Propylene oxidation over poly(azomethines) doped with heteropolyacids

Article abstract of DOI:10.1006/jcat.1999.2678

Kinetic studies of propylene oxidation and isopropanol conversion were made using new catalytic systems, i.e., conjugated polymer-supported heteropolyacid catalysts. Unsubstituted and dimethoxy-substituted aromatic poly(azomethines), PPI and PMOPI, were used as matrices to which selected heteropolyanions (HPA) were introduced via acid-base-type doping consisting of protonation of imine nitrogen. PPI was synthesized via polycondensation of 1,4-phenylenediamine with terephthaldehyde, while PMOPI was synthesized via the reaction of 1,4-phenylenediamine with 2,5-dimethoxyterephthaldehyde. Protonation of PPI and PMOPI with heteropolyacids (H₃PW₁₂O₄₀ and H₃PMo₁₂O₄₀) was done in acetonitrile solutions; the protonation level being controlled by changing the starting HPA/polymer molar ratio in the reaction medium. FTIR studies showed that the structural identity of individual HPA is preserved on introduction into the polymer matrices. Raman studies, however, showed that protonation leads to changes in polymer chain conformation. In propylene oxidation, the PPI(HPA)_y and PMOPI(HPA)_y catalysts yielded hexadiene as the primary product. HPA introduction into the polymer matrices changed the selectivity of the acid since the reaction conducted in the presence of crystalline H₃PMo₁₂O₄₀ yields acrolein. Only products of nondestructive oxidation (hexadiene, acrolein, propionaldehyde, acetone, benzene) were seen in all these catalysts.

Full text of DOI:10.1006/jcat.1999.2678

Journal of Catalysis 189, 297–313 (2000)
doi:10.1006/jcat.1999.2678, available online at http://www.idealibrary.com on
Propylene Oxidation over Poly(azomethines) Doped with Heteropolyacids
�
,1
Wincenty Turek, Edyta-Stochmal-Pomarza n´ ska,† Adam Pro n´ ,‡ and Jerzy Haber§
�
Institute of Physical Chemistry and Technology of Polymers, Silesian Technical University, Strzody 9, 44 100 Gliwice, Poland; †Department
of Materials Sciences and Ceramics, Academy of Mining and Metallurgy, Al.Mickiewicza 30, 30 059 Krak o´ w, Poland; ‡Department
of Chemistry, Warsaw University of Technology, Noakowskiego 3, 00 664 Warszawa, Poland, and CEA Grenoble
DRFMC/SI3M/PMS, 17 rue des Martyrs, 38 054 Grenoble, France; and §Institute of Catalysis and Surface
Chemistry, Polish Academy of Sciences, Niezapominajek, 30 239 Krak o´ w, Poland
Received April 22, 1999; revised September 7, 1999; accepted September 9, 1999
electrochemical, and electrooptic properties (1, 2). How-
Kinetic studies of propylene oxidation and isopropanol con- ever, these systems may also exhibit extremely interesting
version were undertaken using new catalytic systems, namely, catalytic properties provided that they are doped with cata-
conjugated polymer-supported heteropolyacid catalysts. Unsub-
stituted and dimethoxy-substituted aromatic poly(azomethines)
lytically active species, for example, heteropolyanions. The
first successful incorporation of heteropolyanions into poly-
(PPI and PMOPI) were used as matrices to which selected het-
conjugated matriceswasachieved byelectropolymerization
of suitable heterocyclic monomers in the presence of het-
eropolyacids (3, 4). In this procedure heteropolyanions are
introduced into the polymer matrix in situ during the elec-
tropolymerization process.
eropolyanions (HPAs) were introduced via acid–base-type doping
which consists of protonation of imine nitrogen. PPI was synthe-
sized via polycondensation of 1,4-phenylenediamine with tereph-
thaldehyde, whereas PMOPI was prepared via the reaction of
1
,4-phenylenediamine with 2,5-dimethoxyterephthaldehyde. Pro-
We have used a different approach in the preparation
tonation of PPI and PMOPI with heteropolyacids—H3PW12O40 and
H3PMo12O40—was carried out in the acetonitrile solutions. The pro- of conjugated polymer-supported heteropolyacid catalysts.
tonation level was controlled by changing the starting molar ratio First we prepared a neutral matrix of a given conjugated
of HPA : polymer in the reaction medium. Both poly(azomethines) polymer to which heteropolyanions were introduced by
in their base forms and protonated poly(azomethines) were char-
acterized using X-ray diffraction and various spectroscopic meth-
ods (FTIR, Raman, XPS). Thermal stability and BET surface area
were also studied. X-Ray diffraction patterns show that HPA intro-
duced into the polymer matrices is molecularly dispersed and does
We have introduced heteropolyanions into conjugated
not form a separate phase. FTIR studies prove that the structural
identity of individual heteropolyanions is preserved on introduc-
tion into the polymer matrices. However, Raman studies prove that
protonation leads to changes in polymer chain conformation. In the
chemical or electrochemical doping. For example, by ox-
idative doping of neutral polyacetylene with H3PMo12O40
we obtained a new catalytic system extremely active in al-
cohol conversion reactions (5).
polymers such as polyaniline and selected poly(azo-
methines), which contain basic sites, via an acid–base-type
of doping that consists of protonation of imine nitrogen
oxidation of propylene the PPI(HPA)y and PMOPI(HPA)y catalysts with heteropolyacids. Catalysts prepared by this method
give hexadiene as the main reaction product. Introduction of HPA were also tested in alcohol conversion (6, 7).
into the polymer matrices changes the selectivity of the acid since
the reaction carried out in the presence of crystalline H3PMo12O40
results in acrolein as the main product. For all catalysts studied,
only products of nondestructive oxidation can be detected (hexadi-
Aromatic poly(azomethines) of the general formula
( == HCArCH == NArN == )n
ene, acrolein, propionaldehyde, acetone, benzene).
Press
�c 2000 Academic are especially well suited for use in heterogeneous cata-
lysis because they exhibit excellent, for a polymeric sys-
Key Words: poly(azomethines); heteropolyanions; nondestruc-
tem, thermal stability in both the neutral (undoped)
and heteropolyacid-doped forms (7). Taking into account
this enhanced thermal stability we have decided to test
poly(azomethine)-supported heteropolyacid catalysts in
another technologically important reaction, namely, olefin
oxidation.
The process of olefin oxidation may take place on metal
or metal oxide catalysts. In the majority of cases the reaction
is carried out using solid solutions of oxides, mixtures of
oxides, or mixtures of oxysalts as catalysts (8–16).
tive oxidation.
INTRODUCTION
Conjugated polymers have been extensively studied for
the last 20 years mainly due to their unusual electronic,
1
To whom correspondence should be addressed. Fax: (48–12)633–71–
6
1. E-mail: stochmal@uci.agh.edu.pl.
297
0021-9517/00 $35.00
Copyright �c 2000 by Academic Press
All rights of reproduction in any form reserved.

2
98
TUREK ET AL.
In catalytic oxidation of hydrocarbons on oxide catalysts forms of oxygen influences the selectivity of the oxidation
two different intermediate complexes initiate two differ- process.
ent routes of the reaction (17–21). When oxygen is acti-
�
�
vated ionic radical oxygen species O and O2 are formed;
these are the main oxidizing species in the total oxidation
of simple molecules such as H2, CO, and CH4. They may be
considered electrophilic reagents, which in reactions with
olefins attack the molecule in the region of highest electron
EXPERIMENTAL
Two types of polymer matrices were synthesized: the sim-
plest unsubstituted aromatic polyimine
�
�
density. Such electrophilic addition of O2 or O results in
the formation of peroxy or epoxy complexes, respectively,
which under the conditions of heterogeneous catalytic ox-
idation are intermediates of the degradation of the carbon and its dimethoxy derivative
skeleton and total oxidation. These reactions may be classi-
fied as electrophilic oxidation (22–24). The second route of
heterogeneousoxidation startswith activation ofthe hydro-
carbon molecule by abstraction of hydrogen from a given
carbon atom, which becomes prone to nucleophilic addition
of the oxide ion O2 . It should be emphasized that the latter
has no oxidizing properties, but is a nucleophilic reactant.
The consecutive steps of hydrogen abstraction and oxygen
addition may then be repeated to obtain selectively more
oxygenated molecules. These reactions are classified as nu-
cleophilic oxidation (22–24). The role of oxidizing agents
in these steps of the reaction sequence is played by cations
of the catalyst lattice. After the nucleophilic addition of
the lattice oxygen ion the oxygenated product is desorbed,
leaving a vacancy at the catalyst surface.
For a long time heteropolycompounds were studied as
selective oxidation catalysts (22–24). Heteropolyacids with
Keggin structure are particularly attractive for mechanistic
studies since they present a well-characterized surface and
the versatility of their composition allows one to control
easily the acidic and redox properties. They provide nucle-
ophilic oxygen, the reactivity of which may be controlled,
and do not generate electrophilic oxygen.
As already mentioned poly(azomethines) are not in-
ert catalytic supports. The introduction of heteropolyacids
to these polymers occurs via the protonation reaction of
imine nitrogens. As a result catalytically active centers
are molecularly dispersed and strongly chemically bonded
to the polymer (7). The surface of the active phase of
poly(azomethine)-supported catalysts exhibits properties
significantly different from those of crystalline catalysts. As
a consequence of molecular dispersion and of the lack of
lattice defects characteristic of crystalline solids no elec-
trophilic forms of chemisorbed oxygen can exist on the sur-
�
a. Synthesis of Unsubstituted Aromatic Polyimine (PPI)
PPI was prepared from p-phenylenediamine and tereph-
thaldehyde using a modification of the method described
in (25).
Freshly recrystallized p-phenylenediamine (5.4 g) was re-
fluxed with 6.7 g of recrystallized terephthaldehyde in a so-
lution of 10 g of anhydrous sodium acetate in 35 ml of glacial
acetic acid. In the course of the reaction a bright yellow
polymer precipitated. After 2 h of refluxing the precipitate
was separated from the solution; then it was refluxed in
2
50 ml of dimethylformamide to dissolve unreacted mono-
mer and condensation products of low molecular weight.
Finally the polymer was washed with acetone and dried
in vacuo for 3 h.
b. Synthesis of Dimethoxy-Substituted
Polyimine (PMOPI)
PMOPI was prepared from p-phenylenediamine and
2
,5-dimethoxyterephthaldehyde using a modification of
the method described in (26, 27). Since 2,5-dimethoxy-
terephthaldehyde is not available commercially it was syn-
thesized following the procedure reported in (28).
The reaction was carried out at room temperature in
a 1 : 1 NMP : HMPA solution (110 ml) to which 2.3 g of
LiCl was added. To this reaction medium freshly recrys-
tallized diamine (4.1 g) was added followed by 7.5 g of
2,5-dimethoxyformamide. The reaction medium was vig-
orously stirred for 48 h. The polymer obtained was precipi-
tated in methanol, filtered, and then washed with water and
methanol. Finally it was dried under dynamic vacuum.
face. In the catalytic process only O2� (i.e., nucleophilic
form of oxygen) originating from individual Keggin units
can participate.
Thus heteropolyacid-doped poly(azomethines) are inter-
esting not only as new catalysts but also as suitable catalytic
systems for better elucidation of the mechanism of olefin
oxidation. In addition to physicochemical characterization
of these new catalysts, one of the goals of this research was
c. Doping (Protonation) of PPI and PMOPI
with Heteropolyacids
The doping (protonation) of the base form of the syn-
to determine how the reduction in the number of active thesized polymers was carried out at room temperature,

PROPYLENE OXIDATION OVER POLY(AZOMETHINES)
299
TABLE 1
Protonation Levels of Prepared Catalysts Determined by Elemental Analysis: C, H, N, Mo, or W
PMOPI (or PPI) structural
unit : HPA molar ratio in
the protonating medium
Protonating level determined
per polyimine in the
structural unit
Types of polymer matrix
PPI (repeat unit: C14H10N2)a
1 : 2
PPI(H3PMo12O40)0.709
PMOPI (repeat unit: C16H14N2O2)
1 : 0.025
PMOPI(H3PMo12O40)0.018
PMOPI(H3PMo12O40)0.036
PMOPI(H3PMo12O40)0.068
PMOPI(H3PMo12O40)0.198
PMOPI(H3PMo12O40)0.431
PMOPI(H3PW12O40)0.012
PMOPI(H3PW12O40)0.038
PMOPI(H3PW12O40)0.066
PMOPI(H3PW12O40)0.173
PMOPI(H3PW12O40)0.399
1
1
1
1
1
1
1
1
1
: 0.05
: 0.1
: 0.25
: 1
: 0.025
: 0.05
: 0.1
: 0.25
: 1
a
A more detailed characterization of PPI(HPA)y samples is given in Ref. (7).
typically for 1–1.5 h, in 0.02 M solution of a given hetero- gen and propylene, were used in the molar ratio of 1.2. The
polyacid in acetonitrile. The doped polymer was washed substrates were diluted with nitrogen to achieve an initial
repeatedly with pure acetonitrile to remove the unreacted concentration of propylene equal to 0.020 molar fraction
3
heteropolyacid. The following acids were used in this study: in the stationary state. A constant flow rate of 15.0 dm /h
H3PMo12O40 and H3PW12O40. The doping level can be con- was used throughout the entire experiment. The conver-
veniently varied by changing the ratio of polyimine to het- sion levels ranged from 5 to 15% . The measurements were
eropolyacid.
performed in the temperature range 441–545 K.
The kinetics of isopropanol conversion were measured
on two types of catalysts: freshly prepared ones and those
previously used for studies of propylene conversion kinet-
ics. The reaction of isopropanol conversion was carried out
in an oxygen-free atmosphere (6, 31). The concentration of
isopropanol in nitrogen was 1.45 mol% . Conversion levels
ranged from 5 to 20% .
d. Spectroscopic Studies
FTIR studiesofundoped and doped polyimineswere per-
formed in transmission geometry using the pressed pellet
technique on a Digilab FTS 60 spectrometer.
Raman spectra were obtained on a Bruker RFS 100 RT
Raman spectrometer with a 1064-nm excitation line.
X-Ray photoelectron spectra were recorded on a Escalb
Typically 2 g of the catalysts was used for each measure-
ment. Before measurements all catalysts were standarized
in the reactor in a flow of the reaction mixture with the same
chemical composition as the mixture used in the catalytic
2
10 spectrometer.
e. X-Ray Diffraction Studies
�
test. The standarization was performed at 120 C for 2 h.
X-ray diffractograms were obtained on a Seifert-FPM
XRD 7 diffractometer using CuK� radiation.
A detailed characterization of the samples used can be
found in Table 1.
The reaction products were analyzed chromatographi-
cally. Based on the results obtained both selectivity and
f. TG and DSC Studies
TG and DSC studies were carried out using a Setaram activation energy were calculated for each catalyst used.
�
TGDSC 111 apparatus. The heating rate was 10 C/min.
g. BET Specific Surface Area
RESULTS AND DISCUSSION
BET specific surface area measurements were performed
by adsorption of nitrogen at the temperature of liquid N2
on a Carlo Erba Sorpty 1750 apparatus.
First we must specify the system of chemical formula
notation we use throughout the paper for both polyimines
doped with heteropolyacids. Assuming the repeat units
of C14H10N2 and C16H14N2O2 for PPI and PMOPI, re-
spectively, we express the doping level as the fraction of
h. Catalytic Tests
Kinetic studies of propylene oxidation and isopropanol heteropolyacid (HPA) molecules, y, per repeat unit, i.e.,
conversion were undertaken.
(C14H10N2) (HPA)y or in shortened version PPI(HPA)y
Studies of propylene oxidation kinetics were carried out and (C16H14N2O2)(HPA)y or in shortened version
in a differential reactor (29, 30). The substrates, i.e., oxy- PMOPI(HPA)y.

3
00
TUREK ET AL.
Chemical compositions of all samples studied in this re- plicity in the description of the samples studied we neglect
search were determined by elemental analysis (C, H, N, hydration.
Mo, or W). Since heteropolyacids entering the polymer
Interesting information concerning the interactions of
matrix are hydrated a good fit of the analytical data can the inserted polyheteroanions with the polyimine matrix
be obtained only if the hydration process is taken into ac- can be extracted from analysis of FTIR spectra of sam-
count. The final formulas should therefore be expressed ples doped to different levels. In this paper we present the
as PPI(HPA)y(H2O)z and PMOPI(HPA)y(H2O)z. For sim- spectra of the POMPI(H3PW12O40)y system (Fig. 1) but for
FIG. 1. FTIR spectra of (a) unprotonated PMOPI, (b) PMOPI(H3PW12O40)0.038, (c) PMOPI(H3PW12O40)0.066, (d) PMOPI(H3PW12O40)0.399, and
(
e) crystalline H3PW12O40.

PROPYLENE OXIDATION OVER POLY(AZOMETHINES)
301
FIG. 2. Raman spectra of (a) unprotonated PPI, (b) PPI(H3PW12O40)0.038, (c) PPI(H3PW12O40)0.104, (d) PPI(H3PW12O40)0.468, and
(
e) PPI(H3PW12O40)0.804.
other aromatic polyimines and other heteropolyacids the nM == Od, nX–Oa (32)—the intensity of which increases with
doping-induced spectral changes are very similar. increasing doping level. The position of the band acribed to
The first and rather obvious conclusion is that the struc- nM-Oc-M depends on the doping level. For low doping lev-
�
1
tural identity of the Keggin units characteristic of HPA is els it is blue shifted by ca. 25 cm with respect to its position
�
1
preserved on the insertion into the polymer matrix. This is in the crystalline heteropolyacid at 799 cm . With increas-
manifested by the presence of four principal bands ascribed ing doping level it approaches the position characteristic of
to the Keggin structural unit—nM–Oc–M, nM–Ob–M, crystalline HPA. We interpret this shift as a measure of the

3
02
TUREK ET AL.
FIG. 3. Raman spectra of (a) unprotonated PMOPI, (b) PMOPI(H3PW12O40)0.038, (c) PMOPI(H3PW12O40)0.066, (d) PMOPI(H3PW12O40)0.173, and
e) PMOPI(H3PW12O40)0.399.
(
interactions between the inserted dopant and the matrix. for polyimines doped with HPA to different levels. This is
These interactions occuring via protonation of the imine indeed the case (7). In isopropanol conversion PPI(HPA)y
sites cause some structural deformation of the [MO6] struc- is very selective to acetone for low values of y, whereas
tural subunit which in turn is manifested by a shift of the for high values of y it acts as a predominantly acid–base
nM–Oc–M IR band. Thus at low doping levels the proto- catalyst, giving propylene as the main product. Evidently
nation of the matrix by the inserted acid molecules is more for low y all acid molecules are transformed into the conju-
efficient. If this interpretation is correct one should expect gated bases via the protonation reaction with the strongly
different selectivity to the products of acid–base catalysis basic environment of the matrix. As a result all acid–base

PROPYLENE OXIDATION OVER POLY(AZOMETHINES)
303
centers are blocked and the catalyst acts as a redox cata- in basic PPI and PMOPI all nitrogen atoms are spectroscop-
lyst. To the contrary for high values of y the majority of ically equivalent in the N 1s spectrum one line is expected
acid molecules do not participate in the protonation proba- with a binding energy characteristic of imine nitrogen. The
bly due to steric reasons and can be considered as a source spectrum of PPI base is shown in Fig. 4a. Indeed the dom-
of mobile protons necessary for catalysis of the acid–base inant line at 398.7 eV is characteristic of imine nitrogen.
type.
Both PPI and PMOPI in their base form give clear Raman
spectra (Figs. 2a and 3a). Based on the comparison with low-
molecular-weight model compounds (33) and with poly(p-
phenylene vinylene) (34), which is isoelectronic with PPI,
the following attribution of bands is proposed: A weak band
�
1
at 1622 cm observed in both polymers corresponds to
the C == N stretching mode. The corresponding band is also
observed in the IR spectra; however, in the latter case it
is very strong as expected. Two strong bands at 1578 and
�
1
1
551 cm in PPI originate from C == C stretching of the
para-substituted aromatic bands. In PMOPI the latter band
�
1
is blue shifted by ca. 10 cm and as a result both bands
strongly overlap. By analogy with PPV and the model com-
pound the bands at 1363 cm in PPI and at 1359 cm in
PMOPI can be ascribed to C–H deformations in the imine
�
1
� 1
�
1
group. Strong bands at 1192 and 1153 cm in PPI are at-
tributed to C–Car stretching and C–H ring deformations,
respectively. Contrary to PPI, in PMOPI two bands orig-
inating from C–H ring deformations appear at 1162 and
�
1
1
121 cm . This is caused by the presence of two nonequiv-
alent aromatic rings, di- and tetrasubstituted.
The most striking doping-induced Raman spectral
change in PPI and PMOPI (Figs. 2 and 3) is strong line
broadening caused by amorphization of the sample as
evidenced by X-ray diffraction (see below). Due to the pro-
tonation reaction that accompanies the doping significant
changes are observed in the bands corresponding to the
protonation sites (CH == N groupings). The weak band at
�
1
1
622 cm gradually disappears on doping and a new band
�
1
characteristic of the doped polymer appears at 1665 cm
in the spectrum of PPI and at 1651 cm in the spectrum
of PMOPI. The bands due to C == C stretching in the
aromatic ring are also influenced. The band at 1578 cm
in PPI slightly shifts toward higher wavenumbers whereas
the band at 1551 cm significantly weakens. Two modes
observed in IR but inactive in Raman appear at 1467
and 1412 cm . These changes indicate that in addition to
amorphization of the sample, the doping induces significant
conformational changes in the polymer chain.
�
1
�
1
�
1
�
1
At the end of the discussion of the Raman data it should
be stressed that no lines attributable to heteropolyanions
are observed even for high heteropolyanions content. Thus
IR and Raman are perfectly complementary in our case;
the first technique enables us to follow the deformations in
Keggin structural units occuring on insertion into the poly-
mer matrix, whereas the latter method provides evidence
of chain conformational changes induced by doping.
Protonation of polyimine basic sites by acidic dopants
FIG. 4. X-Ray photoelectron N 1s spectra of unprotonated PPI
a) and PPI protonated with H3PW12O40: (b) PPI(H3PW12O40)0.104,
(
should also be evidenced by ESCA N 1s spectroscopy. Since (c) PPI(H3PW12O40)0.804.

3
04
TUREK ET AL.
FIG. 5. X-Ray photoelectron W 4f spectrum of PPI(H3PW12O40)0.468.
However, an additional satellite line of weak intensity is observed in polyaniline protonated with heteropolyacids
also observed at 400.2 eV. We were unable to get rid of (35). The spectrum can be decomposed into three compo-
this peak despite careful purification of the sample. Its ex- nents at 398.7, 400.2, and 401.8 eV. The last two peaks are in
act origin is not known at present. The doping with het- the range characteristic of charged (protonated) nitrogens.
eropolyacids results in a gradual growth of higher-binding- This means that in the doped polymer two spectroscopically
energy components of the spectrum at the expense of the nonequivalent protonated sites coexist.
imine base peak (Figs. 4b and 4c). Such behavior is char-
In Figs. 5 and 6 W 4f and Mo 3d are shown for PPI doped
acteristic of the protonation reaction and has already been with H3PW12O40 and H3PMo12O40, respectively. In the
FIG. 6. X-Ray photoelectron Mo 3d spectrum of PPI(H3PMo12O40)0.452.

PROPYLENE OXIDATION OVER POLY(AZOMETHINES)
305
spectrum of tungsten only one doublet characteristic of where L = the average size of crystalline domains, B = half-
W(VI) is observed which means that the doping with width (in radian units), and k = Scherrer constant (equals
H3PW12O40 is purely acid–base in nature and W does not about 1). We have calculated the average size of the crys-
change its oxidation state on insertion into the polymer talline zones. In the case of PPI it was 390 A˚ (based on the
matrix. On the contrary, in the case of the doping with reflection peaked at 20.3), whereas for PMOPI it was 215 A˚
H3PMo12O40, the acid–base doping is acompanied by a re- (based on the reflection peaked at 13.9).
dox reaction because in the spectrum of Mo 3d, in addition
On doping the reflections characteristic of the base form
to the doublet corresponding to Mo(VI) a weak doublet gradually disappear, indicating increasing amorphization.
due to Mo(V) is observed. Thus molybdenum is partially No reflections originating from crystalline heteropolyacids
reduced on insertion into PPI.
can be detected even for the highest doping levels. This
Doping-induced amorphization ofthe sample manifested means that the doped polymers form a one-phase sys-
by the broadening of IR and Raman lines can be di- tem in which the dopant molecules are molecularly dis-
rectly monitored by X-ray diffraction measurements. In persed. At the highest doping level we observe a new re-
Figs. 7 and 8 X-ray diffractograms recorded for increas- flection at low Bragg angles which we attribute to one-
ing doping levels are shown for PPI(H3PW12O40)y and dimensional ordering of the dopant in the polymer matrix.
PMOPI(H3PW12O40)y. Undoped (basic) polymers are sur- A similar phenomenon was previously observed for hetero-
prisingly highly crystalline. The crystallinity index deter- polyacid-doped polyaniline (37).
mined usingthe method of Hindeleh and Johnson isroughly
Spectroscopic and diffraction studies provide significant
the same for both polyimines, slightly lower than 30% . information concerning the molecular and supramolecular
PMOPI, however, gives broader reflections which means structures of doped polyimines. However, from the cata-
that in this case crystalline domains are smaller. Using the lytic point of view, such properties as the value of specific
Scherrer formula correlating the size of crystalline domains surface or thermal stability are equally important.
with the half-width of the Bragg reflections (36),
Similarly as in the case of heteropolyacid-doped polyani-
line (38), for PPI specific surface area monotonically de-
creases with the increase in doping level (Fig. 9). Ev-
idently heteropolyacids attached to the surface via the
k � �
L =
,
B � cos 2
FIG. 7. X-ray diffractograms of (a) unprotonated PPI, (b) PPI(H3PW12O40)0.003, (c) PPI(H3PW12O40)0.038, (d) PPI(H3PW12O40)0.104,
(
e) PPI(H3PW12O40)0.468, and (f) PPI(H3PW12O40)0.804.

3
06
TUREK ET AL.
FIG. 8. X-ray diffractograms of (a) unprotonated PMOPI, (b) PMOPI(H3PW12O40)0.012, (c) PMOPI(H3PW12O40)0.038, (d) PMOPI(H3PW12O40)0.066,
e) PMOPI(H3PW12O40)0.173, and (f) PMOPI(H3PW12O40)0.399.
(
�
protonation reaction lower the porosity of the system. The drated, the small mass loss around 100–120 C is associ-
behavior of PMOPI is different. In this case the specific sur- ated with partial dehydration of the dopant. As expected
face area increases with y for low doping levels and then be- a weak endotherm is observed in this temperature range
gins to decrease (Fig. 10). Probably light doping causes dis- in the DSC curve. The intensity of this endotherm in-
integration of agglomerates present in the polymer. When creases with increasing doping level. Catalyst degrada-
�
this process is complete further doping leads to a decrease tion starts at ca. 280–300 C and is manifested in the DSC
in the remaining porosity.
line by a set of strongly overlapping exotherms. However,
�
The thermal stability of H3PW12O40-doped PPI is ex- up to 420–440 C no significant mass loss is recorded in
cellent for a polymeric system. A typical TG, DTG, DSC TG.
sets of curves are shown in Figs. 11 and 12. Since het-
Similar features are observed for H3PW12O40-doped
eropolyanions inserted into the polymer matrix are hy- PMOPI; however, in this case the degradation starts at ca.
FIG. 9. Specific surface area (SBET) versus protonation level (y) of
FIG. 10. Specific surface area (SBET) versus protonation level (y) of
PPI protonated with H3PW12O40.
PMOPI protonated with H3PW12O40.

PROPYLENE OXIDATION OVER POLY(AZOMETHINES)
307
FIG. 11. Thermogravimetric (TG), differential gravimetric (DTG), and differential scanning calorimetry (DSC) set of curves for
PPI(H3PW12O40)0.104.
FIG. 12. Thermogravimetric (TG), differential gravimetric (DTG), and differential scanning calorimetry (DSC) set of curves for
PPI(H3PW12O40)0.804.

3
08
TUREK ET AL.
may suggest that the reaction takes place in the diffusion
limit. However, low reaction temperature seems to exclude
this hypothesis.
In principle one can imagine the slowing down of the
transport of oxygen by the polymer matrix which would
make the diffusion process the limiting step of the reaction.
However, if the reaction occurred in the diffusion limit the
activation energy would be lower for catalysts containing
less heteropolyacid (for low protonation levels the interac-
tions between the matrix and the dopant are stronger and
the ratio of matrix molecules to heteropolyacid molecules
is higher). The experiment shows the opposite.
In the reaction of isopropanol conversion on freshly pre-
pared polymer-supported H3PMo12O40 catalysts, contain-
ing different amounts of heteropolyacid, catalytic activity
seems to be independent of heteropolyacid concentration,
i.e., the activity normalized per 1 g of heteropolyacid is very
similar for all catalysts tested. However, for catalysts used
previously for propylene oxidation the measured activity
is higher. This increase in activity is especially pronounced
in the case of the catalyst containing the largest amount of
HPA, i.e., PMOPI(H3PW12O40)0.431.
In all cases studied propylene oxidation reaction carried
out prior to isopropanol conversion causes an increase in
FIG. 13. Arrhenius plots of propylene oxidation over (a) PPI
(
(
H3PMo12O40)0.709, (b) PMOPI(H3PMo12O40)0.431, and (c) PMOPI
H3PMo12O40)0.068.
�
6
0 C lower temperatures. This also applies to H3PM12O40-
doped PPI and PMOPI.
Another question, important from the catalytic point of
view, should be addressed, namely, the stability of the cata-
lyst in the presence of the reagents undergoing the catalytic
reaction. Since heteropolyacid is chemically bound to the
polymer matrix via ionic bonds it cannot be leached out by
isopropanol or ethylene or any other reagent tested. The
action of these reagents does not deprotonate the polymer;
thus heteropolyanions remain fixed to the basic sites of the
polymer matrix in the course of the catalytic reaction.
The results of catalytic tests are shown in Figs. 13–16. The
kinetic data are presented as the logarithm of the observed
reaction rate versus the reciprocal of the absolute reaction
temperature. For comparative reasons all data have been
normalized to 1 g of heteropolyacid in the catalyst. By this
procedure the contribution of catalytically inactive polymer
matrix to the mass of the catalyst is eliminated.
The calculated activation energies and selectivities are
collected in Tables 2–4. As expected in the reaction of
propylene oxidation H3PW12O40 catalysts turned out to be
less active than H3PMo12O40 catalysts. Surprisingly, the less
FIG. 14. Arrhenius plots of propylene oxidation over (a)
active catalysts show much lower activation energy. This PMOPI(H3PW12O40)0.399 and (b) PMOPI(H3PW12O40)0.066.

PROPYLENE OXIDATION OVER POLY(AZOMETHINES)
309
FIG. 15. Arrhenius plots of isopropanol decomposition over catalytic samples: (a) PMOPI(H3PW12O40)0.066, (b) PMOPI(H3PW12O40)0.399.
(
1) Propylene; (2) acetone. (A) Freshly prepared catalysts, (B) catalysts used previously for the propylene oxidation reaction.
activity and simultanously a significant decrease in the acti-
vation energy of the dehydration reaction. This observation of propylene oxidation (272 C) no thermal destruction
maysuggest that some ofthe protonspreviouslyinactive are of the catalyst takes place. This high thermal stability
Despite the rather high temperature of the process
�
activated via propylene oxidation process.
concerns both the polymer matrix and the heteropolyacid
TABLE 2
Selectivities and Activation Energies of the Catalytic Propylene Oxidation Reaction
HPA content in
Selectivity (% )
Activation
energy
(kJ/mol)
the sample
(wt% )
Temperature
(K)
Catalyst
Propionaldehyde
Acetone
Acrolein
Benzene
Hexadiene
PPI(H3PMo12O40)0.709
86.3
71.8
29.8
72.7
38.2
464
—
8.3
—
—
2.8
85.6
—
—
97.2
6.1
33.7
59.6
64.7
16.2
22.8
4
96
495
32
498
39
505
27
487
09
PMOPI(H3PMo12O40)0.431
PMOPI(H3PMo12O40)0.068
PMOPI(H3PW12O40)0.399
PMOPI(H3PW12O40)0.066
—
0.9
4.5
3.9
4.9
18.0
—
—
90.6
77.2
5
—
2.0
1.1
4.4
19.9
20.8
—
—
79.0
72.8
5
—
—
—
—
29.3
25.4
32.5
21.1
38.2
53.5
5
—
—
—
—
11.7
15.8
—
—
88.3
84.2
5

3
10
TUREK ET AL.
FIG. 16. Arrhenius plots of isopropanol decomposition over catalytic samples: (a) PMOPI(H3PMo12O40)0.068, (b) PMOPI(H3PMo12O40)0.431.
1) Propylene, (2) acetone. (A) Freshly prepared catalysts, (B) catalysts used previously for the propylene oxidation reaction.
(
dopant. Evidentlythe polymer matrixstabilizesthe inserted acids would inevitably lead to a decrease in the number of
molecules of heteropolyacids. This conclusion is corrobo- acidic centers (mobile protons). This in turn would result in
rated by the determination of the catalyst activity in iso- a significant decrease in the dehydration reaction rate. The
propanol conversion. Thermal destruction of heteropoly- experiment shows the opposite.
TABLE 3
Selectivities of the Catalytic Isopropanol Decomposition Reaction Measured for
(a) PMOPI(H3PMo12O40)y Catalysts and (b) PMOPI(H3PW12O40)y Catalysts
HPMo/HPW content
Selectivity at 440 K (% )
of sample
Catalysta
(wt% )
Propylene
Acetone
a. PMOPI(H3PMo12O40)y catalysts
A
B
A
B
PMOPI(H3PMo12O40)0.068
PMOPI(H3PMo12O40)0.068
PMOPI(H3PMo12O40)0.431
PMOPI(H3PMo12O40)0.431
29.8
29.8
71.8
71.8
43.7
11.8
75.3
81.3
56.3
88.2
24.7
18.6
b. PMOPI(H3PW12O40)y catalysts
A
B
A
B
PMOPI(H3PW12O40)0.066
PMOPI(H3PW12O40)0.066
PMOPI(H3PW12O40)0.399
PMOPI(H3PW12O40)0.399
38.2
38.2
72.7
72.7
1.1
23.2
97.7
99.6
98.9
76.8
2.3
0.4
a
(
A) Freshly prepared catalysts, (B) catalysts used previously for the propylene oxidation reaction.

PROPYLENE OXIDATION OVER POLY(AZOMETHINES)
311
TABLE 4
Activation Energies of the Catalytic Isopropanol Decomposition Reaction Measured for
(a) PMOPI(H3PMo12O40)y Catalysts and (b) PMOPI(H3PW12O40)y Catalysts
HPMo/HPW content
Activation energy (kJ/mol)
of sample
Catalysta
(wt% )
Dehydration reaction
Dehydrogenation reaction
a. PMOPI(H3PMo12O40)y catalysts
A
B
A
B
PMOPI(H3PMo12O40)0.067
PMOPI(H3PMo12O40)0.067
PMOPI(H3PMo12O40)0.430
PMOPI(H3PMo12O40)0.430
29.8
29.8
71.8
71.8
133.3
79.7
106.3
70.2
57.7
79.3
43.2
46.0
b. PMOPI(H3PW12O40)y catalysts
A
B
A
B
PMOPI(H3PW12O40)0.066
PMOPI(H3PW12O40)0.066
PMOPI(H3PW12O40)0.390
PMOPI(H3PW12O40)0.390
38.2
38.2
72.7
72.7
113.2
84.0
88.4
83.3
5.4
51.1
23.9
22.2
a
(
A) Freshly prepared catalysts, (B) catalysts used previously for the propylene oxidation reaction.
An increase in the activity of the catalysts used previously the catalyst. The reoxidation of the catalyst must be slow in
for propylene oxidation can be rationalized by the assump- this case.
tion of diffusion of the heteropolyacid molecules from the
bulk of the polymer to its surface.
The recombination of two allyl complexes accompanied
by dehydrogenation results in the formation of benzene.
The phenomenon of diffusion of the catalyst active phase We observe small amounts of this product only in the case
toward the catalyst surface, induced by the reagents, has of PMOPI(H3PW12O40)0.399 catalyst.
previously been observed for oxide as well as metal alloy
If during the catalytic process an oxygen atom originating
catalysts (39). This concept of diffusion is additionally cor- from the Keggin unit is attached and one hydrogen atom is
roborated by the fact that the increase in catalytic activity given off, an acrolein molecule is formed. Fast reoxidation
is more pronounced for catalysts with higher content of the of the catalysts favors this process, and as a result under
active phase.
these conditions no hexadiene formation is observed. Also
For all catalysts studied only products of nondestructive no hexadiene is formed if other than nucleophilic forms of
oxidation can be detected. These are acrolein, hexadiene, oxygen are present in the system.
propionaldehyde, acetone, and benzene.
If oxygen originating from the Keggin unit is bound to
In the case of H3PMo12O40 catalysts hexadiene is the the allyl radical, together with a hydrogen atom detached
dominant product at low temperatures. For all catalytic sys- from the propylene molecule in the process of its activa-
tems studied the selectivity to hexadiene decreases with in- tion, propion aldehyde is formed. The selectivity toward
creasing temperature, whereas the selectivity to acrolein this product is, however, very low.
increases. We interpret this phenomenon as the manifes-
tation of faster reoxidation of heteropolyanions at higher molecularly dispersed in PMOPI matrix, i.e., the catalysts
temperatures. that produce predominantly hexadiene, high activation en-
Hexadiene is the product of the recombination of two ergy is measured.
It is interesting to note that for H3PMo12O40 catalysts
pi-allyl complexes which are formed at the first stage of
For H3PW12O40 catalysts molecularly dispersed in
propylene molecule activation on transition metal ions of PMOPI matrix both hexadiene and acrolein are produced.
TABLE 5
Selectivity and Activation Energy of the Catalytic Propylene Oxidation Reaction Measured
for H3PMo12O40 on � -Al2O3
HPMo content
of sample
(wt% )
Selectivity (% )
Activation
energy
(kJ/mol)
Temperature
(K)
Catalyst
Propionaldehyde
Acrolein
Benzene
H3PMo12O40
on � -Al2O3
11.7
445
475
1.7
1.5
95.5
97.8
2.8
0.7
53.6

3
12
TUREK ET AL.
appropriate orientation of the propylene molecule in the
course of catalytic reaction. This fact is strongly manifested
by the influence of the polymer matrix on the selectivity of
the oxidation reaction. For H3PMo12O40 catalysts dispersed
in PMOPI no significant decrease in the selectivity to hexa-
diene is observed with increasing temperature. This is true
even for low contents of H3PMo12O40.
CONCLUSIONS
To summarize, we have prepared a new type of polymer-
supported catalysts by acid–base-type doping of solid aro-
matic poly(azomethines) with heteropolyacids. These new
catalysts are active in propylene oxidation, exhibiting total
selectivity to the products of nondestructive oxidation, with
hexadiene as the dominant product. In the case of oxidation
to hexadiene two factors play an important role: the lack of
electrophilic forms of oxygen and steric effects influenced
by the polymer matrix.
The general conclusion derived from this research can
be formulated as follows: Since in all polymer-supported
catalysts only the products of nondestructive oxidation are
detected the formation of products of destructive oxidation
(
including total combustion) is inherently associated with
the presence of electrophilic forms of oxygen which are
absent in our systems.
ACKNOWLEDGMENT
This work was financially supported by the Polish Commitee for Scien-
tific Research (KBN) Grant 10.10.160.427.
FIG. 17. Arrhenius plots of propylene oxidation over � -Al2O3-
supported, crystalline H3PMo12O40.
REFERENCES
The ratio of acrolein to hexadiene increases with increasing
content of heteropolyacid in the polymer matrix; however,
the activation energy of the oxidation reaction decreases.
The prepared catalysts differ significantly from the cata-
lysts obtained by the deposition of H3PMo12O40 on � -Al2O3
which we studied under the same experimental conditions
for comparative reasons. In this case acrolein is the dom-
inant product (Table 5, Fig. 17). Interesting catalytic be-
havior is reported for the PPI(H3PMo12O40)y sample with
high content of heteropolyacid. In this case the activation
energy of the oxidation reaction is much lower than for
PMOPI(H3PMo12O40)y catalysts. At low temperatures es-
sentially hexadiene is produced with high selectivity. How-
ever, in the case of PPI(H3PMo12O40)0.709 an abrupt change
1
2
. Nalwa, H. S. (Ed.), “Handbook of Organic Conductive Molecules and
Polymers.” Wiley, New York, 1997.
. Skotheim, T. A., Elsenbaumer, R. L., and Reynolds, J. R. (Eds.),
“
1
Handbook of Conducting Polymers.” Marcel Dekker, New York,
998.
3
. Bidan, G., Genies, E. M., and �L apkowski, M., J. Electroanal. Chem.
51, 297 (1988).
4. Keita, B., Bouaziz, D., and Nadjo, L., J. Electroanal. Chem. 255, 303
1988).
2
(
5
. Po z´ niczek, J., Kulszewicz-Bajer, I., Zag o´ rska, M., Krucza �l a, K., Dyrek,
K., Biela n´ ski, A., and Pro n´ , A., J. Catal. 132, 311 (1991).
. Hasik, M., Turek, W., Stochmal, E., L� apkowski, M., and Pro n´ , A.,
J. Catal. 147, 544 (1994).
6
7. Stochmal-Pomarza n´ ska, E., Hasik, M., Turek, W., and Pro n´ , A., J. Mol.
Catal. 114, 267 (1996).
8
9
. Haber, J., and Grzybowska, B., J. Catal. 28, 489 (1973).
. Linde, W. R., Margolis, L. J., and Rogi n´ ski, S. Z., Dokl. Akad. Nauk
SSSR 136, 360 (1961).
in selectivity is observed in the temperature range 191 to
�
2
23 C: acrolein is formed instead of hexadiene. This means
10. Kurina, L. N., Kowal, L. M., and Kotienko, N. F., Z. Fiz. Chim. 49, 1494
that the polymer catalyst begins to behave similarly to clas-
sic heteropolyacid catalysts.
The lack of electrophilic forms of oxygen in the pro-
cess of propylene oxydation is not a sufficient condition for
the formation of hexadiene. Several steric conditions must
also be fulfilled; i.e., the matrix may or may not impose
(1975).
1
1
1. Kowal, L. M., and Kurina, L. N., Z. Fiz. Chim. 51, 889 (1977).
2. Ansic, A. G., Sokolowskij, W. S., and Boreskow, G. K., Kinet. Katal.
15, 1033 (1974).
1
1
3. Matsuura, Y., and Schmit, G. C. A., J. Catal. 20, 19 (1971).
4. Haber, J., Mielczarska, E., and Turek, W., React. Kinet. Catal. Lett. 34,
45 (1987).

PROPYLENE OXIDATION OVER POLY(AZOMETHINES)
313
1
1
1
5. Turek, W., Przem. Chem. 69, 398 (1990).
6. Turek, W., and Sucho n´ , S., Pol. J. Appl. Chem. 36, 225 (1992).
7. Biela n´ ski, A., Haber, J., in “Oxygen in Catalysis.” Marcel Dekker, 27. Yang, C. J., and Jenekhe, S. A., Chem. Mater. 3, 878 (1991).
26. Morgan, P. W., Kwolek, S. L., and Plechter, T. C., Macromolecules 20,
729 (1987).
New York, 1991.
8. Haber, J., Stud. Surf. Sci. Catal. 72, 279 (1992).
9. Haber, J., in “Proceedings, 8th International Congress on Catalysis,
Berlin, 1984,” Vol. 1, p. 85. Dechema, Frankfurt-am-Main, 1984.
0. Haber, J., in “The Role of Solid State Chemistry in Catalysis” (R. K.
28. Wood, J. H., and Gibson, E., J. Am. Chem. Soc. 71, 393 (1949).
29. Schwab, G. M., Chem. Ing. Technol. 39, 1191 (1967).
30. Br u¨ ckman, K., Haber, J., and Turek, W., J. Catal. 114, 196 (1988).
31. Turek, W., �L apkowski, M., and Bidan, G., Mater. Sci. Forum 122, 65
(1993).
1
1
2
2
Grasselli and J. F. Brazdil, Eds.), ACS Symp. Ser. No. 279, p. 3. Am. 32. Rocchiccioli-Deltcheff, C., Thouvenot, F., and Franck, R., Spectro-
Chem. Soc., Washington, DC, 1985.
chim. Acta. A 32, 587 (1976).
1. Haber, J., in “Heterogenous Catalytic Oxidation” (B. K. Warren and
S. Oyama, Eds.), ACS Symp. Ser. No. 638, p. 20. Am. Chem. Soc.,
Washington, DC, 1996.
33. Stochmal-Pomarza n´ ska, E., Ph.D. thesis, University of Mining and
Metallurgy. (AGH), Krakow, unpublished.
34. Buisson, J. P., Lefrant, S., and Louran, G., Synth. Met. 49–50, 305(1992).
35. Hasik, M., Pro n´ , A., Po z´ niczek, J., Biela n´ ski, A., Piwowarska, Z.,
Krucza �l a, K., and Dziembaj, R., J. Chem. Soc. Faraday Trans. 90, 2099
(1994).
2
2
2. Missono, M., Catal. Rev. 29, 269 (1987).
3. Missono, M., in “Proceedings, 10th International Congress on Catal-
ysis, Budapest, 1992” (M. J. Phillips and M. Ternan, Eds.). Akad e´ miai
Kiad o´ , Budapest, 1993.
36. Rabiej, S., Eur. Polym. J. 27, 947 (1991).
2
4. Br u¨ ckman, K., and Haber, J., in “Advances in Catalyst Design” (M. 37. �L u z˙ ny, W., and Hasik, M., Solid State Commun. 99, 685 (1996).
Graziani and C. N. R. Rao, Eds.), Vol. 2, p. 111. World Scientific, 38. Hasik, M., Pro n´ , A., Raynor, J. B., and �L u z˙ ny, W., New J. Chem. 19,
Singapore, 1993.
1155 (1995).
2
5. Manassen, J., and Khalif, S., J. Am. Chem. Soc. 88, 1943
39. Haber, J., Mielczarska, E., and Turek, W., Z. Phys. Chem. 144, 69
(
1966).
(1985).