A facile precursor route to transition metal molybdates using a polyzwitterionic matrix bearing simultaneously charged moieties and complexing groups

Article abstract of DOI:10.1039/b418233b

An unconventional but easily accessible precursor route involving the thermal treatment of hybrid precursors containing an ampholytic polymer matrix is developed to prepare multimetallic oxides of catalytic interest such as transition metal molybdates. A copolymer of diallyldimethylammonium chloride and a functionalized maleamic acid bearing an amine group suited for cation complexation was designed, synthesized and used as a matrix to stabilize inorganic species generated in solution from Ni(NO₃) ₂·6H₂O, Co(NC₃)₂· 6H₂O and/or Mn(NO₃)₂·4H₂O together with (NH₄)₆Mo₇O₂₄· 4H₂O. UV-vis-NIR as well as ¹³C-NMR studies suggest that the interactions between the cations and the polymer in solution are mainly electrostatic. Only minor complexation interactions take place under certain conditions. Homogeneous hybrid blends were prepared from these solutions. The presence of a complexing amine group in addition to the charged betaine moieties in the polymer permits stabilization of more than stoichiometric amounts of the metal species in the blends. XRD measurements suggest that the homogeneity in the solid state can be kept up to about 1.5 mol of each metal that is incorporated (anionic as well as cationic) per mol of repeat units of the copolymer. The blends were calcined under air at 600°C to produce the simple as well as mixed nickel, cobalt and manganese molybdates. Characterization of the final phases by XRD and Raman spectroscopy indicates that the α- as well as the β-molybdate phases can be prepared, and that the mixed structures are solid solutions of the simple NiMoO₄, MnMoO ₄ and CoMoO₄ If the precursors engaged are homogeneous, the pH of the precursor solution, the amount of metal that is incorporated in the matrix, and the nature of the polymer matrix seem to exert only a minor influence on the nature of the final phase, which demonstrates the versatility and facile applicability of the method. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b418233b

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/materials | Journal of Materials Chemistry
A facile precursor route to transition metal molybdates using a
polyzwitterionic matrix bearing simultaneously charged moieties and
complexing groups{
Franc¸ois Rullens,a Nicolas Deligne,a Andre´ Laschewsky*bc and Michel Devillers*a
Received 3rd December 2004, Accepted 3rd February 2005
First published as an Advance Article on the web 18th February 2005
DOI: 10.1039/b418233b
An unconventional but easily accessible precursor route involving the thermal treatment of hybrid
precursors containing an ampholytic polymer matrix is developed to prepare multimetallic
oxides of catalytic interest such as transition metal molybdates. A copolymer of
diallyldimethylammonium chloride and a functionalized maleamic acid bearing an amine group
suited for cation complexation was designed, synthesized and used as a matrix to stabilize
inorganic species generated in solution from Ni(NO3)2?6H2O, Co(NO3)2?6H2O and/or
Mn(NO3)2?4H2O together with (NH4)6Mo7O24?4H2O. UV–vis–NIR as well as 13C-NMR studies
suggest that the interactions between the cations and the polymer in solution are mainly
electrostatic. Only minor complexation interactions take place under certain conditions.
Homogeneous hybrid blends were prepared from these solutions. The presence of a complexing
amine group in addition to the charged betaine moieties in the polymer permits stabilization of
more than stoichiometric amounts of the metal species in the blends. XRD measurements suggest
that the homogeneity in the solid state can be kept up to about 1.5 mol of each metal that is
incorporated (anionic as well as cationic) per mol of repeat units of the copolymer. The blends
were calcined under air at 600 uC to produce the simple as well as mixed nickel, cobalt and
manganese molybdates. Characterization of the final phases by XRD and Raman spectroscopy
indicates that the a- as well as the b-molybdate phases can be prepared, and that the mixed
structures are solid solutions of the simple NiMoO4, MnMoO4 and CoMoO4. If the precursors
engaged are homogeneous, the pH of the precursor solution, the amount of metal that is
incorporated in the matrix, and the nature of the polymer matrix seem to exert only a minor
influence on the nature of the final phase, which demonstrates the versatility and facile
applicability of the method.
tunable properties: solubility in various polar media, number
1. Introduction
of functional groups of each type, presence of additional
New methods for the preparation of bulk or dispersed mixed
substituents able to induce suprastructure or complexation of
oxides exhibiting well-defined characteristics with respect to
the cations, etc. Up to now, much work has been devoted
structure, dispersion of the different components, specific
towards the design of polymers such as poly(betaine)s or
surface area and morphology, are continuously searched for,
poly(ampholyte)s, with the purpose of taking advantage of the
e.g. the production of new efficient catalysts. Among these
presence of electric charges along the chains. These structures
methods, the precursor routes, based on the decomposition
make them particularly suited to stabilize and interact with
of citrate gels,1 oxalates,2 or polyaminocarboxylates like
the anions and cations of a soluble inorganic salt by electro-
EDTA and its analogs,3 were successfully used. Alternatively,
structurally simple organic polymers such as poly(ethylene
glycol),4 poly(amide) and poly(imide) derivatives,5 but also
more elaborate polymer structures, resulting from the associa-
tion of monomers bearing carboxylic acid or amine moieties,6
were occasionally employed to synthesize polymer–metal com-
plexes, which are thermally treated7 to obtain the inorganic
materials. The latter polymers constitute particularly interest-
ing systems for this kind of application because of their
static8–11 interactions, to produce homogeneous hybrid mate-
rials in the solid state after removal of the solvent. However,
the poly(betaine)s designed up to now suffer from a major
drawback: the salt-free polymer structures are only obtained
via multistep reactions, involving tedious purification routes
such as the use of ion exchange resins, for example.12 Next to
these considerations, it was not clear whether the presence of
additional complexing groups next to the charged moieties of
the betaine in a polymer structure could help the stabilization
of simple inorganic salts in solution or in the solid state.
Recently, we published first results concerning the develop-
ment of a new synthetic pathway to multimetallic oxides,
based on the use of hybrid inorganic–organic mixtures made
from simple inorganic salts and an ordered and regular
{ Electronic supplementary information (ESI) available: Electronic
spectra and UV–vis–NIR absorbance maxima for Ni++ solutions. See
http://www.rsc.org/suppdata/jm/b4/b418233b/
*andre.laschewsky@iap.fhg.de (Andre´ Laschewsky)
devillers@chim.ucl.ac.be (Michel Devillers)
1668 | J. Mater. Chem., 2005, 15, 1668–1676
This journal is ß The Royal Society of Chemistry 2005

View Article Online
poly(ampholyte) matrix.13 The polymer used in that previous
work, namely poly(N,N-diallyl-N-hexylamine-alt-maleic acid),
was prepared by multistep synthesis and a complex purifica-
tion procedure to avoid the presence of residual inorganic salts
in the structure.14 We describe in this paper further investiga-
tions carried out in order to optimize this approach, and make
it easier to implement in the context of catalysts design and
preparation. Because the organic polymer used serves as a
sacrificial matrix to prepare oxides, the purpose is to develop a
very facile synthetic route to an appropriate organic polymer
matrix, which is able to stabilize efficiently cationic as well
as anionic metal species. This was achieved by choosing a
commercial cationic monomer together with a functional,
most easily accessible maleamic acid. Different from our first
approach,14 the polymer is made now from a hydrolytically
stable quaternized, i.e. permanently cationic, ammonium
monomer which enables strong electrostatic interactions
with anions at any pH value. Also, the polymer behaves as a
polyzwitterion in a large pH window. Additionally to the
betaine groups, the new polymer prepared exhibits a piper-
azine moiety in its repeat unit, which is known to be efficient
for cation complexation, either if present in the side group or
in the main chain of a macromolecule.15–17 Because the
influence of an additional complexing group on the stabiliza-
tion of inorganic species in the polymer–salt blends was not
clear from the literature, this point was elucidated. A detailed
description of the interactions between the organic matrix
and the inorganic species in solution as well as in the solid
state is given, in order to propose a realistic model for the
incorporation of cationic as well as anionic metal species in
the polymer blend. Our strategy is broadened to the prepara-
tion of various transition metal molybdates (Mn, Ni, Co) that
can be used as catalysts in selective oxidations.18,19 Finally, we
were interested in checking the influence of different experi-
mental parameters on the nature of the final phase that is
obtained, as the nature of the polymer matrix, the relative
amount of metal and polymer and the pH of the precursor
solutions.
2.2. Methods
1
Standard H- and 13C-NMR spectra were taken on a Varian
Gemini 200 MHz spectrometer. 13C experiments (APT,
(semi)quantitative spectra) were performed on a 500 MHz
spectrometer (Bruker, Avance 500). 1H chemical shifts were
calibrated to the residual signals of the solvents for D2O
(4.63 ppm) and CD3OD (3.3 ppm). For 13C-NMR spectra,
sodium 2,2-dimethyl-2-silapentane-5-sulfonate (DSS) was
employed as external reference for chemical shifts in D2O,
and residual proton signal from CD3OD (49.1 ppm) was used
as internal standard. FTIR spectra of the products in KBr
pellets (1 wt.%) were recorded on a Biorad FTS135 spectro-
meter. Elemental analysis was carried out at University
College, London, UK. ICP analysis was performed at the
analytical laboratory of the Fraunhofer-Institute for Applied
Polymer Research. The TGA curves were recorded under air
atmosphere on a Mettler-Toledo 851e thermogravimetric
analyser, applying a heating rate of 10 uC min21. Solution
UV–vis–NIR measurements were performed on a Varian Cary
5E. Powder X-ray diffractograms (XRD) were taken with a
Siemens D5000 diffractometer, using the Cu Ka line (l =
0.1541 nm). Raman spectra were taken on a Bruker RFS100/S
spectrometer, at a wavelength of 1064 nm. Scanning electron
microscopy (SEM) was performed with a Gemini DSM
982 digital scanning microscope. The hybrid materials were
successfully analyzed by deposition of Cr at the surface of the
samples using a plasma, in order to increase their conductivity.
2.3. Synthesis of comonomer and polymerization
(49-Methyl-piperazinyl)-4-oxo-2-butenoic acid. 20.03
g
(200.0 mmol) of N-methyl piperazine were added dropwise
to a solution of 19.62 g (200.1 mmol) of maleic anhydride in
20 ml of acetone at 0 uC. A precipitate formed rapidly. The
solvent was evaporated, the residue was dissolved in 100 ml
of a mixture of ethanol–water [70 : 30 (v : v)]. After several
precipitations into pentane, the precipitate was isolated by
filtration, and dried overnight in vacuo at 50 uC, to give 34.9 g
(yield: 88%) of white powder. 1H-NMR (200 MHz, CD3OD): d
(ppm): 6.20 (d; 1H, LCH–COOH); 6.05 (d; 1H, –N–C(LO)–
CHL); 3.85–3.60 (m; 4H, –C(LO)–N–CH2–); 3.26–3.06 (m;
4H, –C(LO)–N–C–CH2–N); 2.77 (s; 3H, –CH3). 13C-NMR
(50 MHz, CD3OD): d (ppm): 171.8 (–COOH); 170.6 (–N–
C(LO)–); 132.8 (LCH–COOH); 131.6 (–N–C(LO)–CHL); 53.6
(–CH2–N–CH3); 44.9 (–C(LO)–N–CH2–, cis-position of the
amide); 43.7 (N–CH3); 39.8 (–C(LO)–N–CH2–, trans-position
of the amide). MS (APCI): 199.2 [M+1]+.
2. Experimental
2.1. Materials
Solvents used, namely n-pentane (Fischer Scientific) and
methanol (Fischer Chemicals), were analytical reagent grade.
Concentrated hydrochloric acid (37 wt.% in water) was
supplied from Fisher Scientific. Diallyldimethylammonium
chloride (DADMAC, 65 wt.% in water) was purchased from
Aldrich. Maleic anhydride was supplied from Fluka Chemika.
1-Methylpiperazine, succinic acid, cobalt(II) nitrate hexahy-
drate and ammonium heptamolybdate(VI) tetrahydrate were
purchased from Acros Organics and used without further
purification. Nickel(II) nitrate hexahydrate was supplied from
Janssen Chimica. Manganese(II) nitrate tetrahydrate was pur-
chased from Alfa Aesar. 2,29-Azobis[2-methyl-N-(2-hydroxy-
ethyl)propionamide] (VA086) was supplied from Wako. Water
used during synthesis and purification steps was first purified
using an Elgastat Maxima system (resistance: 18.2 MV).
Spectra/Por dialysis membranes (nominal cut-off: 2000 Da)
were employed for polymer purification.
Poly[N,N-dimethyl-N,N-diallylammmonium chloride-alt-
(49-methyl-piperazinyl)-4-oxo-2-butenoic acid] (inner salt).
105.8
g of (49-methyl-piperazinyl)-4-oxo-2-butenoic acid
(533.7 mol) were dissolved in 132.8 g of a 65 wt.% aqueous
solution of DADMAC (533.9 mol), by adding 50 g of water
and adjusting the pH value to 8 with 0.5 M aqueous
NaOH. After the adding of the initiator (VA086, 10 mol.%
to monomers), the homogeneous solution was degassed by six
freeze-and-thaw cycles, and reacted at 75 uC during 48 h. After
cooling, the crude solution was dialyzed, first in aqueous HCl
(pH = 2) for 7 days, and subsequently in pure water for 5 days,
This journal is ß The Royal Society of Chemistry 2005
J. Mater. Chem., 2005, 15, 1668–1676 | 1669

View Article Online
changing the solvent every day. The solution was lyophilized
1
both monomers even at low conversions, as indicative for a
to yield 31 g (18%) of a yellowish solid. H-NMR (300 MHz,
strictly alternating copolymerization. This is in agreement with
the scarce reports on the copolymerization of diallyldimethyl-
ammonium chloride with maleamic acids,20 and matches well
with the alternating copolymerization found for diallylammo-
nium salts or diallylamine derivatives and maleic acid.14,21,22
The extensive dialysis of the copolymers at neutral pH
produces copolymers that are free of inorganic ions, namely
of sodium and chloride, as carefully checked by elemental
analysis and ICP. This purification produced therefore a salt-
free zwitterionic copolymer of the carboxybetaine type, as was
considered important in order to avoid contamination of the
mixed organic and inorganic hybrid materials prepared from
this copolymer by incorporation of selected inorganic species.
Finally, the structure of the copolymer makes it well-suited for
interacting with inorganic compounds via electrostatic (due to
the presence of carboxylate and ammonium moieties), but
also via complexation interactions (due to the presence of the
N-methyl piperazine group).
D2O): d (ppm): 4.20–3.41 (4H; –N+–CH2–); 3.41–2.86 (10H,
N+–CH3 and C(LO)N–CH2); 2.86–0.80 (15 H; backbone,
H3C–N–CH2–). 13C-NMR, APT sequence (125MHz, D2O):
d (ppm): 183.2 (C(LO)OH); 176.8 (–C(LO)N–); 75.4, 73.8
(N+–CH2); 57.5, 55.3, 53.9 (N+–CH3); 47.5, 44.5, 43.6
(C(LO)N–CH2–CH2–); 46.7, 46.1, 45.0, 41.4, 41.0 (–CH–
backbone, –CH–C(LO), .N–CH3); 32.9, 29.2 (–CH2–
backbone). FTIR (KBr) selected bands (cm21): 3025, 2940,
2862, 2809 (nsym+asym CH3, CH2, CH); 1630 (n CLO in
C(LO)NR2), 1577 (nasym CLO in COO2), 1470 (dsym CH2 and
dasym CH3), 1394 (dsym CH3 and nsym CLO in COO2). Elem.
Anal. Calcd. for C17H29N3O3 (M = 324.20) C 63.11 H N
12.99 Cl 0 C/N ratio 4.853; Found C 54.70 H 8.51 N 11.27 Cl
0.01 C/N ratio 4.854. Na content (ICP): below 0.01%.
2.4. Preparation of the hybrid materials
In a typical experiment, 100 mg of polymer were dissolved
in 25 ml of water by stirring at room temperature. The
appropriate amounts of 0.1 M aqueous solutions (pH = 10)
of the inorganic species (Ni(NO3)2?6H2O, Mn(NO3)2?4H2O,
Co(NO3)2?6H2O and (NH4)6Mo7O24?4H2O) were added in
order to reach the desired molar ratio of metal ion(s) per
repeat unit of copolymer. The pH value was adjusted by
adding an appropriate amount of 0.1 M aqueous NH3 or
HNO3. The solution was stirred for 30 min, and finally
lyophilized to produce coloured, finely divided powders.
3.2. Solution studies of precursors
Preliminary miscibility tests in solution of the organic matrix
with the inorganic salts were carried out first. Ni(NO3)2?6H2O,
Mn(NO3)2?4H2O and Co(NO3)2?6H2O were incorporated
alone or together with (NH4)6Mo7O24?4H2O under determined
pH conditions, in presence of the polymer. The homogeneity
of the solutions was checked by a qualitative appreciation of
the transparency of the solution. The general trend indicates
that homogeneous solutions can be produced throughout the
whole pH range. To determine which kind of interactions
stabilize the organic polymer, bearing simultaneously charged
moieties and complexing groups, and the cationic inorganic
species in solution, UV–vis–NIR studies were undertaken,
using different kinds of hypothetical ligands for the Ni(II) and
Co(II) ions. Next to the polymer, the precursor monomer,
N-methyl piperazine and succinic acid were tested. For a given
pH value of about 10.4, a slight shift of the absorption
bands between a reference spectrum of Ni++ in presence of
ammonia on the one hand, and the spectra of Ni++ together
with N-methyl piperazine or (49-methylpiperazinyl)-4-oxo-2-
butenoic acid on the other hand, is observed (cf. Fig. S1 in the
electronic supplementary information (ESI){). This suggests
the existence of complexation interactions between the cation
and the amine moieties of the ligands. On the other hand, no
shift was found in the presence of succinic acid or the polymer
employed as hypothetical ligands, but this finding does not
indicate that no complexation occurs. In order to evidence a
possible influence of pH on the interactions between the matrix
and Ni++ in solution, spectra of the two species at different
pH values were recorded (cf. Table S1 in the ESI{). The
maxima of absorbance from the spectra of Ni++ in presence of
polymer, for different basic pH values in order to ensure
sufficient deprotonation of the amine group, were collected
and compared with the corresponding data from the reference
spectra. Slight shifts between reference and polymer solutions
spectra are observed at pH values of 9.2 and 9.8, suggesting the
occurrence of complexation interactions. On the other hand,
no complexation can be evidenced when the pH value is raised
3. Results and discussion
3.1. Synthesis of polymer matrix
The polymer was prepared in analogy to a literature pro-
cedure,20 by copolymerizing the commercially available diallyl-
dimethylammonium chloride with an unsaturated amino acid,
namely the maleamic acid that is prepared by ring-opening
addition of N-methyl piperazine onto maleic anhydride (Fig. 1).
Copolymer yields are typically in the range of 15–20%, but as
far as the monomers are really easily accessible, this point does
not constitute a major drawback in this context. Whereas
1H-NMR spectra are complex due to the superposition of
broad signals, 13C-NMR spectra of the copolymer are more
instructive, and show the signals of COO2, amide and
ammonium moieties in the structure. The presence of the
functional COO2 and amide moieties is corroborated by
characteristic signals in the FTIR spectra. Elemental analysis
indicates that the copolymer contains equimolar amounts of
Fig. 1 Facile synthetic route to the precursor monomer and
copolymer matrix.
1670 | J. Mater. Chem., 2005, 15, 1668–1676
This journal is ß The Royal Society of Chemistry 2005

View Article Online
or reduced (namely pH = 8.4, 10.4 or 11.6). The same
experiments were carried out using Co++, but the spectra did
not prove the existence of complexation in solution between
the polymer and the cation, at least at pH values from 2 to 9.2.
Precipitation of the chelate occurred if the pH was subse-
quently increased, which limited the obtained information.
Similar studies could not be carried out with the Mn++
solutions, because of the absence of any spin-allowed d–d
transition in high-spin octahedral d5 complexes.
signals appears, due to the presence of a paramagnetic species
which dramatically diminishes the relaxation times of all
nuclei. As a consequence, the general trend from these UV–
vis–NIR and NMR studies suggests that the interactions
between the cations and the polymer in a basic solution are
essentially of electrostatic nature, even if additional complexa-
tion involving the piperazine group is thought to take place
under some conditions.
In addition, solution 13C-NMR has been implemented in
the case of the Ni++ ion, in order to get complementary
information which can not be obtained from UV–vis spectro-
scopy, namely about the nature of the interactions between
the carboxylate moiety and the cation. Fig. 2 compares the
13C-NMR spectra of the prepared monomer and copolymer
with the spectra of the same species in the presence of
Ni(NO3)2?6H2O, at a fixed pH value of 10. Spectra of the
monomer together with Ni++ exhibit a shift downfield of all
peaks in the spectrum, compared with the spectrum of the salt-
free monomer, combined with a broadening of the signals at
178 ppm and 141 ppm, attributed to amide and carboxylate
and the double bond. This means that the paramagnetic Ni++
species preferably interacts in aqueous solution with these
groups, via complexation interactions in the case of COO2.
In the same way, the peak at 183 ppm, attributed to the
carboxylate moiety in the copolymer, broadens when 0.5 mol
of metal per mol of repeat units of the polymer is added,
compared with the spectrum of the salt-free copolymer. Here,
no significant shift of the other peaks in the spectrum can be
evidenced, which suggests that the major interactions between
Ni++ and the matrix are probably mainly electrostatic. If the
amount of salt is increased up to equimolar quantities of metal
cation and repeat units of the polymer, broadening of all the
3.3. Solid state studies of precursors
In order to prepare homogeneous hybrid mixtures made from
the synthetic polymer and the selected inorganic salts, a
systematic screening of the homogeneity of the materials in the
solid state was carried out using powder XRD. Test samples
were at first prepared by mixing and grinding in the solid state
the polymer with each of the inorganic salts, in a molar ratio
of repeat units of polymer relatively to the inorganic metals
of 1 to 0.1, to get powders. Sharp lines which indicate the
presence of phase-separated inorganic salts were observed in
the diffractograms of these powders, i.e. the mixtures were not
homogeneous. This finding demonstrates also that XRD
measurements are appropriate to check the eventual inhomo-
geneity of the samples, even if only a small amount of
inorganic species does not sufficiently interact with the
polymer matrix. In order to obtain homogeneous mixtures,
hybrid materials were prepared in solution, lyophilized and
analyzed by XRD. Fig. 3 compares the XRD pattern of
the polymer with the diffractograms of hybrid materials
made from the polymer and Co(NO3)2?6H2O together with
(NH4)6Mo7O24?4H2O, obtained from solutions prepared at
different pH values, and for different amounts of salts
incorporated. The results indicate that the materials prepared
from aqueous solutions at a pH value of 10 are homogeneous
when the total molar amount of inorganic salts introduced
Fig. 2 (a) Semi-quantitative 13C NMR spectra in D2O (pH = 10) of
(49-methyl-piperazinyl)-4-oxo-2-butenoic acid (1), (49-methyl-piperazi-
nyl)-4-oxo-2-butenoic acid in the presence of Ni(NO3)2?6H2O (molar
ratio: 1 mol of ligand for 0.5 mol of metal cation) (2), the salt-free
copolymer (3), and the copolymer in presence of Ni(NO3)2?6H2O
(molar ratios: 0.5 mol (4) and 1 mol (5) of metal cation per mol of
repeat units of the copolymer); (b) magnified parts (spectra 6, 7 and 8)
of spectra 3, 4 and 5, respectively, illustrating the carbonyl signals from
amide and carboxylic moieties.
Fig. 3 XRD patterns of the polymer (a) and hybrid blends made
from the polymer and different amounts of Co(NO3)2?6H2O and
(NH4)6Mo7O24?4H2O and prepared at different pH values: from (b) to
(e): pH = 2; 0.25, 0.5, 1 and 2 mol of each metal ion per mol of repeat
units of the copolymer; from (f) to (j): pH = 10; 0.25, 0.5, 1, 1.5 and
2 mol of each metal ion per mol of repeat units of polymer.
This journal is ß The Royal Society of Chemistry 2005
J. Mater. Chem., 2005, 15, 1668–1676 | 1671

View Article Online
is up to 1.5 mol of each metal per mol of repeat units of
polymer. A pH value of 2 is not suitable for obtaining
homogeneous materials, unless only 0.25 mol of each metal is
incorporated. This is most probably due to the high degree of
protonation of the carboxylate moiety under these conditions,
but also to the presence of bulky Mo(VI) species such as
[Mo7O24]62 or [Mo8O26]42 under these acidic conditions.23
The difficulty of these bulky species to interact with the matrix
probably contributes to the loss of homogeneity. Attempts
have been made to identify the phase containing the residual
inorganic salt which did not interact with the polymer, by
comparison of the XRD patterns with JCPDS files of the
precursor inorganic materials, and by using Raman spectro-
scopy, but without any success. Globally, the general trend
demonstrates that 1 mol of cationic as well as anionic metal
species per mol of repeat units of polymer can be easily
blended under basic conditions. This is due to the presence of
the carboxylate and ammonium moieties in the matrix
structure, and is in agreement with previous observations on
similar systems.13,24 However, 0.5 additional mol of metal
species per mol of repeat units of the polymer can be blended
in the matrix to produce homogeneous materials. This result
indicates that, in addition to the carboxylate group, another
moiety takes part in the subsequent stabilization of Co(II) in
the solid state. Next to the carboxylates, the amido or more
probably amino groups from the piperazine moieties probably
interact with the cationic inorganic species. This is in line with a
similar observation in the literature25 concerning the auxiliary
stabilization in the solid state of low molar mass salts in
conventional poly(betaine) systems whose structure exhibits
atoms bearing lone electron pairs. For example, it has been
shown that poly(sulfobetaine)s based on poly(acrylamide) allow
higher salt contents than the analogous poly(methacrylate). In
this case, the excess of anionic molybdates are assumed to
rearrange themselves around positively charged entities in the
system, which produces a homogeneous blend. As expected,
the structure of the polymer has indeed a major effect on the
mixing behaviour with inorganic salts. Globally, however,
when 1 mol of each metal is incorporated in the matrix under
basic conditions, a simple model based on electrostatic
interactions between the inorganic species (anionic as well as
cationic) and the charged moieties of the polymer permits the
explanation of the formation of homogeneous materials in
the solid state (Fig. 4). Additionally, SEM was implemented
to look for eventual correlations between the (in)homogeneity
of the materials and the morphology of the particles. Fig. 5
illustrates typical micrographs of (in)homogeneous materials.
Spherical particles and aggregates of sphere-like particles are
mainly detected on the micrographs of the materials prepared
Fig. 5 Typical SEM micrographs (magnification: 1006) of homo-
geneous (a) and inhomogeneous (b) materials made from the polymer
and Ni(NO3)2?6H2O incorporated jointly with (NH4)6Mo7O24?4H2O;
pH of the precursor solution: 10 (a) and 4 (b).
under basic pH conditions. However, the morphology of the
particles is quite different in the case of materials prepared
under acidic conditions: next to the same sphere-like particles
isolated for the materials prepared under basic conditions,
particles exhibiting various shapes are also typically present
in the micrographs. This is a further piece of evidence of the
formation of a separate phase under acidic conditions, leading
to an inhomogeneous material.
3.4. Thermal behaviour of the precursors
Thermal decomposition of the organic matrix and formation
of the oxide phases were followed by thermogravimetric
analysis (TGA). Fig. 6 displays the TGA curves of hybrid
materials made from Ni(NO3)2?6H2O, Co(NO3)2?6H2O and
(NH4)6Mo7O24?4H2O. All curves exhibit a first weight loss
between 50 uC and 100 uC, which is attributed to dehydration
of the material. This is followed by different steps correspond-
ing to the pyrolysis of the matrix above 200 uC, which is
combined with the degradation of the nitrate counter-ions
and the formation of the expected oxide phase. The weight
becomes stable from y570 uC, and up to 850 uC. The pyrolysis
of the matrix is nearly complete, as the residual weights match
well with the values calculated on the basis of the formation of
the corresponding molybdates, which indicates, as expected,
that the main degradation products consist of the pure
simple or mixed Ni and/or Co molybdates which are free of
contaminants. The residual C, H and N contents in the final
Fig. 4 Schematic representation of the incorporation of anionic as
well as cationic metal species in a zwitterionic polymer, leading to the
formation of a homogeneous hybrid material.
1672 | J. Mater. Chem., 2005, 15, 1668–1676
This journal is ß The Royal Society of Chemistry 2005

View Article Online
Fig. 7 Typical X-ray diffractograms of the simple and mixed Ni and/
or Co molybdates (a) and Mn and/or Co molybdates (b). The star
denotes the position of the (220) reflection, corresponding to a- and/or
b-phase.
Fig. 6 TGA curves under air of different hybrid materials containing
Ni(NO3)2?6H2O and (NH4)6Mo7O24?4H2O (0.5 mol of metal ions
per mol of repeat units of the polymer), whose degradation products
correspond to the simple or mixed Ni and/or Co molybdates.
a pH value of 7, and whose composition corresponds to 0.5 mol
of metals per mol of repeat units of copolymer. The nature of
the molybdate phase (a-, b- or mixture) was determined as
described previously,26–31 by the position of specific XRD lines
in the range 20–36u, referring to the corresponding JCPDS files
(a-NiMoO4: file 33-0948, b-NiMoO4: file 45-0142, b-CoMoO4:
file 21-0868). The evolution of the diffractograms as a function
of the relative amount of Ni++ and Co++ for the different series
of nickel and cobalt molybdates is quite similar. Fig. 8a displays
the evolution of the corresponding interreticular spacings
calculated from the positions of the (220) reflections as a
function of the Ni and Co composition in Ni12xCoxMoO4. A
mixture of the a-phase (mainly) and the b-phase of NiMoO4
on the one hand, and the pure b-CoMoO4 on the other, are
obtained in the case of the two simple oxides, whatever the pH
studied and the amount of salt introduced. The mixed nickel
and cobalt molybdates are identified as mixtures of the a- and
b-phases (x = 0.25 and 0.5 for pH values of 7 and 10; x = 0.25,
0.5 and 0.75 for a pH value of 12) or as the pure b-phase
(x = 0.75 at a pH value of 7; x = 0.75 and 0.90 for a pH
value of 10), whatever the amount of inorganic species
introduced in the precursor. Vegard’s law appears to be
respected in two different composition domains. Subsequently,
shifts in the d(220) values calculated from the peaks’ positions
in the diffractograms are observed going from a-NiMoO4
to a-Ni0.5Co0.5MoO4 (pH = 7, 10) or from a-NiMoO4 to
a-Ni0.25Co0.75MoO4 (pH = 12), and from b-NiMoO4 to
products were checked by elemental analysis (Table 1). The C
content is typically below 0.5%, which demonstrates that the
method is efficient to produce inorganic materials really free of
organic contaminants.
3.5. Preparation and characterization of the final molybdate
phases
According to the thermal behaviour of the hybrid precursors,
inorganic materials were prepared upon calcination of the
precursors under an air flow of 1 L min21 at 600 uC, namely
about 30 uC higher than the final decomposition temperature
of the hybrid blends determined by TGA, in order to ensure
the complete pyrolysis of the organic matrix and favour the
formation of the crystalline phases.
3.5.1. Simple and mixed nickel and cobalt molybdates.
Different sets of simple and mixed nickel and cobalt
molybdates with the general formula Ni12xCoxMoO4, where
x = 0, 0.25, 0.5, 0.75, (0.9) and 1, were prepared according to
the developed method. Systematic preparations were per-
formed in order to check the influence of the pH of the
precursor solution and the amount of inorganic salt introduced
under defined pH conditions on the nature of the formed oxide
phase. Fig. 7a illustrates the XRD patterns of a series of final
oxide materials, obtained from precursor solutions prepared at
b-CoMoO4. This behaviour is consistent with
a lattice
expansion of the molybdate network with increasing Co
content, reflected by the linear increase of the d(220) spacing
with the Co–Ni ratio across each region, and is in agreement
with the increase of the ionic radius of the six-fold coordinated
ions, going from Ni++ to Co++. Since both the cell dimensions
and the relative intensities vary linearly with the Ni–Co atomic
ratio, application of Vegard’s law led to the conclusion that
the structures of the mixed nickel–cobalt molybdates in each
domain are solid solutions of a-NiMoO4 and a-CoMoO4 on
Table 1 Residual C, H and N contents for some prepared simple and
mixed Ni and/or Co molybdates.
Sample
% C
% H
% N
NiMoO4
0.23
0.22
0.34
0.21
0.35
0
0
0
0
0
0
Ni0.75Co0.25MoO4
Ni0.5Co0.5MoO4
Ni0.25Co0.75MoO4
CoMoO4
0.03
0.05
0.03
0.06
This journal is ß The Royal Society of Chemistry 2005
J. Mater. Chem., 2005, 15, 1668–1676 | 1673

View Article Online
Fig. 9 Typical Raman spectra of the simple and mixed Ni and/or Co
molybdates (a) and Mn and/or Co molybdates (b).
are in perfect accordance with those obtained by XRD
measurements. In conclusion, the pH of the precursor solution
and the amount of metal introduced under defined pH
conditions have, under the chosen preparation conditions,
little influence on the nature of the generated phase of the
molybdates.
3.5.2. Simple and mixed manganese and cobalt molybdates.
The same procedure was applied to the preparation of simple
and mixed manganese and cobalt molybdates, corresponding
to the formulation Mn12xCoxMoO4, where x = 0, 0.25, 0.5,
0.75 and 1. The materials were obtained from precursor
solutions prepared at a pH value of 10, and of relative
composition of 0.5 mol of metals per mol of repeat units of
copolymer. The X-ray diffractograms (Fig. 7b) demonstrate
that the simple CoMoO4 and the mixed Mn and Co
molybdates are obtained in the form of their b-phase. The
structure of the simple MnMoO4 was identified in the same
way in agreement with the JCPDS data (file 82-2166),35 to be
the a-MnMoO4, which is in fact isomorphous to b-NiMoO4
and b-CoMoO4.27 Fig. 8b displays the evolution of the
interreticular spacing d(220) as a function of the composition
of Mn12xCoxMoO4. Vegard’s law is respected through the
whole composition range. Subsequently, shifts in the d(220)
values are observed going from a-MnMoO4 to b-CoMoO4. In
this case, the behaviour is consistent with a lattice contraction
of the molybdate network with increasing Co content, reflected
by the linear decrease of the d(220) spacing with increasing
Co–Mn ratio, and is in agreement with the decrease of the
six-fold coordinated ionic radii from Mn++ to Co++. Since
both the cell dimensions and the relative intensities vary
linearly with the Mn–Co atomic ratio, application of Vegard’s
law led in this case to the conclusion that the structures of the
mixed manganese–cobalt molybdates are solid solutions of
a-MnMoO4 and b-CoMoO4 in a common molybdate lattice.36
Again, Raman spectroscopy confirmed the nature of the
phases. Spectra of the materials are given at Fig. 9b. The
spectrum of the pure a-MnMoO4 exhibits well-defined lines at
Fig. 8 Verification of Vegard’s law on the basis of the interreticular
distances corresponding to (220) reflections for different sets of simple
and mixed molybdates, as a function of the experimental conditions:
(a) Ni12xCoxMoO4 phases prepared at different pH values of the
precursor solution (pH = 7, 10 and 12) and from precursors containing
different amounts of metal ions (0.5 and 1 mol of metal ions per mol of
repeat units of the copolymer); (b) Mn12xCoxMoO4 phases prepared
at a pH value of the precursor solution of 10 and from precursors
containing 0.5 mol of metal ions per mol of repeat units of the
copolymer.
the one hand, and b-NiMoO4 and b-CoMoO4 on the other, in
a common molybdate lattice. In addition to the XRD
measurements, the materials were investigated by Raman
spectroscopy, in order to confirm the nature of the phases. The
Raman spectra of the materials prepared at a pH value of 10
and from precursors containing 0.5 mol of metals per mol of
repeat units of polymer are given at Fig. 9a. These data were
analyzed with the help of literature data.32–35 NiMoO4 is
identified by the presence of lines at 961 cm21, 713 cm21 and
705 cm21 on the spectra, corresponding to the so-called
a9-phase, which is a distorted form of a-NiMoO4, obtained
due to the grinding of the sample. Spectra of the mixed
molybdates also exhibit intense lines at around 700 cm21
,
which are normally typical of the a-phase. In the same way, the
absence of this band in the spectrum of CoMoO4 suggests that
the b-phase is produced. Finally, the two neighbour lines at
about 940 cm21 resulting in a poorly-resolved broad band,
reveal the occurrence of the b-phase. As a whole, these results
1674 | J. Mater. Chem., 2005, 15, 1668–1676
This journal is ß The Royal Society of Chemistry 2005

View Article Online
942 cm21, 931 cm21, 882 cm21, 822 cm21, 354 cm21, 323 cm21
and 281 cm21, which is in perfect agreement with the literature
data.37 Spectra of the mixed Mn and Co molybdates show
typical lines in the range 936 cm21–932 cm21 (two neighbour-
ing lines resulting in a broad band with poor resolution),
and by a Concerted Research Action of the ‘‘Communaute´
Franc¸aise de Belgique’’.
Franc¸ois Rullens,a Nicolas Deligne,a Andre´ Laschewsky*bc and
Michel Devillers*a
aUnite´ de Chimie des Mate´riaux Inorganiques et Organiques, Universite´
catholique de Louvain, Place Louis Pasteur, 1/3, B-1348
Louvain-la-Neuve, Belgium. E-mail: devillers@chim.ucl.ac.be;
Fax: +32 10 47 23 30; Tel: +32 10 47 28 27
882 cm21–878 cm21, 823 cm21–817 cm21, 364 cm21–358 cm21
,
333 cm21–326 cm21 and 291 cm21–278 cm21
.
bFraunhofer-Institute for Applied Polymer Research, Geiselbergstraße,
69, D-14476 Potsdam-Golm, Germany.
4. Conclusions
E-mail: andre.laschewsky@iap.fhg.de; Fax: +49 331 568 3000;
Tel: +49 331 568 1327
An unconventional synthetic approach, based on the use of
hybrid precursors made from inorganic salts and a water-
soluble organic polymer, was applied to the preparation of
simple and mixed transition metal molybdates. Because the
polymer serves as sacrificial matrix, a facile synthetic route
to an appropriate polymer was developed. Hybrid precursor
blends were prepared from aqueous solutions, the polymer
stabilizing the inorganic species generated in solution by
the incorporation of Ni(NO3)2?6H2O, Mn(NO3)2?4H2O,
Co(NO3)2?6H2O and (NH4)6Mo7O24?4H2O under the appro-
priate conditions. The interactions in solution between the
polymer matrix and the inorganic species (anionic as well as
cationic) are mainly of an electrostatic nature, even if com-
plexation can be evidenced under some determined conditions.
Solid homogeneous organic–inorganic hybrid blends can be
prepared from these solutions, incorporating up to 1.5 mol of
each metal (anionic as well as cationic species) per mol of
repeat unit of the copolymer. This finding clearly demonstrates
the advantage of using a copolymer bearing an additional
complexing group as organic matrix, compared to previously
used standard poly(betaine) systems. Subsequent calcination
of the homogeneous mixtures under air produces the
corresponding simple and mixed nickel, cobalt and/or manga-
nese molybdates, in the form of their a- and/or b-phases. The
mixed structures are solid solutions of the simple NiMoO4,
CoMoO4 and MnMoO4. As far as the mixtures considered
are homogeneous in the solid state, the nature of the polymer
matrix, the pH of the precursors solution and the relative
amount of salt per repeat unit of copolymer seem to have
no major effect on the nature of the final phase produced.
This result underlines the versatility of the method.
Remarkably, the final materials exhibit a very low C content,
in spite of using a C-rich organic precursor. Therefore,
the approach described here appears as a widely applicable,
rather general, method for the preparation of multicomponent
oxide-type materials of catalytic or whatsoever interest.
It presents an attractive alternative synthetic strategy to
materials that could not be prepared easily via precipitation,
sol–gel or other more classical precursor methods. In
addition, compared to other non-conventional methods, the
present route offers the advantage of proceeding in aqueous
solution.
cUniversita¨t Potsdam, Institut fu¨r Chemie, Karl Liebknecht-Straße,
24-25, D-14476 Potsdam-Golm, Germany.
E-mail: laschews@rz.uni-potsdam.de; Fax: +49 331 977 5054;
Tel: +49 331 977 5225
References
1 M. Alifanti, B. Baps, N. Blangenois, J. Naud, P. Grange and
B. Delmon, Chem. Mater., 2003, 15, 395–403.
2 H. Yamamura, A. Watanabe, S. Shirasaki, Y. Moriyoshi and
M. Tanada, Ceram. Int., 1985, 11, 17–22.
3 H. Wullens, D. Leroy and M. Devillers, Int. J. Inorg. Mater., 2001,
3, 309–321.
4 H.-B. Park, H.-J. Kweon, Y.-S. Hong, S.-J. Kim and K. Kim,
J. Mater. Sci, 1997, 32, 57–65.
5 J. C. W. Chien, B. M. Gong, Y. Yang, Y. Cabrera, J. Effing and
H. Ringsdorf, Adv. Mater., 1990, 2, 305–309.
6 K. Naka, Y. Tanaka, K. Yamasaki, A. Ohki, Y. Chujo and
S. Maeda, Bull. Chem. Soc. Jpn., 2001, 74, 571–577.
7 G. Ca´rdenas and J. Acuna, Int. J. Polym. Mater., 1995, 29,
147–155.
8 M. Galin, E. Marchal, A. Mathis, B. Meurer, Y. M. Monroy Soto
and J. C. Galin, Polymer, 1987, 28, 1937–1944.
9 P. Ko¨berle and A. Laschewsky, Macromolecules, 1994, 27,
2165–2173.
10 P. Favresse and A. Laschewsky, Polymer, 2001, 42, 2755–2766.
11 N. Bonte, A. Laschewsky, B. Mayer and V. Vermylen, Macromol.
Symp., 1996, 102, 273–280.
12 Ph. Favresse and A. Laschewsky, Macromol. Chem. Phys., 1999,
200, 887–895.
13 F. Rullens, M. Devillers and A. Laschewsky, J. Mater. Chem.,
2004, 14, 3421–3426.
14 F. Rullens, M. Devillers and A. Laschewsky, Macromol. Chem.
Phys., 2004, 205, 1155–1166.
15 S. Kumaresan and P. Kannan, Indian J. Chem. Technol., 2003, 10,
180–185.
16 P. Ferruti, M. A. Marchisio and R. Barbucci, Polymer, 1985, 26,
1336–1348.
17 L. H. Gan, N. K. Goh, B. Chen, C. K. Chu, G. Roshan Deen and
C. H. Chew, Eur. Polym. J., 1997, 33, 615–620.
18 D. L. Stern and R. K. Grasselli, J. Catal., 1997, 167, 550–559.
19 L. M. Madeira, M. F. Portela and C. Mazzocchia, Catal. Rev.,
2004, 46, 53–110.
20 A. F. Thu¨nemann, K. Sander, W. Jaeger and R. Dimova,
Langmuir, 2002, 18, 5099–5105.
21 M. Hahn, W. Jaeger, R. Schmolke and J. Behnisch, Acta Polym.,
1990, 41, 107–112.
22 M. Hahn, J. Ko¨tz, K.-J. Linow and B. Philipp, Acta Polym., 1989,
40, 37–43.
23 E. R. Braithwaite and J. Haber, Molybdenum: An Outline of its
Chemistry and Uses, Elsevier, Amsterdam, 1994.
24 P. Favresse and A. Laschewsky, Macromol. Chem. Phys., 1999,
200, 887–895.
25 N. Bonte, A. Laschewsky and V. Vermylen, Macromol. Symp.,
1997, 117, 195–206.
26 T. M. Yanushkevich, V. M. Zhukovskii and V. M. Ust’yantsev,
Russ. J. Inorg. Chem., 1974, 19, 1056–1058.
27 T. M. Yanushkevich, V. M. Zhukovskii and V. M. Ust’yantsev,
Russ. J. Inorg. Chem., 1976, 21, 417–420.
28 J. A. Rodriguez, S. Chaturvedi, J. C. Hanson, A. Albornoz and
J. L. Brito, J. Phys. Chem. B, 1998, 102, 1347–1355.
Acknowledgements
The contribution of D. Chapon (13C-NMR measurements) is
greatly acknowledged. FR thanks FNRS (Fonds National de
la Recherche Scientifique, Brussels) for a FRIA fellowship.
The work was supported in parts by the grant INTAS-00-113,
This journal is ß The Royal Society of Chemistry 2005
J. Mater. Chem., 2005, 15, 1668–1676 | 1675

View Article Online
29 A. Maione and M. Devillers, J. Solid State Chem., 2004, 177,
2339–2349.
33 M. Woydt, A. Skopp, I. Do¨rfel and K. Witke, Wear, 1998, 218,
84–95.
30 D. Vie, E. Martinez, F. Sapina, J.-V. Folgado, A. Beltran,
R. X. Valenzuela and V. Corte´s-Corberan, Chem. Mater., 2004,
16, 1697–1703.
34 S. Sheik Salem, Infrared Phys., 1987, 27, 309–315.
35 G. Lautenschla¨ger, H. Weitzel and H. Fuess, Z. Kristallogr., 1994,
209, 936–940.
31 A. Kaddouri, E. Tempesti and C. Mazzocchia, Mater. Res. Bull.,
2004, 39, 695–706.
36 E. I. Get’mana and V. I. Marchenko, Russ. J. Inorg. Chem., 1986,
31, 1207–1210.
32 E. Payen, P. Dhamelincourt, J. Dhamelincourt, J. Grimblot and
J. P. Bonnelle, Appl. Spectrosc., 1982, 36, 30–37.
37 U. S. Ozkan, M. R. Smith and S. A. Driscoll, J. Catal., 1992, 134,
24–35.
1676 | J. Mater. Chem., 2005, 15, 1668–1676
This journal is ß The Royal Society of Chemistry 2005