Iron(III), cobalt(II) and copper(II) complexes bearing 8-quinolinol encapsulated in zeolite-Y for the aerobic oxidation of styrene

Article abstract of DOI:10.1002/aoc.1752

A series of Fe(III), Co(II) and Cu(II) complexes of 8-quinolinol were encapsulated into the supercages of zeolite-Y and characterized by X-ray diffraction, SEM, N₂ adsorption/desorption, FT-IR, UV-vis spectroscopy, elemental analysis, ICP-AES and TG/DSC measurements. The encapsulation was achieved by a flexible ligand method in which the transition metal cations were first ion-exchanged into zeolite Y and then complexed with 8-quinolinol ligand. The metal-exchanged zeolites, metal complexes encapsulated in zeolite-Y plus non-encapsulated homogeneous counterparts were all screened as catalysts for the aerobic oxidation of styrene under mild conditions. It was found that the encapsulated complexes always showed better activity than their respective non-encapsulated counterparts. Moreover, the encapsulated iron complex showed good recoverability without significant loss of activity and selectivity within successive runs. Heterogeneity test for this catalyst confirmed its high stability against leaching of active complex species into solution.

Full text of DOI:10.1002/aoc.1752

Full Paper
Received: 10 July 2010
Revised: 12 September 2010
Accepted: 19 September 2010
Published online in Wiley Online Library: 7 March 2011
(wileyonlinelibrary.com) DOI 10.1002/aoc.1752
Iron(III), cobalt(II) and copper(II) complexes
bearing 8-quinolinol encapsulated in zeolite–Y
for the aerobic oxidation of styrene
Ying Yanga, Hong Dinga, Shijie Haob, Ying Zhangb∗ and Qiubin Kana∗
A series of Fe(III), Co(II) and Cu(II) complexes of 8-quinolinol were encapsulated into the supercages of zeolite–Y and
characterized by X-ray diffraction, SEM, N2 adsorption/desorption, FT-IR, UV–vis spectroscopy, elemental analysis, ICP-AES
and TG/DSC measurements. The encapsulation was achieved by a flexible ligand method in which the transition metal cations
were first ion-exchanged into zeolite Y and then complexed with 8-quinolinol ligand. The metal-exchanged zeolites, metal
complexes encapsulated in zeolite–Y plus non-encapsulated homogeneous counterparts were all screened as catalysts for the
aerobic oxidation of styrene under mild conditions. It was found that the encapsulated complexes always showed better activity
than their respective non-encapsulated counterparts. Moreover, the encapsulated iron complex showed good recoverability
without significant loss of activity and selectivity within successive runs. Heterogeneity test for this catalyst confirmed its high
c
stability against leaching of active complex species into solution. Copyright ꢀ 2011 John Wiley & Sons, Ltd.
Supporting information may be found in the online version of this article.
Keywords: zeolite-Y; encapsulation; transition metal complexes; 8-quinolinol; oxidation of styrene
Introduction
be efficient for the oxidation of cyclohexane with TBHP.[14] Other
metallo-phthalocyanine complexes (like iron phthalocyanine,
FePc) were also encapsulated into zeolite–Y by the template
method. Methods II and III are relatively more adaptive to
encapsulation of a wide range of complexes into zeolites. Both
approaches have been extensively used to encapsulate Schiff
bases, metalloporphyrins, metallophthalocyanines complexes,
etc.[15,16] Usually, when ligand size is larger than the opening
diameter of zeolite channels, method II is used, wherein the
ligand is constructed within the micropores of transition metal-
exchanged zeolite and then complexed with the resident metal
ions.[17] On the other hand, when the ligand size is smaller than the
opening diameter of zeolite, and the ligand is flexible to diffuse
into the zeolite to react with the previously exchanged metals in
the zeolite cavity, method III is used.[18]
The oxidation of styrene is of great interest because styrene
oxide and benzaldehyde are important and versatile synthetic
intermediates in chemical industry.[1,2] Currently, a wide range of
homogeneous metal complexes for styrene oxidation exist, but
mostofthemsufferfromseveredeactivationduetoeasyformation
of dimeric peroxo- and µ-oxo species. Therefore, much attention
has been devoted to the heterogenization of the homogeneous
catalysts onto solid supports to increase catalyst stability and
allow for catalyst recycling and product separation.[3–8] Generally,
the homogeneous catalysts are heterogenized either through
covalent attachment or encapsulation into porous supports, like
MCM-41,[9] SBA-15[10] and zeolite–Y.[11–13] The encapsulation
strategy is found to be convenient and advantageous because
the metal complex, once formed inside the pore cavity, does
not easily diffuse out and enter the liquid phase during catalytic
reaction. Faujasite zeolite, including zeolite X and Y, is frequently
chosen as a support to encapsulate metal complexes because of
its huge supercage of 13 Å with large pore opening of ca 8 Å
and three-dimensional pore system desirable for easy reactant
accessibility and product diffusion.[11]
8-Quinolinol, a strong bidentate ligand with a dimeter smaller
than 7 Å and its metal complexes larger than 8 Å, can be
encapsulated into zeolite Y by method III. However, as far as
we know, 8-quinolinol and its encapsulated metal complexes are
always ignored, probably due to its moderately conformationally
To encapsulate homogeneous metal complexs into faujasite,
three methods have been developed: (i) the template method;
(ii) the ship-in-bottle method; and (iii) the flexible ligand method.
For method I, preformed transition metal complex is used as
a template around which zeolite is crystallized, and thus it
requires the metal complex to be stable under harsh zeolite
synthesis conditions such as high pH and high temperature.
Therefore, there are limited successful examples of encapsulating
metal complexes using this approach. For example, Balkus
and coworkers synthesized a NaX encapsulated ruthenium
perfluorophthalocyanine complex (RuF16Pc–X) and it proved to
∗
Correspondence to: Qiubin Kan, College of Chemistry, Jilin University,
Changchun 130023, Jilin, People’s Republic of China.
E-mail: catalyticscience@yahoo.com.cn
∗∗
Ying Zhang,DepartmentofMaterialsScienceandEngineering,ChinaUniversity
of Petroleum, Changping District, Beijing, 102249, People’s Republic of China.
E-mail: yingzh1977@163.com
a
CollegeofChemistry,JilinUniversity,Changchun130023,Jilin,People’sRepublic
of China
b
Department of Materials Science and Engineering, China University of
Petroleum, Changping District, Beijing, 102249, People’s Republic of China
c
Appl. Organometal. Chem. 2011, 25, 262–269
Copyright ꢀ 2011 John Wiley & Sons, Ltd.

Iron(III), cobalt(II) and copper(II) complexes bearing 8-quinolinol
flexible structure and their wide application in the field of
luminescent materials.[19] Little attention has been focused on
their catalytic performances except for a pioneering work of MCM-
41 encapsulated [Fe(Q)3]Cl2 (Q = 8-quinolinol). The encapsulation
significantly improved the catalytic activity of [Fe(Q)3]Cl2 in
phenol hydroxylation with H2O2 but the leaching of Fe(Q)3
species occurred due to the same diameters of pore size and
pore entrance.[20] Moreover, there are few studies examining the
leaching process of encapsulated catalysts carefully. Herein, we
report the encapsulation of Fe(III), Co(II) and Cu(II) complexes of
8-quinolinol into the zeolite–Y by method III. These encapsulated
metalcomplexeswerefullycharacterizedandscreenedascatalysts
for the aerobic oxidation of styrene in the presence of a sacrificial
co-reductant isobutyraldehyde. Heterogeneity tests for these
catalysts were also performed to investigate their stability against
leaching of active complex species into solution.
was mixed with 2.18 g of 8-quinolinol (15 mmol) in 50 ml of
tetrahydrofuran and aqueous NH3 –NH4+ buffer was added to the
solution to maintain a pH of 7.0–8.0. In parallel, 5 ml of aqueous
Co(NO3)2·4H2O or Cu(CH3COO)2 solution (1 M) was mixed with
1.45 g of 8-quinolinol (10 mmol) in 50 ml of tetrahydrofuran. The
resulting mixture was stirred and refluxed for 2 h. After cooling,
these solid products were separated by filtration, washed with
excess ethanol and dried under vacuum. The spectroscopic data
of these complexes are given in the Supporting Information.
Catalytic Oxidation
The oxidation of styrene with air was carried out in a batch reactor.
In a typical run, 10 mmol of styrene along with 10 ml of CH3CN,
25 mmol of isobutyraldehyde and certain amounts of catalyst
were added into a 100 ml two-necked flask equipped with a
condenser and an air pump. The reaction was started by bubbling
the dry air with a stable flow rate of 80 ml min−1 controlled by
a flowmeter into the bottom of reactor at reaction temperature.
After the reaction was finished, the catalyst was filtered, washed
with CH3CN, dried at 100 ◦C overnight and reused directly without
further purification. The liquid organic products were quantified
by using a gas chromatography (Shimadzu, GC-8A) fitted with FID
detector and HP-5 capillary column. The liquid organic products
wereidentifiedbycomparisonwithauthenticsamplesandverified
by GC-MS coupling.
Experimental
Preparation of Metal-exchanged Zeolite, M–Y (M = Fe(III),
Co(II) or Cu(II))
Firstly, Na–Y was prepared according to the procedure reported
with slight modification.[21] In a typical synthesis, 1.6 g of sodium
hydroxide and 2.5 g of sodium aluminate (41 wt%) were dispersed
in 25 ml of deionized water under constant stirring, and then
18 ml of silica sol (25 wt%) was added. The mixture with a molar
composition of 13.2 SiO2 –Al2O3 –16.4 Na2O–354 H2O was stirred
at room temperature for 1 h to form gel slurry, and then was
hydrothermally crystallized at 100 ◦C for 24 h. Subsequently, the
resultant precipitate was separated from the mother liquor by
filtration, washed with deionized water and dried at 150 ◦C for
12 h.
Metal-exchanged zeolite–Y (M–Y) was prepared as follows:
an amount of 2.0 g Na–Y zeolite was suspended in 40 ml
distilled water containing 0.32 mmol of metal salt [Fe(NO3)3·9H2O,
Co(NO3)2·4H2O or Cu(CH3COO)2]. The mixture was stirred at
room temperature for 24 h. The M–Y materials were obtained
by filtration, washed with hot distilled water till the filtrate was
free from any metal ion content and dried for 12 h at 150 ◦C in air.
Results and Discussion
Morphological and Textural Properties of Materials
The X-ray powder diffractograms of Na–Y, Fe–Y and FeQ3 –Y are
depicted in Fig. 1 and indicate a well crystalline FAU structure of
cubic symmetry.[22] The appearance of peaks in iron-exchanged
zeolite (Fe–Y) and encapsulated one (FeQ3 –Y) suggests that the
supercages of the zeolite are able to accommodate metal com-
plexes and the crystallinity of the zeolite–Y has been preserved
during encapsulation. Only a small decrease was observed in the
intensities of the peaks at 2θ < 20◦ due to the encapsulation of
metal complexes. No new peaks assigned to pure iron complexes
were detected in the encapsulated zeolite probably due to very
PreparationofEncapsulatedMetalComplexesof8-quinolinol,
MQn –Y, (Q = 8-quinolinol, n = 2 or 3)
This process was performed with excess 8-quinolinol ligands
(nligand/nmetal = 3), and allowed the insertion of the ligand in
the cavity of the zeolite followed by its complexation with metal
ions. The complexation was carried out in MeOH at 80 ◦C for
24 h. Uncomplexed ligands and the complexes adsorbed on the
exterior surface were removed by full Soxhlet extraction with
MeOH until the eluate became colorless. The extracted sample
was ion-exchanged with aqueous 0.01 M NaCl solution at 80 ◦C for
8 htoremoveuncoordinatedmetalions,followedbywashingwith
deionized water until no Cl− anions could be detected with AgNO3
aqueous solution. The MQn –Y samples were then dried at 150 ◦C
for several hours to constant weight. This procedure brought
aboutbrilliantcolorchange, suggestingthatthepropertiesofpure
compounds have been dramatically changed upon encapsulation.
Synthesis of Pure Metal Complexes of 8-Quinolinol
The procedures for the preparation of pure metal complexes
are as follows: 5 ml of aqueous Fe(NO3)3·9H2O solution (1 M)
Figure 1. XRD patterns of (a) Na–Y, (b) Fe–Y and (c) FeQ3 –Y.
c
Appl. Organometal. Chem. 2011, 25, 262–269
Copyright ꢀ 2011 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/aoc

Y. Yang et al.
Figure 2. SEM images of (a) Fe–Y, (b) FeQ3 –Y, (c) CoQ2 –Y and (d) CuQ2 –Y.
Table 1. Chemical composition and textual properties of various materials prepared
BET surface
Microporous
External surface
Materials
Chemical composition
area (m2 g−1
)
volumea (cm3 g−1
)
area (m2 g−1
)
Na–Y
Na25[Al25Si167O384] · 109H2O
707.5
599.1
429.8
645.7
316.2
424.7
314.6
0.320
0.255
0.188
0.296
0.140
0.187
0.140
17.7
50.8
16.8
12.3
16.1
32.8
14.2
Fe–Y
Na15.7Fe3.1[Al25Si167O384] · 110H2O
Na15.7Fe1.0(FeQ3)2.1[Al25Si167O384] · 110H2O
Na22.8Co1.1[Al25Si167O384] · 107H2O
Na22.8Co0.2(CoQ2)0.9[Al25Si167O384] · 107H2O
Na23.2Cu0.9[Al25Si167O384] · 111H2O
Na23.2Cu0.1(CuQ2)0.8[Al25Si167O384] · 111H2O
FeQ3 –Y
Co–Y
CoQ2 –Y
Cu–Y
CuQ2 –Y
a From the nitrogen adsorption isotherm at 77 K, calculated by the t-method.
low percentage loading. Similar results were also obtained from
the XRD patterns of cobalt and copper catalysts (Fig. S1).
and pore volumes indicates the presence of metal 8-quinolinol
complexes within the cavities of the zeolites and not on the
external surface. Similar observations was also found for inclusion
of Mn(III) Schiff base complexes into zeolite–Y.[23] Nevertheless,
the decreasing degree depends on the amount of incorporated
complexes as well as their molecular size and geometrical con-
formation inside the zeolitic host. As a result, compared with their
respectiveM–Y,areductionbyabout30,25and50%wasobserved
upon encapsulation of Fe, Co and Cu complexes of 8-quinolinol
into zeolite–Y, respectively. On the other hand, the external
surface areas determined from the t-plot increase with the incor-
poration of metal ions by comparing Fe–Y and Cu–Y with Na–Y.
This increase can be explained in terms of the zeolite structural
deformation occurring during the cation-exchanged process.[24]
SEM micrographs of the modified zeolites exemplified by Fe–Y,
FeQ3 –Y, CoQ2 –Y and CuQ2 –Y are depicted in Fig. 2. From these
micrographs, Fe–Y possesses the octahedron shape of crystal
(Fig. 2a), similar to that of zeolite–Y. The SEM images of FeQ3 –Y
(Fig. 2b), CoQ2 –Y (Fig. 2c) and CuQ2 –Y (Fig. 2d) indicate the pres-
ence of well-defined zeolite crystals free from any shadow of metal
complexes present on their external surface. Both X-ray diffraction
and SEM results demonstrate that good crystallinity of zeolites is
kept during the encapsulation of metal complexes by method III.
The nitrogen adsorption isotherms for the parent (Na–Y) and
modified zeolites (M–Y and MQn –Y) are typical of microporous
materials (not shown); BET surface areas, micropore volumes and
external surface areas were calculated by fitting the adsorption
data in the corresponding theories; the values are given in Table 1.
As expected, the surface areas and the micropore volumes of M–Y
and MQn –Y show an apparent reduction when compared with
their respective parent materials. The lowering of the surface areas
Spectroscopic Characterization
Figure 3 shows the IR spectra of Na–Y, Fe–Y and FeQ3 –Y. It is
evident that framework vibration bands of zeolite–Y dominate
c
wileyonlinelibrary.com/journal/aoc
Copyright ꢀ 2011 John Wiley & Sons, Ltd.
Appl. Organometal. Chem. 2011, 25, 262–269

Iron(III), cobalt(II) and copper(II) complexes bearing 8-quinolinol
These bands are attributed to intra-ligand transition, ligand-to-
metal change transfer (CT) and d–d transitions, respectively.[25]
Compared with the pure CoQ2, however, these bands in the region
of 200–400 nm were blue-shifted upon encapsulation of cobalt
complexes in the Na–Y matrix, indicating the restructuring of the
ligand after coordination to the cobalt ion fixed in the Na–Y.
However, the d–d band considerably red-shifted to 598 nm for
CoQ2 –Y as compared with 577 nm in the spectrum of CoQ2.
This perhaps originates from the interaction of cobalt complex
with zeolite framework, although this complex was located in the
supercages in a neutral form. Similar results can be also obtained
from iron and copper catalysts (see Fig. 4B and C).
Quantification
The Si–Al ratios of modified zeolites are in the range of those
expected for faujasite and are almost in the same ratio as in the
parent zeolite (not shown). The results of chemical elemental
analyses give a Si–Al ratio of ca 6.60, which corresponds
to a unit formula Na25[Al25Si167O384] · 109H2O for Na–Y and
Na15.7Fe3.1[Al25Si167O384] · 110H2O for Fe–Y (Fe and H2O contents
were estimated by ICP-AES and TG/DSC, respectively; Table 1). The
Na and Fe in Fe–Y material give a degree of Na exchange of
37.2%, and the unit cell formula of iron-exchanged zeolites shows
an iron dispersion of 3.1 per unit cell. For these exchanged irons,
only 32.3% remains in the FeQ3 –Y, corresponding to a unit cell
formula Na15.7Fe1.0(FeQ3)2.1[Al25Si167O384] · 110H2O (see Tables 1
and 2), indicating that the exchanged irons were not removed by
refluxing with 0.01 M NaCl. Similar results are obtained for other
modified zeolites and their chemical compositions are also listed
in Table 1.
The chemical composition analyses also confirmed the purity
and stoichiometry of the pure and encapsulated complexes. The
chemicalanalysesoftheencapsulatedsamplesrevealthepresence
oforganicmatterwithaC : Nratioroughlysimilartothetheoretical
value of 7.80 for pure complexes (Table 2), indicating that the 8-
quinolinol ligands are intact resident in zeolite Y. Since these
ligands were all coordinated to the exchanged transition metal
ions in zeolite Y on the condition that the free ligands were fully
removed, itisreasonablethattheamountofcomplexedmetalions
was determined by that of ligands encapsulated. Therefore, the
amounts of metal complexes or ligands encapsulated in zeolite
matrix were determined by the N elemental analysis. Nitrogen
elemental analysis also reveals that the presence of ca 6.3 moieties
of 8-quinolinol per unit cell for FeQ3 –Y and the iron complexes
Figure 3. FT-IR spectra of (a) Na–Y, (b) Fe–Y and (c) FeQ3 –Y.
the spectra below 1200 cm−1 for all samples. The bands at 462,
572, 692 and 785 as well as 997 and 1138 cm−1 are attributed
to T–O bending mode, double ring, symmetric stretching and
asymmetric stretching vibrations, respectively.[25] No shift is
observed upon introduction of metal ions and inclusion of metal
complexes, further substantiating that the zeolite framework
remains unchanged. In spite of this, there is obviously a difference
in the range of 1200–1600 cm−1 among the three samples. It
can be seen that the prominent bands, assigned to C–O, C–N
and aromatic ring vibrations of ligands, are present at 1276, 1325,
1376, 1466 and 1500 cm−1 in the spectra of FeQ3 –Y (Fig. 3c) and
FeQ3 (Fig. S2a). However, these bands are absent in the case of
Na–Y and Fe–Y samples. In line with the FeQ3 –Y sample, the
framework IR spectra of other MQn –Y samples (Fig. S3) are also in
agreement with their pure analogs (Fig. S2), verifying that various
metal complexes have been encapsulated inside zeolite–Y.
The UV–vis spectra of Co–Y, CoQ2 –Y and CoQ2 are shown in
Fig. 4(A).Clearly,thereisaweakandbroadadsorptionintheregion
of400–600 nmforCo–Y,whichcanbeassignedtotheoverlapping
4
4
due to T1 –4T2 and T1 –4A2 transitions.[26] On the other hand,
three intense bands were observed for CoQ2 –Y and CoQ2 in
the regions of 218–320, 320–375 and 550–650 nm, respectively.
Table 2. Selected characteristics of catalysts
Ma (mmol g−1
Exchangeable
)
C/N analysis (wt%)
H analysis (wt%)
H in waterd
Catalysts
As Qb
C
N
C/N
Totalc
H in Q
Fe–Y
0.204
0.074
0.080
0.019
0.065
0.005
–
–
4.95
–
–
0.63
–
–
7.90
–
1.57
2.82
1.57
2.64
1.58
2.02
1.57
1.48
1.53
1.50
1.58
1.56
–
1.34
–
FeQ3 –Y
Co–Y
0.150
–
CoQ2 –Y
Cu–Y
0.090
–
2.05
–
0.25
–
8.20
–
1.14
–
CuQ2 –Y
0.040
0.89
0.11
8.12
0.46
a M = Fe, Co or Cu.
b The amount of metal coordinated with 8-quinolinol (Q).
c Estimated by elemental analysis and total amount of H = H in water + H in Q.
d Quantification based on TG/DSC results.
c
Appl. Organometal. Chem. 2011, 25, 262–269
Copyright ꢀ 2011 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/aoc

Y. Yang et al.
Table 3. Thermogravimetric analysis data of catalysts
Temperature
range (◦C)
Weight
loss (%)
Group
lossa
Type of
loss
Materials
Fe-Y
150–280
150–266
266–750
150–270
150–263
363–750
150–269
150–257
257–750
14.1
13.3
4.0
nH2O
nH2O
Q
Endothermic
Endothermic
Exothermic
Endothermic
Endothermic
Exothermic
Endothermic
Endothermic
Exothermic
FeQ3-Y
Co-Y
13.8
13.5
3.0
nH2O
nH2O
Q
CoQ2-Y
Cu-Y
14.2
14.0
3.0
nH2O
nH2O
Q
CuQ2-Y
a Q stands for 8-quinolinol.
Figure 5. TG/DSC curves of Fe–Y (solid lines) and FeQ3 –Y (dashed lines).
loadings are always higher or similar for the MQn –Y zeolites as
compared with the exchanged M–Y (Table 2), suggesting that the
encapsulated metal complexes are stable in zeolite matrix and
there is no metal leaching during complex encapsulation. This is
quite different from the Mn leaching after the introduction of salen
ligands into the zeolite–Y as reported in the literature.[27]
The differential scanning calorimetry (DSC) data along with the
percentage weight loss at different steps and their assignments
are presented in Table 3. TG and DSC profiles of two representative
samples, Fe–Y and FeQ3 –Y, are reproduced in Fig. 5. An endother-
mic loss◦of intra-zeolite water occurs between temperature 150
and 280 C in Fe–Y and between 150 and 266 ◦C in FeQ3 –Y,[28] al-
though the catalysts were dried at 150 ◦C. As illustrated in Table 3,
14.1 and 13.3% weight loss of water was detected in Fe–Y and
FeQ3 –Y, respectively, which corresponds to 110 water molecules
in their respective unit cell. The second step involves the slow but
exothermic weight loss in a wider temperature range due to de-
composition of the chelating ligand.[29,30] A very small percentage
weight loss (4.0% for FeQ3 –Y) indicates the presence of only small
amounts of metal complexes in zeolite cavities. It is slightly lower
than the ligand content estimated by N elemental analysis (6.5%).
Similar results can be obtained from cobalt and copper catalysts
(see Table 3 and Fig. S4).
Figure 4. UV-vis spectra of (A) (a) Co–Y, (b) CoQ2 –Y and (c) CoQ2 (inset);
(B) (a) Fe–Y, (b) FeQ3 –Y and (c) FeQ3 (inset); and (C), (a) Cu–Y, (b) CuQ2 –Y
and (c) CuQ2 (inset).
formed in the supercages are rationally 1 : 3 (Fe : Q) stoichiometric
compounds.Ontheotherhand,1.8and1.6moietiesof8-quinolinol
per unit cell for CoQ2 –Y and CuQ2 –Y demonstrate that the cobalt
and copper complexes formed in the supercages are rationally
1 : 2 stoichiometric compounds. It is also found that the metal
Entrapment of metal complexes inside the zeolite cages and
not on their external surfaces was previously demonstrated by
c
wileyonlinelibrary.com/journal/aoc
Copyright ꢀ 2011 John Wiley & Sons, Ltd.
Appl. Organometal. Chem. 2011, 25, 262–269

Iron(III), cobalt(II) and copper(II) complexes bearing 8-quinolinol
Table 4. Effect of different catalysts on the aerobic oxidation of styrene
Product selectivity (mol%)c
M (mmol
g
Styrene
TOF
(h−1 b
Entry
Catalysts
−1) total
conversion (%)a
)
So
Bza
Others
1
2
Fe–Y
0.204
2.048
0.224
0.217
0.518
0.080
2.880
0.109
0.065
2.840
0.045
60.6
80.1
83.4
74.2
66.5
63.3
51.0
56.7
54.2
61.8
89.3
74.3
89.4
30.1
43.2
24.5
37.4
45.1
33.2
46.2
39.2
30.8
40.4
31.0
61.7
56.8
58.2
60.7
51.5
58.5
53.0
59.6
62.8
59.6
59.0
8.2
0
FeQ3
3
FeQ3 –Y
FeQ3 –Y (2nd)
93.1
17.3
1.9
3.4
8.3
0.8
1.2
6.4
0
4
85.5
d
5
Fe–Y + FeQ3
Co–Y
74.7
6
197.8
117.0
130.0
208.5
441.4
637.9
7
CoQ2
8
CoQ2 –Y
Cu–Y
9
10
11
CuQ2
CuQ2 –Y
10.0
a Reaction conditions: supported catalyst 50.0 mg (FeQ3 5.5 mg, CoQ2 2.0 mg and CuQ2 1.0 mg using the same molar concentration of metal ions as
compared with FeQ3 –Y, CoQ2 –Y and CuQ2 –Y), styrene 1.14 ml (10 mmol), CH3CN 10 ml, flow of air 80 ml min−1, isobutyraldehyde 2.28 ml (25 mmol),
temperature 80 ◦C and duration 8 h.
b TOF (turnover frequency): moles of substrate converted per mole metal ion per hour.
c So, styrene oxide; Bza, benzaldehyde; and Others, include benzoic acid, phenylacetaldehyde and 1-phenylthane-1,2-diol.
d Fe–Y, 18.0 mg and FeQ3, 3.7 mg.
nitrogen adsorption and SEM analyses. However, TG/DSC analysis
can substantiate this result more accurately. Taking into account
the content of Na in the zeolite and the fact that only 1%
of the cationic exchange sites are on the external surface of
the zeolite,[27,31] it can be calculated that, if all the external
ionic exchange sites of the zeolite were completely occupied
by transition metal ions followed by the complete complexing
with 8-quinolinol, the weight loss due to the combustion of these
complexes present on the surface would be about 0.14% for
FeQ3 –Y, 0.13% for CoQ2 –Y and 0.15% for CuQ2 –Y, respectively.
Actually, a Q weight loss of 3.0–4.0% was found, and thus most
of the coordination compounds should be in the cavities of the
zeolite.
more active than the resulting CoQ2 –Y and shows a higher
TOF value (Table 4, entries 6 and 8). The activity reduction
of CoQ2 –Y may be ascribed to the constraints and steric
hindrance after encapsulation of cobalt complexes in the zeolite
supercages.
The encapsulated metal complexes show many advantages
over their homogeneous counterparts. Firstly, the catalytic activity
of MQn –Y is always found to be better than their respective non-
encapsulated complexes (Table 4, entries 2, 3, 7, 8, 10 and 11).
The reason for the higher activity of encapsulated complexes is
the site-isolation of the metal complexes.[32] Secondly, the pure
complexeswerecompletelydestroyedduringthefirstrunsincethe
color was significantly changed. Taking CoQ, for example, during
thereaction,thecolorofsolutionchangedfromlightyellowtodark
brown. This may be related to the irreversible deactivation due
to formation of µ-oxo and µ-peroxo dimeric and other polymeric
species, especially when using oxidant.[33] On the other hand,
zeolite encapsulated metal complexes did not undergo any color
change during the reaction and could be easily separated and
reused. As illustrated in Table 4, the recycled FeQ3 –Y displays
similar activity and product distribution as compared with the
fresh one.
To further verify the heterogeneity of the zeolite encapsulated
catalysts, leaching tests for FeQ3 –Y, CoQ2 –Y and CuQ2 –Y
catalysts were performed. We filtered the catalyst at the reaction
temperature (80 ◦C), e.g. the FeQ3 –Y after 1.0 h (24.9% styrene
conversion). At this time, half the volume was filtered and the
resulting clear solution was allowed to react. The percentage
of leaching was estimated by comparing the time–conversion
plot of the twin reactions with and without solid. It was
found that styrene could just be converted at a very low rate
(Fig. 6a) and this trend is almost parallel to the blank experiment
(not shown). The filtrate neither shows any activity towards
oxidation reaction nor does the ICP-AES analysis show the
presence of iron in the filtrate. However, as for CoQ2 –Y and
CuQ2 –Y catalysts, it was found that, after the hot filtration,
the mother liquor reacted further at roughly the same rate
as that observed when the catalyst was not filtered (Fig. 6b
Catalytic Properties
To ensure a fair comparison, the molar ratios of metal ions
to styrene were the same in the pure and encapsulated
complexes for each reaction and relevant control experiments
were performed. Table 4 summarizes the catalytic results for the
aerobicoxidationofstyreneovervariouscatalysts.ExceptforNa–Y
(not listed), all the metal-exchanged zeolites and metal complexes
whether encapsulated or not exhibit activity. This shows that
the active sites are exchanged metal ions and encapsulated
metal complexes. Actually, according to the elemental analysis,
the majority of the exchanged transition metals in MQn –Y are
coordinated with 8-quinolinol, i.e. 67.7% for FeQ3 –Y, 81.8% for
CoQ2 –Y and 88.9% for CuQ2 –Y. This demonstrates that the
activity of MQn –Y is mainly conducted by the encapsulated
metal complexes, not exchanged metal ions. Both FeQ3 –Y
and CuQ2 –Y are more active than their corresponding metal-
exchanged zeolites and pure metal complexes. The mechanical
mixture of Fe–Y (18.0 mg) and FeQ3 (3.7 mg) shows 66.5%
styrene conversion under the same reaction conditions, which
is lower than that obtained from FeQ3 –Y, although the total
amount of iron was equivalent to that of FeQ3 –Y. It is probable
that there exists synergism between the exchanged metal
ions and encapsulated metal complexes. However, Co–Y is
c
Appl. Organometal. Chem. 2011, 25, 262–269
Copyright ꢀ 2011 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/aoc

Y. Yang et al.
Figure 6. (aˆ-¡): Time-conversion plot for aerobic oxidation of styrene in combination with isobutyraldehyde as co-reductant. (aˆ−3): Heterogeneous
reaction check by continuing the reaction after removing the catalyst after 1.5 h (removing FeQ3 –Y after 1.0 h). Reaction conditions: 10 mmol of styrene,
10 ml of CH3CN, 25 mmol of isobutyraldehyde, 50 mg of catalyst (a, FeQ3 –Y; b, CoQ2 –Y and c, CuQ2 –Y), flow of air 80 ml/min and temperature 80 ◦C.
and c), which leads to the inevitable conclusion that all of
the observed activity can be attributed to homogeneous Co(II)
or Cu(II) species leached from the zeolite. According to the
literature,[34] we suppose that the remarkable leaching occurring
for CoQ2 –Y and CuQ2 –Y can be attributed to the change in the
unstable coordinate environment of the active sites during the
reaction.
isolation as well as the synergism effect between the exchanged
metal ions and encapsulated metal complexes. Leaching tests
confirmed only FeQ3 –Y was stable against leaching of active
species into solution and can be recycled without significant loss
of activity and selectivity within successive runs. Moreover, the
selectivity to benzaldehyde was higher than to styrene oxide in all
cases due to the acidic nature of zeolite matrix.
Additionally, we found that the acidic nature of hydrophilic
zeolite was reflected in the product distribution. The catalytic
results over homogeneous complex catalysts show that benzalde-
hyde (Bza) is the major oxidation product, followed by styrene
oxide (So), similar to the literature.[35,36] The encapsulated com-
plexes show similar product distributions to the homogeneous
counterparts, except for the presence of phenylacetaldehyde and
1-phenylethane-1,2-diol, formed via ring opening of phenyloxi-
rane on acid sites. Lower styrene oxide selectivity for MQn –Y and
M–Y is also connected with this ring opening reaction, suggesting
that the acidity of the zeolite support decreases the selectivity
to phenyloxirane. Comparing these results with our earlier data
on the epoxidation of styrene over tethered iron Schiff base
complexes on SBA-15 under identical reaction conditions,[37] we
observed that here the selectivity to phenyloxirane was lower,
further indicating that the acidity of the zeolite probably acceler-
ates further isomerization of styrene oxide to phenylacetaldehyde.
However, this selectivity to styrene oxide is still higher than that
obtained by zeolite–Y encapsulated Mn complexes,[23] which can
probably be explained by the higher Si–Al molar ratio of zeolite–Y
in this study associated with lower acidity.
Supporting information
Supporting information may be found in the online version of this
article.
Acknowledgment
This work was supported by the National Basic Research Program
of China (2004CB217804) and the National Natural Science
Foundation of China (20673046).
References
[1] G. D. Yadav, A. A. Pujari, Org. Process Res. Dev. 2000, 4, 88.
[2] B. S. Lane, K. Burgess, Chem. Rev. 2003, 103, 2457.
[3] A. Corma, H. Garcia, Chem. Rev. 2002, 102, 3837.
[4] T. Sato, J. Dakka, R. A. Sheldon, J. Chem. Soc. Chem. Commun. 1994,
16, 1887.
[5] M. A. Camblor, A. Corma, A. Martinez, J. Perez Parienta, J. Chem. Soc.
Chem. Commun. 1992, 8, 589.
[6] G. Tozzola, M. A. Mantegazza, G. Ranghino, G. Petrini, S. Bordiga,
G. Ricchiardi, C. Lamberti, R. Zulian, A. Zecchina, J. Catal. 1998, 179,
64.
[7] A. Singh, T. Selvam, J. Mol. Catal. A: Chem. 1996, 113, 489.
[8] B. F. Sels, D. De Vos, P. A. Jacobs, Tetrahedron Lett. 1996, 37, 8557.
[9] T. Joseph,D. Srinivas,C. S. Gopinath,S. B. Halligudi,Catal.Lett.2002,
83, 209.
Conclusions
Fe(III), Co(II) and Cu(II) complexes of 8-quinolinol were successfully
encapsulated in the supercages of zeolite–Y by a flexible
ligand method, as evidenced by FT-IR, UV–vis spectroscopy,
N2 adsorption/desorption techniques. The loadings of metal
and ligand were quantified by ICP-AES, TG/DSC and CHN
elemental analysis, and the internal location of active species was
demonstrated. The structural integrity of zeolite throughout the
encapsulation procedure was confirmed by XRD and SEM studies.
These encapsulated metal complexes displayed better activity
than their corresponding homogeneous analogs due to the site-
[10] T. Joseph, S. B. Halligudi, J. Mol. Catal. A: Chem. 2005, 229, 241.
[11] R. A. Sheldon, I. W. C. E. Arends, A. Dijksman, Catal. Today 2000, 57,
157.
[12] C. R. Jacob, S. P. Verkey, P. Ratnasamy, Micropor. Mesopor. Mater.
1998, 22, 465.
[13] G. J. Hutchings, Chem. Commun. 1999, 999, 301.
[14] K. J. Balkus, M. Eissa, R. Levado, J. Am. Chem. Soc. 1995, 117, 10753.
[15] N. Herron, Inorg. Chem. 1986, 25, 4714.
[16] C. Bowers, K. P. Dutta, J. Catal. 1990, 122, 271.
[17] D. Srinivas, S. Sivasanker, Catal. Surv. Asia 2003, 7, 121.
[18] R. Vijayalakshmi, S. K. Kulshreshtha, Micropor.Mesopor.Mater. 2008,
111, 449.
c
wileyonlinelibrary.com/journal/aoc
Copyright ꢀ 2011 John Wiley & Sons, Ltd.
Appl. Organometal. Chem. 2011, 25, 262–269

Iron(III), cobalt(II) and copper(II) complexes bearing 8-quinolinol
[19] H. S. Wang, J. H. Huang, S. J. Wu, C. Xu, L. H. Xing, L. Xu, Q. B. Kan,
Mater. Lett. 2006, 60, 2662.
[30] Y. Yang, Y. Zhang, S. J. Hao, Q. B. Kan, Catal.Commun. 2010, 11, 808.
[31] E. Pa´ez-Mozo, N. Cabrimas, F. Lucaccioni, D. D. Acosta, P. Patron,
A. L. Cmestra, P. Ruiz, B. Delmon, J. Phys. Chem. 1993, 97, 12819.
[32] Y. Yang, Y. Zhang, S. J. Hao, J. Q. Guan, H. Ding, F. P. Shang, P. P. Qiu,
Q. B. Kan, Appl. Catal. A: Gen. 2010, 381, 274.
[20] C. B. Liu, Y. K. Shan, X. G. Yang, X. K. Ye, Y. Wu, J. Catal. 1997, 168, 35.
[21] M. M. Htay, M. M. Oo, Wld Acad. Sci. Eng. Technol. 2008, 48, 114.
[22] J. A. Kaduk, J. Faber, The Rigaku J. 1995, 12, 14.
[23] M. Salavati-Niasari, Micropor. Mesopor. Mater. 2006, 95, 248.
[24] J. Sebastian, K. Mohan, R. V. Jasra, J. Catal. 2006, 244, 208.
[25] C. Jin, W. B. Fan, Y. J. Jia, B. B. Fan, J. H. Ma, R. F. Li, J. Mol. Catal. A:
Chem. 2006, 249, 23.
[26] X. Y. Qi, G. J. Wang, X. K. Ye, Y. Wu, Chem. J. Chin. Univ. 1996, 17, 939.
[27] M. Silva, C. Freire, B. de Castro, J. L. Figueiredo, J. Mol. Catal. A:Chem.
2006, 258, 327.
[33] X. L. Wang, G. D. Wu, J. P. Li, N. Zhao, W. Wei, Y. H. Sun, Catal. Lett.
2007, 119, 87.
[34] T. Kovalchuk, H. Sfihi, V. Zaitsev, J. Fraissard, J. Catal. 2007, 249, 1.
[35] V. Indira, S. B. Halligudi, S. Gopinathan, C. Gopinathan, React. Kinet.
Catal. Lett. 2001, 73, 99.
´
[36] A. Zsigmond, A. Horva´th, F. Notheisz, J. Mol. Catal. A: Chem. 2001,
171, 95.
[28] M. R. Maurya, S. J. J. Titinchi, S. Chand, J. Mol. Catal. A: Chem. 2003,
201, 119.
[37] Y. Yang, J. Q. Guan, P. P. Qiu, Q. B. Kan, Transition Met. Chem. 2010,
35, 263.
[29] P. K. Saha, B. Dutta, S. Jana, R. Bera, S. Saha, K. Okamoto, S. Koner,
Polyhedron 2007, 26, 563.
c
Appl. Organometal. Chem. 2011, 25, 262–269
Copyright ꢀ 2011 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/aoc