Dehydration of 1-propanol using H3PW12O40 as catalyst

Article abstract of DOI:10.1039/b005256f

The activation of H₃PW₁₂O₄₀ under different thermal conditions has been evaluated by calorimetry in acetonitrile solution using pyridine as a probe, and the results correlated to the dehydration of 1-propanol. Treatments at 200, 300, and 400 °C show substantial changes in the catalytic activity of 12-tungstophosphoric acid as a function of activation temperature and time. Analysis of the samples by XRD, FTIR, TGA and DSC was performed to determine structural changes and thermal stability of the heteropolytungstic acid. The Royal Society of Chemistry 2001.

Full text of DOI:10.1039/b005256f

View Article Online / Journal Homepage / Table of Contents for this issue
Dehydration of 1-propanol using H PW O as catalyst
3
12 40
a
a
b
José A. Dias,* Sílvia C. L. Dias* and Nicholas E. Kob
a
Laboratório de Catálise - Instituto de Química-Universidade de Brasilia, Caixa Postal 04478,
Brasília-DF, 70919-970, Brazil. E-mail: jdias@unb.br; scdias@unb.br
DuPont Experimental Station Route 141 Wilmington, DE 19880, USA.
E-mail: nicholas.e.kob@usa.dupont.com
b
Received 30th June 2000, Accepted 9th November 2000
First published as an Advance Article on the web 9th January 2001
The activation of H PW O under different thermal conditions has been evaluated by calorimetry in acetonitrile
3
12 40
solution using pyridine as a probe, and the results correlated to the dehydration of 1-propanol. Treatments at 200,
00, and 400 ЊC show substantial changes in the catalytic activity of 12-tungstophosphoric acid as a function of
3
activation temperature and time. Analysis of the samples by XRD, FTIR, TGA and DSC was performed to
determine structural changes and thermal stability of the heteropolytungstic acid.
Introduction
in correlation of solid acidity to acid catalysed reactivity for
24,25
different solid acids.
Acid catalysis involving heteropoly tungstophosphoric com-
In this study the calorimetric method has been used to
evaluate the acidity of the protons for HPW calcined at
temperatures of 200, 300 and 400 ЊC for 2 hours in air. The
results along with powder XRD, FTIR, TGA and DSC were
correlated to the activity for dehydration of 1-propanol.
1–3
pounds has attracted much interest. Heteropolyanions such
3Ϫ
as PW O
have very low negative charge on the surface
oxygens, which make their acid forms, H PW O (abbreviated
1
2
40
3
12 40
4
as HPW), strong acids. HPW is known to have a primary and
secondary structure. The primary structure comprises the P,
W and O atoms. Twelve WO octahedra surround a central
6
PO tetrahedron, which include four kinds of oxygen atoms.
Experimental
Materials
4
Four edge-shared units of W O13 are connected by corners
3
accounting for the total amount of atoms in the anion. This
arrangement is known as the Keggin structure (α type), and
geometrical isomers (β and γ type) are also known. The
Phosphotungstic acid (HPW) was obtained from the Aldrich
chemical company, acetonitrile and pyridine from Fisher.
Further purification of acetonitrile involved first drying over
molecular sieves 4 Å for 24 hours, distilling over P O , and
finally storing over molecular sieves 4 Å in a dry bag. The
solvent was used next day to avoid any water contamination.
5–9
secondary structure depends upon the degree of hydration and
ϩ
is formed by H and H O molecules. A typical structure of
2
2
5
3Ϫ
H PW O ؒ6H O is comprised of four polyanions of PW O
connected by H (H O) bridges by hydrogen bonding at the
terminal oxygens atoms of the anions.
3
12 40
2
12 40
ϩ
2
2
5–9
Pyridine was distilled over CaH . A solution of pyridine in
2
Many studies have demonstrated the usefulness of HPW as
acetonitrile was prepared with both liquids dried, and stored
in the dry bag. 1-Propanol used in the dehydration reaction was
obtained from Aldrich and used as received. All the other
reactants used were of analytical grade.
an acid catalyst, being more effective than ordinary protonic
1
0–19
acids.
However, few have been done on the effects of cal-
20
cination pretreatment on acidity and reactivity of HPW. The
primary Keggin structure is known to be stable up to 500 ЊC,
and after that it decomposes to P O and WO . Significant
2
5
3
Calcination procedures
changes are known to occur to the secondary structure at
temperatures below that of decomposition. For example, HPW
is capable of losing some lattice oxygen without collapsing its
HPW was treated under a dry air flow at 200, 300 and 400 ЊC,
during 2 hours, and 10 hours (300 ЊC), after each temperature
had been reached. Then, calcined HPW was stored immediately
in a dry box. The sample drying apparatus and technique has
11
primary Keggin structure at 450 ЊC. This is accomplished by
some slight rearrangement of the secondary structure. The
26
16
successfully been used to dry HZM-5 previously.
complete thermal decomposition of HPW has been studied.
Changes to the secondary structure below 450 ЊC will affect not
only the whole structure, but the acidity and reactivity of HPW.
Therefore, the effect of pretreatment calcination is of interest.
Acidity determinations of a heterogeneous solid acid are
most commonly done by use of Hammett indicators, and
infrared analysis of an adsorbed base. Recently, these methods
have come under criticism due to their inability accurately to
Calorimetric measurements
Samples of 0.4 g of calcined HPW were weighed inside
the dry box, and transferred to an isothermal calorimetric cell
23
containing a stirrer bar. All details of the calorimeter and the
18
calorimetric procedure were described previously.
21
determine the acid site number and strength for a solid acid.
1
-Propanol dehydration
Other common measurements involve TPD (temperature-
programmed desorption), and microcalorimetry of gaseous
The reaction was carried out using a one-time flow through
reactor over 0.5 g of the catalyst. Nitrogen was used as the
carrier gas, which was passed over a saturator containing 1-
propanol, heated to 90 ЊC. The pre and post gas samples were
analysed by gas chromatography (Hewlett Packard model 5890).
22
bases. More recently, a calorimetric method known as
23
Cal-ad, which involves measuring the heat of interaction
between a probe base and the acid sites of the solid in a non-
interacting solvent, has been used. This method was successful
2
28
J. Chem. Soc., Dalton Trans., 2001, 228–231
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b005256f

View Article Online
Thermal analysis
DSC samples were analysed in a DuPont TA instruments
Ϫ1
model 2910 DSC using a heating rate of 10 ЊC min . TGA was
obtained on the same instrument using 2950 module. Both
experiments were performed using dry air as the purge gas.
Powder XRD and FTIR
Samples were scanned at 40 kV with a Philips X-ray diffract-
Ϫ1
ometer at 20.00 s deg (Fig. 1) and a Rigaku D/MAX-2A/C
with Cu-Kα radiation at 40 kV and 30 mA (Fig. 2). A 2θ range
of 5 to 50Њ was scanned. Infrared spectra were obtained with a
Ϫ1
Bomem MB-100 FTIR spectrophotometer (4 cm resolution),
using pressed ground samples with dry KBr (pellets).
Results and discussion
Structural data: XRD and FTIR
In order to analyse possible modifications in the HPW
structure, X-ray diffraction and FTIR spectra were taken from
samples calcined under different conditions. Powder XRD
patterns indicate that the structures of samples calcined at 200
and 300 ЊC for 2 hours are identical (Fig. 1). The positions
and intensities of the main peaks (2θ = 10.3, 25.3 and 34.6Њ) are
Fig. 1 XRD patterns of HPW calcined under air for 2 hours at:
a) 400; (b) 300; (c) 200 ЊC.
(
14
similar to those expected for the Keggin structure. Also, the
sample calcined at 400 ЊC for 2 hours (Fig. 1) has almost the
same d spacing, indicating the Keggin structure, but the relative
intensities are different. The peaks are slightly broadened
compared to those of samples treated at 200 and 300 ЊC.
The sample calcined at 400 ЊC displays a heterogeneous color
on its surface, which may be indicative of partial decomposition
of HPW on the surface (see below). It is known that this
decomposition is affected not only by temperature but also by
the time the solid is exposed to heating. Therefore, XRD of
samples calcined at 300 ЊC for 4 and 10 hours and 400 ЊC for
4
hours were taken (Fig. 2). It is clear from the XRD pattern
that the sample treated at 400 ЊC for 4 hours underwent large
decomposition forming a solid with a low degree of crystal-
linity containing some WO3 phase (new broad peaks at
2
θ = 23.5, 34.6, and 41.2Њ). Also, HPW calcined at 300 ЊC for 10
hours exhibits a XRD pattern with lower intensities (Fig. 2c)
compared to the sample treated for 4 hours. It should be
mentioned that both samples (300 ЊC for 10 hours, and 400 ЊC
for 4 hours) had, at the end of calcination, a substantial portion
of the solid surface colored (i.e. a mixture of yellow, green,
and blue particles). The results demonstrate the kinetic
phenomenon of decomposition of the Keggin structure, which
will be discussed later.
Fig. 2 XRD patterns of HPW calcined under air at: (a) 400 (4 hours);
b) 300 (4 hours); (c) 300 ЊC (10 hours).
(
The primary structure of HPW was examined by FTIR.
Ϫ1
Since the region between 1200 and 700 cm is the fingerprint
for the Keggin structure, infrared spectra of samples calcined
(
200, 300, and 400 ЊC) for 2 and 4 hours (Fig. 3) were obtained.
It can be observed that the characteristic bonding absorptions
Ϫ1
of the anion, P–O (1081), W᎐O (984), W–O –W (892 cm ),
᎐
t
c
Ϫ1
and W–O –W (798 cm ), are maintained for all samples,
e
except for the sample calcined at 400 ЊC for 4 hours. These
bands are broader and of lower intensities, evidencing a
relatively large amount of decomposition for the primary
Keggin structure.
Fig. 3 FTIR spectra of HPW calcined under air at (a) 200, (b) 300 and
c) 400 ЊC for 4 hours. Sample pellets were 0.25% (w/w) with dry KBr.
(
decomposition of the primary structure was observed for short
treatments at the studied temperatures.
It has been shown in previous reports that XRD patterns of
HPW differ with calcination pretreatment, due to changes
in the unit cell caused by different hydration degrees. However,
no mention was made as to the effects of this difference on
acidity and reactivity. It is likely that a slight structural
rearrangement of HPW (changed because of differences in
degree of protonation) when calcined for longer periods at
higher temperatures (у 300 ЊC) results in solids having
decreased acidities (see below). Differences in acidity are
probably related to the secondary structure, since no evident
Thermal analysis
Thermal analysis of our samples was performed to ascertain if
differences in hydration could account for the discrepancies
in our results from those reported previously. Thermal analysis
of HPW shows three endothermic peaks in the DSC corre-
sponding to a mass loss in the TGA. The TGA of HPW shows
that mass loss occurs in three steps. The structure has been
J. Chem. Soc., Dalton Trans., 2001, 228–231
229

View Article Online
shown to be a 24 hydrate (H PW O ؒ24H O). Previous studies
Table 1 Thermal treatments and heat of reaction (25 ± 1 ЊC) of HPW
3
12 40
2
11,27
with pyridine in CH CN solution
have also shown that mass loss occurs in three steps.
The
3
complete loss of hydration to form an anhydrous sample of
HPW is a two step process. The first loss corresponds to a
hexahydrate of HPW. After the second step is complete the
structure is in the anhydrous form. Since our samples were
calcined for 2 hours, isothermal TGA was performed at each of
the various temperatures to investigate the effects of time on
the thermal properties of HPW. Isothermal TGA studies for
a HPW sample at 200 ЊC in an air flow for 2 hours demonstrate
that HPW is in the anhydrous form. This is consistent with
Total heat
evolved /cal
a
b
c
Ϫ1
T/ЊC, t /h
∆Hlim /kcal mol
2
3
3
4
00, 2
00, 2
00, 10
00, 2
5.543 ± 0.102
4.459 ± 0.107
3.832 ± 0.151
1.401 ± 0.165
18.8 ± 0.2
17.5 ± 0.2
11.5 ± 0.3
8.4 ± 0.4
a
Samples of HPW treated at different temperatures and time exposed
b
to a flow of dry air. Upon complete titration of the three protons
of HPW. Errors are the standard deviation of three experiments. The
relatively large values reflect the cumulative process of adding heats
28
other findings when drying HPW with P O . Isothermal TGA
2
5
analysis of samples at 300 and 400 ЊC gave identical thermo-
grams confirming all samples are in the anhydrous form.
Further, coupled with FTIR spectra of all samples calcined
for 2 hours, it shows that no apparent decomposition of the
bulk Keggin structure has occurred after thermal treatment,
c
after each addition of pyridine. Calculated enthalpy obtained for the
reactionofonlythefirstprotonofHPWwithpyridine(H PW ϩ Py
H PW ϩ HPy ). The errors are based on variance–covariance matrix
3
Ϫ
ϩ
2
Ϫ1
analysis, but are within the estimated error of ±0.5 kcal mol using the
18
whole calorimetric curve (see text). Conversion: 1 kcal = 4.184 kJ.
29
which had been observed previously.
Ϫ
diprotonated (H PW ). Therefore, the effect of other equilibria
can be simplified for calculation. It was discussed before that
2
Thermal decomposition of HPW
18
There are many complicating factors involving drying hetero-
poly compounds such as HPW. Generally, any polyprotic
oxoacid behaving as a Brönsted acid (i.e. has dihydroxy func-
tionality) can be converted into an anhydride (i.e. µ-oxo bridge
between metals) by thermal heating. For hydrated HPW, the
strongest acid is the anhydrous form, while the anhydride is
weaker and formed by loss of protons connecting the Keggin
units in the acid (nH PW O → PW O ϩ ³nH O). Thus,
in the three step process of deprotonation of HPW in CH CN
3
solution the first step is the most important to correlate with
catalytic activity. The difference of this method of calculation
of enthalpies for the samples in this study from the one
18
described earlier (considering all three steps) is less than 0.5
Ϫ1
kcal mol for the first deprotonation. From the total heat
evolved in the reaction with pyridine, independently of the
calculation method for the enthalpy, it is evident in Table 1 that
the acid strength of samples calcined at 200 and 300 ЊC in air
for 2 hours is superior to that of a sample calcined at 400 ЊC for
2 hours in air. These results are in contrast to another study in
3
12 40
12 38.5
2
¯
²
it is important to use the anhydrous acid in most catalytic
studies involving acidity dependence.
In this study XRD, FTIR, and color change were used pri-
marily to check decomposition of the bulk Keggin structure. Of
note is the kinetic formation of the anhydride phase according
to the results of XRD and FTIR for HPW calcined at different
times (Figs. 2 and 3). The color change must be considered
with caution. The formation of the constituent oxides of HPW
17
which the acidity was determined using Hammett indicators,
but they agree well with microcalorimetric measurements of
14–16
NH sorption.
It was claimed that the enhanced acidity
3
at 400 ЊC was due to the HPW being anhydrous which would
17
not be achieved by lower calcination temperatures. Thus,
HPW was treated at 300 ЊC for 2 and 10 hours, and then titrated
with pyridine (Table 1). It can be observed that the increased
amount of time did not enhance the acidity. Actually, as stated
before, the sample calcined for 10 hours showed traces of
decomposition through the mixture of colors on the solid
surface.
(
WO and P O ) occurs as a consequence of decomposition
3 2 5
16
from the anhydride. Since this anhydride is metastable, only a
fraction of the total calcined HPW produces WO (yellow) and
3
reduced W O
(blue), which accounts for the green color
n
3n-1
observed on some samples. Changes in the color of HPW with
11
calcination temperature have been reported before. Although
the surface of HPW is probably changed with calcination at
higher temperatures, the bulk is retained as indicated by XRD
and FTIR results. However, the surface amounts of the
anhydride phase as well as the oxides increase with time of
calcination, and are well revealed by XRD and FTIR spectra
Therefore, the differences in acidity for the HPW samples
must be due to other factors. First, the surface area of the solids
may account for a difference in measured acidity. It is known
that calcination of a solid can reduce its surface area. This is
not the case in this study since the samples (200, 300, and
400 ЊC calcined for 2 hours) have surface areas of 6.0, 5.4, and
(
Figs. 2 and 3). It can be concluded that this kinetic decom-
2
Ϫ1
position of the anhydrous HPW with consequent loss of acidity
is not always sensitive to monitoring by powerful methods
such as XRD or FTIR. Despite that, calorimetry is well suited
to detect decomposition of anhydrous HPW. Reaction with
4.9 m g respectively. In this case the surface area reduction is
not very significant. Since most of the HPW protons are in the
bulk of the solid (only about 0.008 mmol of protons per gram
19
2
Ϫ1
of HPW is located on the surface for a solid about 5 m g ),
the surface area will not bias these calorimetric results.
18
16
pyridine or ammonia indicates loss of protonic acidity
through the decreased heat evolved by the samples calcined at
higher temperatures.
As mentioned earlier, heating of HPW at temperatures below
its complete decomposition temperature of 500 ЊC might result
in some reordering of the Keggin structure, or in a mass loss of
the acidic protons. Reordering is not necessarily accompanied
by a great mass loss. Slight rearrangement of the proton posi-
tions as well as partial decomposition forming the anhydride
and/or oxide phases at higher temperatures or over longer
periods probably causes the lower enthalpies calculated. To
make sure that the difference in the enthalpies was not due to
constraints on pyridine reacting with the protons in solution,
a gas phase reaction with HPW was tested.
Acidity of HPW by calorimetric titration
Calorimetric results for determination of the acidic proton
strength of HPW calcined at different temperatures are
shown in Table 1. Titration of HPW with pyridine in a solution
of CH CN was done three times for each sample, and the
3
average heats used to calculate the enthalpies of interaction.
It should be mentioned that the enthalpy for each sample is
calculated by taking the data up to the limit of one proton
Ϫ
ϩ
equivalence (H PW ϩ Py
H PW ϩ HPy ), so that the
3
2
Dehydration of 1-propanol
second and third deprotonation were not considered as con-
comitant equilibria. For the first points of titration most
It can be argued that the difference in measured acidity is due
to the inability of the pyridine probe effectively to get to the
of the acid in solution is either totally protonated (H PW) or
3
2
30
J. Chem. Soc., Dalton Trans., 2001, 228–231

View Article Online
probably due to a slight structural rearrangement of HPW and/
or a small amount of decomposition of the anhydrous HPW
forming the anhydride phase. This results in a solid of lower
acidity. The acidity of HPW calcined at different temperatures
has been measured by solution calorimetry, and correlated
to the dehydration of 1-propanol as a model reaction. The
coupled methods can be used to monitor loss of acidity by
HPW after thermal treatments.
Acknowledgements
The authors would like to extend their deepest respect and
gratitude to Russell S. Drago (1928–1997), who directed part of
this work at the University of Florida.
References
1
2
3
4
5
6
T. Yamada, Petrotech., 1990, 13, 627.
Y. Izumi, J. Mol. Catal., 1983, 84, 402.
T. Hibi, Appl. Catal., 1986, 24, 69.
N. Mizuno and M. Misono, J. Mol. Catal., 1994, 86, 319.
G. A. Tsigdinos, Top. Curr. Chem., 1978, 76, 4.
K. I. Mateev and I. V. Kozhevnikov, Russ. Chem. Rev., 1982, 51,
1075.
Fig. 4 Conversion of 1-propanol catalysed by HPW, calcined at differ-
ent temperatures (ЊC)/times (h).
proton in the slightly changed HPW structure calcined at
4
00 ЊC. To investigate this possibility, samples of calcined
HPW were tested for activity in 1-propanol dehydration. Since
-propanol is a small molecule diffusional constraints due to
7 M. Misono, Catal. Rev. Sci. Eng., 1987, 29, 269.
8
T. Okuhara, N. Mizuno and M. Misono, Adv. Catal., 1996, 41,
13.
1
1
structural rearrangements of the Keggin ion are not a factor.
Fig. 4 shows the results for conversion of 1-propanol into pro-
pene at 170 ЊC. It is evident from these results that the activity
of the sample calcined at 400 ЊC is inferior to that of samples
calcined at 200 and 300 ЊC (2 and 10 hours). The acid strength,
as measured by calorimetry, shows the same trend as the
reaction results. Correlation of the calculated enthalpies and
9
M. T. Pope, Heteropoly and Isopoly Oxometalates, Springer-Verlag,
Berlin, 1983.
1
0 O. Shingo, K. Junichi and M. Sugimoto, US Pat., 5321196, 1994.
11 H. Hayashi and J. Moffat, J. Catal., 1982, 77, 473.
12 Y. Izumi K. Matsuo and K. Urabe, J. Mol. Catal., 1983, 18, 299.
1
3 Y. Onoue, Y. Mizutani, S. Akiyama and Y. Izumi, Chemtech., 1978,
8
, 432.
2
14 L. C. Jozefowicz, H. G. Karge, E. Vasilyeva and J. B. Moffat,
Microporous Mater., 1993, 1, 313.
5 G. I. Kapustin, T. R. Brueva, A. L. Klyachko, M. N. Timofeeva,
the percent conversion of 1-propanol is linear (r = 0.991).
Activation energies for the dehydration of 1-propanol were
1
investigated for the calcined samples. Activation energies of
S. M. Kulikov and I. V. Kozhevnikov, Kinet. Catal., 1990, 31, 896.
Ϫ1
1
3
7–18 kcal mol were found for the samples calcined at 200,
00 and 400 ЊC. These results are in the range for other solid
16 F. Lefebvre, F. X. Liu-Cai and A. Auroux, J. Mater. Chem., 1994, 4,
1
25.
30
17 A. K. Ghosh and J. B. Moffat, J. Catal., 1986, 101, 238.
acids used for dehydration of 1-propanol, and are consistent
1
8 R. S. Drago, J. A. Dias and T. O. Maier, J. Am. Chem. Soc., 1997,
19, 7702.
with no diffusional constraints.
1
1
2
2
2
2
2
2
9 J. A. Dias, J. P. Osegovic and R. S. Drago, J. Catal., 1999, 183, 83.
0 H. Hayashi and J. B. Moffat, J. Catal., 1983, 83, 192.
1 T. Xu, E. Munson and J. Haw, J. Am. Chem. Soc., 1994, 116, 1962.
2 A. Corma, Chem. Rev., 1995, 95, 559.
3 C. Chronister and R. S. Drago, J. Am. Chem. Soc., 1993, 115, 4793.
4 R. S. Drago, N. Kob and V. Young, Inorg. Chem., 1997, 36, 5127.
5 N. Kob and R. S. Drago, Catal. Lett., 1997, 49, 229.
Conclusion
It has been shown that calcination treatment has a profound
effect on the acidity of HPW. Samples calcined in air at 200,
3
00 and 400 ЊC for two hours are all in the anhydrous form of
HPW. No evident decomposition of the bulk Keggin structure
was observed, although the sample calcined at 400 ЊC showed
color modifications on its surface. Longer treatments at
temperatures higher than 300 ЊC clearly indicated changes of
its primary structure. Samples of HPW calcined, for the same
amount of time, at 200 and 300 ЊC have higher acidities and
reactivities than a sample calcined at 400 ЊC. The difference is
26 R. S. Drago, S. C. L. Dias, M. Torrealba and L. de Lima, J. Am.
Chem. Soc., 1997, 119, 4444.
2
7 C. Xian-e, D. Daichun, N. Jianping, J. Youming, Z. Jing and
Q. Yixiang, Thermochim. Acta, 1997, 292, 45.
2
8 J. Keggin, Proc. R. Soc. London, Ser. A, 1934, 144, 75.
9 B. W. L. Southward, J. S. Vaughan and C. T. O’Connor, J. Catal.,
2
1
995, 153, 293.
30 M. Bowker, W. Petts and K. Waush, J. Catal., 1986, 99, 53.
J. Chem. Soc., Dalton Trans., 2001, 228–231
231