Bio-inspired Mo(II) complexes as active catalysts in homogeneous and heterogeneous olefin epoxidation

Article abstract of DOI:10.1016/j.apcata.2010.06.011

New hydrotalcite clay materials functionalized with molybdenum(II) complexes holding potential catalytic properties were prepared by a stepwise procedure. The material was first calcined at 823 K for 4 h, to eliminate all the carbonate ions; the layered structure was reconstructed after treatment with a solution of either tryptophane (trpH) or phenylalanine (pheH) in a KOH aqueous solution of dmf at 343 K. Subsequent impregnation with a solution of the organometallic precursor [MoI₂(CO)₃(NCCH ₃)₂] led to substitution of the nitriles by the ligands affording the organometallic derivatized clays. The equivalent homogeneous phase complexes [MoI₂(CO)₃L] (L = trpH or pheH) were also prepared for comparison purposes. All the materials were characterized by powder X-ray diffraction, FTIR and ¹³C CP MAS-DD solid state NMR. The complexes were also characterized by FTIR, UV/vis, elemental analysis, and ¹H and ¹³C solution NMR accordingly. Both the complexes and these new materials, the latter containing 3.90 wt.% and 5.42 wt.% of Mo, respectively for HT-trp-Mo and HT-phe-Mo materials, catalyzed the epoxidation of cyclooctene and styrene at 328 K. The heterogeneous catalytic systems performed better showing that the presence of the clay is not innocent. In addition, further oxidation of styrene yielding benzaldehyde has occurred. A possible mechanism is proposed based on the detection of two reaction intermediaries by GC/MS. The heterogeneous catalysts also show superior handling capabilities due to Mo increased resistance to moisture and to separation from reaction medium.

Full text of DOI:10.1016/j.apcata.2010.06.011

Applied Catalysis A: General 384 (2010) 84–93
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Bio-inspired Mo(II) complexes as active catalysts in homogeneous and
heterogeneous olefin epoxidation
a
a
a
b
a,∗
Cristina I. Fernandes , Nuno U. Silva , Pedro D. Vaz , Teresa G. Nunes , Carla D. Nunes
a
Departamento de Química e Bioquímica, CQB, Faculdade de Ciências da Universidade de Lisboa, 1749-016 Lisboa, Portugal
Centro de Química Estrutural, Instituto Superior Técnico, Av. Rovisco Pais, 1049-001 Lisboa, Portugal
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 9 April 2010
Received in revised form 2 June 2010
Accepted 4 June 2010
Available online 11 June 2010
New hydrotalcite clay materials functionalized with molybdenum(II) complexes holding potential cat-
alytic properties were prepared by a stepwise procedure. The material was first calcined at 823 K for
4
h, to eliminate all the carbonate ions; the layered structure was reconstructed after treatment with a
solution of either tryptophane (trpH) or phenylalanine (pheH) in a KOH aqueous solution of dmf at 343 K.
Subsequent impregnation with a solution of the organometallic precursor [MoI2(CO)3(NCCH3)2] led to
substitution of the nitriles by the ligands affording the organometallic derivatized clays. The equivalent
homogeneous phase complexes [MoI2(CO)3L] (L = trpH or pheH) were also prepared for comparison pur-
Keywords:
Molybdenum
Amino acids
1
3
poses. All the materials were characterized by powder X-ray diffraction, FTIR and C CP MAS-DD solid
state NMR. The complexes were also characterized by FTIR, UV/vis, elemental analysis, and 1H and
13
C
Clays
Heterogeneous catalysis
Oxidation catalysis
solution NMR accordingly. Both the complexes and these new materials, the latter containing 3.90 wt.%
and 5.42 wt.% of Mo, respectively for HT–trp–Mo and HT–phe–Mo materials, catalyzed the epoxidation
of cyclooctene and styrene at 328 K. The heterogeneous catalytic systems performed better showing that
the presence of the clay is not innocent. In addition, further oxidation of styrene yielding benzaldehyde
has occurred. A possible mechanism is proposed based on the detection of two reaction intermediaries
by GC/MS.
The heterogeneous catalysts also show superior handling capabilities due to Mo increased resistance
to moisture and to separation from reaction medium.
©
2010 Elsevier B.V. All rights reserved.
1
. Introduction
Among the possible strategies for introducing anions, rehydra-
tion of a calcined precursor has proved advantageous compared to
direct anion-exchange [3], as competition between anions can be
efficiently prevented, as long as carbonate, the anion found in the
minerals, is excluded. For this reason manipulation of the materials
under an inert atmosphere is usually crucial. In another interesting
route, the synthesis of the double hydroxide is carried out in the
presence of the anions chosen to be part of the final material [7,8].
Despite the nature of the anions can be wide-ranging, if carrying
proper functional groups, it can undergo further reactions, namely
coordination to a metal center [6]. In addition calcined hydrotal-
cite has also been shown to catalyze reactions on its own, such as
the aziration of oxiranes [9–11], or aldol condensation [12] serv-
ing as an environmentally friendly catalyst, which can undergo
recyclability easily.
Heterogeneous catalysis is advantageous as a process itself since
it allows an easy separation of the products from the catalyst
in a given reaction, being therefore favored in many industrial
processes. The combination of organometallic complexes and a par-
ticular support has received much attention in recent years owing
to the new perspectives it opens, since the first are known for their
capabilities in selective and efficient promotion of many reactions
Layered double hydroxides are an important class of natural
occurring minerals whose syntheses in the lab are easily acces-
sible [1]. Such materials, also known as anionic clays have the
capability of anion swelling making it very versatile. In addition,
despite the rich intercalation chemistry that can be played with,
it also shows a wide range of variations of its composition. This
latter aspect is based on the versatility of the general formula
2
1
+
−x
3+
m−
2+
3+
[
M
M
(OH) ](A )x/m · nH O where both M and M can be
x
2
2
m−
−
−
varied [2], and A
is usually a simple anion such as Cl , NO or
3
2
−
3+
CO3 . In its general formula, x measures the amount of M relative
to (M3+ + M ), correlating with the spacing between layers. More-
over, the spacing changes concomitantly with both the size of the
intercalated anions and the degree of hydration, and many works
dealt already with this ranging from the intercalation of simple
organic and inorganic anions to complicated biological ones [3–6].
2+
∗ Corresponding author. Tel.: +351 217 500 876; fax: +351 217 500 088.
E-mail address: cmnunes@fc.ul.pt (C.D. Nunes).
0
926-860X/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2010.06.011

C.I. Fernandes et al. / Applied Catalysis A: General 384 (2010) 84–93
85
in solution, while the latter optimize recovery and recycling of the
catalysts [13,14].
correlation given by Becke’s three-parameter hybrid functional
with Lee, Yang and Parr’s gradient-corrected correlation functional
[18]. The basis set for Mo and I consisted of the standard SDD basis
set [19] augmented with a f polarization function in all calculations
[20], while the 6-311G** basis set [21] was used on all remaining
atoms. The starting geometries of 2 and of 3 were based on spectro-
scopic and Cambridge Structural Database (CSD) [22] data and were
optimized without any symmetry constraints. Frequency calcula-
tions were performed at the same level of theory to confirm the
nature of the stationary points, yielding no imaginary frequencies
for the minima.
The tailoring of the catalyst based on both the supporting mate-
rial and the incorporation of the organometallic precursor must be
continuous. In particular, porous materials are very promising can-
didates for such purposes, as they offer a large surface area inside
the pores, but must be designed having in mind the need to afford
a stable (and reactive) material without blocking the access to the
catalytic active sites, so that the reaction can proceed and a good
contact between reactants is promoted.
In this work we have prepared amino acid containing clays,
which were consequently derivatized with organometallic Mo
cores. The resulting materials were tested in the catalytic epoxi-
dation of olefins where the heterogeneous performed better than
the corresponding complexes in homogeneous phase catalysis. All
this will be reported and discussed in more detail in the following
sections.
2.2. Preparation of the complexes and clay materials
2.2.1. General procedure for preparation of complexes
[MoI (CO) L] (L = trpH or pheH) (2, 3)
2
3
A solution of [MoI (CO) (NCCH ) ] (0.50 mmol) in MeOH
2
3
3 2
(10 mL) was treated with a solution of the desired ligand, either
trpH or pheH (1 mmol), in MeOH (5 mL). The resulting solution was
stirred for 14 h at room temperature. The desired complex pre-
cipitated and the solvent was filtered off. The complex was then
washed with hexane and dried under vacuum. All manipulations
were carried out under N2 atmosphere.
2
. Experimental
2.1. Materials
All preparations and manipulations were performed using
standard Schlenk techniques under nitrogen. Commercial grade
solvents were dried and deoxygenated by standard procedures,
distilled under nitrogen and kept over 4 Å molecular sieves
2.2.2. General procedure for preparation of amino acid
intercalated clay materials HT–L (L = trp or phe) (4, 5)
A
mixture of trpH (7.30 mmol) or pheH (7.30 mmol) in
(
3 Å for CH CN). Tryptophane (trpH) and phenylalanine (pheH),
freshly distilled dmf (10 mL) and deionized water (20 mL) with
1 equivalent of KOH was stirred until complete solubilization. This
solution was then added to a suspension of calcined HT (1.00 g)
in freshly distilled dmf (25 mL) at 343 K. The reaction mixture was
stirred at the same temperature for 48 h. All manipulations were
carried out under N2 atmosphere to prevent the uptake of car-
bonate anions. The resulting material was then filtered off, washed
three times with deionized water (3× 20 mL), and dried in a desic-
cator under vacuum.
3
Mg,Al-hydrotalcite (HT), tert-butylhydroperoxide (TBHP), and
cyclooctene (cy8) and styrene (sty), were purchased from Aldrich
and used as received. Prior to the intercalation experiments com-
mercial hydrotalcite was calcined (823 K for 4 h) in order to
eliminate all the carbonate anions [9]. The precursor organometal-
lic complex [MoI (CO) (NCCH ) ] (1) was prepared according to
2
3
3 2
the literature methods [8,15].
Powder XRD data were collected on a Phillips PW1710 diffrac-
tometer using Cu K␣ radiation filtered by graphite. FTIR spectra
were measured with a Nicolet Nexus 6700 FTIR spectrometer using
KBr pellets (for complexes) in transmission mode and also using dif-
fuse reflectance (for clay materials). All FTIR spectra were measured
2.2.3. General procedure for preparation of organometallic clay
materials HT–L–Mo (L = trp or phe) (6, 7)
A solution of [MoI (CO) (NCCH ) ] (1) (0.34 g, 0.65 mmol) in dry
2
3
3 2
−1
using 2 cm resolution.
CH Cl (5 mL) was added to a suspension of 1.00 g of the HT inter-
2 2
UV/vis spectra were measured with a Shimadzu UV-2450PC
spectrometer equipped with a temperature controlled cell. All
calated material HT–L (4 or 5) in dry CH Cl2 (20 mL). The reaction
mixture was stirred under a N2 atmosphere, at room temperature,
for 24 h. The resulting material was then filtered off, washed twice
2
−4
spectra were measured as 10 M solutions using CH CN as solvent
3
in the 190–900 nm range.
with CH Cl2 (2× 20 mL), and dried in a desiccator under vacuum.
2
1H and 13C solution NMR spectra were obtained at 400.13 MHz
and 100.62 MHz, respectively, with a Bruker Avance 400 spectrom-
2.3. Characterization of complexes and clay materials
eter using CD OD or (CD ) SO as solvent. Chemical shifts are quoted
3
3 2
in ppm from TMS.
2.3.1. [MoI (CO) (trpH)] (2)
2
3
13C solid state NMR measurements were performed at room
temperature on a Bruker MSL 300P spectrometer operating at
Yield: 92%. C14H12N O5I Mo (639.95) calcd.: C 26.26, H 1.89, N
2 2
4.38; found: C 26.17, H 2.12, N 4.54.
7
5.47 MHz. The standard magic angle spinning (MAS) cross-
IR (KBr, ꢀ cm−1): 3455 (vs), 1986 (vw) (s), 1932 (m), 1892 (s),
1754 (s), 1627 (vs), 1496 (w), 1455 (w), 1030 (m).
polarization–dipolar decoupling RF pulse sequence (CP–DD) was
used under about 3.8 kHz spinning rate. 13C spectra were recorded
UV/vis [10 M in CH CN, ꢁ nm (ε M cm )]: 282 (30460), 345
−4
−1
−1
3
◦
with a pulse duration of 5 ␮s (corresponding to 90 magnetiza-
(5640), 440 (1870).
1
tion tip angle), 2 ms contact time, 4 s recycling delay and a number
of scans of about 15000. The Hartmann–Hahn condition was opti-
mized using glycine, also the external reference to set the chemical
H NMR (400.13 MHz, CD OD, r.t., ı ppm): 7.70 (d, H ,
3
9
J8,9 = 8.6 Hz), 7.35 (d, H , J = 8.7 Hz), 7.16 (s, H5), 7.09 (t, H ), 7.01
6
6,7
8
(t, H7, J7,8 = 8.0 Hz), 3.58–3.62 (m, H ), 3.33 (m, H3b), 2.86–2.98 (m,
2
1
3
shift scale ( CO at 176.1 ppm).
H3a).
13
Microanalyses for CHN and Mo quantification were performed
at CACTI, University of Vigo. CHN analyses were performed on a
Fisons EA 1108; Mo quantification was performed on a Perkin Elmer
Optima 4300DV using In as internal standard.
C NMR (100.62 MHz, CD OD, r.t., ı ppm): 178.9 (COOH), 136.9
3
(C11), 127.5 (C10), 123.3 (C5), 121.0 (C ), 118.3 (C ), 118.1 (C7), 110.9
8
9
(C ), 110.2 (C ), 56.2 (C ), 30.2 (C ).
6
4
2
3
DFT calculations [16] were performed using the Gaussian03 [17]
program, revision C02, with the B3LYP hybrid functional, which
includes a mixture of Hartree–Fock exchange with DFT exchange-
2.3.2. [MoI (CO) (pheH)] (3)
2 3
Yield: 89%. C12H11NO5I Mo (600.78) calcd.: C 23.97, H 1.85, N
2
2.33; found: C 23.58, H 1.88, N 2.21.

8
6
C.I. Fernandes et al. / Applied Catalysis A: General 384 (2010) 84–93
IR (KBr, ꢀ cm−1): 3418 (vs), 2065 (w), 1982 (m), 1932 (vs), 1745
m), 1626 (vs), 1593 (s), 1456 (m), 1415 (m), 1356 (m), 747 (s).
(
−4
−1
−1
UV/vis [10 M in CH CN, ꢁ nm (ε M cm )]: 288 (21600), 353
3
(
8740), 436 (2300).
1
H NMR (400.13 MHz (CD ) SO, r.t., ı ppm): 7.26 (s bd, 5H, Ph),
3
2
4
.12 (s bd, H ), 3.82 (s, H3a), 3.17 (s, H3b).
2
13
C NMR (100.62 MHz (CD ) SO, r.t., ı ppm): 182.6.9 (COOH),
3
2
1
38.1 (C ), 129.0 (C
), 128.6 (C5,5 ), 127.0 (C7), 59.8 (C ), 38.0 (C ).
ꢀ ꢀ
6,6
2 3
4
2.3.3. HT–trp (4)
Elemental analysis found: C 21.52, N 4.20, H 4.85.
−1
IR (KBr ꢀ cm ): 3404 (vs), 1664 (s), 1586 (s), 1458 (m), 1359
s), 933 (s), 745 (s) 671 (s).
(
13
−
C CP MAS NMR (ı ppm): 177.0 (COO ), 136.8 (C11), 127.0
C10/C5), 121.2 (C ), 119.7 (C7/C ), 112.0 (C ), 109.7 (C ), 56.3 (C ),
(
9
8
6
4
2
2
9.8 (C ).
3
2.3.4. HT–phe (5)
Elemental analysis found: C 18.79, N 2.02, H 4.96.
−1
IR (KBr, ꢀ cm ): 3471 (vs), 1571 (s), 1496 (m), 1456 (m), 1367
s), 913 (s), 763 (s).
(
Scheme 1.
13
−
C CP MAS NMR (ı ppm): 178.6 (COO ), 136.0 (C ), 130.4 (C
ꢀ
),
4
6,6
1
28.9 (C5,5 /C7), 57.1 (C ), 38.0 (C ).
ꢀ
2 3
3. Results and discussion
3.1. Characterization of the complexes
2.3.5. HT–trp–Mo (5)
Elemental analysis found: C 14.58, N 2.74, H 4.05, Mo 3.90.
IR (KBr, ꢀ cm ): 3446 (vs), 2035 (m), 1893 (m), 1592 (vs), 1365
−1
The precursor complex [MoI2(CO)3(NCCH3)2] (1) [15]
was allowed to react with 1 equivalent of either trpH or
pheH (Scheme 1), affording the complexes formulated as
(
s).
13
−
C CP MAS NMR (ı ppm): 181.7 (COO ), 136.4 (C11), 127.2
[
MoI (CO) (trpH)] (2) and [MoI (CO) (pheH)] (3), respectively,
2 3 2 3
(
C5/C10), 122.2 (C7/C /C ), 111.4 (C /C ), 56.1 (C ), 27.3 (C ).
8
9
4
6
2
3
following the procedure outlined in Scheme 1 (numbering scheme
of ligands is also presented).
2
.3.6. HT–phe–Mo (6)
DFT calculations using the B3LYP functional as implemented
in the G03 package have also been carried out in order to eval-
uate the geometry of complexes in terms of coordination mode
and vibrational wavenumbers (discussed later). Another reason
for conducting such simulation was that this type of Mo-centered
heptacoordinated complexes are difficult to crystallize as evi-
denced by the reduced number of structures in the Cambridge
Structural Database (CSD) [22]. The calculated geometries, rep-
resented in Fig. 1, were found to be true minima without any
imaginary harmonic wavenumbers, after calculation of second
derivatives.
Elemental analysis found: C 11.09, N 1.12, H 4.00, Mo 5.42.
−1
IR (KBr, ꢀ cm ): 3429 (vs), 2088 (w), 1912 (m), 1753 (m), 1589
vs), 1366 (vs), 929 (vs).
(
(
13
−
C CP MAS NMR (ı ppm): 180.1 (COO ), 136.3 (C ), 128.7
4
C5,5 /C6,6 /C7), 56.9 (C ), 37.6 (C ).
ꢀ ꢀ
2 3
2.4. Catalytic tests
Cyclooctene (0.943 mL, 7.3 mmol) or styrene (0.880 mL,
.7 mmol) and tert-butylhydroperoxide (200 mol%, 2.65 mL for cy8
7
In fact, these geometries were found to match some reported
structures after conducting a search in the CSD, with ligands hold-
ing a carboxylate group. In this way, given that amino acids behave
as amphoteric species it is quite reasonable that at a pH around 7
both amino acids bear deprotonated carboxylate and protonated
amine groups with an overall neutral charge. Therefore coordina-
tion to the metal centers should be rationalized by means of the
or 2.80 mL for sty, 5.5 M solution in decane) were mixed in the
presence of complexes or materials (1 mol%) in a glass reactor, at
3
28 K, and vigorously stirred. Dibutylether (1.047 mL, 6 mmol) was
2
added to the reaction and used as internal standard and CH Cl2
(
2 mL) as solvent. The final volume of the reaction is ca. 6 mL,
corresponding to a catalyst concentration of 12.2 mM or 12.8 mM
for the case of cy8 and sty, respectively.
−
COO group, thus corroborating with the structures found in the
Conversion and product yields were monitored by taking sam-
ples and analyzing them using a Shimadzu QP2100-Plus GC/MS
system and a capillary column (Teknokroma TRB-5MS). All reac-
tions were conducted under normal atmosphere. Some control
experiments were also conducted under N2 atmosphere to eval-
CSD.
The 1H NMR spectrum of [MoI (CO) (trpH)] (2) shows clear
2
3
differences from those of the precursor complex 1, namely the
absence of a resonance at ca. 2.0 ppm assigned to the acetoni-
trile ligands of 1. After coordination to trpH the 1H NMR spectrum
uate the effect of the presence of O . Such experiments yielded
2
evidences the signals from the ligand which are shifted from the
original free ligand, reflecting the coordination of the Mo-center.
The changes experienced by the protons are in some cases very
clear as shown in Table 1 which summarizes the observed chem-
ical shifts for complexes 2 and 3 and comparison with the free
ligands.
similar results to those carried out under regular atmosphere.
TBHP consumption was determined by taking two samples at
the end (24 h) of each catalytic experiment. One of them was ana-
lyzed directly by GC/MS while the remaining was added to MnO2
to destroy TBHP yielding tert-butanol. The difference in the area
of tert-butanol detected represents the remaining TBHP. The con-
sumption is determined by subtracting this from the initial value.
The characterization accomplished by vibrational spectroscopy
was also useful to track the formation of the complexes,

C.I. Fernandes et al. / Applied Catalysis A: General 384 (2010) 84–93
87
Fig. 1. DFT calculated geometries for complexes [MoI2(CO)3(trpH)] (2) (top) and
MoI2(CO)3(pheH)] (3) (bottom).
[
according to Fig. 2. In both complexes [MoI (CO) (trpH)] and
2
3
[
MoI (CO) (pheH)] the ꢀC N modes are not observed confirming
2 3
the displacement of the labile CH CN ligands from 1.
3
The strong absorptions of the ꢀC O modes are clearly observed
confirming the presence of the [MoI (CO) ] core; this observation
2
II
3
is also indicative that the initial Mo oxidation state is preserved.
In addition, the observed positions of the ꢀC O bands of both com-
plex 2 (observed at 1986 cm , 1932 cm and 1892 cm ) and 3
Fig. 2. FTIR spectra of trpH, HT–trp (4), [MoI2(CO)3(NCMe)2] (1), [MoI2(CO)3(trpH)]
−1
−1
−1
(
2) and HT–trp–Mo (6).
(
observed at 2065 cm 1, 1982 cm−1 and 1932 cm ) are shifted
− −1
−
1
relatively to the precursor complex 1 (observed at 2072 cm
,
3.2. Intercalation of the ligands and organometallic fragment in
−1
−1
and 1921 cm ). In complex 2 further changes are
2
016 cm
also observed. A drastic change is observed in the ꢀC O mode
hydrotalcite
−
1
of the trpH ligand which shifts from 1668 cm
(in free trpH)
The synthetic procedure for the preparation of the hybrid
organometallic intercalated materials is summarized in Scheme 2.
In this work the reconstruction method was employed, by taking
−1
to 1754 cm
predict a shift of 99 cm
agreement with the observed shift of 86 cm . In the case of
complex 3 the shift predicted from DFT values is 92 cm which
compares with a shift of 66 cm observed experimentally, from
in complex 2. DFT calculated wavenumbers also
−1
for this mode, which is in very good
−
1
−
1
−1
−1
−1
bands appearing at 1560 cm (in free pheH) and 1626 cm in 3
(
not shown).
Electronic spectra were also measured to check the oxidation
state of the metal centers. Fig. 3 shows the UV/vis spectra in ace-
II
tonitrile of Mo based complexes 1 and 3 and comparison with the
VI
Mo based [MoO Cl ] complex.
2
2
It is evident that the electronic spectra of both Mo complexes (1
and 3) are quite similar. Both display a medium intensity band in
the 350–400 nm range and a very intense band at ca. 300 nm which
is due to metal-to-ligand charge transfer (MLCT) between Mo and
CO ligands [23]. This is indicative that in both complexes the molyb-
denum centers are in the same oxidation state. A different profile is
detected in the [MoO Cl ] complex spectrum. This complex shows
2
2
only a band at ca. 260 nm assigned to MLCT between Mo and O
23].
Fig. 3. UV/vis spectra in acetonitrile of [MoI2(CO)3(NCMe)2] (1, –·–),
[
[
MoI2(CO)3(pheH)] (3, —) and [MoO2Cl2] (– –).

8
8
C.I. Fernandes et al. / Applied Catalysis A: General 384 (2010) 84–93
Scheme 2.
advantage of the memory effect displayed by the hydrotalcite mate-
rial (hereafter denoted as HT). In this way HT lamellar material was
first submitted to a calcination process at 823 K for 4 h to eliminate
structure leading to a multiphase sample would take place [26].
Alternatively, the high content of hydrogen found may be explained
by the uptake of OH groups. In this way on the basis of the amount of
nitrogen for the two materials, the contents of trp and phe incorpo-
rated are 71% and 63%, respectively, the rest being fulfilled by OH
anions from the solution [8]. In this way the reconstructed clays
can be rationalized as [Mg Al (OH)16](trp)1.42(OH)0.58·4H O for
the CO32 anions from the interlayer region, as described in our
−
previous work [24].
Briefly, during this process the structure of the lamellar material
collapses, and is thereafter rebuilt when the deprotonated amino
acids trp and phe are intercalated inside the interlayer space by
using dimethylformamide (dmf) at 343 K [8,25]. The deprotonation
of the amino acids is carried out in deionized water in the presence
of KOH and the resulting solution is then added to the HT suspen-
sion leading to the preparation of the composite materials HT–trp
6
2
2
HT–trp (4) and [Mg Al (OH)16](phe)1.26(OH)0.74·4H O for HT–phe
6
2
2
(5).
After being dehydrated for some time these hybrid clay
materials were then reacted with the precursor complex
[MoI (CO) (NCMe) ] (1) in dry CH Cl during 24 h at room temper-
2
3
2
2
2
(
4) and HT–phe (5). These materials have been recently reported by
ature, under inert atmosphere, affording the new organometallic
hybrid materials HT–trp–Mo (6) and HT–phe–Mo (7) after filtering
and drying.
us [24]. Results from elemental analyses for materials HT–trp (4)
and HT–phe (5) show that the N content is below of what would be
expected based on a complete reconstruction with the amino acids.
The N contents found for 4 and 5 were 4.20% and 2.02%, respectively.
These values are slightly lower than the maximum expected values
of 5.90% and 3.21%, by the same order. It could suggest that incor-
poration of some carbonate during reconstruction of the layered
The molybdenum loadings for the organometallic compos-
ite materials 6 and 7 were found to be 3.90 wt.% and 5.42 wt.%
−
1
−1
(0.41 mmol g and 0.56 mmol g ), respectively, showing similar
values to those reported earlier for related materials [8], though
higher metal quantities were incorporated in comparison to the
Table 1
1
H and 13C NMR chemical shifts for trpH and pheH ligands and, complexes 2 and 3.
H2
H3a
H3b
H5
H6
H7
H8
H9
trpH
2
3.53
3.60
3.04
2.93
3.33
3.33
7.26
7.16
7.36
7.35
6.96
7.01
7.05
7.09
7.58
7.70
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
COOH
trpH
2
56.1
56.2
27.2
30.2
108.7
110.2
125.8
123.3
112.7
110.9
119.3
118.1
120.3
121.0
122.9
118.3
127.7
127.5
137.3
136.9
174.8
178.9
H2
H3a
H3b
H5,5ꢀ
H6,6ꢀ
H7
pheH
3
3.99
4.12
3.13
3.17
3.29
3.82
7.33
7.43
7.26 (5H)
7.38
C2
C3
C4
C5,5ꢀ
C6,6ꢀ
C7
COOH
pheH
3
56.9
59.8
37.2
38.0
136.3
138.1
130.2
128.6
129.9
129.0
128.4
127.0
174.2
182.6
For numbering scheme, please refer to Scheme 1.

C.I. Fernandes et al. / Applied Catalysis A: General 384 (2010) 84–93
89
Scheme 3.
immobilization in a mesoporous support [27]. Based on these Mo
loading values it is likely that the organometallic Mo-cores inside
the gallery are coordinated to two ligands, as the Mo loading in
6
and 7 is less than half the concentration of intercalated ligands
detected in the precursor materials 4 and 5. It should be mentioned
that the coordination of the Mo-centers to the intercalated ligands
must be rationalized on a different basis from that formulated for
complexes 2 and 3. From the results shown above it comes that
the confined Mo species are most probably coordinated to the N
atoms of the amine groups from the intercalated amino acids, as
evidenced in Scheme 3.
This is based on the rationale that the carboxylate moieties are
most probably interacting with the Mg/Al oxide layers leaving the
Fig. 4. XRD powder patterns of HT–trp (4), HT–phe (5), HT–trp–Mo (6) and
HT–phe–Mo (7) materials.
NH groups free for coordination. To this respect, since the interca-
2
lation of the amino acids is carried out at pH values around 10 the
amine groups are most probably completely deprotonated, there-
fore being able to bind to Mo.
posite material 6 (Fig. 2) the metal carbonyl groups (ꢀC O) at
2035 cm and 1893 cm are shifted relatively to those of com-
−
1
−1
All the new materials were characterized by powder XRD, FTIR
1
3
and solid state C NMR techniques.
plex [MoI (CO) (trpH)] 2 probably owing to specific host guest
2
3
The XRD powder patterns obtained for these lamellar materi-
als are typical of this type of reasonably well-ordered materials,
exhibiting sharp and symmetric 0 0 3 reflections (Fig. 4) [28].
The XRD powder patterns of materials 4 and 5 were found to be
similar to those reported recently [24]. The calculated d parameters
were found to be 18.6 Å and 18.7 Å for both HT–trp and HT–phe
materials, respectively. Assuming the HT sheets to be 4.8 Å thick
interactions between the [Mo(CO) ] core and the layers of the HT
3
derivatized host material. Another reason for such observation may
be the fact that inside the clay the Mo-core is possibly coordinated
to two amino acid anions while the complex is formulated with
only one, as described earlier in this section. The same trends are
observed in HT–phe–Mo (7), with the ꢀC O modes of the [Mo(CO)3]
−
1
−1
and 1912 cm . No bands that might be
fragment at 2088 cm
−1
assigned to the ꢀC N modes, observed at 2304 cm and 2227 cm
−
1
[
1,3], it is possible to estimate the gallery height of HT–trp and
HT–phe to be 13.8 Å and 13.9 Å, respectively, indicating that such
guest molecules are in the upright position.
in the precursor 1, from the acetonitrile ligands, were detected in
either HT–trp–Mo (6) or HT–phe–Mo (7).
After treatment of materials 4 and 5 with complex 1, the inten-
sity of the 0 0 3 reflections is significantly reduced in materials
HTC–trp–Mo (6) and HT–phe–Mo (7), obtained after reaction with
complex [MoI (CO) (NCCH ) ].
The presence of the Mo-centers in high loading (ca. 4–6 wt.%)
at the midpoint of the interlayers may be responsible for a marked
increase in the electron density. This is in line with previous reports
In both materials HT–trp–Mo and HT–phe–Mo the spectral fea-
tures are more feasible. The absorptions of the ꢀC O modes are
more hardly observed due to small loading of the [MoI (CO) ] cores
2
3
within the interlayer spacing of the composite materials.
2
3
3 2
13
The C CP MAS NMR spectra of the lamellar materials 4 and 6
(Fig. 5) exhibit at least three isotropic peaks below 100 ppm along
with several resonances in the 120 < ı < 150 ppm range assigned
to carbon atoms of the aromatic moieties, and a signal at ı ca.
177 ppm and 182 ppm, respectively for 4 and 6, assigned to the
carboxylate carbon atoms, which is more clearly detected in the
spectrum of 6 due to the presence of less intense overlapping
spinning side-bands. This result indicates that the deprotonated
amino acid ligand is inside the interlamellar space of hydrotalcite
in material 4. The sharp signals observed for material HT–trp (4), are
[
8,29,30].
The FTIR spectrum of the HT host material is similar to
that of other materials of the same type. In the high frequency
−1
region (4000–2000 cm ) it is dominated by broad bands at about
3
−
1
400 cm , assigned to the O–H stretches of weakly hydrogen
bonded water molecules and layer hydroxyl groups. Some carbon-
ate groups, from the air and humidity, may be responsible for the
−1
band at around 1400 cm
.
similar to those found in solution for the trpH ligand, being assigned
−
The derivatized materials HT–trp (4) and HT–phe (5) have been
described elsewhere recently by us and will not be considered
in much detail in this work [24]. Briefly the main features were
as ı = 177.0 (COO ), 136.8 (C11), 127.0 (C10/C5), 121.2 (C ), 119.7
9
(C7/C ), 112.0 (C ), 109.7 (C ), 56.3 (C ) and 29.8 (C ) ppm. In mate-
8
6
4
2
3
rial 5, three peaks at ı = 136.0 (C ), 130.4 (C
ꢀ
) and 128.9 (C5,5 /C7),
ꢀ
4
6,6
−1
−1
−
observed at 1664 cm and 1571 cm due to the ꢀCOO group for
materials 4 and 5, respectively, which confirms the intercalation of
the amino acid moieties.
are attributed to the phenyl ring of phe ligand, the aliphatic carbons
appearing at ı = 57.1 (C ) and 38.0 (C ) ppm while another signal
2
3
−
at ı = 178.6 ppm is due to the COO group (spectrum not shown).
In Mo-containing material 6, no resonances of the acetonitrile
ligands of precursor complex 1 are detected, and signals from the
Treatment of the composite materials 4 and 5 with the
organometallic complex [MoI (CO) (NCCH ) ] (1) led to the new
2
3
3 2
materials HT–trp–Mo (6) and HT–phe–Mo (7) (not shown). In com-
C O ligands are not visible, since these nuclei have a long relax-

9
0
C.I. Fernandes et al. / Applied Catalysis A: General 384 (2010) 84–93
Table 2
Catalytic results for the oxidation of olefins using catalysts 1–3, 6 and 7.
Olefin
Catalysts
Mo (␮mol)
TOFa
Conversionb
Yieldc
cy8ox
Selectivityd
cy8
[MoI2(CO)3(NCCH3)2] (1)e
244
9
2
10
8
17
81
81
100
100
100
100
100
100
[
[
MoI2(CO)3(trpH)] (2)
MoI2(CO)3(pheH)] (3)
9
13
26
20
9
9
13
26
20
9
7.3
HT–trp–Mo (6)
f
HT–phe–Mo (7)
Olefin
sty
Catalysts
Mo (␮mol)
TOF
Conversion
Yield
styox
Selectivity
bzcho
[MoI2(CO)3(NCCH3)2] (1)e
220
5
6
68
16
13
13
20
4
8
5
11
8
48
10
5
8
7
30
25
62
38
61
29
[
[
MoI2(CO)3(trpH)] (2)
MoI2(CO)3(pheH)] (3)
7.7
HT–trp–Mo (6)
12
7
f
.5
18
HT–phe–Mo (7)
6
28
20
Catalytic reactions using 1:100:200 mol ratio of catalyst:substrate:TBHP carried out in CH2Cl2 at 328 K.
a
b
c
d
e
f
−1
−1
Calcd. after 10 min reaction as mol mol [Mo] h
Calcd. after 24 h.
.
cy8ox = cyclooctene oxide, styox = styrene oxide, bzcho = benzaldehyde.
Selectivity towards epoxide calcd. as “epoxide yield”/conversion.
Results from Ref. [31].
Recycle experiment under batch conditions (new loading of substrate and oxidant after 24 h reaction).
ation time. On the other hand, the signals of the trp ligand in 6 are
clear exhibiting little or no shift compared to the precursor mate-
−
rial 4: 181.7 (COO ), 136.4 (C11), 127.2 (C5/C10), 122.2 (C7/C /C ),
8
9
1
11.4 (C /C ), 56.1 (C ), 27.3 (C ). In this material the peaks present
4 6 2 3
larger widths at half-height which may be due to the different envi-
ronments probed. This may arise from amino acid molecules that
are coordinated to [MoI (CO) ] cores while others are not.
2
3
In material 7, only four signals are resolved, namely the signals
at ı = 136.3 (C ), 128.7 (remaining aromatic nuclei), 56.9 (C ), 37.6
4
2
−
(
C3), and that of the COO group at 180.1 ppm (not shown).
A word should be given on the stability of the organometallic
materials 6 and 7. It is known that clays can hold water in its intra-
gallery spacing. Such situation would bring some disadvantages to
the use of Mo organometallic cores. This was avoided by thoroughly
dehydrating materials 4 and 5, prior to the introduction of the Mo
cores. To this respect and in a similar way to previous reported stud-
Fig. 5. 13C CP MAS NMR spectra for materials HT–trp (4) and HT–trp–Mo (6); *
denotes spinning side-bands. The dashed line is eye-guide for the COO resonance.
ies on similar [MoI (CO) ] systems [8], these were actually found
−
2
3
to be resistant for quite some while. This is true either in terms of
aspect to the naked eye (light tan color persistence) as well as on
the observation of the ꢀC O bands in the FTIR spectrum for at least
six months.
Fig. 6. Reaction kinetics during the transformation of cyclooctene to cyclooctene oxide (cy8ox) or styrene to styrene oxide (styox) and/or benzaldehyde (bzcho) achieved
over homogeneous and heterogeneous catalyst: (a) [MoI2(CO)3(trpH)] (2) and [MoI2(CO)3(pheH)] (3) homogeneous; (b) HT–trp–Mo (6) and HT–phe–Mo (7) heterogeneous
catalysts. The insets refer to the initial hours of the catalytic cycles. From Catalytic reactions using 1:100:200 mol ratio of catalyst:substrate:TBHP carried out in CH2Cl2 at
3
28 K.

C.I. Fernandes et al. / Applied Catalysis A: General 384 (2010) 84–93
91
Scheme 4.
3
.3. Catalytic studies on epoxidation of olefins
the cleavage of the C–C bond, as outlined in Scheme 4, based on pre-
vious results [32,33]. Furthermore, on this catalytic system some
other products were equally identified, in very low yields though,
i.e. below 2%. These comprised 2-phenylacetaldehyde, 2-hydroxy-
1-phenylethanone and 2-oxo-2-phenylacetaldehyde, which may
originate from further oxidation of the epoxide.
The results achieved in this work follow the trend previously
reported by Maurya for polymer bound Mo complexes for the epox-
idation of styrene [32,33]. The formation of benzaldehyde was
The activity of both the Mo derivatized HT materials 6, 7 and
the complexes [MoI (CO) (NCCH ) ] (1), [MoI (CO) L] (L = trpH or
2
3
3
2
2
3
pheH) (2, 3) as catalyst precursors for the liquid-phase epoxidation
of olefins was investigated using cyclooctene (cy8) and styrene (sty)
as substrates with tert-butylhydroperoxide (TBHP) as the oxygen
source, at 328 K, in air, and 3 mL of dichloromethane as solvent
(
Table 2). No reaction took place neither without a metal containing
catalyst nor in the presence of a catalyst without TBHP.
All catalysts tested exhibited similar initial activities indepen-
dently of either being homogenous or heterogeneous and also
independently of the substrate. Nevertheless TOF values are not
outstanding as some results recently reported for systems based
also on the [MoX (CO) ] core (X = Br, I) [31].
detected when the oxidant was either H O2 or TBHP, though in
2
lower yields in the latter case. The authors propose an oxidative
cleavage such as that depicted in Scheme 4, pathway A. In fact the
evolution of formaldehyde has not been detected. In spite of that, it
was possible to detect tert-butylformate and 2-(tert-butylperoxy)-
1-phenylethanol, which make possible to propose the mechanism
as schematized in Scheme 4, pathway B. The observation of tert-
butylformate may be regarded as a formaldehyde “trapped” species
(based on the mechanism proposal), thus confirming the oxidative
cleavage mechanism, based on the evolution of formaldehyde.
It should be mentioned that oxidative cleavage in styrene lead-
ing to benzaldehyde is known to occur due to the presence of
radical species. However, in this case it is known that oxidation
reactions where Mo carbonyl complexes are used in the presence
of TBHP does not proceed via radical species as it was reported
very recently [34]. In this way the formation of benzaldehyde must
be rationalized on the ability of the catalytic species to undergo
such transformation. We have conducted also a catalytic exper-
iment under N2 atmosphere using material 7, TBHP and styrene
yielding similar results.
2
3
The conversion for both cy8 and sty, calculated after 24 h are
comparable to others obtained in MoVI systems [6]. These observed
conversions are fair being possible to state that the four catalytic
systems tested presented conversions ranging from ca. 10% to as
II
much as 30%, stressing the efficiency of such Mo based catalysts.
Recycle experiments were also carried out using material 6 with
both cy8 and sty. The performance after 48 h reaction was found
to be quite similar to the first 24 h. Consumption of TBHP by the
catalysts was found to reach an average value of 76% of the initial
value. This is clearly more than the initial olefin loading meaning
that catalysts most probably decompose TBHP in a parallel cycle to
the catalytic cycle itself.
All catalytic systems led to oxidation products, though in poor to
fair yields as evidenced in Table 2. According to Fig. 6, the kinetics
shows that there is a continuous yield increase. The heterogeneous
systems performed better than the homogeneous counterparts,
evidencing a synergetic beneficial effect of the clay support, in
both terms of conversion and yield. In fact material 6 was found
to be superior in cy8 epoxidation, whereas material 7 was the best
performing in sty oxidation.
In related homogeneous systems previously reported by us
the presence of catalytic systems where the metal complexes
hold NH moieties perform worse than the counterparts with
diazabutadiene-type ligands [35–39]. A comparison of cy8 conver-
sion for reported systems and catalysts 2, 3, 6 and 7 is presented in
Fig. 7.
If the epoxide was the only product detected in the cy8 system,
meaning 100% selectivity, this was not the case for the oxidation
of styrene where both the epoxide and benzaldehyde have been
detected. The presence of benzaldehyde results from a process of
oxidative cleavage where the epoxide is further oxidized and conse-
quently originating benzaldehyde and formaldehyde as a result of
In fact results achieved in this work for complexes 2 and 3
(homogeneous phase systems) resemble the former systems with
fair conversions clearly not reaching completeness of the reaction.
The last point (catalyst 6) stresses the beneficial effect of the immo-
bilization of the catalyst. The same effect was achieved with catalyst
7 for sty oxidation according to Table 2.

9
2
C.I. Fernandes et al. / Applied Catalysis A: General 384 (2010) 84–93
trary, besides formation of styrene oxide, oxidation of styrene
also led to benzaldehyde due to further oxidation of the epoxide.
The detection of two relevant compounds, tert-butylformate and
2
-(tert-butylperoxy)-1-phenylethanol, make evidence of the oxida-
tive cleavage pathway making possible the proposal of a possible
mechanism towards the formation of benzaldehyde starting from
styrene oxide. This is in agreement with previous reports found in
the literature and was more evident in the case of the phenylalanine
containing systems.
A comparison with previously reported catalytic systems evi-
dences that the poor performance of the homogeneous catalysts
may be due to the ligand nature which is imposed by the presence
of NH moieties. The heterogenization of the catalyst precursors
seemed to circumvent this by producing more efficient catalytic
systems showing that the presence of the clay is not innocent, in
agreement with previous reports [40]. Despite this another major
advantage of the heterogenization process is the fact that separa-
tion of catalyst from the reaction medium is accomplished in a more
undemanding fashion and that the catalytic active species may be
more robust towards air and moisture.
Acknowledgments
The authors thank FCT, POCI and FEDER (project
PTDC/QUI/71576/2006) for financial support. Special thanks
would be addressed to Prof. José M. Lopez Nieto, editor of this
journal, and to the referees for critical suggestions during the peer
review process concerning the quality of the discussion.
References
[
[
[
1] S.P. Newman, W. Jones, New J. Chem. 22 (1998) 105–115.
2] A.I. Khan, D. O’Hare, J. Mater. Chem. 12 (2002) 3191–3198.
3] S.P. Newman, S.J. Williams, P.V. Coveney, W. Jones, J. Phys. Chem. B 102 (1998)
6
710–6719.
4] (a) S.K. Jana, Y. Kubota, T. Tatsumi, J. Catal. 255 (2008) 40–47;
b) T. Tatsumi, K. Yamamoto, H. Tajima, H. Tominaga, Chem. Lett. 21 (1992)
15–818.
5] (a) M. Del Arco, S. Gutiérrez, C. Martín, V. Rives, J. Rocha, J. Solid State Chem.
77 (2004) 3954–3962;
b) L. Desigaux, M. Ben Belkacem, P. Richard, J. Cellier, P. Leone, L. Cario, C.
[
[
(
8
1
(
Taviot-Gueho, B. Pitard, Nano Lett. 6 (2006) 199–204.
6] S. Gago, M. Pillinger, A.A. Valente, T.M. Santos, J. Rocha, I.S. Gon c¸ alves, Inorg.
Chem. 43 (2004) 5422–5431.
[
[
[
[
7] S. Gago, M. Pillinger, T.M. Santos, J. Rocha, I.S. Gon c¸ alves, Eur. J. Inorg. Chem.
(
2004) 1389–1395.
8] M. Vasconcelos-Dias, C.D. Nunes, P.D. Vaz, P. Ferreira, M.J. Calhorda, Eur. J. Inorg.
Chem. (2007) 2917–2925.
9] G. Negrón, N. Guerrra, L. Lomas, R. Gavi n˜ o, J. Cárdenas, ARKIVOC xi (2003)
Fig. 7. Structure–catalytic activity relationship from MoII based complexes for cy8
conversion. The numbers adjacent to ligand names refer to the respective reference,
the last two ligands corresponding to this work.
1
79–184.
[10] J.M. Fraile, J.I. García, C.I. Herrerías, J.A. Mayoral, E. Pires, Chem. Soc. Rev. 38
(2009) 695–706.
[
11] K. Ebitani, K. Motokura, K. Mori, T. Mizugaki, K. Kaneda, J. Org. Chem. 71 (2006)
440–5447.
5
4
. Conclusions
[12] V. Gianotti, M. Benzi, G. Croce, P. Frascarolo, F. Gosetti, E. Mazucco, M. Bottaro,
M.C. Gennaro, Chemosphere 73 (2008) 1731–1736.
[
[
13] M.H. Valkenberg, W.F. Hölderich, Catal. Rev. 44 (2002) 321–374.
14] A. Taguchi, F. Schüth, Micropor. Mesopor. Mater. 77 (2005) 1–45.
The intercalation of heptacoordinated halocarbonyl Mo(II) com-
plexes in organic–inorganic hybrid clay materials holding amino
acids as ligands was successfully accomplished. The amino acids
were intercalated, under basic conditions, in the interlayer spacing
of previously calcined hydrotalcite clay and afterwards introduc-
ing an organometallic fragment. Several techniques confirmed this
proposal. For instance, FTIR spectra exhibited the bands assigned
to ꢀC O modes reflecting the incorporation of the organometal-
lic fac-[MoI (CO) ] core and others showing coordination of Mo to
[15] P.K. Baker, S.G. Fraser, E.M. Keys, J. Organomet. Chem. 309 (1986) 319–321.
[
16] R.G. Parr, W. Yang, Density Functional Theory of Atoms and Molecules, Univer-
sity Press, Oxford, NY, 1989.
17] Gaussian 03, Revision C.02, http://www.gaussian.com/g misc/g03/citation g03
.htm.
[
[18] (a) C. Lee, W. Yang, G. Parr, Phys. Rev. B 37 (1988) 785–789;
b) A.D. Becke, J. Chem. Phys. 98 (1993) 5648–5652.
(
[
19] D. Andrae, U. Haeussermann, M. Dolg, H. Stoll, H. Preuss, Theor. Chim. Acta 77
(1990) 123–141.
[20] A.W. Ehlers, M. Böhme, S. Dapprich, A. Gobbi, A. Höllwarth, V. Jonas, K.F.
Köhler, R. Stegmann, A. Veldkamp, G. Frenking, Chem. Phys. Lett. 208 (1993)
2
3
the ligand. X-ray powder patterns, on the other hand, revealed the
reconstruction of the HT layered structure.
The Mo-containing materials were tested in the oxidation
of cyclooctene and styrene. Results show that although con-
versions are fair, selectivity is only achieved for cy8 (100%) as
evidenced by detecting cy8ox as the only product. On the con-
111–114.
[
21] (a) A.D. McLean, G.S. Chandler, J. Chem. Phys. 72 (1980) 5639–5648;
(b) R. Krishnan, J.S. Binkley, R. Seeger, J.A. Pople, J. Chem. Phys. 72 (1980)
650–654;
(
(
c) A.J.H. Wachters, J. Chem. Phys. 52 (1970) 1033–1036;
d) P.J. Hay, J. Chem. Phys. 66 (1977) 4377–4384;
(e) K. Raghavachari, G.W. Trucks, J. Chem. Phys. 91 (1989) 1062–1065;

C.I. Fernandes et al. / Applied Catalysis A: General 384 (2010) 84–93
93
(
f) L.A. Curtiss, M.P. McGrath, J.-P. Blaudeau, N.E. Davis, R.C. Binning Jr., L. Radom,
J. Chem. Phys. 103 (1995) 6104–6113;
g) M.P. McGrath, L. Radom, J. Chem. Phys. 94 (1991) 511–516.
[32] M.R. Maurya, U. Kumar, P. Manikandan, Dalton Trans. (2006) 3561–3575.
[33] M.R. Maurya, M. Kumar, S. Sikarwar, React. Funct. Polym. 66 (2006) 808–818.
[34] P.J. Costa, M.J. Calhorda, F. Kuhn, Organometallics 29 (2010) 303–311.
[35] J.C. Alonso, P. Neves, C. Silva, A.A. Valente, P. Brandão, S. Quintal, M.J.V. de Brito,
P. Pinto, V. Felix, M.G.B. Drew, J. Pires, A.P. Carvalho, M.J. Calhorda, P. Ferreira,
Eur. J. Inorg. Chem. (2008) 1147–1156.
(
[
[
22] F.H. Allen, Acta Crystallogr. B 58 (2002) 380–388.
23] (a) M. Takuma, Y. Ohki, K. Tatsum, Organometallics 24 (2005) 1344–1347;
(
b) P. Nag, R. Bohra, R.C. Mehrotra, R. Ratnani, Synth. React. Inorg. Met.-Org.
Nano-Metal Chem. 32 (2002) 1549–1559.
[36] J.C. Alonso, P. Neves, M.J. Pires da Silva, S. Quintal, P.D. Vaz, C. Silva, A.A. Valente,
P. Ferreira, M.J. Calhorda, V. Félix, M.G.B. Drew, Organometallics 26 (2007)
5548–5556.
[37] M.S. Saraiva, N.L. Dias Filho, C.D. Nunes, P.D. Vaz, T.G. Nunes, M.J. Calhorda,
Micropor. Mesopor. Mater. 117 (2009) 670–677.
[
[
24] P.D. Vaz, C.D. Nunes, New J. Chem. 34 (2010) 541–546.
25] S. Aisawa, H. Kudo, T. Hoshi, S. Takahashi, H. Hirahara, Y. Umetsu, E. Narita, J.
Solid State Chem. 177 (2004) 3987–3994.
[
[
26] S. Carlino, Solid State Ionics 98 (1997) 73–84.
27] J. Gimenez, C.D. Nunes, P.D. Vaz, A.A. Valente, P. Ferreira, M.J. Calhorda, J. Mol.
Catal. A: Chem. 256 (2006) 90–98.
[38] M.S. Saraiva, C.D. Nunes, T.G. Nunes, M.J. Calhorda, J. Mol. Catal. A 321 (2010)
92–100.
[
28] J.-C. Dupin, H. Martinez, C. Guimon, E. Dumitriu, I. Fechete, Appl. Clay Sci. 27
[39] M.S. Saraiva, S. Quintal, F.C.M. Portugal, T.A. Lopes, V. Félix, J.M.F. Nogueira,
M. Meireles, M.G.B. Drew, M.J. Calhorda, J. Organomet. Chem. 693 (2008)
3411–3418.
(
2004) 95–106.
[
29] P. Beaudot, M.E. De Roy, J.P. Basse, Chem. Mater. 16 (2004) 935–945.
30] S. Gago, M. Pillinger, R.A. Sá Ferreira, L.D. Carlos, T.M. Santos, I.S. Gon c¸ alves,
Chem. Mater. 17 (2005) 5803–5809.
[
[40] A. Podgor sˇ ek, M. Zupan, J. Iskra, Angew. Chem. Int. Ed. 48 (2009) 8424–8450.
[31] M. Vasconcellos-Dias, C.D. Nunes, P.D. Vaz, P. Ferreira, P. Brandão, V. Félix, M.J.
Calhorda, J. Catal. 256 (2008) 301–311.