
peaks corresponding to pure PWA are observed
in samples containing 10 and 20% w/w, whereas
the catalyst containing 30% w/w PWA exhibits a
crystalline material.
It is worth mentioning here that the disappearance
of the characteristics peaks of PWA for the samples
with low loading (≤10% w/w) and the appearance of
these peaks in the diffractograms of samples containing
20% w/w PWA reflect the high dispersion of PWA in
the porous silica, and the composition behaves as an
amorphous material.31
The information acquired by IR, XRD, TG and
DTA suggests a strong interaction between the host
silica network and the guest PWA molecules. It
follows that PWA in the silica samples prepared here
is well dispersed and strongly bound in the porous

Figure 2. IR spectra of pure SiO2 , pure PWA and PWA supported on


silica (different % w/w loadings): (a) calcined at 350 ◦ C in air for 4 h;
(b) calcined at 500 ◦ C in air for 4 h.

X-ray diffraction
XRD patterns determined for the various silica
supported PWA samples with different loadings and
calcined at 350 and 500 ◦ C are illustrated in Figs 3(a)
and 3(b). As shown in Fig. 3(a), the XRD patterns
revealed PWA present in an amorphous state on the
SiO2 support. However, it has been reported that at
low loading (≤10 wt %), PWA forms a finely dispersed
species on a silica surface, and interacting species such
as (≡SiOH2 )+ (H2 PW12 O40 )− may form,8 while at
high loading (≥20% w/w) a PWA crystal phase is
developed on silica. These results are consistent with
those reported by Izumi et al.4 and Kozhevnikov et al.2
They did not obtain the characteristics diffraction
peaks of PWA until the loading exceeded 20% w/w.
Therefore, the XRD patterns obtained for pure PWA
and SiO2 containing 10, 15, 20 and 30% w/w of Figure 3. X-ray diffraction of pure SiO2 , pure PWA and PWA
PWA calcined at 350 ◦ C show only one broad peak supported on silica (different % w/w loadings). (a) calcined at 350 ◦ C
as observed for pure SiO2 and supported PWA up in air for 4 h; (b) calcined at 500 ◦ C in air for 4 h.

516 J Chem Technol Biotechnol 82:513–523 (2007)


DOI: 10.1002/jctb
Perspective

Table 1. Variation of specific surface area (SBET ), total pore volume (TPV) and average pore diameter (APD) with calcination temperature for
PWA-SiO2 catalysts

300 ◦ C 350 ◦ C 400 ◦ C 500 ◦ C

Calcination Temp. SBET TPV APD SBET TPV APD SBET TPV APD SBET TPV APD
PWA % Mixing m2 g−1 cc/g Å m2 g−1 cc/g Å m2 g−1 cc/g Å m2 g−1 cc/g Å

Pure SiO2 288 0.351 397 285 0.360 376 283 0.345 379 281 0.361 373
5 270 0.361 328 274 0.378 384 280 0.384 353 285 0.394 342
7 258 0.353 363 264 0.374 397 272 0.380 363 281 0.394 332
10 254 0.342 341 259 0.312 358 268 0.350 382 277 0.386 331
15 236 0.333 330 245 0.296 345 255 0.328 375 265 0.359 350
20 208 0.227 423 211 0.214 433 217 0.208 365 247 0.340 413
30 166 0.224 430 170 0.182 484 178 0.172 373 218 0.309 409

silica network. These observations are in accordance Acidity and basicity determination
with earlier findings indicating the reliability of the Determination of the surface acidic and basic sites
preparation method.32 – 34 Results for the catalytic dehydration–dehydrogenation
of IPA over PWA supported on silica calcined at 350
and 500 ◦ C are shown in Fig. 4. IPA dehydration has
Surface area and porosity
been used by several authors35,36 as a test reaction
The surface area of PWA supported on silica gel
to determine the acidity of different catalysts and
samples calcined at 300, 350, 400 and 500 ◦ C
it proceeds quickly on weak acid sites.37 In the
was determined on the basis of nitrogen adsorption
working conditions used, it was found that IPA reacts
and desorption measurements. The isotherms show
on the bare support and on the PWA supported
type II BDDT classification closed hysteresis loops
catalysts selectively to propene. In samples calcined
at intermediate relative pressures. The pore size
at 350 ◦ C, the addition of 3% w/w PWA, as indicated
distribution of these samples was determined from
in Fig. 4, sharply increases the conversion of IPA to
the desorption data. The specific surface area of the
propene. This increase reaches a maximum for the
investigated catalysts, SBET , was calculated from these
addition of 10% w/w, and remains constant up to
measurements and is shown in Table 1.
an addition of 30% w/w. Figure 4 also shows the
It is clear from Table 1 that pure SiO2 calcined
results of the decomposition reaction of IPA catalyzed
at 300–500 ◦ C has SBET values between 288 and
by PWA/SiO2 calcined at 500 ◦ C. It is seen that the
281 m2 g−1 . Results also show that the addition of
addition of PWA into SiO2 increases the yield of
PWA catalyst into the SiO2 support led to a continuous
propene steadily, passing through a maximum at 7%
decrease in SBET values up to the addition of 30% w/w
w/w, followed by a continuous decreases up to 30%
PWA for all calcination temperatures. In contrast,
w/w addition. It is clear from the above results that
samples containing 20 or 30% w/w exhibited a
the catalysts calcined at 350 ◦ C exhibits higher activity
remarkable decrease in surface area. At even higher
towards propene formation than the catalysts calcined
coverage, crystallization of PWA takes place, filling
at 500 ◦ C.
the pores of the support. The much lower surface
area at 20 or 30% w/w addition is due to the strong
influence on the number of pores and the formation
of a monolayer PWA surface, which has lower surface 100 350 °C
area. It is worth noting from the above results that the 500 °C
SBET of the prepared catalysts increased on increasing
80
the calcination temperature. This behavior may be
% Conversion

attributed to the creation of new pores accompanying


the removal of protons and the decomposition of the 60
Keggin structure to simple oxide.
The total pore volume also decreased, while an 40
increase in the average pore diameter is associated
with increase in PWA loading. The decrease in SBET
can be explained by the different distribution of PWA 20
on the SiO2 surface. These changes are suspected
to be caused by plugging of the pores due to the 0
agglomeration of PWA. A pore size analysis of the SiO2 1 3 5 7 10 15 20 30 PWA
tested samples (results not shown) showed a single % Mixing of PWA
peak distribution that is effectively located at the Figure 4. Activity variation of IPA with percentage mixing (i.e,
average pore radius, indicating the presence of a different % w/w loadings). of PWA supported on SiO2 calcined at
microporous structure. 350◦ and 500 ◦ C in air for 4 h.

J Chem Technol Biotechnol 82:513–523 (2007) 517


DOI: 10.1002/jctb
Perspective

The above results show that catalyst containing −−−→ C3 H6 + + H+ .. O2− ku (4)
10% w/w PWA has sufficient acid sites to produce
the dehydration pathway for IPA better than that In this mechanism, 2-propanol is first protonated at
of unsupported acid. The difference between the the acidic sites (step 1) of the heteropoly oxometalat
performances of the supported catalysts is explained leading to perturbation of the electron density
on the basis of PWA interaction with the support via distribution in the protonated species and a weakening
the creation of more acidic sites, which encourage ultimate scission of the C–O bond.30,43 The resulting
the dehydration reaction and the effect of percentage carbocation C3 H7 + (step 2) is found to be associated
loading on the dispersion of the acid and in turn, with the (presumably terminal) oxygen atoms of the
the catalyst performance. The selectivity to propene heteropoly oxometalate anions30,43 (step 3). After that
as a function of percentage mixing of PWA is 100%. the dehydration cycle may be completed via one
Selectivities were not influenced by IPA conversion in reaction path, i.e. propene (step 4) production.
the conversion range covered (1–30% w/w). The following conclusions can be drawn from
Significant changes in the number and nature of catalytic dehydration of IPA over PWA supported
(Brønsted or Lewis) acid sites have been observed on silica:
with changes in PWA content. Yet, IPA dehydration
selectivities were not affected by these structural (i) The catalysts calcined at 350 ◦ C possess higher
changes, suggesting either that only one type of acid activity towards propene formation than the
site (Brønsted or Lewis) catalyzes the dehydration catalysts calcined at 500 ◦ C; (ii) the decrease in
reaction or that the different types of sites present acidity of the catalysts calcined at 500 ◦ C may
lead to identical selectivities. However, the ability of be attributed to decomposition of the Keggin
the catalyst to break the C–OH bond during the structure to WO3 ; and (iii) the prepared catalysts
chemisorption itself by polarizing the C–O bond and exhibit an acidic character with selectivity 100%,
leaving the O–H bond intact may be important in which reflects that there are no basic sites available
catalyzing the dehydration reaction. This is facilitated on the catalyst surfaces.
by the presence of acidic groups.29 In agreement with
such a suggestion; Okuhara et al.38 reported that, on Adsorption of pyridine and 2,6-dimethyl pyridine over
Nax H3−x PW12 O40 , the formation of ethylene from 10% w/w PWA supported on SiO2 calcined at 350 and
ethyl alcohol requires fission of the C–O bond on 500 ◦ C
the Brønsted acid sites. Moreover, it is known that It is known that the chemisorption of pyridine (PY)
propene formation requires Brønsted acid sites.39 and 2,6-dimethyl pyridine (DMPY) can be used
Based on the above discussion, one may conclude as basic probe molecules to determine the acidity
that, the dehydration activity of the catalysts is mainly of catalyst.44 It was reported that PY is selectively
due to; (i) the Brønsted acidity of the catalysts, aided adsorbed on both Brønsted (B) and Lewis (L) acid
to a large extent by the water adsorption, which sites.45,46 On the other hand, DMPY is selectively
is known to create Brønsted sites;39 (ii) the flexible adsorbed on Brønsted acid site47,48 but not Lewis sites
nature of the secondary structure of heterocompounds, because of the steric hindrance of two methyl groups.
which allows the polar molecule, e.g. isopropanol, to So the difference between PY and DMPY adsorption
penetrate the bulk of the solid;38 and (iii) the existence is a measure of the Lewis acid sites.
of the Keggin structure, which has an influence in The poisoning of the active surface sites of 10% w/w
stabilizing the reaction intermediates. PWA supported on silica catalysts (SPWA10) in IPA
Accordingly, it is highly probable that the active conversion was performed through saturation of the
sites in the conversion of 2-propanol are Brønsted in acid sites with PY or DMPY according to the following
nature and it is most likely that the reaction proceeds procedures. After measuring the conversion activity
via a carbonium ion mechanism. Such a mechanism of the SPWA10 catalyst at 185 ◦ C the catalyst was
was recommended by Hayashi and Moffat40 – 42 for the injected with different volumes of PY or DMPY in the
dehydration of alcohols over heteropoly compounds. stream of the reactants using N2 as a carrier gas. The
According to the above discussion, the dehydration results obtained are shown in Figs 5 and 6. Figure 5
reaction mechanism may be systemized as: shows the influence of PY or DMPY additions on
the distribution products of IPA reaction on SPWA10
C3 H7 OH + H+ . . . . O2− ku calcined at 350 ◦ C. The results indicate that PY and
DMPY suppressed IPA dehydration activity. Thus,
−−−→ C3 H7 OH2 + . . . O2− ku (1)
the chemisorbed PY and DMPY decrease the yield of
+
C3 H7 OH2 the propene by ≈90 and 80%, respectively.
From the above results, we see that the volume of PY
−−−→ C3 H7 + + H2 O (2)
required to inhibit the activity of the catalysts towards
+
C3 H7 + O 2−
ku propene is higher than that of DMPY by about 10%.
+ This value corresponds to the presence of Lewis acid
−−−→ C3 H7 + . . . . . . O2− ku (3)
sites. Thus, the SPWA10 catalyst is an acidic catalyst
+
C3 H7 + . . . . . . O 2−
ku with major Brønsted acid sites and minor Lewis

518 J Chem Technol Biotechnol 82:513–523 (2007)


DOI: 10.1002/jctb
Perspective

available sites on the catalyst surface which process


100 the dehydration reaction to propene formation.
PY In conclusion, the chemisorption of PY and DMPY
DMPY indicates that the concentration of acid sites, and
80 especially of Brønsted sites (measured with DMPY),
grows on increasing the PWA content up to the
% Conversion

addition of 10% w/w. The distribution of these sites is


60 (≈80%) Brønsted (≈10%) strong Lewis and (≈10%)
weak Lewis acid sites.
40
Effect of catalytic reaction temperature on the
decomposition of IPA over 10% w/w PWA/SiO2 catalyst
20 presaturated with PY and DMPY
The poisoning of active sites of the most active catalyst
(SPWA10) calcined at 350 ◦ C in the IPA conversion
0 was performed through the previous saturation of acid
0 2 4 6 8 10 12
Volume of PY & DMPY, mL sites with PY or DMPY according to the following
procedure. The catalyst at room temperature was
Figure 5. Activity variation of IPA with volume of PY and DMPY over saturated with base probe molecules after nitrogen
SPWA10 catalyst calcined at 350 ◦ C in air for 4 h.
gas was bubbled through PY or DMPY saturators.
The total flow rate was fixed at 120 mL min−1 at room
temperature (the time of saturation was 15 h). The
100 PY
conversion activity of the catalyst after saturation for
DMPY IPA decomposition was measured at different reaction
temperatures using the conditions that were used for
80 the decomposition of IPA on unsaturated catalyst.
The results of conversion activity are shown in Fig. 7.
% Conversion

The results obtained reveal that the adsorption of PY


60 retards the conversion activity of IPA towards propene
more so than that of DMPY. This means that PY was
adsorbed on the acid sites more strongly than DMPY.
40
On increasing the temperature, the removal of PY
adsorbed on the Lewis site was shown to be faster
20 than that on the Brønsted acid site.
The results obtained demonstrate that the removal
of PY or DMPY pre-adsorbed on weak acid sites
0 occurs at less than 150 ◦ C whereas the intermediate
0 2 4 6 8 10 12
strength and strong acid sites are desorbed at
Volume of PY & DMPY, mL temperatures above 150 ◦ C. Therefore, the above
Figure 6. Activity variation of IPA with volume of PY and DMPY over
results also indicate that the investigated catalysts have
SPWA10 catalyst calcined at 500 ◦ C in air for 4 h. weak, intermediate and strong acid sites.

acid sites. These experiments confirmed that 2,6-


dimethyl pyridine does not saturate Lewis acid sites
100
or even adsorb competitively with IPA on SPWA10
catalyst. The addition of PY, however, decreases IPA
80
dehydration activity to a greater extent than DMPY,
% Conversion

confirming that PY saturates the available Brønsted 60


and Lewis acid sites. Even after saturation of the
SPWA10 surface with PY, IPA dehydration activity untreated catalyst
40
remained at about 10% of the initial value. This may treated with PY
treated with DMPY
reflect the competitive adsorption of PY and IPA 20
on Lewis acid sites that is too weak to adsorb PY
irreversibly. In addition, Fig. 6 shows the variations 0
of conversion and propene yield with additions of PY 100 150 200 250 300
or DMPY on SPWA10 catalyst calcined at 500 ◦ C. Reaction Temperature, oC
It is clear from this figure that there is no difference Figure 7. Activity variation of IPA with reaction temperature on
between the amount chemisorbed from PY or DMPY. unsaturated and saturated 10% w/w PWA/SiO2 catalyst with PY and
This means that the Brønsted sites are the only DMPY, calcined at 350 ◦ C in air for 4 h.

J Chem Technol Biotechnol 82:513–523 (2007) 519


DOI: 10.1002/jctb
Perspective

100% selectivity only for PWA added up to 10% w/w.


100 However, the activity and selectivity of the prepared
Conv. catalyst towards the formation of ethyl acetate depend
Y. ethyl acetate on the distribution of the acid sites available on the
% Conversion, yield & selectivity

80 Y. acetaldehyde catalyst surfaces. In addition, the behavior of the


S. ethyl acetate activity products suggests that the role of catalysts
S. acetaldehyde
calcined at 350 ◦ C, which possess a Keggin structure,
60
differs from that of catalysts calcined at 500 ◦ C, in
which the WO3 is the active component. Also, it is
40 important to mention here that both activity and yield
of ester are correlated with two factors: the calcination
temperature and the percentage loading of PWA.
20 Therefore, if most attention is paid to product
selectivity, it is clearly indicated that intermediate and
strong acid sites led to a direct dehydration (i.e. ester)
0 result on the supported catalysts in the range (1–10%
SiO2 1 3 5 7 10 15 20 30 PWA
w/w) for catalysts calcined at 350 ◦ C. A sharp decrease
% Mixing of PWA
in selectivity towards ester formation, but an increase
Figure 8. Catalytic esterification of acetic acid with ethanol over in selectivity towards the by-products, acetaldehyde,
PWA/SiO2 catalysts calcined at 350 ◦ C in air for 4 h. is noticed with increasing percentage loading of PWA
up to 30% w/w. These results demonstrate that when
PWA is highly dispersed on the support (as indicated
100
by XRD and FT-IR spectra), it behaves as an active
Conv.
and selective catalyst towards ethyl acetate formation.
Y. ethyl acetate On the other hand, when the percentage loading of
% Conversion, yield & selectivity

80 Y. acetaldehyde PWA >10% w/w, its selectivity decreases sharply. It is


S. ethyl acetate interesting to mention here that pure PWA possesses
S. acetaldehyde
≈15% yield of ester and ≈85% yield of acetaldehyde.
60 This means that the proton (Brønsted acid site) is
not a major selective site towards the production of
40
ester.16 In contrast with Brønsted acids, Lewis acids
in the catalyst support give very different results. This
is evidenced by the continuous decrease in catalyst
20 selectivities when PWA loading exceeds 10% w/w.
The use of zirconia-supported HPA catalysts in the
esterification of acetic acid and isoamyl alcohol showed
0 high selectivity to ester formation.49 Moreover, it was
SiO2 1 3 5 7 10 15 20 30 PWA
observed that the catalytic activities of these supported
% Mixing of PWA
HPA catalysts for esterification are higher than that
Figure 9. Catalytic esterification of acetic acid with ethanol over of bulk HPA. This could be explained by taking into
PWA/SiO2 catalysts calcined at 500 ◦ C in air for 4 h. account that the catalytically active surface increases
when the acid is dispersed on the support. These
Catalytic activity results are in agreement with the present study and
The catalytic estrification of acetic acid with ethyl with that reported by Ghosakikali et al.50 These results
alcohol on silica supported PWA calcined at 350 and support the view of dependence of the activity and
500 ◦ C for 4 h was carried out at 185 ◦ C and the selectivity on the acidity of the catalysts, and on
results are shown in Figs 8 and 9. It is seen that the ability of the catalyst to activate and incorporate
pure SiO2 support is inactive towards ethyl acetate gaseous oxygen into crystal lattice oxygen (O2− ) by
formation. In addition, a significant increase in the electron transfer from the surface14,51 accompanied by
yield of ethyl acetate was observed on SiO2 supported a release of electrons (from alcohol) to the surface of
PWA reaching a maximum value (≈85 and 56%) for the catalysts:
the addition of 10% w/w, then an observable decrease
2e−
for additions up to 30% w/w for catalysts calcined at ↔
350 and 500 ◦ C, respectively. Although the yield of Surface oxygen in the gas phase
O2−
ester decreases, the yield of acetaldehyde increases for
additions up to 30% w/w. From the acidity and basicity measurement, Fig. 4,
Further, comparison between selectivities of the it is clear that the prepared samples exhibit only
catalysts calcined at 350 and 500 ◦ C shows that acidic catalysts. Thus, the dehydrogenation step of
supported catalysts calcined at 500 ◦ C exhibits 100% ethyl alcohol during the esterification reaction (i.e.
selectivity whereas catalysts calcined at 350 ◦ C exhibit formation of acetaldehyde) is rather easy because

520 J Chem Technol Biotechnol 82:513–523 (2007)


DOI: 10.1002/jctb
Perspective

both electrophilic and nucleophilic oxygen species can CONCLUSIONS


abstract hydrogen. The main conclusions that can be derived from the
Further increase in the calcination temperature of results obtained are:
the supported catalysts to 500 ◦ C, Fig. 9, results in
1- The Keggin structure is active and selective only at
an observable decrease in the activity conversion
lower loadings ( ≤10% w/w PWA). The maximum
to ethyl acetate. It is worth mentioning here that
yield of ester (≈85%) is obtained at a loading
although the yield of ester is less than that obtained
of 10% w/w calcined at 350 ◦ C. Above this
for catalysts calcined at 350 ◦ C, the selectivity of ratio, at which a monolayer formed, the yield of
all supported catalysts reaches 100%. Moreover, the ester decreases whereas the yield of acetaldehyde
selectivity of pure PWA calcined at 500 ◦ C is also increases.
improved. These results reveal that the decomposition 2- The acidity measurements revealed that three types
of keggin structure to WO3 within the silica support of acid site existed on the prepared catalysts,
possesses acidic catalyst with only Brønsted acid sites namely weak and strong Lewis and Brønsted acid
as predicted from the acidity measurement, Fig. 4. sites.
However, Wang et al.52 have shown that the 3- The Brønsted acid site resulting from hydroxylation
development of acid sites was related to the of tungsten oxide plays a main role in the formation
coordinatively unsaturated W6+ species on the surface of ester whereas the Brønsted acid site from the
(Lewis acid site)53 and to the hydroxyl groups formed proton of unsupported PWA does not play a main
by protonating the bridging Si–O–W bonds on the role in the production of ester.
sample surface (Brønsted acid sites). The results 4- Decomposition of the Keggin structure to WO3
suggested that tungsten oxide was selectively dispersed within the SiO2 support exhibits less activity but
on the surface of silica with a dispersion capacity gives 100% selectivity towards ester formation.
of 0.5 W6+ nm−2 based on the weaker interaction This means that the Brønsted acid sites created
between tungsten oxide species and supports. The on the surface of these catalysts encourage the
states of the tungsten oxide species on the surface dehydration process in the reaction of ethyl alcohol
of silica were strongly dependent on the loading with acetic acid.
amount of tungsten oxide, the nature of support and
the calcination temperature. In samples with loading
amounts of tungsten oxide lower than the dispersion REFERENCES
capacity, tungsten oxide species on the catalyst were 1 Izumi Y, Otake M and Kagaku S, Catalysts with acid-base
and redox activities – heteropoly acid catalysis. Chem Rev
mainly in the tetrahedrally coordinated state. When 34:116–141 (1982).
the WO3 loadings were increased and were higher 2 Kozhevnikov IV and Mateveev KI, Homogeneous catalysts
than the dispersion capacity, distorted octahedrally based on heteropoly acid. Appl Catal 5:135–150 (1983).
coordinated tungsten oxide species formed on the 3 Misono M, Acidic and catalytic properties of heteropoly
compounds. Catal Rev Sci Eng 29:269–282 (1987).
surface of the catalyst. Increase of calcination 4 Izumi Y, Urabe K and Onaka M, Zeolite, Clay and Heteropoly
temperature of the samples to 500 ◦ C led to the Acid in Organic Reactions, Kodansha, Tokyo/VCH, New York,
conversion of tungsten oxide species on catalyst from p. 990 (1992).
the tetrahedrally coordinated state to the octahedrally 5 Okahara T, Mizuno N and Misono M, Catalytic chemistry of
heteropoly compounds. Adv Catal 41:113–252 (1994).
coordinated state.
6 Izumi Y, Hasebe R and Urabe K, Catalysis by heterogeneous
Hua and Sommer54 have shown that the activity supported heteropoly acid. J Catal 84:402–409 (1983).
is strongly dependent on WOx concentration. Both 7 Haber J, Pamin K, Matachowski L and Mucha D, Catalytic
catalysts are slightly active below 5% w/w W loading. performance of the dodecatungestophosphoric acid on
Increasing WOx concentration first leads to a large different supports. Appl Catal A 256:141–152 (2003).
8 Fricke R and Ohlman G, Electron spin resonance studies of free
increase in catalytic activity and then a small decline. It and supported 12-heteropoly acids PartI: Air and vacuum
was reported that the acidity and number of Brønsted dehydration of H3 [PMo12 O40 ]xH2 O. J Chem Soc Faraday
acid sites present on PWA/SiO2 sample increased Trans I 82:263–271 (1986).
with the tungsten oxide loading up to monolayer 9 Izumi Y and Urabe K, Catalysis of heteropoly acids entrapped
in activated carbon. Chem Lett 5:663–666 (1981).
coverage.55,56 As the surface WOx density increases
10 Schwegle MA, Van Bekkum H and De Munck NA,
above monolayer coverage, formation of polytungstate Molybdenum-tungsten interchange of heteropolyanions as
species is responsible for generating Brønsted acid a measure of stability a phosphorus-31NMR. Appl Catal A
sites from H2 on PWA/SiO2 via the partial reduction 74:191–200 (1991).
11 Izumi Y, Urabe K and Onaka M, Zeolite, Clay and Heteropoly
of W6+ Lewis acid centers and delocalization of the
Acids in Organic Reactions. VCH, Weinheim, p. 131 (1992).
negative charge, as suggested by Baertch et al.56 12 Mizuno N and Misono M, Heterogeneous catalysis. Chem Rev
The catalytic activity measurements of PWA 98:199–217 (1998).
supported on silica systems towards the esterification 13 Said AA, Gas phase esterification of acetic acid with ethanol over
of acetic acid with ethanol provide consistent MoO3 supported on AlPO4 and the effect of modification
with phosphomolybdic acid and Ce4+ ions. J Chem Technol
information about changes taking place in the surface Biotechnol 78:733–742 (2003).
properties of silica supports after the incorporation 14 Said AA, Abd El-Wahab MMM and El-Shobaky GA, Study on
of PWA. the alkali metal-promoted V2 O5 supported on silica used as a

J Chem Technol Biotechnol 82:513–523 (2007) 521


DOI: 10.1002/jctb
Perspective

catalyst for the esterification of acetic acid with ethyl alcohol. 33 Izumi Y, Ono M, Kitagawa M, Yoshida M and Urabe K, Silica-
Oxid Commun 27:402–412 (2004). included heteropoly compounds as solid acid catalysts.
15 Abd El-Wahab MMM and Said AA and El-Shihry ShS, Struc- Microporous Mater 5:255–262 (1995).
tural and catalytic activity of V2 O5 -supported on AlPO4 34 Izumi Y, Urabe K and Onaka M, Advances in liquid-phase
catalysts. Montashefte Chemie 135:357–370 (2004). organic reactions using heteropolyacid and clay. Microporous
16 Said AA and Abd El-Wahab MMM, Surface properties and Mesoporous Mater 21:227–233 (1998).
catalytic behavior of MoO3 /SiO2 in esterification of acetic 35 Sohn JR and Jang HJ, Characterization of zirconia-silica
acid with ethanol. J Chem Technol Biotechnol 81:329–335 unmodified and modified with sulfuric acid and acid catalysis.
(2006). J Mol Catal 64:349–360 (1991).
17 Abd El-Wahab MMM and Said AA, Phosphomolybdic acid 36 Yamaguchi O, Uegaki T, Miyata Y and Shimizu K, Formation
supported on silica gel and promoted with alkali metal ions as of AlVO4 [aluminum vanadate] solid solution from alkoxides.
catalysts for the esterefication of acetic acid by ethanol. J Mol J Am Ceram Soc 70:198–200 (1987).
Catal 240:109–118 (2005). 37 Kakaoka T and Dumesic JA, Acidity of unsupported and silica
18 Hondnett BK and Moffat JB, Application of temperature- supported vanadia, molybdena and titania as studied by
programmed desorption to the study of heteropoly com- pyridine adsorption. J Catal 112:66–79 (1988).
38 Okuhara T, Kasai A, Hayakawa N, Yoneda Y and Misono M,
pounds: Desorption of water and pyridine. J Catal
Catalysis by heteropoly compounds VI. The role of the bulk
88:253–263 (1984).
acid sites in catalytic reactions over Nax H3−x PW12 O40 . J
19 Moffat JB and Imelik B (Eds), Catalysis by Acids and Basses.
Catal 83:121–130 (1983).
Elsevier, Amesterdam, p. 157 (1985).
39 Akimoto M, Shima K, Ikeda H and Echigoya E, 12-Hetero-
20 Blasco T, Corma A, Martinez A and Martinez-Escolano PJ,
polymolybdates as catalysts for vapor-phase oxidative dehy-
Supported heteropoly acid (HPW). catalysts for the continu-
drogenation of isobutyric acid: 2 Group Ib, IIb, IIIb, and VIII
ous alkylation of isobutane with 2-butene: The benefit of using metal salts. J Catal 86:173–186 (1984).
MCM-41 with larger pore diameters. J Catal 177:306–313 40 Hayashi H and Moffat JB, The properties of heteropoly acids
(1998). and the conversion of methanol to hydrocarbons. J Catal
21 Moffat JB and Kasztelan S, The oxidation of methane on 77:473–484 (1982).
heteropolyoxometalates. J Catal 109:206–211 (1988). 41 Hayashi H and Moffat JB, Methanol conversion over metal salts
22 Hodjati S, Petit C, Pitchon V and Kiennemann A, Removal of 12-tungstophosphoric acid J. Catal 81:61–66 (1983).
of NOx from a lean exhaust gas by absorption on 42 Hayashi H and Moffat JB, Conversion of methanol into
heteropolyacids: Reversible sorption of nitrogen oxides in hydrocarbons over ammonium 12-tungstophosphate. J Catal
H3 PW12 O40 ·6H2 O. J Catal 197:324–334 (2001). 83:192–204 (1983).
23 Rocchiccioli–Deltcheff C, Aouissi A, Launay S and Fournier M, 43 Highfield JG and Moffat JB, Characterization of sorbed inter-
Silica-supported 12-molybdophosphoric acid catalysts: Influ- mediates and implications for the mechanism of chain growth
ence of the thermal treatments and of the Mo contents on in the conversion of methanol and ethanol to hydrocarbons
their behavior, from IR, Raman, X-ray diffraction studies and over 12-tungstophosphoric acid using infrared photoacoustic
catalytic reactivity in the methanol oxidation. J Mol Catal spectroscopy. J Catal 98:245–258 (1986).
114:331–342 (1996). 44 Corma A, Fornes V, Kolodziejski W and Martineztriguero IJ,
24 Rocchiccioli-Deltcheff C, Amirouche M and Fournier M, Struc- Orthophosphoric acid interactions with ultrastable zeolite-y:
ture and catalytic properties of silica-supported polyoxo- infrared and NMR studies. J Catal 145:27–36 (1994).
molybdates III 12-molybdosilicic acid catalysts: vibrational 45 Ziolek M, Kujawa J, Saur O, Aboulayt A and Lavalley JC,
study of the dispersion effect and nature of the Mo species Influence of sulfur dioxide adsorption on the surface
in interaction with the silica support. J Catal 138:445–456 properties of metal oxides. J Mol Catal A 112:125–132
(1992). (1996).
25 Essayem N, Holmqvist A, Gayraud PY, Vedrine JC and Ben 46 Aramendia MA, Borau V, Jimenez C, Marinas JM, Porras A
Taarit Y, In situ FTIR studies of the protonic sites of and Urbano FJ, Magnesium oxides as basic catalysts for
H3 PW12 O40 and its acidic cesium salts Mx H3−x PW12 O40 . organic processes: Study of the dehydrogenation-dehydration
J Catal 197:273–280 (2001). of 2-propano. J Catal 161:829–838 (1996).
26 Mohana Rao K, Gobetto R, Iannibello A and Zecchina A, 47 Fei Shen Y, Suib SL, Deeba M and Koermer GS, Lumines-
Solid state NMR and IR studies of phosphomolybdenum cence and IR characterization of acid sites on alumina. J
and phosphotungsten heteropoly acids supported on SiO2 , Catal 146:483–490 (1994).
γ Al2 O3 , and SiO2 Al2 O3 . J Catal 119:512–516 (1989). 48 Bautista FM, Campelo JM, Garcia A, Luna D, Marinas JM,
Moreno MC, et al., Acidity and catalytic activity of
27 Nowinska K, Fiedorow R and Adamiec J, Catalytic activity of
AlPO4 -B2 O3 and Al2 O3 -B2 O3 (5–30 wt% B2 O3 ) systems
supported heteropoly acids for reactions required strong acid
prepared by impregnation. Appl Catal A 170:159–168 (1998).
centers. J Chem Soc Faraday Trans 87:749–753 (1991).
49 Pizzio L, Vazquez P, Caceres C and Blanco M, Tungstophos-
28 Highfield JG and Moffat JB, Characterization of 12-tungsto-
phoric and molybdophosphoric acids supported on zirconia
phosphoric acid and related salts using photoacoustic
as esterification catalysts. Catal Lett 77:233–239 (2001).
spectroscopy in the infrared region: I Thermal stability and
50 Ghanbari-Siahkali A, Philippou A, Dwyer J and Anderson MW,
interactions with ammonia. J Catal 88:177–187 (1984). The acidity and catalytic activity of heteropoly acid on MCM-
29 Abu-Zied BM, Synthesis, characterization and catalytic activity 41 investigated by MAS NMR, FTIR and catalytic tests. Appl
studies on acid-type catalysts. PhD thesis, Assiut University Catal A 192:57–69 (2000).
(1997). 51 Ai M, The relationship between the oxidation activity and the
30 Highfield JG and Moffat JB, Elucidation of the mechanism acid-base properties of Fe2 O3 -aseed mixed oxides. J Catal
of dehydration of methanol over 12-tungstophosphoric 60:306–315 (1979).
acid using infrared photoacoustic spectroscopy. J Catal 52 Wang Y, Chen Q, Yang W, Xie Z, Xu W and Huang D, Effect
95:108–119 (1985). of support nature on WO3 /SiO2 structure and butene-1
31 Kukovecz A, Balogi Zs, Konya Z, Toba M, Lentz P, Niwa SI, metathesis. Appl Catal A 250:25–37 (2003).
et al., Synthesis, characterisation and catalytic application of 53 Martin C, Malet P, Solana G and Rives V, Structural analysis
sol-gel derived silica-phosphotungstic acid composites. Appl of silica-supported tungstates. J Phys Chem B 102:2759–2768
Catal A 228:83–94 (2002). (1998).
32 Molnar A, Keresszegi C and Torok B, Heteropoly acids immo- 54 Hua W and Sommer J, Alumina-doped Pt/WOx /ZrO2 catalysts
bilized into a silica matrix: characterization and catalytic for n-heptane isomerization. Appl Catal A 232:129–135
applications. Appl Catal A 189:217–224 (1999). (2002).

522 J Chem Technol Biotechnol 82:513–523 (2007)


DOI: 10.1002/jctb
Perspective

55 Sheinthauer M, Cheung TK, Jentoft RE, Grasselli RK, Gates 56 Baertsch CD, Soled SL and Iglesia E, Isotopic and chemical
BC and Knozinger H, Characterization of WOx /ZrO2 titration of acid sites in tungsten oxide domains supported on
by vibrational spectroscopy and n-Pentane isomerization zirconia. J Phys Chem B 105:1320–1330 (2001).
catalysis. J Catal 180:1–13 (1998).

J Chem Technol Biotechnol 82:513–523 (2007) 523


DOI: 10.1002/jctb

Você também pode gostar