The isomerization of 1-butene on stoichiometric and nonstoichiometric boron phosphate: The dependence of the acidity on stoichiometry

Article abstract of DOI:10.1006/jcat.1998.2267

The isomerization of 1-butene to cis- and trans-2-butene and isobutene has been studied on boron phosphate (BP) catalysts with P/B ratios from 0.8 to 1.4 and at various reaction temperatures and contact times. Isobutene forms on BP with P/B ratios of 1.2 or higher on which relatively strong Broensted acid sites exist and increases with reaction temperature up to 400°C. Although deactivation of the catalyst occurs, the activity and selectivity to isobutene can be maintained by the continuous injection of small quantities of water into the feed stream. The conversion and selectivities are related to the formation of carbocations on Broensted acid sites on the high P/B compositions.

Full text of DOI:10.1006/jcat.1998.2267

JOURNAL OF CATALYSIS 180, 142–148 (1998)
ARTICLE NO. CA982267
The Isomerization of 1-Butene on Stoichiometric and Nonstoichiometric
Boron Phosphate: The Dependence of the Acidity on Stoichiometry
1
S. Gao and J. B. Moffat
Department of Chemistry, University of Waterloo, The Guelph-Waterloo Centre for Graduate Work in Chemistry, Waterloo, Ontario N2L 3G1, Canada
Received March 24, 1998; revised August 20, 1998; accepted August 24, 1998
coupling of methane has been investigated on this as well
The isomerization of 1-butene to cis- and trans-2-butene and as a number of other phosphates (30).
isobutene has been studied on boron phosphate (BP) catalysts with
P/B ratios from 0.8 to 1.4 and at various reaction temperatures and
contact times. Isobutene forms on BP with P/B ratios of 1.2 or higher
on which relatively strong Br o¨ nsted acid sites exist and increases
The present study reports on the isomerization of 1-
butene to the 2-butenes and isobutene on boron phosphate
with P/B ratios from 0.8 to 1.4 at various reaction tempera-
tures and contact times and relates the skeletal isomeriza-
tion to the presence of relatively strong Br o¨ nsted acid sites
on compositions with P/B greater than unity.
�
with reaction temperature up to 400 C. Although deactivation of
the catalyst occurs, the activity and selectivity to isobutene can be
maintained by the continuous injection of small quantities of water
into the feed stream. The conversion and selectivities are related
to the formation of carbocations on Br o¨ nsted acid sites on the high
EXPERIMENTAL
P/B compositions.
�c 1998 Academic Press
Materials
The BP catalysts were prepared from boric (Fisher cer-
tified ACS grade) and orthophosphoric (Fisher 85% ) acids
with preparative P/B ratios of 0.8, 1.0, 1.2, 1.3, 1.4, and 1.6
INTRODUCTION
This isomerization of 1-butene is of both fundamental
and practical interest (1–12). Isobutene, formed from the
skeletal isomerization of 1-butene, is employed as the pre-
cursor of methyl tertiary butyl ether and its formation is
characteristic of the presence of relatively strong acid sites
on heterogeneouscatalysts. Recently(1) bimolecular mech-
anisms have ben proposed as operative in the skeletal iso-
merization of 1-butene and most recently additional exper-
iments have provided evidence to support the contention
(
denoted as BPX with X equal to the ratio). Appropriate
�
amounts of the acids were stirred continuously at 70 C for
6
�
h followed by drying in vacuum at 150 C for 24 h. The
dried solids were ground and the 100–200 mesh portions
were employed for the subsequent experiments.
Measurements
The isomerization of 1-butene was affected in a flow re-
that the by-products are formed bimolecularly (8). The 1- action system equipped with a quartz reactor containing
butene isomerization process is also of importance in the 200 mg of sample, except as otherwise indicated. The cata-
evaluation of the acidic properties of catalysts and the de- lysts were pretreated in situ in a flow (20 ml/min) of helium
pendence of these on preparative, pretreatment and reac- at Tpt for 2 h after which the temperature was increased to
tion conditions (1).
the reaction temperature (Tr) prior to exposure to the feed-
Phosphates have been studied for their catalytic prop- stream. The reaction system was connected to a gas chro-
erties for many years (13, 14). The properties of boron matograph (HP 5890) equipped with a TCD and a 2 m �
phosphate (abbreviated as BP), of particular interest in 1/8 in. OD Carbopack column. No conversion of 1-butene
this laboratory, have been assessed from studies of the con- wasobserved in the absence ofcatalyst at anyofthe reaction
versions of alcohols (15–21), deuterium exchange (22–24), temperatures employed in the present work. Conversions
temperature-programmed desorption (25) and infrared and selectivities are calculated on a carbon basis:
spectroscopy (26–28). More recently, the conversion of
Products � 100
2
-methylbutanal to isoprene has been studied on stoichio-
Conversion (% ) =
metric and nonstoichiometric BP (29) and the oxidative
Products + Residual butene
Product �
Selectivity to �(% ) =
� 100.
1
To whom correspondence should be addressed.
Total products
142
0021-9517/98 $25.00
Copyright �c 1998 by Academic Press
All rights of reproduction in any form reserved.

ISOMERIZATION OF 1-BUTENE
143
Temperature-programmed desorption experiments were
TABLE 1
performed in the reaction system. Effluents were either led
directly to the TCD detector or were passed to a GCMS
Equilibrium Composition Percentages for 1-Butene
and Its Isomers
system (HP5890, HPMS5970).
1
�
H MAS NMR spectra were obtained on a Bruker AMX- T( C)
1-Butene
Cis-2-butene
Trans-2-butene
Isobutene
5
00 with an external reference of benzene at room temper-
1
2
00
00
3
6
10
12
11
14
16
17
26
28
28
28
60
52
46
43
ature. The spinning rate ranged between 6.5 and 8 kHz,
depending upon the sample.
300
50
3
RESULTS
peratures between 300 and 1400 K (Table 1). On increase of
the P/B ratio to 1.0 the conversion increases but, as with the
0.8 composition, no isobutene is observed until a temper-
ature of 300 C is reached. The cis/trans ratio for 2-butene
remains greater than unity.
The effects of changes in the P/B ratio and of the reac-
tion temperature, the latter for values of 100, 200, 300, and
�
3
50 C are summarized in Fig. 1. With the BP 0.8 sample the
�
conversion of butene is less than 5% at all temperatures
�
while up to 350 C no isobutene is observed. The cis/trans
The observations with the compositions for which P/B is
1.2 or higher are significantly different from those where
P/B is lower. With the former the conversions at the low-
ratio of 2-butene for this composition is greater than unity
�
�
up to 300 C but slightly less at 350 C. In comparison, at
equilibrium the cis/trans ratio is less than unity at all tem-
�
est temperature (100 C) are significantly larger, reaching as
high as 40% , the cis/trans ratios are approaching unity while
selectivities to isobutene remain vanishingly small. With in-
�
crease in temperature to 200 C and higher the conversions
increase markedly to values of 80–90% , the cis/trans ratio
decreases to 0.6–0.7 and both are relatively independent
of both temperature and composition. Concomitantly the
�
skeletal isomerization product becomes significant at 200 C
�
and reaches 15–25% at 350 C, showing dependence on the
catalyst composition at the two highest temperatures.
The temperature at which the catalysts with P/B greater
than 1 are pretreated has little or no effect on the con-
�
version and selectivities until 450 C is reached at which
pretreatment temperature the conversion and selectivity
to isobutene both decrease for a reaction temperature of
�
4
00 C (not shown).
In addition to the isomers of 1-butene, small quantities
of propylene, n-butane, isobutane, pentenes, and octenes
were detected in the effluent (Table 2).
With the catalysts of P/B greater than 1 the selectivity to
isobutene increases with residence time to ultimately reach
a plateau (Fig. 2), indicative of equilibrium with respective
to both double bond and skeletal isomerization. Concomi-
tantly, as expected, the selectivities to t2B and c2B decrease
with the former remaining slightly larger than the latter.
TABLE 2
a
b
Selectivity to Minor Products
�
�
�
Product
350 C
400 C
450 C
C3
0.9
0.4
0.5
1.4
0.5
0.8
2.1
0.5
0.4
Isobutane
>
C4
FIG. 1. Effect of the P/B (molar ratio) and reaction temperature on
the conversion of butene and selectivities to the products. W = 50 mg,
F = 13.7 ml/min.
a
Percentage.
P/B = 1.6, W = 200 mg, F = 13.7 ml/min.
b

1
44
GAO AND MOFFAT
TABLE 3
Ammonia Desorbed from B-P-O Catalyst a
Catalyst
�mol/g
BP08
BP10
BP12
BP14
BP16
38
33
1293
2925
4452
a
Samples were pretreated in helium flow at
�
4
00 C for 2h and 3 � 10 ml of NH3 was injected
�
upstream at 100 C.
water, while the higher temperature peak which is seen in
�
the 400–600 C range is due to the associative desorption
of water formed from protons and hydroxyl groups on the
surface of the solid (25). The latter peak can be seen to
increase in intensity as the P/B ratio increases, indicating
an increase in the number of protons with P/B ratio. Prior
FIG. 2. Conversion of 1-butene and selectivities to products at various
�
residence times. Reaction temperature, 400 C, W = 200 mg, F = 13.7 ml/
min, P/B = 1.6.
�
exposure of the samples to NH3 at 100 C followed by TPD
With increasing time-on-stream changes in the product
composition are observed. As illustrated with PB 1.4 in
Fig. 3 both the conversion and the selectivity to isobutene
decrease while that to t2B and c2B increase. However, with
the continuous addition of small partial pressures of wa-
ter vapour to the reaction feed stream the aforementioned
changes are considerably reduced.
Little or no visual evidence for the formation of carbon
was found under the conditions and times on-stream in the
present work.
produces peaks resulting from the desorbed NH3 which be-
gin to emerge at 300 C and whose magnitude increases with
�
P/B ratio (Fig. 5). It is of interest to note that with BP 1.6,
for example, sizeable quantities of NH3 continue to des-
�
orb at temperatures as high as 600 C. The quantities of
NH3 desorbed from the various samples are summarized in
Table 3.
1H MAS NMR spectra of 0.8, 1.0, and 1.4 P/B composi-
tions show a number of peaks, some of which are of rela-
tively insignificant intensities (Fig. 6). With increasing P/B
both the chemical shifts and the relative intensities of the
two principal peaks increase. Exposure of the 1.4 P/B sam-
ple to water vapour producesrelativelylittle change in these
peaks. However, after exposure to 1-butene the relative in-
tensities of the peaks from the 1.4 P/B sample decrease al-
though the chemical shifts are relatively unchanged. After
the latter sample is exposed to water vapour the intensity
of the principal peak has approximately regained its value
prior to exposure to 1-butene.
Temperature-programmed desorption (TPD) of the BP
samples show the presence of at least two overlapping
peaks, both due to the desorption of water (Fig. 4). The
�
lower temperature peak at 150–300 C results from the des-
orption of water which exists on the catalysts as molecular
DISCUSSION
BP with P/B ratios less than or equal to 0.8 are evi-
dently not effective in either the conversion of 1-butene
or its isomerization to isobutene while those with P/B
greater than one produce significant quantities of iso-
butene. Since infrared spectroscopic studies of pyridine,
2
,6-dimethylpyridine, 2,5-dimethylpyridine, and 2,6-ditert-
butylpyridine on stoichiometric and nonstoichiometric
boron phosphate have shown that BP with P/B ratios
greater than unity possesses predominantly Br o¨ nsted acid
sites while those with P/B less than one are largely Lewis
in nature (26), it appears that the isomerization process
on BP catalysts is largely occurring on Br o¨ nsted acid sites,
FIG. 3. Changes in conversion and selectivities for various times-on-
stream and the effect of the introduction of small quantities of water to
the feedstream. Reaction conditions as in Fig. 2 except for the addition of
water.

ISOMERIZATION OF 1-BUTENE
145
�
FIG. 4. Temperature-programmed desorption patterns of the BP samples after pretreatment at 110 C for 2 h in a flow of helium.
consistent with the consensus in the literature (1), although
Lewis acidic and Lewis acidic–Lewis basic sites have also
been mentioned (12). Earlier work (13, 14) has provided
evidence for the increase in the number of strong Br o¨ nsted
acidic sites as the P/B ratio increases above 1. The results
of the TPD studies with NH3 reported in the present work,
1
together with the H MAS NMR chemical shifts provide
additional support for this contention. The observation of
a vanishingly small selectivity to isobutene with the P/B
ratio of 1.0 which, however, increases as the P/B ratio
increases, at least up to 1.4 while the selectivities to the
2
-butenes decrease, is consistent with the requirement of
strong Br o¨ nsted acid sites for the generation of isobutene
in contrast with a relatively facile production of the
2
-butenes.
The variance of values of the conversion with P/B ra-
�
tio (where P/B > 1) and reaction temperature (200–350 C)
suggests that equilibrium with respect to the double-bond
shift process has been achieved at these temperatures and
with these catalyst compositions, but not with respect to
skeletal isomerization. The observations that the selectiv-
ities to the 2-butenes decrease while that to isobutene in-
creases with increasing temperature and increasing contact
time suggest that the 2-butenes are precursors to isobutene.
The aforementioned may be rationalized by postulating
that a secondary carbocation is formed from 1-butene, cis-,
and trans-2-butene in the presence of weak acidic sites
and an equilibrium established between these four species
(Fig. 7). Given the appropriate conditions of sufficiently
FIG. 5. Temperature-programmed desorption of NH3 from BP sam-
�
ples. Samples were pretreated in situ in a helium flow at 400 C for 2 h
�
followed by exposure to 30 ml (STP) of NH3 at 100 C. TPD conditions:
�
carrier gas, helium; heating rate 12 C/min. The signals from BP 0.8 and
BP 1.0 were enlarged by a factor of 2.

1
46
GAO AND MOFFAT
1
�
FIG. 6. H MAS NMR of BP samples with P/B equal to 0.8, 1.0, and 1.4 pretreated at 400 C for 2 h in indicated environment.
high temperature and strength of acidic sites the secondary for the conversion of a secondary to a primary carbocation
carbocation may be converted to the primary analogue involves a 1,2-methyl shift to generate a methylcyclopropyl
which forms the isobutene by loss of a proton. Thus, as intermediate.
the isobutene is formed the 2-butenes will be depleted. It
Although polemics concerning the bimolecular mech-
should be recalled that the mechanism usually proposed anism for the skeletal isomerization of 1-butene have
FIG. 7. Carbocation mechanism for isomerization of 1-butene.

ISOMERIZATION OF 1-BUTENE
147
consistent with the aforementioned rationalization of the
regenerative effect of introducing water vapour.
Lastly, a few remarks concerning the structure of stoi-
chiometric and nonstoichiometric BP may be of value. BP
has been shown to be isostructural with one or more forms
of SiO2 with both the phosphorus and boron surrounded by
a tetrahedron of oxygen atoms (35). Although the P-O and
B-O bonds in the network of tetrahedra may be considered
as �-bonds as in SiO2 the P-O and B-O internuclear sep-
arations are 1.55 and 1.44 A˚ , respectively, and the former
corresponds to approximately 1/3 �-bond per �-bond. The
network structure of tetrahedra may thus be represented
as one in which the boron atom and the PO4 group are pos-
itively and negatively charged, respectively, superimposed
on the aforementioned �-bonded structure (Fig. 9). Such a
representation may explain, at least in part, the increase in
Br o¨ nsted acidity as the P/B ratio increases.
FIG. 8. Dehydration–hydration of BP.
recently appeared (31–33) the evidence in the present work
is consistent with a predominantly monomolecular mecha-
nism as advocated by Ponec (1), although the formation of
C3 hydrocarbons, albeit in small quantities, suggests that a
bimolecular process may be operative in the production of
byproducts.
Earlier temperature programmed desorption studies on
BP of various P/B ratios have shown that four peaks, each
due to the desorption of water and having activation ener-
gies of 6, 14, 29, and 56 kcal/mol are observed at temper-
�
atures of approximately 100, 125, 200, and 250 C, respec-
tively (25). The first two peaks have been attributed to the
desorption of molecular water, probably hydrogen-bonded
to the surface while the latter two result from the associative
ACKNOWLEDGMENTS
The financial support of the National Sciences and Engineering Re-
desorption of water which existed as hydroxyl groups on the search Council of Canada through Operating, Equipment, and Industrially
surface. The number of surface hydrogen atoms has been Oriented Research Grants, the Province of Ontario through its Univer-
1
8
2
sity Research Incentive fund and Imperial Oil through the Environmental
Science and Technology Alliance Canada is gratefully ackowledged.
estimated as 8 � 10 atoms/m from the exchange of deu-
terium from the gas phase and similar numbers were found
for the number of empty sites (22). Thus the decrease in
activity of the catalysts with P/B greater than 1 after pre-
REFERENCES
�
treatment at 450 C can be ascribed to lossofsurface protons
in the form of desorbed water.
1
2
. Houzvicka, J., and Ponec, V., Catal. Rev.-Sci. Eng. 39, 319 (1997).
. Seo, G., Jeong, H. S., Hong, S. S., and Uh, Y. S., Catal. Lett. 36, 249
(1996).
The regeneration of the catalyst during the isomeriza-
tion process by the concurrent onstream addition of small
partial pressures of water vapour can be attributed to the
dissociative adsorption of water to regenerate the surface
hydroxyl groups (Fig. 8). Infrared spectra have shown that
water adsorbed on dehydrated BP is partially dissociated
to form BOH and POH groups (28).
3. Guisnet, G., Andy, P., Snep, N. S., Benazzi, E., and Travers, C., J. Catal.
159, 551 (1996).
4. Asensi, M. A., Corma, A., and Martinez, A., J. Catal. 159, 561 (1995).
5. Boronat, M., Viruela, P., and Corma, A., J. Phys. Chem. 100, 633(1996).
6. Meriaudeau, P., Bacaud, R., Hung, L. N., and Vu, A. T., J. Mol. Catal.
A-Chem. 177, 110 (1996).
7. Woo, H. C., Lee, K. H., and Lee, J. S., App. Catal. A-General 147, 134
(1996).
1
The H MAS NMR spectra contain two peaks for the
8. Pellet, R. J., Casey, D. G., Huang, H. M., Kessler, R. V., Kuhlman,
E. J., Dyoung, C. L., Sawicki, R. A., and Ugolini, J. R., J. Catal. 423,
samples with P/B ratios of 0.8, 1.0, and 1.4. As the P/B ratio
increases at least one of the peaks intensifies and the chem-
ical shift increases consistent with the expected increase in
the number of protons as well as the acidic strength (34).
On exposure of the BP (1.4) sample to water vapour the
159 (1995).
9. Houzvicka, J., Diefenbach, O., and Ponec, V., J. Catal. 164, 288 (1996).
0. Gao, S., and Moffat, J. B., Catal. Lett. 42, 105 (1996), and references
therein.
1
1
1
1
1. Cheng, Z. X., and Ponec, V., Appl. Catal. 118, 127 (1994).
2. Cheng, Z.-X., and Ponec, V., J. Catal. 148, 607 (1994).
3. Moffat, J. B., Catal. Rev.-Sci. Eng. 18, 199 (1978).
1
intensity of one of the H peaks increases, indicative of the
sorption of water. Exposure of the same sample to 1-butene
reduces the intensity of the principal peak while the in- 14. Moffat, J. B., “Topics in Phosphorus Chemistry” (M. Grayson and
E. J. Griffith, Eds.), Vol. 10, p. 285. Wiley, Interscience, New York,
troduction of water vapour largely restores the intensity,
1980.
1
1
1
5. Moffat, J. B., and Riggs, A. S., J. Catal. 28, 157 (1973).
6. Moffat, J. B., and Riggs, A. S., J. Catal. 42, 388 (1976).
7. Jewur, S. S., and Moffat, J. B., J. Chem. Soc. Chem. Commun., 801
(
1978).
1
1
2
2
2
8. Jewur, S. S., and Moffat, J. B., J. Catal. 57, 167 (1979).
9. Moffat, J. B., and Jewar, S. S., J.C.S. Faraday I 76, 746 (1980).
0. Moffat, J. B., and Jewur, S. S., J. Catal. 75, 94 (1982).
1. Grisebach, J., and Moffat, J. B., J. Catal. 80, 350 (1983).
2. Moffat, J. B., and Scott, L. G., J. Catal. 45, 310 (1976).
FIG. 9. Structure of BP.
23. Moffat, J. B., J. Mol. Spectrosc. 61, 211 (1976).

1
48
GAO AND MOFFAT
2
2
4. Moffat, J. B., and Scott, L. G., J. Chem. Soc. Faraday Trans. 75, 503 31. Houzvicka, J., and Ponec, V., Ind. Eng. Chem. Res. 36, 1424 (1997).
(
1979).
5. Moffat, J. B., Chao, E., and Nott, B., J. Colloid Interface Sci. 67, 240
1978).
32. Guisnet, M., Andy, P., Gnep, N. S., Travers, C., and Benazzi, E., Ind.
Eng. Chem. Res. 37, 300 (1998).
33. Houzvicka, J., and Ponec, V., Ind. Eng. Chem. Res. 37, 303 (1998).
34. Pfeifer, H., in “Acidity and Basicity of Solids: Theory, Assessment and
Utility” (J. Fraissard and L. Petrakis, Eds.), p. 255. Kluwer, Dordrecht,
1994.
(
26. Miyata, H., and Moffat, J. B., J. Catal. 62, 357 (1980).
27. Moffat, J. B., J. Chem. Soc. Faraday I 77, 2493 (1981).
28. Moffat, J. B., and Neeleman, J. F., J. Catal. 34, 376 (1974).
29. Moffat, J. B., and Schmidtmeyer, A., Appl. Catal. 28, 161 (1986).
30. Ohno, T., and Moffat, J. B., Catal. Lett. 9, 23 (1991).
35. Van Wazer, J. R., in “Phosphorous and its Compounds,” Interscience,
New York, 1990.