G. Trautwein et al. / Journal of Catalysis 331 (2015) 110–117
111
oxidation reactions and/or to the over-oxidation of the nitroso
2.2. Catalytic tests: aniline oxidation
derivative formed [8].
Polyoxometalates (POMs) are anionic early transition metal-oxo
clusters of nanometric size that show several relevant properties to
fields such as catalysis, material science, biomedicine and nan-
otechnology. From the catalytic point of view, POMs can act as
both strong Brønsted acids (stronger than mineral acids) and oxi-
dation catalysts showing fast and reversible redox reactions with-
out any structural change [9–12]. In addition, POMs also display
good thermal stability and low cost and they are easily recoverable
from the reaction media in high amounts. Compared to mineral
acids, these well-known clusters offer remarkable environmental
and economic advantages derived from their more efficient and
cleaner catalytic processes, while they do not decompose or deac-
tivate in the presence of water in contrast to what is commonly
observed for the Lewis acidic chloride metals. Among the vast
POM family, lanthanide-containing species constitute one of the
largest and most active groups at present because combination of
POM building blocks with rare-earth metals has been shown to
be a powerful tool for designing new architectures and introducing
additional properties to the POM system [13–15]. With regard to
catalysis, the incorporation of rare-earth metals into the POM
framework allows the resulting cluster to act as a Lewis acid cata-
lyst [16–19]. Moreover, combination of Lewis acid lanthanide sites
and nucleophilic POM surfaces with Lewis base character can
result in bifunctional catalysts suitable for cyanosilylation or oxi-
mation reactions owing to their ability to simultaneously activate
complementary substrates [20,21].
Considering the above, we decided to systematically explore the
use of lanthanide(III)-containing POMs (Ln-POMs) as heteroge-
neous catalysts for the selective oxidation of organic substrates
with H2O2 in place of the traditional mineral or Lewis acids to
avoid the acidic waste drain produced by the latter. More specifi-
cally, we have first focused our studies on those Ln-POMs showing
dilacunary Keggin-type frameworks because, unlike plenary Keg-
gin heteropolyacids or monolacunary Ln-POMs, they have been
scarcely applied in catalytic reactions, and hence their potential
applications in this area are still to be developed. Hereby, we report
our results on the heterogeneous catalytic activity towards the
organic phase oxidation of aniline with H2O2 of the sodium salts
of four isomorphic dimeric anions composed of lanthanide-
stabilised dilacunary Keggin subunits with b-type skeletons,
The oxidation of aniline was conducted in a 10 mL glass flask.
The following chemicals were successively introduced: 35 L of
aniline, 85 L of hydrogen peroxide (30%), 4.9 mL of dichlor-
oethane and 2 mol of the corresponding catalyst previously dried
l
l
l
at 100 °C (15 mg of Na-bb-Ln4; 4 mg of the heteropolyacids
H3PMo12O40 or H3PW12O40). The optimum amount of catalyst
was determined as shown in Fig. A.2 in the Supplementary Infor-
mation. Initial experiments were performed using fresh catalysts;
for subsequent tests, the catalysts were previously treated with a
hydrogen peroxide solution (30%) for 1.5 h. After adding a small
Teflon magnet, the mixture was stirred at room temperature for
9 h. Both shorter (0.5 h) and longer reaction times (up to 24 h)
were also tested for comparative purposes. Regarding the solvent,
preliminary tests demonstrated that dichloroethane leads to yields
higher than those obtained with other conventional solvents such
as chloroform, dichloromethane, toluene and dimethylsulphoxide.
The solvent seems to play an important role both stabilising the
dimeric POM framework against dissociation or decomposition
and contributing with protons to the development of the reaction.
Polar solvents should also provide protons to the reaction medium,
but unfortunately, they work in favour of the catalyst solubilisa-
tion, and hence we limited our studies to a non-polar solvent in
this work. To determine the performance of the heterogeneous cat-
alysts in consecutive runs, they were recovered from the mixture
by filtration upon completion of the reaction time and dried at
100 °C for 24 h before being reused in a new reaction process.
The amount of catalyst solubilised during the reaction was esti-
mated from the weight losses between consecutive runs. Analo-
gous experiments were also performed using catalysts previously
calcined at 200 °C in air.
2.3. Characterisation
For the preliminary identification of the freshly prepared
Na-bb-Ln4 compounds, powder X-ray diffraction (PXRD) patterns
were collected from 2h = 4 to 60° (0.02606° step size, 30 s per step)
using a Philips X’PERT PRO automatic diffractometer operating at
40 kV–40 mA in h–h configuration with monochromated Cu K
a
radiation (k = 1.5418 Å) and a PIXcel solid state detector (3.347°
active length in 2h). Infrared spectra between 400 and 4000 cm–1
were obtained on a SHIMADZU FTIR-8400S spectrometer operating
in transmittance mode using KBr pellets (20 scans per spectrum,
resolution of 4 cm–1).
namely
Na12[Ln4(H2O)6(b-GeW10O38)2]Áꢀ44H2O
(Na-bb-Ln4,
Ln = Dy, Ho, Er, Tm) [22]. The reaction process is represented in
Scheme 1.
The results of the elemental analyses for the title Na-bb-Ln4 cat-
alysts are listed in Table B.1 in the Supplementary Information.
They show metal contents very close to the expected theoretical
values in all cases. The experimental Ge/Ln/Na ratios are in good
agreement with the exclusive presence of lanthanide-containing
dilacunary Keggin subunits and indicate that no significant decom-
position takes place during the oxidation process.
Thermogravimetric (TG) analyses were also performed in order
to determine the degree of hydration and thermal stability.
They were carried out from room temperature to 700 °C at a rate
of 5 °C min–1 using a TA Instruments 2960 SDT thermobalanceunder
a 100 cm3 min–1 flow of air (Fig. A.3 in the Supplementary Informa-
2. Experimental
2.1. Materials
All starting materials were purchased from commercial sources
and used without further purification. The Na12[Ln4(H2O)6(b-
GeW10O38)2]Áꢀ44H2O (Na-bb-Ln4, Ln = Dy, Ho, Er, Tm) POM
sodium salts used in this work were prepared according to the lit-
erature [22]. The specific synthetic procedure was as follows: to a
solution of the corresponding lanthanide chloride salt (0.98 mmol;
DyCl3Á6H2O, 0.369 g; HoCl3Á6H2O, 0.372 g; ErCl3Á6H2O, 0.374 g;
TmCl3Á6H2O, 0.376 g) in aqueous 0.5 M NaOAc buffer (40 mL) solid
GeO2 (0.093 g, 0.89 mmol) and Na2WO4Á2H2O (2.640 g, 8.00 mmol)
were successively added. The reaction mixture was stirred at room
temperature for 1 h. Compounds Na-bb-Ln4 were isolated as highly
gathered needles from slow evaporation of the resulting in limpid
solution at room temperature and identified by powder X-ray
diffraction and infrared spectroscopy (Fig. A.1 in the Supplemen-
tary Information).
tion). The experimental and calculated mass losses (Dm) for the
dehydration process, together with the decomposition tempera-
tures (Td) are given in Table B.2 in the Supplementary Information.
Potential modifications of the catalysts during their thermal
treatment were analysed in situ by diffuse reflectance infrared
Fourier transform spectroscopy (DRIFTS) coupled to mass spec-
trometry.
equipped with
SpectraTech diffuse reflectance accessory was used. Prior to all
A
Mattson Infinity MI60 DRIFT spectrophotometer
a
chamber of controlled environment and
a