Facile Immobilization of a Lewis Acid Polyoxometalate onto Layered Double Hydroxides for Highly Efficient N-Oxidation of Pyridine-Based Derivatives and Denitrogenation

Article abstract of DOI:10.1002/cctc.201500594

N-Oxides are a class of highly important compounds that are used widely as synthetic intermediates. In this paper, we demonstrate for the first time the use of a polyoxometalate-based composite material as a highly efficient heterogeneous catalyst for the N-oxidation of pyridines and its derivatives in the presence of H₂O₂ at room temperature. The composite was prepared by the intercalation of the [La(PW₁₁O₃₉)₂]^11- anion into a layered double hydroxide (LDH) modified with tris(hydroxymethyl)aminomethane (Tris). Additionally, the Tris-LDH-La(PW₁₁)₂-based catalyst has been employed for the denitrogenation of a model oil mixture in the presence of 1-butyl-3-methylimidazolium tetrafluoroborate and H₂O₂. Denitrogenation can be achieved in 40 min at 75 C. Finally, the heterogeneous catalyst can be recovered easily and reused at least 10 times without a measurable decrease of the catalytic activity and disintegration of the Tris-LDH-La(PW₁₁)₂ structure. Between the sheets: We demonstrate for the first time the use of a polyoxometalate-based composite material as a highly efficient heterogeneous catalyst for the N-oxidation of pyridines and its derivatives in the presence of H₂O₂ at room temperature. The composite is prepared by the intercalation of the [La(PW₁₁O₃₉)₂]^11- anion into a layered double hydroxide modified with tris(hydroxymethyl)aminomethane.

Full text of DOI:10.1002/cctc.201500594

DOI: 10.1002/cctc.201500594
Full Papers
Facile Immobilization of a Lewis Acid Polyoxometalate
onto Layered Double Hydroxides for Highly Efficient
N-Oxidation of Pyridine-Based Derivatives and
Denitrogenation
Kai Liu,[a] Zhixiao Yao,[a] Haralampos N. Miras,*[b] and Yu-Fei Song*[a]
N-Oxides are a class of highly important compounds that are
used widely as synthetic intermediates. In this paper, we dem-
onstrate for the first time the use of a polyoxometalate-based
composite material as a highly efficient heterogeneous catalyst
for the N-oxidation of pyridines and its derivatives in the pres-
ence of H2O2 at room temperature. The composite was pre-
pared by the intercalation of the [La(PW11O39)2]11À anion into
a layered double hydroxide (LDH) modified with tris(hydroxy-
methyl)aminomethane (Tris). Additionally, the Tris-LDH-
La(PW11)2-based catalyst has been employed for the denitroge-
nation of a model oil mixture in the presence of 1-butyl-3-
methylimidazolium tetrafluoroborate and H2O2. Denitrogena-
tion can be achieved in 40 min at 758C. Finally, the heteroge-
neous catalyst can be recovered easily and reused at least 10
times without a measurable decrease of the catalytic activity
and disintegration of the Tris-LDH-La(PW11)2 structure.
Introduction
[15]
In recent years, heterocyclic N-oxides have received much at-
tention because of their wide use as versatile synthetic inter-
mediates, their biological importance, their use as protecting
groups, auxiliary agents, oxidants, catalysts, surrogates for het-
erocyclic boronic acids, and ligands in metal complexes.[1–4] It is
known that the conventional preparation of pyridine N-oxides
employs a noncatalytic oxidation process using dioxiranes,[5]
activated H2O2,[6] H2SO5 (Caro’s acid),[7] and magnesium mono-
phthalate[8] or peracids[9,10] as oxidants. However, these re-
agents are not only highly corrosive but also generate large
amounts of waste during the reaction, which require frequent
and laborious work-up treatments. In an effort to develop al-
ternative synthetic routes and minimize the waste production
of the N-oxidation process, the promotion of the N-oxidation
of pyridines using homogeneous catalysts, which include
AcOH/H2O2,[11] MeReO3/H2O2,[12] trifluoroacetic anhydride
(TFAA)/H2O2,[13] and Tf2O/Na2CO3·1.5H2O2,[14] has been studied.
Unfortunately, these homogeneous catalysts are not recyclable,
which restrict their application in practice. To overcome such
problems, a number of heterogeneous catalysts, which include
VxSi4xO6.4x
,
GrO-MTO (GrO: Graphene oxide; MTO: methyl-
trioxorhenium),[16] Ru(PVP)/c-Al2O3 (PVP: polyvinylpyrroli-
done),[17] and Ti-MWW (Ti modified aluminosilicate with MWW
topology),[18] were applied to the N-oxidation of pyridines. The
drawbacks of these reported heterogeneous catalytic systems
are the high reaction temperature and the highly toxic organic
solvents that are usually required for N-heterocycles that are
resistant to oxidation because of their substantially high oxida-
tion potentials. Recently, it was shown that polyoxometalates
(POMs) such as [(R4N)8][BW11O39H][19] and [(C18H37)2(CH3)2N]7-
[PW11O39][20] are able to catalyze N-oxidation reactions. Howev-
er, these systems exhibit issues with catalyst leaching and ther-
mal stability. Consequently, it is fundamentally important to
develop efficient, recyclable, and environmentally benign cata-
lytic systems for N-oxidation.
In this paper, the catalytic efficacy of a lanthanum(III)-con-
taining sandwich POM, K11[La(PW11O39)2]·15H2O (denoted as K-
La(PW11)2), has been evaluated for the N-oxidation of pyridine-
based substrates. Initial tests showed that the catalytic reaction
proceeds efficiently under mild conditions (room temperature)
in the presence of H2O2 as the oxidant. However, if we consider
the cost and environmental impact associated with large-scale
applications, it is of fundamental importance to design and
prepare a recyclable and reusable heterogeneous catalyst. It
has been demonstrated that layered double hydroxides (LDHs)
are excellent candidates for the immobilization of anionic
POMs,[21–23] in which the LDHs not only function as the support
but also provide multiple interactions between the host layers
and the guest POM anions. The intercalation of POMs into
LDHs prevents the POM species form leaching into the reac-
tion mixture and increases the selectivity of the catalytic reac-
tion. Based on these observations, we prepared a composite
[a] K. Liu, Z. Yao, Prof. Y.-F. Song
State Key Laboratory of Chemical Resource Engineering
Beijing University of Chemical Technology
Beijing 100029 (P.R. China)
Fax: (+86)10-64431832
E-mail: songyf@mail.buct.edu.cn
[b] Dr. H. N. Miras
WestCHEM, School of Chemistry
University of Glasgow
Glasgow, G12 8QQ (UK)
E-mail: harism@chem.gla.ac.uk
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201500594.
ChemCatChem 2015, 7, 3903 – 3910
3903
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
material based on the catalytically active [La(PW11O39)2]11À spe-
cies by intercalating it into LDHs modified with tris(hydroxyl)-
aminomethane (Tris).[24,25] The heterogeneous Tris-LDH-
La(PW11)2 catalyst shows excellent catalytic activity in N-oxida-
tion and denitrogenation reactions under mild conditions.
and (012) reflections of Tris-LDH-La(PW11)2 shift to 2q=8.9,
19.5, and 29.48, which indicates that the [La(PW11O39)2]11À ions
were intercalated successfully into the interlayer of the Tris-
modified LDH.
FT-IR spectra of K11-La(PW11)2, Tris-LDH-CO3, and Tris-LDH-
La(PW11)2 are presented in Figure 1B. For Tris-LDH-CO3, the
peaks at n˜ =1086 and 1134 cmÀ1 can be ascribed to the char-
acteristic stretching vibrations of MÀOÀC and CÀCÀO, respec-
tively. For Tris-LDH-La(PW11)2, the asymmetric and symmetric vi-
brations at n˜ =1080 and 940 cmÀ1 are assigned to WÀO
stretching, whereas these vibrations appear at n˜ =1093 and
950 cmÀ1 in the spectrum of K11-La(PW11)2. The redshift indi-
cates the presence of strong electrostatic interactions and hy-
drogen bonding between the host layers and the guest POM
anions.[26] The peak at n˜ =1051 cmÀ1 can be assigned to the
stretching vibration of the PÀO bond, which appears at n˜ =
1049 cmÀ1 in the spectrum of K11-La(PW11)2. The other charac-
teristic peaks of K11-La(PW11)2 at n˜ =882, 835, and 768 cmÀ1 are
associated with the stretching vibration of WÀOÀW, which
shift to n˜ =889, 826, and 771 cmÀ1 in the spectrum of Tris-
LDH-La(PW11)2. Additionally, the peak at n˜ =451 cmÀ1 in the
spectra of both Tris-LDH-CO3 and Tris-LDH-La(PW11)2 is assigned
to the OÀMÀO vibration in the brucite-like layers of the LDH.[27]
Meanwhile, the peak at n˜ =1359 cmÀ1 of Tris-LDH-CO3 caused
Results and Discussion
The ion exchange of K-La(PW11)2 with Tris-LDH-CO3 under am-
bient conditions without the necessity to degas CO2 results in
the formation of a new intercalated assembly of Tris-LDH-
La(PW11)2 (Scheme 1). The XRD pattern of Tris-LDH-CO3 shows
the characteristic (003), (006), (012), (110), and (113) reflec-
tions at 2q=11.6, 23.6, 34.9, 61.1, and 62.38, respectively
(Figure 1). After ion exchange, the characteristic (003), (006),
2À
by the vibration of CO3 disappears in the spectrum of Tris-
LDH-La(PW11)2, which suggests the successful ion exchange of
Scheme 1. Intercalation of [La(PW11O39)2]11À into Tris-stabilized LDHs.
2À
CO3 by POM anions.
The 13C cross polarization magic-angle spinning (CP/MAS)
NMR spectrum of Tris-LDH-CO3 (Figure 2A) exhibits a strong
2À
signal at d=171 ppm ascribed to the interlayer CO3
anions,[24] which is absent in the spectrum of Tris-LDH-
2À
La(PW11)2, caused by the complete ion exchange of the CO3
anions by the [La(PW11O39)2]11À clusters during the intercalation
process. Furthermore, a set of signals centered at d=À5 ppm
in the 31P NMR spectrum of Tris-LDH-La(PW11)2 implies that the
[La(PW11O39)2]11À anions were incorporated successfully into the
Tris-modified LDHs (Figure 2B).
X-ray photoelectron spectroscopy (XPS) of Tris-LDH-La(PW11)2
is shown in Figure 3A. The W4f spectrum can be deconvolut-
ed into doublets, which consists of W4f7/2 at a binding energy
(BE) of 35.6 eV and W4f5/2 at BE=37.7 eV, respectively. The
doublets are ascribed to the W in the WÀO bond configuration
and are observed typically for W6+.[28] This result is in accord-
ance with the oxidation state of W of K-La(PW11)2. Thermog-
ravimetry with differential thermal analysis (TG-DTA) was per-
formed for Tris-LDH-La(PW11)2. Two weight-loss stages can be
observed with a temperature increase from 32 to 8008C (Fig-
ure 3B). The first weight loss is 8.29% between 40 and 2448C,
which can be ascribed to the removal of water molecules ab-
sorbed on the surface and interlayer space of Tris-LDH-
La(PW11)2. The second weight loss step of 9.95% between 244
and 8008C corresponds to the decomposition of Tris and the
collapse of the layered structure. Based on the TG-DTA and ele-
mental analysis results (Mg=7.152%, Al=3.969%, W=
54.41%, N=0.588%), the elemental composition of the com-
posite material can be identified as: Mg0.66Al0.33(C4H8NO3)0.093
-
Figure 1. A) XRD patterns of Tris-LDH-CO3 and Tris-LDH-La(PW11)2; B) FTIR
spectra of a) Tris-LDH-CO3, b) Tris-LDH-La(PW11)2, and c) K11-La(PW11)2.
(OH)1.72La0.015(PW11O39)0.03 ·1.03H2O (Table S1).
ChemCatChem 2015, 7, 3903 – 3910
www.chemcatchem.org
3904
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 3. A) XPS spectra for the W4f core level of Tris-LDH-La(PW11)2; B) TG-
DTA profile of Tris-LDH-La(PW11)2.
Figure 2. A) 13C CP/MAS NMR spectra of Tris-LDH-CO3 and Tris-LDH-
La(PW11)2; B) 31P CP/MAS NMR spectra of Tris-LDH-CO3 and Tris-LDH-
La(PW11)2.
TEM images of Tris-LDH-La(PW11)2 (Figure 4) revealed that
the material consists of 50~70 nm layered areas. The dark dots
and blocks observed in the samples of Tris-LDH-La(PW11)2 can
be identified as the intercalated POM species. The energy-dis-
persive X-ray spectroscopy (EDX) data (Figure 4C) obtained
from the regions of the black dots observed in the TEM re-
vealed the presence of La, P, and W, consistent with the com-
position of the POM in the molecular catalyst Tris-LDH-
K11[La(PW11O39)2]. In addition, the size of the black spots falls in
the range of 1–2 nm, which is in good agreement with the size
of the POM clusters. Based on these data, we conclude that
the black spots observed in TEM images are the POM anions.
BET measurement was performed on Tris-LDH-CO3 and Tris-
LDH-La(PW11)2 to obtain detailed information on the structural
features of the as-prepared materials. The specific surface area,
pore volume, and average pore diameter (estimated from N2
adsorption–desorption isotherms) are presented in Table S2.
Tris-LDH-CO3 and Tris-LDH-La(PW11)2 display H4-type hysteresis
loops (Figure S1), which indicates that the pores are produced
by the aggregation of slit-shaped microsized pores,[29] similar
to the Mg3 Al-POMs composites but significantly different from
the unmodified Mg3 Al-CO3 LDH, which exhibits H3-type hyste-
resis loops.[30]
Figure 4. A, B) TEM images of Tris-LDH-La(PW11)2; C) EDX spectrum of Tris-
LDH-La(PW11)2.
ChemCatChem 2015, 7, 3903 – 3910
www.chemcatchem.org
3905
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
N-Oxidation of pyridine and its derivatives
Table 2. Oxidation of 4-picoline[a] with H2O2 in different solvents.[b]
To evaluate the catalytic performance of Tris-LDH-La(PW11)2, we
selected the oxidation of 4-picoline as a model reaction using
30 wt% H2O2 as the oxidant (Scheme 2). The reaction proceeds
efficiently in water in the presence of a catalytic amount of
Tris-LDH-La(PW11)2 at room temperature (Table 1), and the cor-
Entry
Solvent
Yield[c]
[%]
TON[d]
1
2
3
4
5
6
7
8
methanol
ethanol
acetone
acetonitrile
ethyl acetate
toluene
trichloromethane
water
12
15
4
30
38
10
8
20
52
81
42
94
130
203
105
235
[a] 4-Picoline is readily soluble in all solvent systems. [b] Reaction condi-
tions: 1 mmol 4-picoline, 5 mmol (30 wt%) H2O2, 4 mmol catalyst of Tris-
LDH-La(PW11)2, 1 mL solvent, room temperature, 8 h. Assignment of prod-
ucts by 1H and 13C NMR spectroscopy. [c] Isolated yields. [d] TON=moles
of product/moles of catalyst used.
Scheme 2. Model reaction of the N-oxidation of 4-picoline.
Table 1. Oxidation of 4-picoline in water by different catalysts.[a]
tween 4-picoline and the solvents and the different solubility
of 4-picoline in various solvents. Thus, the optimum solvent for
the model reaction is water.
Entry
Catalyst
Oxidant
Yield[b]
[%]
TON[c]
To test the general applicability of the catalytic efficacy of
the composite on this type of reaction, a series of pyridine de-
rivatives were subjected to the optimized reaction conditions.
Most of the substrates were oxidized successfully to the corre-
sponding N-oxides in high yields at room temperature
(Table 3). Pyridine could be oxidized to pyridine-N-oxide within
only 4 h in 99% yield. For the mono- or dimethyl substituted
pyridine substrates, 2-picoline, 3-picoline, 4-picoline and 2,3-lu-
tidine, and 2,6-lutidine, yields of 96, 95, 94, 93, and 92% were
obtained within 6–8 h, respectively (entries 2–6). For the elec-
tron-deficient substrate 2-chloropyridine, the N-oxidation pro-
cess exhibits a relatively low yield of 24% (entry 7). Further-
more, the quinoline can be oxidized successfully to the corre-
sponding quinoline-N-oxide with a yield of 91% in 4 h at 658C
(entry 8). Isoquinoline and 8-hydroxyquinoline can be oxidized
efficiently to the corresponding products at slightly elevated
temperatures (658C).
1
2
3
4
5
6
7
8
9
Na-PW12
K-P2W18
K-PW11
La2O3
LaCl3
K-La(PW11)2
Tris-LDH-CO3
Tris-LDH-La(PW11)2
–
H2O2
H2O2
H2O2
H2O2
H2O2
H2O2
H2O2
H2O2
H2O2
72
76
44
31
17
97
6
180
190
110
78
43
243
15
94
12
235
–
[a] Reaction conditions: 1 mmol 4-picoline, 5 mmol H2O2 (30 wt%),
4 mmol catalyst, 1 mL H2O, room temperature, 8 h. Assignment of prod-
ucts by 1H and 13C NMR spectroscopy. [b] Isolated yields. [c] Turnover
number (TON)=moles of product/moles of catalyst used.
responding 4-picoline-N-oxide can be obtained in high yields
of ~95% (entry 6). In contrast, the Keggin POM of
Na3[PW12O40]·15H2O (Na-PW12), the Dawson POM of
K6[P2W18O62]·19H2O (K-P2W18), and the monolacunary Keggin
POM of K7[PW11O39]·13H2O (K-PW11) provide yields of 72, 76,
and 44%, respectively (entries 1–3). If La2O3 and LaCl3 are used
as catalysts, the reaction is sluggish (entries 4–5). Notably, if K-
La(PW11)2 is applied to the reaction, 97% yields can be ob-
tained, which is higher than the 95% yield given by Tris-LDH-
La(PW11)2. This is because K-La(PW11)2 is a highly efficient homo-
geneous catalyst, whereas Tris-LDH-La(PW11)2 is a heterogene-
ous catalyst for which diffusion phenomena have a considera-
ble contribution. If the unmodified LDHs (Tris-LDH-CO3) is used
as a catalyst, a very low yield of 6% is obtained. Moreover, in
the absence of a catalyst, the use of only H2O2 results in 12%
oxidation of 4-picoline in water.
Denitrogenation by Tris-LDH-La(PW11)2
Petrochemical streams contain N/S-containing heterocyclic
species in addition to a series of mono- and polyaromatic com-
pounds. As a result of the environmental impact of these com-
ponents and the induced problems in numerous refinery pro-
cesses, industry employs various processes (denitrogenation
and desulfurization) for their removal. Thus, we made an effort
to investigate the efficacy of our heterogeneous catalyst under
experimental conditions used in refinery processes. Specifically,
we investigated the efficacy of the Tris-LDH-La(PW11)2 hetero-
geneous catalyst in the denitrogenation processes. Initially, we
studied the oxidation of quinoline in model oil at different
temperatures to identify the effect of the temperature on the
denitrogenation process (Table 4). It is observed that 53, 59,
66, and 78% N removal can be achieved at 25, 35, 45, and
558C, respectively, in 90 min. A further temperature increase to
65 or 758C leads to almost 100% denitrogenation of quinoline
An investigation of the influence of the solvents on the oxi-
dation reaction (Table 2) shows that the oxidation reaction
gives relatively low yields of 12, 15, 4, and 8% in methanol,
ethanol, acetone, and acetonitrile, respectively. Meanwhile, the
oxidation reaction gives higher yields of 52, 81, 42, and 94% in
ethyl acetate, toluene, trichloromethane, and water. The differ-
ent yields could be attributed to the different interactions be-
ChemCatChem 2015, 7, 3903 – 3910
www.chemcatchem.org
3906
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Table 3. Oxidation of pyridines with H2O2 and Tris-LDH-La(PW11)2 in
water.[a]
Table 4. Effect of temperature on N removal by the Tris-LDH-La(PW11)2/
[bmim]BF4 system at 758C.[a]
Entry Substrate
Product
t
Yield[b] TON[c]
[h] [%]
Entry
T
[8C]
t
N removal
[%]
[min]
1
2
3
4
5
6
25
35
45
55
65
75
90
90
90
90
55
40
53
59
66
78
100
100
1
4
6
99
96
248
240
2
[a] Reaction conditions: 122 mL H2O2, 17 mL model oil (N content
100 ppm), 1 mL [bmim]BF4, H2O2/quinoline/Tris-LDH-La(PW11)2 =300:30:1.
3
4
6
8
95
94
238
235
5
6
7
8
8
93
92
233
230
60
16 24
Figure 5. Effect of the H2O2/quinoline molar ratio on the N removal from
quinoline. Reaction conditions: T=758C, 17 mL model oil (N content
100 ppm), 1 mL [bmim]BF4, quinoline/Tris-LDH-La(PW11)2 =30:1.
8[c]
4
4
4
91
92
98
228
230
245
the N removal increases from 46 to 85% in 90 min. If the
molar ratio is increased to 8:1, 100% N removal can be ach-
ieved in 65 min. A further increase of the H2O2/quinoline ratio
to 10:1 or 15:1 led to complete denitrogenation in only 40 or
30 min, respectively. Therefore, the selected H2O2/quinoline
ratio for our subsequent experiments is the optimum value of
10:1.
9[c]
10[c]
Further investigation of the quinoline/Tris-LDH-La(PW11)2
molar ratios on the N-removal process was performed at 758C.
The time for complete denitrogenation increases from 40 to
90 min if the quinoline/Tris-LDH-La(PW11)2 molar ratio decreas-
es from 60:2 to 60:1 in the presence of H2O2 (Table 5, entry 1).
If the quinoline/Tris-LDH-La(PW11)2 molar ratio reaches 60:4 and
60:8, N removal reaches 100% in 50 and 60 min, respectively.
Furthermore, a slight increase of the amount of H2O2 to
0.183 mL leads to almost 100% denitrogenation in only 30 min
at 758C with a H2O2/quinoline/Tris-LDH-La(PW11)2 molar ratio of
900:60:8 (entry 9).
[a] Reaction conditions: 1 mmol substrate, 5 mmol H2O2 (30 wt%), 4 mmol
Tris-LDH-La(PW11)2, 1 mL H2O, room temperature. Assignment of products
by 1H and 13C NMR spectroscopy. [b] Isolated yields. [c] TON=moles of
product/moles of catalyst used.
in 55 and 40 min, respectively. Therefore, we used a tempera-
ture of 758C for subsequent experiments.
The optimum amount of H2O2 for the N-oxidation of the re-
moved N-containing compounds was identified in a series of
control experiments by applying different H2O2/quinoline
molar ratios under the experimental conditions of T=758C,
N=100 ppm, and quinoline/Tris-La(PW11)2 =30:1. Upon increas-
ing the H2O2/quinoline molar ratio from 3:1 to 5:1 (Figure 5),
The proportion of N removal and ln(Ct/C0) are plotted
against the reaction time in Figure 6, in which C0 and Ct are
the initial quinoline concentration and the quinoline concen-
tration at time t, respectively. The linear fit of the data reveals
ChemCatChem 2015, 7, 3903 – 3910
www.chemcatchem.org
3907
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Table 5. Effect of quinoline/Tris-LDH-La(PW11)2 molar ratio on N removal
by the Tris-LDH-La(PW11)2/[bmim]BF4 system at 758C.[a]
Entry
H2O2/quinoline/catalyst
H2O2
[mL]
t
N removal
[%]
[min]
1
2
3
4
5
6
7
8
9
600:60:1
122
122
122
122
122
183
183
183
183
90
40
50
40
60
70
30
30
30
100
100
100
96
100
100
100
100
100
600:60:2=300:30:1
600:60:4=150:15:1
600:60:8=150:15:2
600:60:8=150:15:2
900:60:1
900:60:2=450:30:1
900:60:4=450:30:2
900:60:8=450:30:4
[a] Reaction conditions: T=758C, 17 mL model oil (N=100 ppm), 1 mL
[bmim]BF4.
Figure 7. N removal from a) quinoline (10 ppm, 17 mL), b) quinoline
(100 ppm, 500 mL), c) pyridine (100 ppm, 17 mL), and d) indole (100 ppm,
17 mL). Reaction conditions: T=758C, a) 17 mL model oil (N=10 ppm),
H2O2/quinoline/Tris-LDH-La(PW11)2 =300:3:1; b) 500 mL model oil
(N=100 ppm), H2O2/quinoline/Tris-LDH-La(PW11)2 =300:30:1; c) 17 mL model
oil (N=100 ppm), H2O2/pyridine/Tris-LDH-La(PW11)2 =300:30:1; d) 17 mL
model oil (N=100 ppm), H2O2/indole/Tris-LDH-La(PW11)2 =300:30:1.
The time required for the complete N removal in the case of
pyridine, quinoline, and indole is 25, 40, and 35 min, respec-
tively. Thus, the denitrogenation activity decreases in the order
of pyridine> indole> quinoline in agreement with the elec-
tron density on the N atoms of the aromatic ring and steric
hindrance considerations.
To confirm that the catalyst is truly heterogeneous, the N-ox-
idation of 4-picoline was performed using Tris-LDH-La(PW11)2 as
the catalyst and H2O2 as the oxidant at 258C in water. The cata-
lyst was separated from the reaction mixture when the conver-
sion was 80% after 125 min. The monitoring of the reaction
mixture by GC showed that the substrate conversion stopped
immediately. The addition of the catalyst back to the reaction
mixture restarted the reaction, and the conversion reached
99% (Figure 8). Furthermore, the catalyst was separated from
the reaction mixture by centrifugation, washed with water, and
reused at least 10 times without a clear loss of activity (Fig-
ure S2). Moreover, the structure of Tris-LDH-La(PW11)2 remains
intact after the N-oxidation reaction as evidenced by FTIR spec-
troscopy, XRD, 13P NMR spectroscopy, and XPS (Figure S2).
Figure 6. N removal from quinoline and ln(Ct/C0) as a function of the reac-
tion time at 758C. H2O2 =0.122 mL, 17 mL model oil (N=100 ppm),
[bmim]BF4 =1 mL, H2O2/quinoline/Tris-LDH-La(PW11)2 =300:30:1.
that the catalytic reaction exhibits pseudo-first-order kinetics
for the denitrogenation of quinoline (R2 =0.9959). The rate
constant k of the oxidation reaction was determined to be
0.0985 minÀ1 based on Equations (1) and (2). The oxidation of
quinoline to quinoline-N-oxide can be completed in 40 min.
Thus, the catalyst exhibits high efficacy for the oxidation of
quinoline to quinoline-N-oxide, and the catalytic reaction
obeys pseudo-first-order kinetics with 100% selectivity for
quinoline-N-oxide.
dCt
dt
ð1Þ
ð2Þ
k ¼ À
ꢀ
ꢁ
C0
Ct
kt ¼ ln
With an N content as low as 10 ppm, 100% N removal can
be achieved in only 8 min under mild conditions (Figure 7).
With an increase in the volume of the model oil to 500 mL,
complete denitrogenation can be achieved in 40 min. The
result shows that highly efficient denitrogenation for different
substrates, which include pyridine, quinoline, and indole, can
be achieved in the presence of the Tris-LDH-La(PW11)2/1-butyl-
3-methylimidazolium tetrafluoroborate ([bmim]BF4) system.
Figure 8. Recycling and reuse of Tris-LDH-La(PW11)2.
ChemCatChem 2015, 7, 3903 – 3910
www.chemcatchem.org
3908
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
These data confirm further that the catalysis of Tris-LDH-
La(PW11)2 is truly heterogeneous.
Experimental Section
Materials
A suggested mechanism for the denitrogenation of the oil
phase and subsequent catalytic oxidation of 4-picoline in the
aqueous phase by Tris-LDH-La(PW11)2 in the presence of H2O2 is
presented Figure 9. The 4-picoline substrate is firstly extracted
by [bmim]BF4 to the aqueous phase as [bmim]BF4 is water
soluble. The intermolecular interaction at the interphase of the
two solvent systems promotes the extraction process. Based
on the N-containing products in the oil stream, the nature of
the ionic liquid can vary accordingly.[31–34] The presence of H2O2
and the POM component of the Tris-LDH-La(PW11)2 catalyst
promote the formation of POM-based peroxo species responsi-
ble for the oxygen transfer and subsequent oxidation of the
substrate to the corresponding 4-picoline-N-oxide. The ability
for oxygen transfer by the generation of peroxo species in
POM systems is well known and has been studied extensively
by several groups.[35–37]
Quinoline (99%), pyridine (99%), indole (99%), 2-picoline (98%), 3-
picoline (99%), 4-picoline (99%), 2,3-lutidine (98%), 2,6-lutidine
(99%), 2-chloropyridine (99%), isoquinoline (99%), 8-hydroxyquino-
line (99%), n-octane (98%), H2O2 (30 wt%), and [bmim]BF4 were
purchased from Sigma–Aldrich and used directly without further
purification. All solvents were of analytical grade, purchased from
Alfa Aesar, and used without further purification.
Characterization
Powder XRD patterns were recorded by using a Rigaku XRD-6000
diffractometer under the following conditions: 40 kV, 30 mA, CuKa
radiation (l=0.154 nm). FTIR spectra were recorded by using
a Bruker Vector 22 infrared spectrometer using KBr pellets. The
solid-state NMR experiments were performed at 75.6 MHz for 13C
and at 121.0 MHz for 31P by using a Bruker Avance 300M solid-
state spectrometer equipped with a commercial 5 mm MAS NMR
probe. The N2 adsorption–desorption isotherms were measured by
using a Quantachrome Autosorb-1 system at 77 K, and the samples
were degassed at 1208C for 6 h before the measurements. TEM
images were recorded by using a JEOL JEM-2010 electron micro-
scope operated at 200 kV. TG-DTA was performed by using a TGA/
DSC 1/1100 SF from Mettler Toledo in a flow of N2 with a heating
rate of 108C minÀ1 from 32 to 8008C. XPS was performed with
monochromatized AlKa excited X-radiation (PHI Quantera SXM). An
inductively coupled plasma atomic emission spectroscope (ICP-
AES, Shimadzu ICPS-7500) was used to measure the concentration
of W in the catalysts. GC was performed by using an Agilent 7820A
GC system using a 30 m 5% phenylmethyl silicone capillary
column with an ID of 0.32 mm and a 0.25 mm coating (HP-5).
Figure 9. Proposed mechanism for the denitrogenation of the oil phase and
subsequent catalytic oxidation of 4-picoline in the aqueous phase by Tris-
LDH-La(PW11)2 in the presence of H2O2.
Preparation of Tris-LDH-CO3
Tris-LDH-CO3 was prepared as described previously.[24] The catalyst
was synthesized by an ion-exchange method.[25]
Conclusions
Oxidation of pyridines to N-oxides
We have successfully prepared a Lewis acid composite material
in which a polyoxometalate (POM) was intercalated into a lay-
ered double hydroxide (LDH) modified with tris(hydroxymethy-
l)aminomethane (Tris). The resulting nanocomposite Tris-LDH-
La(PW11)2 was characterized fully by various techniques and in-
vestigated as a heterogeneous catalyst for the N-oxidation of
pyridines and the denitrogenation of model oil. Notably, the N-
oxidation of pyridines to the corresponding N-oxides can be
achieved easily at room temperature with an excellent selectiv-
ity, high yield, and high efficacy. Complete denitrogenation can
be achieved within 40 min at 758C. In addition, the catalyst
can be recovered easily and recycled for more than 10 times
without a clear loss of catalytic activity. Therefore, the catalyst
shows a good potential for practical and industrial application.
Substrate (1 mmol), H2O2 aqueous solution (30 wt%), catalyst
(0.4 mol%; Tris-LDH-La(PW11)2 contained 4 mmol POMs), and H2O
(1 mL) were mixed in a 20 mL glass bottle at 25 or 658C under vig-
orous stirring. The reaction was quenched after 4–12 h. The result-
ing organic products were separated from the aqueous phase by
extraction, and the extracts were analyzed by GC with flame ioniza-
tion detection (FID). The conditions were as follows: injection port
temperature 2908C; detector temperature 2908C; oven tempera-
ture 708C; carrier gas: ultrapure nitrogen; sample injection volume
1 mL.
Denitrogenation of model oil
Typically, a solution of quinoline, pyridine, and indole in n-octane
was used as model oil with a N content of 100 ppm. The catalytic
oxidative denitrogenation experiments were performed in a 25 mL
two-necked flask, to which H2O2 (0.122 mL, 30 wt%), model oil
(17 mL), [bmim]BF4 (1 mL), and Tris-LDH-La(PW11)2 (H2O2/quinoline/
Cat=300:30:1) were added. The reaction mixture was stirred at
758C. During the reaction, the upper layer of the model-oil phase
ChemCatChem 2015, 7, 3903 – 3910
www.chemcatchem.org
3909
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[14] X. Zhu, K. D. Kreutter, H. Hu, M. R. Player, M. D. Gaul, Tetrahedron Lett.
2008, 49, 832–834.
was withdrawn periodically and analyzed by GC–FID. Quinoline,
pyridine, and indole were identified using reference standards.
[15] L. Rout, T. Punniyamurthy, Adv. Synth. Catal. 2005, 347, 1958–1960.
[16] P. K. Khatri, S. Choudhary, R. Singh, S. L. Jain, O. P. Khatri, Dalton Trans.
2014, 43, 8054–8061.
[17] P. Veerakumar, S. Balakumar, M. Velayudham, K.-L. Lu, S. Rajagopal,
Catal. Sci. Technol. 2012, 2, 1140–1145.
Acknowledgements
This work was supported by the National Basic Research Pro-
gram (2014CB932104), the National Science Foundation of China
(21222104, U1407127), the Fundamental Research Funds for the
Central Universities (RC1302, YS1406), and the Beijing Engineering
Centre for Hierarchical Catalysts. H.N.M. acknowledges the finan-
cial support from the University of Glasgow, the Royal Society of
Edinburgh, and Marie Curie actions.
[18] W. Xie, Y. Zheng, S. Zhao, J. Yang, Y. Liu, P. Wu, Catal. Today 2010, 157,
114–118.
[19] W. Zhao, C. Yang, Y. Ding, B. Ma, New J. Chem. 2013, 37, 2614–2618.
[20] Y. Ding, W. Zhao, J. Mol. Catal. A 2011, 337, 45–51.
[21] P. Jacobs, B. Sels, D. D. Vos, M. Buntinx, F. Pierard, A. Kirsch-De Mes-
maeker, Nature 1999, 400, 855–857.
[22] P. Liu, H. Wang, Z. Feng, P. Ying, C. Li, J. Catal. 2008, 256, 345–348.
[23] S. Zhao, J. Xu, M. Wei, Y.-F. Song, Green Chem. 2011, 13, 384–389.
[24] Y. Kuroda, Y. Miyamoto, M. Hibino, K. Yamaguchi, N. Mizuno, Chem.
Mater. 2013, 25, 2291–2296.
[25] Y. Chen, D. Yan, Y. F. Song, Dalton Trans. 2014, 43, 14570.
[26] Y. Dou, J. Han, T. Wang, M. Wei, D. G. Evans, X. Duan, Langmuir 2012,
28, 9535–9542.
Keywords: heterogeneous catalysis
lanthanum · oxidation · polyoxometalates
·
immobilization
·
[27] Z. P. Xu, H. C. Zeng, Chem. Mater. 2001, 13, 4555–4563.
[28] L. Salvati, L. E. Makovsky, J. M. Stencel, F. R. Brown, D. M. Hercules, J.
Phys. Chem. 1981, 85, 3700–3707.
[29] B. J. Aronson, C. F. Blanford, A. Stein, Chem. Mater. 1997, 9, 2842–2851.
[30] S. Brunauer, P. H. Emmett, E. Teller, J. Am. Chem. Soc. 1938, 60, 309–319.
[31] J. Eßer, P. Wasserscheid, A. Jess, Green Chem. 2004, 6, 316–322.
[32] A. E. Visser, R. D. Rogers, J. Solid State Chem. 2003, 171, 109–113.
[33] J. G. Huddleston, H. D. Willauer, R. P. Swatlowski, A. E. Visser, R. D.
Rogers, Chem. Commun. 1998, 1765–1766.
[34] A. R. Hansmeier, G. W. Meindersma, A. de Haan, Green Chem. 2011, 13,
1907–1913.
[35] I. V. Kozhevnikov, in Polyoxometalate Molecular Science, Vol. 98 (Eds.: J. J.
Borrµs-Almenar, E. Coronado, A. Müller, M. Pope), NATO sciences series,
Elsevier, 2012, pp. 351–380.
[1] F. F. Bamoharram, M. M. Heravi, M. Roshani, N. Tavakoli, J. Mol. Catal. A
2006, 252, 219–225.
[2] A. V. Malkov, M. Bell, F. Castelluzzo, P. Kocovsky, Org. Lett. 2005, 7,
3219–3222.
[3] L. C. Campeau, S. Rousseaux, K. Fagnou, J. Am. Chem. Soc. 2005, 127,
18020–18021.
[4] J. Sniekers, N. R. Brooks, S. Schaltin, L. Van Meervelt, J. Fransaer, K. Bin-
nemans, Dalton Trans. 2014, 43, 1589–1598.
[5] M. Ferrer, F. Sµnchez-Baeza, A. Messeguer, Tetrahedron 1997, 53, 15877–
15888.
[6] G. B. Payne, P. H. Deming, P. H. Williams, J. Org. Chem. 1961, 26, 659–
663.
[7] G. J. Robke, E. J. Behrman, J. Chem. Res. Synop. 1993, 412–413.
[8] P. Brougham, M. S. Cooper, D. A. Cummerson, H. Heaney, N. Thompson,
Synthesis 1987, 1987, 1015–1017.
[36] M. V. Vasylyev, R. Neumann, J. Am. Chem. Soc. 2004, 126, 884–890.
[37] N. Mizuno, M. Misono, Chem. Rev. 1998, 98, 199–217.
[9] H. S. Mosher, L. Turner, A. Carlsmith, Org. Synth. 1953, 33, 79–81.
[10] L. Skatteboel, B. Boulette, J. Org. Chem. 1969, 34, 4150–4153.
[11] V. Boekelheide, W. J. Linn, J. Am. Chem. Soc. 1954, 76, 1286–1291.
[12] C. CopØret, H. Adolfsson, T. A. V. Khuong, A. K. Yudin, K. B. Sharpless, J.
Org. Chem. 1998, 63, 1740–1741.
Received: May 29, 2015
Revised: July 3, 2015
[13] D. Rong, V. A. Phillips, R. S. Rubio, M. ngeles Castro, R. T. Wheelhouse,
Tetrahedron Lett. 2008, 49, 6933–6935.
Published online on September 28, 2015
ChemCatChem 2015, 7, 3903 – 3910
www.chemcatchem.org
3910
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim