You are on page 1of 11

Applied Catalysis A: General 175 (1998) 43±53

Skeletal isomerization of 1-butene on 12-tungstophosphoric


acid supported on zirconia
E. LoÂpez-Salinasa,*, J.G. HernaÂndez-CorteÂza, Ma.A. CorteÂs-JaÂcomea, J. Navarretea,
Ma.E. Llanosa, A. VaÂzqueza, H. ArmendaÂriza, T. LoÂpezb
a
Instituto Mexicano del PetroÂleo, SubdireccioÂn de TransformacioÂn Industrial, Eje Central LaÂzaro CaÂrdenas 152, 07730 MeÂxico DF, Mexico
b
Universidad AutoÂnoma Metropolitana-Iztapalapa, PO Box 55-534, 09340 MeÂxico DF, Mexico

Received 3 January 1998; received in revised form 1 June 1998; accepted 1 June 1998

Abstract

A series of 0±25 wt% H3[W12PO40] (TPA) impregnated on freshly precipitated Zr(OH)4 were prepared. The solids were
characterized by a Hammett indicator method, speci®c surface area and pore size measurements, pyridine adsorption FTIR
and later tested as catalysts in the isomerization of 1-butene. Maximum acid strength of 5±20 wt% TPA/ZrO2 calcined at
673 K is H0ˆÿ9.3, but dried samples (393 K) showed near superacid strength, i.e. H0ÿ13.75. The sites are mainly strong
Lewis acids, but TPA supported on stabilized ZrO2 (calcined at 773 K) shows both strong Brùnsted and Lewis acidity. TPA
addition to hydrated Zr(OH)4 stabilizes the surface area of the ®nal calcined material in comparison with that of pure ZrO2; the
greater the TPA content the higher the resulting surface area. Skeletal rearrangement of 1-butene to isobutylene proceeds on
5±25 wt% TPA/ZrO2 but not on pure ZrO2, the greater the TPA content the higher the initial selectivity towards isobutene. In
contrast, 20 wt% TPA supported on SiO2 formed no isobutylene. # 1998 Elsevier Science B.V. All rights reserved.

Keywords: 12-Tungstophosphoric acid; Zirconia; Solid acid; Skeletal 1-butene isomerization

1. Introduction stabilized zirconia (e.g. depletion of OH groups).


For instance, the extensive studies of Hino and Arata
One of the main reasons that has drawn great [4±8] showed that sulfate addition to some freshly
attention to the use of hydrated zirconia as a precursor precipitated metal oxides such as Fe2O3, TiO2, ZrO2
of a catalyst carrier is the fact that its surface OH and HfO2, yielded solids with very strong acid proper-
groups are able to undergo a chemical reaction or ties. These solids which have an acid strength higher
strong interaction with incorporated second compo- than 100% H2SO4 (H0ˆÿ11.9) are called solid super-
nents such as oxo-anions (as described below) and acids (SSA). Most attention has been paid to sulfated
transition metal cations [1±3], respectively. Upon zirconia, ®rst reported by Holm and Bailey [9] in
thermal activation, hydrated zirconia yields catalysts 1962, because it has shown the highest acid strength
with very different properties from those obtained by among sulfated metal oxides (H0ÿ16.04). Jin et al.
incorporating the second component on previously [10] ascribed the strong acidity of these solids to the
electron withdrawing surface-bonded (SO4)ˆ anion,
*Corresponding author. E-mail: esteban@tsekub.imp.mx creating a coordinatively unsaturated and electron

0926-860X/98/$ ± see front matter # 1998 Elsevier Science B.V. All rights reserved.
PII: S0926-860X(98)00198-7
44 E. LoÂpez-Salinas et al. / Applied Catalysis A: General 175 (1998) 43±53

de®cient metal center that behaves as a strong Lewis addition to hydrated ZrO2 could yield strong solid
acid site. The addition of sulfate to ZrO2 yields not acids.
only a solid acid 104 times stronger than 100% H2SO4 Attempting to develop strong solid acid catalysts,
[11] but also stabilizes the tetragonal phase of zirconia we selected the 12-tungstophosphoric acid (TPA),
and consequently its speci®c surface area [12]. How- since it shows the highest acidity among Keggin-type
ever, Babou et al. [13] using quantum chemistry HPAs, and impregnated it on freshly precipitated
calculations suggested that the acidity of sulfated ZrO2. In this study we report on the preparation
zirconia corresponds, roughly speaking, to that of pure method of x wt% H3[W12PO40]/ZrO2 (xˆ0±25 wt%),
H2SO4, i.e. this solid is not really superacidic but only their characterization and catalytic evaluation in the
strongly acidic. Recently, we have corroborated these skeletal isomerization reaction of 1-butene.
results by DRIFTS and 1 H MAS NMR spectroscopy
studies on sulfated zirconias [14].
Instead of sulfate, other oxo-anions like tungstate 2. Experimental
and molybdate anions impregnated on hydrated zir-
conia produced SSA: WO3/ZrO2 (H0ÿ14.52) and 2.1. Preparation of H3[W12PO40]/ZrO2
MoO3/ZrO2 (H0ˆÿ12.7) [15]. The addition of Fe and
Mn to SOˆ 4 =ZrO2 has a promoting effect on its acid The solids precipitated from zirconium salts
strength, being three orders of magnitude more active are complexes with a general formula (ZrOx-
in the isomerization of n-butane than SO4/ZrO2 (OH)4ÿ2xyH2O)n, where the relative amount of oxy-
[16,17]. A recent report on the synthesis of B2O3/ gen, hydroxide and water depends mainly on the rate
ZrO2 (H0ˆÿ13) obtained by impregnation of boric of precipitation [24] and the precursor salt [25,26]. For
acid, which has low acidity (pKaˆ5.0), has con®rmed the sake of clarity hereinafter they will be referred to
that the interaction of oxo-anions with hydrated zir- simply as hydrated or freshly precipitated ZrO2. The
conia indeed results in the formation of stronger acid hydrated ZrO2 in this study was prepared by slow
sites than those of the algebraic sum of the single precipitation of an aqueous ZrO(NO3)2xH2O solution
components [18]. with aqueous NH4OH. The pH of the mixture was kept
Heteropolyacids (HPAs) are another class of ef®- as 10 by controlling the addition of the two solutions.
cient catalysts for various reactions in the liquid- The white gel obtained was repeatedly washed with
phase, e.g. dehydration, etheretion, esteri®cation deionized water and separated in a centrifuge. After
and isomerization. They usually exhibit much higher this, the white paste was dried in an oven at 373 K for
catalytic activities than ordinary solid acids such as 12 h. The impregnation of the heteropolyacid on the
zeolite and silica±alumina [19]. HPAs are formed by hydrated zirconia was carried out as follows: suf®cient
the condensation of more than two different oxo- amounts of an ethanolic solution of 2.510ÿ3 M
anions, for instance, reagent grade H3[W12PO40]6H20 to obtain 5.0,
10.0, 16.6, 20.0 and 25.0 wt% on ZrO2, were poured
12‰WO4 Šˆ ‡ ‰HPO4 Šˆ ‡ 23H‡ onto hydrous ZrO2. The materials were transferred to a
! ‰W12 PO40 Š3ÿ ‡ 12H2 O rotary evaporator where ethanol was evaporated at
353 K. After this, the solids were stored in a desiccator
The HPAs having the Keggin structure are ther- containing both anhydrous CaO and silica gel. These
mally more stable and easy to synthesize than other solids will be hereinafter referred to as 5TPA/Z,
HPAs. Due to their relatively small surface areas (1± 10TPA/Z, 16.6TPA/Z, 20TPA/Z and 25TPA/Z.
10 m2 gÿ1) [20], HPA compounds are commonly
supported on SiO2, SiO2±Al2O3, TiO2 or activated 2.2. Preparation of H3[W12PO40]/SiO2
carbon when used as catalysts for gas±solid phase
reactions [21]. Supports such as alumina and magnesia Since SiO2 has been the preferred catalyst support
which exhibit surface basicity bring about the decom- of a variety of heteropoly compounds, SiO2 was
position of HPAs [22,23]. Due to the low acidity of prepared by a sol±gel method for comparison. First
ZrO2 and its potential interaction with HPAs, their 74 cm3 of tetraethyl orthosilicate (TEOS) was dis-
E. LoÂpez-Salinas et al. / Applied Catalysis A: General 175 (1998) 43±53 45

solved in 48 cm3 of anhydrous ethanol, heated at was quanti®ed by comparing the areas under the curve
353 K and stirred for 2 h in re¯ux conditions. After in the sample with those of known amounts of NH3
this, the solution was cooled down to 268 K in an ice- injected after each run. The NH3-TPD analysis was
salt bath. A mixture made up of 48 cm3 of H2O, carried out in an AMI-3 apparatus from Altamira
37 cm3 of ethanol and 2 cm3 of 1 N H2SO4 was added Instruments. The amount of acid sites per Keggin unit
dropwise to the TEOS±ethanol solution. The resulting (KU) in TPA/Z catalysts was calculated by dividing
mixture was kept for 3 h at 268 K. Then, the sol was the difference between the amount of total acid sites
heated at 353 K at re¯ux conditions for 12 h. After and that of pure ZrO2 by the amount of (KU) contained
this, the mixture was dried in a rotary evaporator at in each catalyst, i.e., assuming that all TPA were
393 K. 16.6 and 20.0 wt% TPA were impregnated on accessible to ammonia molecules and that the acid
dried SiO2 as described above for Zr(OH)4. These sites in pure ZrO2 would be added to those generated
solids will be referred to as 16.6TPA/S and 20TPA/S. by incorporating the TPA.
Speci®c surface areas using the BET method with
2.3. Characterization techniques N2 adsorption at 77 K and pore size distribution by
means of N2 adsorption±desorption isotherms were
The measurement of Hammett acid strength (H0) measured in an ASAP-2000 apparatus from Micro-
was carried out by exposing previously evacuated meritics.
samples (0.1 g) at 393 K for 4 h to a benzene solution Catalytic evaluation of TPA/Z, and other solids for
of selected Hammett indicators: (2-bromo-4,6-di- comparison, was carried out in a ®xed bed ¯ow
nitroaniline, pKaˆÿ6.6; anthraquinone, pKaˆÿ8.2; microreactor connected to a gas chromatograph
2,4,6-trinitroaniline, pKaˆÿ10.1; p-nitrotoluene, equipped with a silica fused capillary column (KCl/
pKaˆÿ11.35; p-nitrochlorobenzene, pKaˆÿ12.7; Al2O3, 50 m, i.d. 0.32 mm) and a ¯ame ionization
2,4-dinitrotoluene, pKaˆÿ13.75). detector. 1-Butene, N2 and H2 were of reagent grade
The nature of the acid sites (Brùnsted or Lewis) was (99.99% Matheson and Infra, respectively) and were
determined in a Fourier transform infrared (FTIR) further puri®ed by passing them through a dryer
Nicolet 710SX spectrometer by means of pyridine containing molecular sieve. The previously calcined
adsorption. The equipment was furnished with a cell catalysts were heat-treated in situ at 673 or 773 K, as
with interchangeable windows in which the samples, indicated, in a stream of N2 for 1 h before reaction and
previously pressed into thin self-supported wafers, then cooled to reaction temperature. Isomerization of
were placed in order to be evacuated (110ÿ6 Torr) 1-butene was carried out with 0.1 g (mesh 150±200) of
in situ at the calcination temperature for 30 min, and catalyst at 673 K, 101.33 KPa, 1-butene/H2
then, after cooling down to room atmosphere, to (molar)ˆ1 and a total WHSVˆ25.2 hÿ1. The conver-
adsorb pyridine. sion of 1-butene is de®ned as the percentage of 1-
The thermal programmed desorption of NH3 (NH3- butene converted to all products.
TPD) was used to estimate the amount and strength of
acid sites formed on the surface of TPA/Z. First, a
sample of previously calcined TPA/Z was heated 3. Results and discussion
(10 K minÿ1) in ¯owing helium (Linde UHP,
99.999%) until reaching the indicated temperature 3.1. Acid strength enhancement
and kept for 60 min. After this, the sample was
allowed to cool down to room temperature and The highest acid strength of ZrO2 calcined in air at
exposed to ¯owing NH3 (20% in He) for 30 min. 773 K for 3 h is H0 between ‡1.5 and ‡3.3 [27].
Later, the system was purged at 303 K for 30 min Accordingly, ZrO2 is considered a weakly acidic oxide
with helium in order to eliminate weakly adsorbed made up mainly of Lewis-type acid sites [28]. On the
NH3. The temperature of the sample rose linearly other hand, solid H3[W12PO40]6H2O contains acid
(10 K minÿ1) from ambient to the indicated tempera- sites stronger than H0ˆÿ8.2 [29]. Table 1 shows the
tures. NH3 concentration was monitored by means of a Hammett acid strength of various TPA/Z catalysts heat
thermal conductivity detector. The amount of NH3 treated at indicated temperatures. The TPA/Z materi-
46 E. LoÂpez-Salinas et al. / Applied Catalysis A: General 175 (1998) 43±53

Table 1
Hammett acid strength of TPA/Z catalysts

TPA/Z catalysts Hammett acid strength (H0)

TPA (wt%) Calc. T (K) ÿ6.6 ÿ8.1 ÿ9.3 ÿ10.1 ÿ11.35 ÿ12.70 ÿ13.75
16.6 393 ‡ ‡ ‡ ‡ ‡ ‡ ‡
5.0 673 ‡ ‡  ± ± ± n.a.
10.0 673 ‡ ‡  ± ± ± n.a.
16.6 673 ‡ ‡ ‡ ± ± ± n.a.
20.0 673 ‡ ‡ ‡ ± ± ± n.a.
10.0 773 ‡ ± ± n.a. n.a. n.a. n.a.
16.6 773 ‡ ± ± n.a. n.a. n.a. n.a.
20.0 773 ‡ ± ± n.a. n.a. n.a. n.a.
n.a.: not analyzed.

als show superacid strength (H0ÿ13.75) when dried


at 393 K. However, when the samples are calcined at
673 K the maximum acid strength decreases to
H0ˆÿ9.3. Calcination at 773 K further decreases
the maximum acid strength to H0ÿ6.6.

3.2. Nature of acid sites

In order to throw light on the type of acid sites


(Lewis or Brùnsted) on TPA/Z, a 20TPA/Z calcined at
673 K was exposed to pyridine and the IR spectra were
recorded from 1700 to 1400 cmÿ1 (Fig. 1). Spectrum
(a) taken after pyridine adsorption at ambient tem-
perature, shows:
(1) very weak bands at 1637, 1546 and 1533 cmÿ1,
ascribed to Brùnsted acidity [30],
(2) intense bands at 1610 and 1442 cmÿ1, ascribed
to strong Lewis acidity [30] and
(3) two weak bands at 1595 and 1577 cmÿ1, arising
from weak Lewis acid sites.
The band at 1488 cmÿ1 has been ascribed to both Fig. 1. FTIR spectra of pyridine adsorbed on 20 wt% TPA
Lewis and Brùnsted acid sites [30]. When the sample supported on hydrated zirconia and then calcined at 673 K.
is evacuated at 373 K for 15 min (spectrum (b)) the Spectrum recorded after evacuating at: (a) room temperature, (b)
bands ascribed to Brùnsted and weak Lewis sites 373 K, (c) 473 K, (d) 573 K, and (e) 673 K. (f) Reference spectrum
mostly disappear. However, the bands at 1610 and taken before pyridine adsorption.
1422 cmÿ1 are visible when evacuating at 473, 573
and 673 K for 15 min (spectra (c), (d) and (e), respec- 773 K. Accordingly, the TPA Keggin structure can
tively). These results indicate that these bands arise be retained below 723 K [31,32].
from very strong Lewis acid sites. A broad band at ca. Nevertheless, it has been reported that pure TPA is
1540 cmÿ1 appears when the sample is heat treated at characterized by mainly Brùnsted-type acidity [19].
673 K (spectrum (e)) and is probably associated to the The impregnation of TPA on a freshly precipitated
structural transformation of metastable tetragonal to ZrO2 which upon further calcination undergoes dehy-
monoclinic ZrO2, which occurs between 673 and droxylation and concomitant structure transformation
E. LoÂpez-Salinas et al. / Applied Catalysis A: General 175 (1998) 43±53 47

Fig. 2. FTIR spectra of pyridine adsorbed on 20 wt% TPA


Fig. 3. NH3 temperature programmed desorption patterns of: (a)
supported on zirconia previously calcined at 773 K. Spectrum
ZrO2, (b) 5TPA/Z, (c) 10TPA/Z, (d) 16.6TPA/Z, (e) 20TPA/Z and
recorded after evacuating at: (a) room temperature, (b) 373 K, (c)
(f) 25TPA/Z.
473 K, (d) 573 K, and (e) 673 K. (f) Reference spectrum taken
before pyridine adsorption.

implies that some (OH) groups may react with TPA WˆO bonds of TPA might react with partially dehy-
protons to form water. In fact, careful experiments by droxylated …Zr…ÿO†3 †‡ species to form anchored TPA
NH3-TPD, temperature programmed reduction and species, e.g. the formation of Zr±O±W bonds. These
temperature programmed exchange, have shown that TPA species would exert an electron withdrawing
protons can be removed from the anhydrous effect on surface Zr4‡ cations, making them to behave
H3[W12PO40], but only when accompanied by loss as strong Lewis acid sites. Accordingly, the strongly
of anionic oxygen atoms [32,33]. To clarify this acidic properties of TPA/Z could be attributed to
behavior, a 20 wt% TPA supported on a previously electron-withdrawing effects similar to those ascribed
calcined ZrO2 (at 773 K) was examined by pyridine to Zr-bonded sulfate groups [10±12].
adsorption FTIR. Fig. 2 shows that the bands ascribed
to Brùnsted acid sites (1639 and 1538 cmÿ1) are not 3.3. Amount of acid sites on H3[W12PO40]/ZrO2
only more prominent than the corresponding ones in
Fig. 1, but also can be clearly observed up to 673 K, The relationship between the amount of supported
indicating that these Brùnsted sites are strong. In this TPA and the total number of acid sites on ZrO2 was
case, the near absence of surface (OH) groups on examined by means of the NH3-TPD method. Fig. 3
ZrO2, allowed protons to remain attached to the shows the NH3-TPD patterns of 0, 5, 10, 16.6, 20 and
heteropolyanion unit until reaching a temperature at 25TPA/Z calcined at 673 K for 4 h. Single ZrO2 shows
which protons react with oxometalate anionic oxygen a desorption pattern made up of two broad peaks with
and depart as water [32]. Assuming that some TPA maximum intensity at 443 and 543 K (pattern (a)). The
protons react with (OH) groups in hydrated zirconia, it desorption patterns of 5-20TPA/Z show similar shapes
would be reasonable to expect that some terminal to that of ZrO2, but the intensity of desorption peaks
48 E. LoÂpez-Salinas et al. / Applied Catalysis A: General 175 (1998) 43±53

Table 2
Total amount and density of acid sites in TPA/Z catalysts calcined at 673 K, as a function of the TPA content

TPA/Z catalysts Acid sites per KU Total acid sitesa Surface areaa Density of acid sitesb
(wt% TPA) (mol NH3/mol KU) (mmol NH3 gÿ1) (m2 gÿ1) (mmol NH3 mÿ2)

0.0 ± 617 132 4.7


5.0 9.1 776 149 5.2
10.0 5.7 812 159 5.1
12.5 5.2 843 173 4.9
16.6 5.3 922 183 5.2
20.0 5.6 1007 181 5.6
20.0c ± 518 132 3.9
25.0 11.8 1628 n.a. ±
a
Measured after calcining at indicated temperatures for 4 h. KU: Keggin unit. n.a.: not analyzed.
b
Total acid sites divided by surface area.
c
TPA impregnated on previously calcined ZrO2.

increases as the TPA content rises. Additionally, the in the density of acid sites is observed (Table 3). These
peak at high temperature becomes more prominent as results can be rationalized by assuming that the KU
the TPA content increases (compare pattern (d) with decomposes between 623 and 773 K [31,32]. These
(a)). Clearly, the addition of the TPA to hydrated results suggest that TPA addition does not increase the
zirconia increases the number of acid sites expressed total number of acid sites per unit area, but modi®es
per gram of solid. For instance, the number of acid their distribution.
sites on 25TPA/Z is 2.6 times greater than that in pure In order to elucidate the contribution of ZrO2 sup-
ZrO2. In Table 2, the amount of acid sites per KU in port on the acidic properties of TPA/Z catalysts,
10±20 TPA/Z calcined at 673 K is approximately ®ve. 16.6 wt% TPA was impregnated on a freshly prepared
Accordingly, unsupported TPA absorbed ®ve ammo- SiO2. The number and the density of acid sites present
nia molecules per KU, after heating and evacuating at on 16.6TPA/S as well as those in pure SiO2 are
473 K [32]. Since only three ammonia molecules are reported in Table 4. Unlike in the case of the TPA/
needed to neutralize the TPA polyanion, the excess Z catalysts, the addition of TPA to SiO2 neither
ammonia may correspond to acid sites formed on increases the total number of acid sites nor their
zirconia because of TPA interaction. However, when density. These results strongly suggest that the inter-
TPA/Z catalysts were calcined at 773 K a dramatic action of TPA with ZrO2 plays a crucial role in the
increase in the number of acid sites per KU as well as development of acid sites.

Table 3
Total amount and density of acid sites in TPA/Z catalysts calcined at 773 K, as a function of the TPA content

TPA/Z catalysts Acid sites per KU Total acid sitesa Surface areaa Density of acid sitesb
(wt% TPA) (mol NH3/mol KU) (mmol NH3 gÿ1) (m2 gÿ1) (mmol NH3 mÿ2)

0.0 ± 500 82 6.1


5.0 14.8 757 91 8.3
10.0 15.4 1029 110 9.4
12.5 15.0 1150 141 8.2
16.6 12.5 1218 123 9.9
20.0 6.1 924 134 6.9
20.0a ± 369 82 4.5
25.0 10.9 1435 n.a. ±
a
Measured after calcining at indicated temperatures for 4 h. n.a.: not analyzed.
b
Total acid sites divided by surface area.
E. LoÂpez-Salinas et al. / Applied Catalysis A: General 175 (1998) 43±53 49

Table 4
Total amount and density of acid sites in SiO2 and 16.6TPA/S catalysts as a function of the calcination temperature

TPA/S catalysts Total acid sitesa (mmol NH3 gÿ1) Surface areaa (m2 gÿ1) Density of acid sitesb (mmol NH3 mÿ2)

673 K 773 K 673 K 773 K 673 K 773 K

SiO2 79 109 383 288 0.2 0.4


16.6TPA/S 80 68 469 345 0.2 0.2
a
Measured after calcining at indicated temperatures for 4 h.
b
Total acid sites divided by surface area.

3.4. Textural properties pure SiO2, calcined at 673 and 773 K, respectively. It
has been reported that the addition of sulfate anions to
Table 5 shows the effect of the addition of the TPA freshly precipitated ZrO2 results in solids with much
on the speci®c surface area, pore volume and mean higher surface areas when calcined at 773±873 K in
pore diameter of TPA/Z catalysts. Pure ZrO2 calcined comparison with unsulfated ZrO2, thus showing a
at 673 K shows a surface area of 132 m2 gÿ1. The stabilizing effect on the structure of the calcined
addition of the TPA results in a considerable stabiliza- zirconia [34].
tion of the surface area of ZrO2. For instance, 20TPA/ Pore size distributions of ZrO2 and those of
Z calcined at 673 K shows a 37% increase in surface 5±20TPA/Z calcined at 673 K are shown in Fig. 4.
area in comparison with that of ZrO2. A similar Here, ZrO2 shows an almost unimodal distribution
stabilization effect can be observed in a series of of 75 AÊ mean pore diameter. Upon TPA additions to
TPA/Z calcined at 773 K, in spite of the fact that in ZrO2, the resulting solids become bimodal with a
comparison with those calcined at 673 K, a 40±60% clear increase in the population of 30±40 A Ê pores.
reduction in speci®c surface area occurred. A similar These results can be explained considering that
stabilization effect takes place on a freshly precipi- one KU has about 10 A Ê diameter, and its location
tated SiO2, dried at 393 K and later impregnated with in the bigger pores (ca. 75 A Ê ) could yield smaller
the TPA solution. For instance in Table 5, 16.6TPA/S pores, e.g. those in the lower mesoporous region
shows a surface area 22% and 20% higher than that of (30±40 A Ê ).

Table 5
Specific surface area, pore volume and mean pore diameter of TPA/Z and TPA/S catalysts as a function of TPA content and calcination
temperature

TPA Calc. Support Surface area Pore volume Mean pore


(wt%) T (K) (m2 gÿ1) (cm3 gÿ1) Ê)
diameter (A

0.0 673 ZrO2 132 0.289 69


5.0 673 ZrO2 149 0.277 63
10.0 673 ZrO2 159 0.287 61
16.6 673 ZrO2 183 0.279 55
20.0 673 ZrO2 181 0.277 57
0.0 773 ZrO2 82 0.203 73
5.0 773 ZrO2 91 0.246 87
10.0 773 ZrO2 110 0.254 77
16.6 773 ZrO2 123 0.249 69
20.0 773 ZrO2 134 0.249 65
0.0 673 SiO2 383 0.139 23
16.6 673 SiO2 469 0.181 25
0.0 773 SiO2 216 0.104 25
16.6 773 SiO2 345 0.167 25
50 E. LoÂpez-Salinas et al. / Applied Catalysis A: General 175 (1998) 43±53

arrangements depending on the catalyst type and


reaction conditions. Rearrangements of the carbon
skeleton (Wagner±Meerwein arrangements), e.g. to
obtain isobutylene from 1-butene, require very strong
acid sites.
Table 6 shows the results of 1-butene isomerization
using 0±25 TPA/Z and 16.6TPA/S (as a comparison).
Pure ZrO2 (calcined at 673 K) shows 50% conversion
and a cis/trans 2-butene ratio of 1.43. No isobutylene
formed on ZrO2, indicating the absence of strong acid
sites. It has been reported that a sulfated zirconia
shows more than twice the conversion of 1-butene
to 2-butene at 394 K compared with unsulfated zirco-
nia [35]. The overall conversion increased to 71% on
5TPA/Z and brought about the formation of isobuty-
lene (6.8% selectivity at 5 min). Increasing the amount
of the TPA on ZrO2 from 10 to 20 wt% increases the
isobutylene selectivity from 9.5 to 17.1%, respec-
tively. Composition of n-butenes at 673 K after
Fig. 4. Pore size distribution of: (a) ZrO2, (b) 5TPA/Z, (c) 10TPA/
5 min using 20TPA/Z catalyst is (1-Cˆ ˆ ˆ
4 : t-C4 : c-C4 )
Z and (d) 20TPA/Z. 1:1.3:1, while that at equilibrium is 1:1.8:1 [36].
The increase in isobutylene selectivity as a function
of the TPA content is associated with relatively more
3.5. Isomerization of 1-butene strong acid sites, as discussed in Section 3.3, which
are responsible for the double bond isomerization. In
Isomerization of n-butenes (1-, cis- and trans-but-2- spite of its higher acid site density and acid site
ene) can occur by both double bond shift and skeletal number per KU, the 25TPA/Z catalyst did not show

Table 6
Conversion of 1-butene and selectivity towards products using TPA/Z and TPA/S catalysts
a
TPA Calc. Support 1-Cˆ
4 conversion Selectivity (mol%)a
(wt%) T (K) (mol%)

3 iC4 nC4 iCˆ
4 t2Cˆ
4 c2Cˆ
4 Othersc

0.0 673 ZrO2 50 0.4 0.0 0.0 0.0 42.1 57.5 0.1
5.0 673 ZrO2 71 0.5 0.3 0.4 6.8 49.1 42.9 0.0
10.0 673 ZrO2 72 1.5 0.4 2.5 9.5 46.6 39.5 0.2
16.6 673 ZrO2 75 3.4 1.3 1.3 16.2 41.6 34.1 2.2
16.6 673 ZrO2b 71 1.1 0.2 0.6 13.0 45.0 40.1 0.0
20.0 673 ZrO2 75 2.7 1.3 1.3 17.1 42.2 35.4 2.8
25.0 673 ZrO2 76 3.6 1.6 1.6 17.1 39.1 33.6 3.4
10.0 773 ZrO2 72 0.0 0.0 0.3 6.9 50.0 42.4 0.1
16.6 773 ZrO2 73 2.0 0.3 0.7 13.0 45.2 38.0 0.9
20.0 773 ZrO2 74 2.1 0.5 1.2 13.2 44.9 37.9 0.2
16.6 673 SiO2 56 0.0 0.0 0.0 0.0 50.1 48.2 1.7
20.0 673 SiO2b 32 0.0 0.0 0.0 0.0 50.7 49.3 0.0
a
Taken after 5 min of reaction.
b
TPA was impregnated on a previously calcined ZrO2 (SiO2) at 773 K for 4 h.
c ÿ1
Mainly pentanes. Reaction conditions: 673 K, 1-Cˆ ˆ ˆ
4 =H2 (molar)ˆ1, 101.3 KPa, WHSVˆ2 h , 0.1 g-cat. 1-C4 : 1-butene, C3 : propene, iC4:
isobutane, nC4: n-butane, iCˆ 4 : isobutylene, t2C ˆ
4 : trans-2-butene, c2Cˆ
4 : cis-2-butene.
E. LoÂpez-Salinas et al. / Applied Catalysis A: General 175 (1998) 43±53 51

any further increase in isobutylene selectivity beyond ison, a sulfated alumina catalyzes the skeletal
that of 20TPA/Z. The reason may be that a great isomerization of 1-butene at 698 K yielding about
portion of the total acid sites developed in 25TPA/Z 27% isobutene after 2 h [41].
is weak (see Fig. 3(f)) and does not contribute to the The product distribution as a function of time-on-
formation of isobutylene. A similar ceiling value has stream for 20TPA/Z is shown in Fig. 5. Though over-
been reported for 25 wt% H3[W12PO40] and 25 wt% all conversion does not diminish markedly, isobuty-
H3[W12SiO40] supported on SiO2 in the alkylation of lene selectivity falls sharply with time, probably
benzene with ethylene at 473 K [28]. Recently, San- associated to selective poisoning of strong acid sites.
tiesteban et al. [37] reported that in highly acidic WOx/ In addition to butenes, 2.7% of propene, 1.3% of i-
ZrO2 catalysts only a small portion (0.005 meq per butane and 1.3% of n-butane, were detected at 5 min,
gram of catalyst) of the total strong acid sites were but they almost disappeared after 1 h of reaction.
responsible for the n-pentane isomerization activity. These results suggest that two types of catalytic sites
Calcination at 773 K of TPA/Z decreased the selec- may initially operate on TPA/Z catalysts. Site (A) is
tivity towards isobutylene by 27%, 20%, and 23% for responsible for the skeletal rearrangement of 1-butene
10.0, 16.6 and 20.0TPA/Z, respectively, relative to and for the cracking reactions. This site has strong acid
calcination at 673 K. This decline is probably asso- properties and deactivates rapidly due to coke deposits
ciated with the loss of the stronger acid sites as a making both isobutylene and cracking by-products
consequence of the partial crystalline structure mod- decrease with time-on-stream. On the other hand,
i®cation from the metastable tetragonal to the mono- double-bond isomerization does not demand a strong
clinic phase from 673 to 773 K [38]. The acid site and proceeds on site (B). Thus when coke
decomposition of the Keggin structure at 773 K deactivates (A) sites, (B) sites continue to participate
[31], can also decrease the catalytic activity. Although in the double-bond shift isomerization in agreement
the fragments of the KU yield a greater acid site with the increase of 2-butene formation as isobutylene
density (see Table 3) on TPA/Z calcined at 773 K, and other by-products decrease. Accordingly, propy-
they are weaker than those in TPA/Z calcined at lene and C‡ 5 by-products are produced via dimeriza-
673 K. tion and cracking reactions …2Cˆ 4 ! octenes !
Heteropolyacids and their salts are commonly sup- isomerized octenes ! cracking products† [42]. On
ported on acidic or neutral carriers, e.g. SiO2, activated the other hand, recent studies strongly point out that
carbon, or TiO2, since basic materials, e.g. MgO, isobutylene is formed via a monomolecular mechan-
Al2O3, decompose them, damaging the catalytic activ- ism in which the precise intermediate species is not yet
ity [28,39]. In Table 4, 16.6TPA/S, unlike the TPA/Z well stabilized [43].
catalysts, shows no indication of isobutylene forma-
tion and overall activity is considerably lower than that
of 5±25TPA/Z catalysts. 4. Conclusions
To generate superacid properties on SO4/ZrO2,
sulfation has to be carried out on freshly precipitated The impregnation of H3[W12PO40] on ZrO2 yields
Zr(OH)4 and not on calcined ZrO2 [10]. Accordingly, strong solid acids (H0ÿ9.3, calcined at 673 K). The
surface OH groups are necessary to exchange with sites are predominantly Lewis acids when TPA was
[SO4]= and ®x them to ZrO2 to generate superacid supported on hydrated ZrO2 and then calcined at
centers [40]. To examine whether a similar effect 673 K, but both Brùnsted and Lewis acid sites are
occurs on the preparation of TPA/Z catalysts, observed when TPA was supported on a previously
16.6 wt% TPA was impregnated on a freshly precipi- calcined ZrO2. The addition of the TPA stabilizes the
tated zirconia and on one previously calcined at surface area of the resulting solid acid in comparison
673 K. TPA/ZrO2 is less active and less selective with that of pure ZrO2. For instance, 16.6TPA/Z shows
for isobutylene than TPA/Zr(OH)4. Likewise, 1.4 times more surface area than that of ZrO2. 5-
20 wt% TPA/calcined SiO2 at 673 K is less active 25TPA/Z exhibit high catalytic activity in the 1-butene
than 16.6 wt%/uncalcined SiO2, the overall conver- isomerization at 673 K. Particularly, skeletal isomer-
sions being 32% and 56%, respectively. As a compar- ization, e.g. isobutylene formation, increases with
52 E. LoÂpez-Salinas et al. / Applied Catalysis A: General 175 (1998) 43±53

Fig. 5. Product distribution as a function of time-on-stream in the isomerization of 1-butene using 20TPA/Z at 673 K. (*) trans-2-butene,
(*) cis-2-butene, () isobutylene.

TPA content until leveling off at 20 wt% TPA. TPA [5] M. Hino, K. Arata, J. Chem. Soc., Chem. Commun. (1979)
1148.
supported on SiO2 shows no isobutylene formation.
[6] M. Hino, K. Arata, J. Am. Chem. Soc. 101 (1979) 6439.
[7] M. Hino, K. Arata, J. Chem. Soc., Chem. Commun. (1980)
851.
Acknowledgements [8] K. Arata, M. Hino, React. Kinet. Catal. Lett. 25 (1984) 143.
[9] Holm, Bailey, US Patent 3 032 599 (1962).
This study was funded in part by a FIES 95-02-III [10] T. Jin, T. Yamaguchi, K. Tanabe, J. Phys. Chem. 90 (1986)
4794.
project. [11] K. Tanabe, H. Hattori, T. Yamaguchi, Crit. Rev. Surf. Chem. 1
(1990) 1.
[12] C.J. Norman, P.A. Goulding, P.J. Moles, Stud. Surf. Sci. Cat.
References 90 (1994) 269.
[13] F. Babou, B. Bigot, G. Coudurier, P. Sautet, J.C. VeÂdrine,
[1] Y. Amenomiya, Appl. Catal. 30 (1987) 57. Stud. Surf. Sci. Cat. 90 (1994) 519.
[2] T. Iizuka, M. Kojima, K. Tanabe, J. Chem. Soc., Chem. [14] H. Armendariz, C. Sanchez-Sierra, F. Figueras, B. Coq, C.
Commun. (1983) 638. Mirodatos, F. Lefebvre, D. Tichit, J. Catal. 171 (1997) 85.
[3] L.A. Bruce, G.J. Hope, J.F. Mathews, Appl. Catal. 8 (1983) [15] K. Arata, Adv. Catal. 37 (1990) 165.
349. [16] C.-Y. Hsu, C.R. Heimbuch, C.T. Armes, B.C. Gates, J. Chem.
[4] M. Hino, K. Arata, Chem. Lett. (1979) 1259. Soc., Chem. Commun. (1992) 1645.
E. LoÂpez-Salinas et al. / Applied Catalysis A: General 175 (1998) 43±53 53

[17] A. Jatia, C. Chang, J.D. MacLeod, T. Okubo, M.E. Davis, Cat. [30] E.P. Parry, J. Catal. 2 (1963) 374.
Lett. 25 (1994) 21. [31] J.B. Moffat, Appl. Catal. A 146 (1996) 65.
[18] H. Matsuhashi, K. Kato, K. Arata, Stud. Surf. Sci. Cat. 90 [32] J.B. Moffat, in: B. Imelik et al. (Eds.), Catalysis by Acids and
(1994) 251. Bases, Elsevier, Amsterdam, 1985, p. 157.
[19] K. Tanabe, M. Misono, Y. Ono, H. Hattori, in: New Solid [33] V.M. Mastikhin, S.M. Kulikov, A.V. Nosv, I.V. Kozhevnikov,
Acids and Bases, Their Catalytic Properties, Stud. Surf. Sci. I.L. Mudrakovsky, M.N. Timofeva, J. Mol. Catal. 60 (1990)
Catal. 51 (1989) 169. 65.
[20] K. Tanabe, M. Misono, Y. Ono, H. Hattori, in: New Solid [34] P. Nascimento, C. Akratopoulou, O. Oszagyan, G. Coudurier,
Acids and Bases, Their Catalytic Properties, Stud. Surf. Sci. C. Travers, J.F. Joly, J.C. Vedrine, in: L. Guczi et al. (Eds.),
Catal. 51 (1989) 166. New Frontiers in Catalysis, Proceedings of the Tenth
[21] Y. Izumi, K. Matsuo, K. Urabe, J. Mol. Catal. 18 (1983) 299. International Congress on Catalysis, Budapest, Hungary,
[22] S. Igarashi, T. Matsuda, Y. Ogino, J. Jpn. Petrol. Inst. 22 1993, p. 1185.
(1979) 331. [35] G.P. Khare, I. Ahmed, US Patent 5 182 247 (1994).
[23] S. Igarashi, T. Matsuda, Y. Ogino, J. Jpn. Petrol. Inst. 23 [36] J.E. Kilpatrick, E.J. Prosen, K.S. Pitzer, F.D. Rossini, J. Res.
(1980) 30. Natl. Bur. Standards 36 (1946) 559.
[24] A. Clearfield, G.P.D. Serrette, A.H. Khazi-Syed, Catal. Today [37] J.G. Santiesteban, J.C. Vartuli, S. Han, R.D. Bastian, C.D.
20 (1994) 295. Chang, J. Catal. 168 (1997) 431.
[25] R. Srinivasan, M.B. Harris, S.F. Simpson, R.J. De Angelis, [38] G.M. Brown, M.R. Neo-Spiret, W.R. Busing, H.A. Levy, Acta
B.H. Davis, J. Mater. Res. 3 (1988) 787. Cryst. B 33 (1977) 1038.
[26] R. Srinivasan, B.H. Davis, Catal. Lett. 14 (1992) 165. [39] Y. Izumi, R. Hasabe, K. Urabe, J. Catal. 84 (1983) 402.
[27] K. Shibata, T. Kiyoura, J. Kitagawa, T. Sumiyoshi, K. Tanabe, [40] J. Livage, K. Doi, C. Mazieres, J. Am. Ceram. Soc. 51 (1968)
Bull. Chem. Soc. Jpn. 46 (1973) 2985. 349.
[28] Y. Nakano, T. Iizuka, H. Hattori, K. Tanabe, J. Catal. 57 [41] E.J. Kuhlmann, J.R. Pascoe, J.E. Browne, K.J. Martin, US
(1978) 1. Patent 5 182 247 (1993).
[29] Y. Saito, P.N. Cook, H. Niiyama, E. Echigoya, J. Catal. 95 [42] T.K. Cheung, J.L. d'Itri, B.C. Gates, J. Catal. 151 (1995) 464.
(1985) 49. [43] J. Houzvicka, V. Ponec, Catal. Rev.-Sci. Eng. 39 (1997) 319.

You might also like