New heptacoordinate tungsten(II) complexes with α-diimine ligands in the catalytic oxidation of multifunctional olefins

Article abstract of DOI:10.1016/j.ica.2021.120263

New tungsten(II) and molybdenum(II) heptacoordinate complexes [MX₂(CO)₃(L_Y)] (MXL_y: M = W, Mo; X = Br, I; L_Y = C₅H₄NCY = N(CH₂)₂CH₃ with Y = H (L₁), Me (L₂), Ph (L₃)) were synthesized and characterized by spectroscopic techniques and elemental analysis. The two tungsten complexes WXL₁ (X = Br, I) were also structurally characterized by single crystal X-ray diffraction. The metal coordination environment is in both a distorted capped octahedron. The complexes with L₁ and L₂ ligands were grafted in MCM-41, after functionalization of the ligands with a Si(OEt)₃ group. The new materials were characterized by elemental analysis, N₂ adsorption isotherms, ²⁹Si MAS and ¹³C MAS NMR. The tungsten(II) complexes and materials were the first examples of this type reported. All complexes and materials were tested as homogeneous and heterogeneous catalysts in the oxidation of multifunctional olefins (cis-hex-3-en-1-ol, trans-hex-3-en-1-ol, geraniol, S-limonene, and 1-octene), with tert-butyl hydroperoxide (TBHP) as oxidant. The molybdenum(II) catalyst precursors are in general very active, reaching 99% conversion and 100% selectivity in the epoxidation of trans-hex-3-en-1-ol. Their performance is comparable with that of the [Mo(η³-C₃H₅)X(CO)₂(L_Y)] complexes, but it increases with immobilization. On the other hand, most of the W(II) complexes display an activity similar or inferior to that of the Mo(II) analogues and it decreases after they are supported in MCM-41. DFT calculations show that tungsten complexes and iodide ligands are more easily oxidized from M(II) to M(VI) than molybdenum ones, while the energies of relevant species in the catalytic cycle are very similar for all complexes, making the theoretical rationalization of experimental catalytic data difficult.

Full text of DOI:10.1016/j.ica.2021.120263

Inorganica Chimica Acta 519 (2021) 120263
Contents lists available at ScienceDirect
Inorganica Chimica Acta
journal homepage: www.elsevier.com/locate/ica
Research paper
New heptacoordinate tungsten(II) complexes with
catalytic oxidation of multifunctional olefins
α
-diimine ligands in the
Maria Vasconcellos-Diasa, Carla D. Nunesa , Vítor Felix , Paula Brandao ,
,
b,
*
c
c
´
˜
a,d,*
´
Maria Jose Calhorda
a
ˆ
Centro de Química e Bioquímica, Departamento de Química e Bioquímica, Faculdade de Ciencias, Universidade de Lisboa, Campo Grande, 1749-016 Lisboa, Portugal
b
ˆ
Centro de Química Estrutural, Faculdade de Ciencias, Universidade de Lisboa, 1049-001 Lisboa, Portugal
c CICECO - Aveiro Institute of Materials, Department of Chemistry, University of Aveiro, 3810-193 Aveiro, Portugal
d
ˆ
BioISI -Biosystems & Integrative Sciences Institute, Departamento de Química e Bioquímica, Faculdade de Ciencias, Universidade de Lisboa, Campo Grande, 1749-016
Lisboa, Portugal
A R T I C L E I N F O
A B S T R A C T
New tungsten(II) and molybdenum(II) heptacoordinate complexes [MX2(CO)3(LY)] (MXLy: M = W, Mo; X = Br, I;
LY = C5H4NCY = N(CH2)2CH3 with Y = H (L1), Me (L2), Ph (L3)) were synthesized and characterized by spec-
troscopic techniques and elemental analysis. The two tungsten complexes WXL1 (X = Br, I) were also structurally
characterized by single crystal X-ray diffraction. The metal coordination environment is in both a distorted
capped octahedron. The complexes with L1 and L2 ligands were grafted in MCM-41, after functionalization of the
ligands with a Si(OEt)3 group. The new materials were characterized by elemental analysis, N2 adsorption iso-
therms, 29Si MAS and 13C MAS NMR. The tungsten(II) complexes and materials were the first examples of this
type reported. All complexes and materials were tested as homogeneous and heterogeneous catalysts in the
oxidation of multifunctional olefins (cis-hex-3-en-1-ol, trans-hex-3-en-1-ol, geraniol, S-limonene, and 1-octene),
with tert-butyl hydroperoxide (TBHP) as oxidant. The molybdenum(II) catalyst precursors are in general very
active, reaching 99% conversion and 100% selectivity in the epoxidation of trans-hex-3-en-1-ol. Their perfor-
Dedicated to Maurizio Peruzzini on the occa-
sion of his 65th birthday
Keywords:
Molybdenum
Tungsten
MCM-41
Catalysis
2-iminomethylpyridine
DFT calculations
mance is comparable with that of the [Mo(
η
3-C3H5)X(CO)2(LY)] complexes, but it increases with immobilization.
On the other hand, most of the W(II) complexes display an activity similar or inferior to that of the Mo(II) an-
alogues and it decreases after they are supported in MCM-41. DFT calculations show that tungsten complexes and
iodide ligands are more easily oxidized from M(II) to M(VI) than molybdenum ones, while the energies of
relevant species in the catalytic cycle are very similar for all complexes, making the theoretical rationalization of
experimental catalytic data difficult.
1. Introduction
templated silica; both species were active in olefin epoxidation, but
the activity was smaller in the material [1]. [Mo(0)(CO)4(R1,R2-IMP)]
The number of tungsten complexes is in general much smaller than
that of molybdenum complexes and usually it is not possible to compare
the role of the metal in their applications, such as catalysis. Unlike Mo,
(R1 = H, Me, Ph, R2 = COOH, SO3Na; [2] R1 = R2 = Me [3]) complexes
were obtained and some catalyzed the polymerization of methyl meth-
acrylate (MMA) [2]. [Mo(CO)4(IMP)] was grafted in MCM-41 and the
authors studied the activity of the complex and the material in olefin
epoxidation in ionic liquids [4]. The Mo(0) complexes were also used to
prepare [Mo(II)(η3–C3H5)Cl(CO)2(R1,R2-IMP)] (R1 = Me, R2 = Me, Ph,
Ph-2-OMe and R1 = Ph, R2 = Me, Ph,) [5] by reaction with allyl halides,
but the tungsten(0) analogues [6] did not react to yield the analogous
allyl complexes [5]. However, the synthesis of W(IV) complexes
[WBr4(R1,R2-IMP)] (R1 = H, R2 = Me, Et, nPr, tBu) was reported, as well
as their magnetic properties which were studied by 1H NMR [7].
W complexes with the very versatile bidentate α-diimines ligands are
few. These ligands give the metal a variety of environments and sub-
stituents for the two nitrogen atoms. In particular, in N-isopropyl 2-imi-
nomethylpyridine (IMP, Scheme 1), the pyridyl and the imine confer to
the ligand an asymmetry different from what substitution may cause in
2,2′-bipyridyl or 1,10-phenanthroline. Many molybdenum and some
tungsten complexes of IMP have been described. The Mo(VI) complex
[MoO2Cl2(IMP)] was synthesized and immobilized in a micelle-
* Corresponding authors.
https://doi.org/10.1016/j.ica.2021.120263
Received 25 October 2020; Received in revised form 15 January 2021; Accepted 15 January 2021
Available online 2 February 2021
0020-1693/© 2021 Elsevier B.V. All rights reserved.

M. Vasconcellos-Dias et al.
Inorganica Chimica Acta 519 (2021) 120263
monodentate N-donor ligands.
2. Results and discussion
2.1. Chemical studies. Complexes
Tungsten and molybdenum complexes [MX2(CO)3(LY)] (MXLy: M =
Mo, W; X = Br, I; LY = L1–3,) were obtained by substitution of the nitriles
in [MX2(CO)3(CH3CN)2] (X = Br, I) by the ligands L1, L2, and L3,
respectively, in dichloromethane during 4–6 h (Scheme 2). All the new
Scheme 1. Representation of the IMP ligands.
Complexes of other metals containing IMP ligands have been used as
catalysts and exhibited high activity. Some examples include Cu(I) for
MMA polymerization [8], palladium nanoparticles immobilized in
functionalized silica for the Suzuki-Miyaura cross coupling reaction and
hydrogenation in water [9], or arene-Ru(II) complexes with function-
alized chains (R2) for oxidative cleavage of 1-octene [10]. Complexes
and clusters with IMP ligands often display interesting photochemical
properties, as reported recently for [Os3(CO)19(IMP)] clusters [11]. We
synthesized [Mo(η3–C3H5)X(CO)2(L)] (L = IMP (L1), Me-IMP (L2) and
Ph-IMP (L3), Scheme 1) and prepared their heterogeneous counterparts
by grafting them on MCM-41. For this system, immobilization did not
always lead to improved efficiency [12].
complexes were characterized by elemental analysis, FTIR, 1H and 13
C
NMR spectroscopy, and complexes WBrL2 and WIL2 also by single
crystal X-ray diffraction (vide infra). The two Mo(II) complexes with L1,
MoBrL1 and MoIL1, were previously reported [13]. The characteriza-
tion will be more detailed for the new tungsten complexes of L2.
In the FTIR spectra of both WBrL2 and WIL2 complexes in the solid
–
–
state, the νC N vibrational modes of the precursor complexes have
–
–
–
disappeared, while the νC O stretching frequencies of the coordinated
–
carbonyl groups were shifted from 1916 and 2028 cmꢀ 1 to 1902, 1949
and 2033 cmꢀ 1 in WBrL2, and from 1889, 1906 and 1996 cmꢀ 1 to 1925,
2003, 2019, and 2065 cmꢀ 1 in WIL2 (Table S1). Three or four
ν
–
–
C O
–
bands may be observed depending on the isomers present [17]. The
Following our previous works on allylic Mo(II) complexes with IMP
ligands, we report the synthesis of tungsten complexes [W
(X)2(CO)3(IMP)] (L = IMP (L1), X = I, Br). Moreover, given that Mo(II)
analogues with L1 showed higher activity when acting as heterogeneous
catalysts supported in MCM-41, than in homogeneous conditions, in the
olefin epoxidation reaction [13], we also decided to prepare the M
(X)2(CO)3 complexes with the Me-IMP (L2) and Ph-IMP (L3) ligands (M
= Mo and W). All these complexes were further immobilized on
MCM–41. The catalytic activity of these new complexes and materials
was tested in the oxidation reactions of several olefins, some of them
ν
C=N stretching modes of the free ligand at 1655 cmꢀ 1 were shifted to
1603 (WBrL2) and 1605 cmꢀ 1 (WIL2) upon metal coordination. Similar
trends were observed for the Mo(II) complexes (all values are collected
in Table S1 of Supporting Information).
–
–
There is a tendency for the W complexes to have higher νC O
–
stretching frequencies than the Mo analogues, reflecting less back
donation to the carbonyl ligands.
The 1H NMR spectra of complexes MoXL2, MoXL3, and WXLn (n =
1,2,3; X = Br, I) only show the peaks of the protons in the α-diimine,
which are shifted relatively to those of the free ligand. For example, in
the tungsten complexes WBrL2 and WIL2, the protons next to the donor
nitrogen atoms, H3 and H9, appear in the free ligand at 3.56 and 8.55
ppm, respectively, and are shifted to higher (downfield) δ values at 3.84
(H3) and 8.76 ppm (H9) in WBrL2) and 3.78 (H3) and 9.56 ppm (H9) in
WIL2. The other three aromatic protons H6, H7 and H8 are shifted from
8.03 to 8.40 ppm (H6), 7.21 to 8.52 ppm (H7), and 7.64 to 7.82 ppm (H8)
in complex WIL2. The assignments were made with resort to bidimen-
sional COSY and HMQC experiments.
–
bearing the C C bond and other functionalities.
–
Although oxidation catalysts are typically Mo(VI) complexes, it has
been proved that [Mo(II)(
η
5-C5R5)(CO3)X ] could be used as catalyst
precursors to produce in situ the very active species [Mo(VI)(
η
5-C5R5)
O2X] [14], which could be also isolated and characterized. Their activity
could approach that of the highly active [MeReO3]. [Mo(
η
3-C3H5)Cl
(CO)2(bian)]
(bian)
1,4-(4-chloro)phenyl-2,3-naph-
thalenediazabutadiene) was the first of another type of Mo(II) com-
plexes, which, in the presence of TBHP, also originated catalysts almost
The 13C NMR spectra of WBrL2 and WIL2 exhibit peaks assigned to
the C3 and C9 carbons at 54.2 and 149.1 ppm, and at 44.0 and 151.0
ppm, respectively. C9 experiences the largest shift relative to the free
ligand L2. These and other peaks arising from the other carbon atoms are
also observed for all the complexes (Experimental).
as active as [Mo(VI)(
η
5-C5R5)O2X] [15]. Their main advantage was that
the activity in the second run was as high as in the first. This was not
observed in the Mo(VI)(
η
5-C5R5) catalysts. Finally, some W(II) com-
plexes behaved as oxidation catalysis precursors, as shown with the
[MX2(CO)3(2-amino-1,3,4-thiadiazole)2] (M = Mo, W) complexes,
which were synthesized and grafted in MCM-41. The W(II) precursors
led to more active catalysts than Mo(II), and both became more active
after immobilization in MCM-41 [16]. We want, in the present work, to
The molecular structures of complexes WBrL1 and WIL1 were
determined by single crystal X–ray diffraction analysis and are shown in
Figs. 1 and 2, respectively.
These two complexes have a coordination number 7, for which three
main alternative structures are possible, namely pentagonal bypiramid,
explore the effect of having a bidentate
α-diimine instead of two
Scheme 2. The synthesis of molybdenum(II) and tungsten(II) [MX2(CO)3(LY)] (MXLy) complexes and the numbering scheme for NMR of carbons and respec-
tive hydrogens.
2

M. Vasconcellos-Dias et al.
Inorganica Chimica Acta 519 (2021) 120263
Fig. 1. Molecular structure of [WBr2(CO)3(L1)] (WBrL1) in two different views: thermal ellipsoids (50% probability level) showing the overall structure (left);
capped octahedral tungsten coordination geometry (right).
Fig. 2. Molecular structure of [WBr2(CO)3(L1)] (WIL1) in two different views: thermal ellipsoids (50 %probability level) showing the overall structure (left); capped
octahedral tungsten coordination geometry (right).
Scheme 3. Procedure for the immobilization of complexes MoXL2, WXL1 and WXL2 (X = Br, I) in MCM.
3

M. Vasconcellos-Dias et al.
Inorganica Chimica Acta 519 (2021) 120263
capped trigonal prism, and capped octahedron. The coordination ge-
ometries of WBrL1 and WIL1 can be described as a distorted capped
octahedron. The carbonyl ligand C(5) is capping the triangular face C
(3)–N(18)–C(1), with bond angles C(5)-W-C(3), C(5)-W-N(18) and C(5)-
W-C(1) ranging from 71.7(2) to 75.9(2)◦ for WBrL1 and from 71.8(2) to
77.8(2) for WIL1. This face is opposite to the X(1)–N(11)–X(2) one (X =
Br, WBrL1 and X = I, WIL1, Figs. 1 and 2 right). Moreover, the chelation
of the bidentate ligand IMP, through the two nitrogen donors, one halide
(X(2) with X = I or Br) and one carbonyl ligand C(1), occurs in the
equatorial plane of the metal coordination sphere (Figs. 1 and 2 left). In
the molecular structure of WBrL1, the W–N(11) and W-N(18) distances
are 2.217(4) and 2.172(4) Å, respectively and in WIL1 the same atom
distances are 2.216(4) and 2.180(4) Å, respectively. These distances
lead to N(11)–W–N(18) bite angles, for the five membered chelate ring,
of 72.9 (2) and 72.7(1)˚for WBrL1 and WIL1, respectively. Bond lengths
and bond angles around the tungsten center are given in Table S2 (SI) for
both complexes.
Table 1
Textural parameters of host and composite materials from powder XRD and N2
isotherms at 77 K, for MCM, MCM-L2, MCM-MoBrL2, MCM-MoIL2, MCM-
WBrL2, and MCM-WIL2 materials.
Sample
d100
S
BET m2.
ΔSaBET
Vp cm3.
ΔVpb
dcBJH
(Ǻ)
gꢀ 1
(%)
gꢀ 1
(%)
(nm)
MCM
35.7
36.2
35.9
36.4
36.0
35.9
1037
674
628
609
578
458
–
0.97
0.42
0.39
0.44
0.39
0.27
–
3.60
2.87
2.67
2.86
2.86
2.65
MCM–L2
35.0
39.4
41.3
44.3
55.8
56.7
59.8
54.6
59.8
72.2
MCM–MoBrL2
MCM–MoIL2
MCM–WBrL2
MCM–WIL2
and MCM–WIL1 and were previously reported for MCM–MoBrL1 and
MCM–MoIL1 [13].
The intensity of the peaks is reduced as the stepwise functionaliza-
tion takes place, but it is more significant in the materials with the Mo
and W cores. This effect does not derive from a decreased crystallinity of
the materials, but to a reduction of the X–ray scattering contrast between
the MCM silica walls and the metal complexes inside the pores, as has
been demonstrated before [19,20].
2.2. Chemical studies. Materials
Nitrogen sorption/desorption studies, performed at 77 K, revealed
for all the materials a reversible type IV isotherm (Fig. 3), typical of
mesoporous MCM-41 materials (pore width between 2 nm and 50 nm,
according to IUPAC) [21]. The textural parameters SBET and VP were
calculated for all materials (Table 1) and showed values comparable
with those in literature data [22]. The capillary condensation/evapo-
ration step is observed in all the materials between 0.33 and 0.42 rela-
tive pressures, and they have a uniform pore size distribution as
reflected in the sharpness of the step.
The new complexes WXL1 and WXL2 (X = Br, I) and MoXL2 were
immobilized in MCM-41 (MCM) according to the procedure reported for
MoXL1 (X = Br, I) and described in Scheme 3 for the new compounds
[12]. MCM-41 (MCM) was synthesized by a template approach [18].
The silylated ligands L1Si and L2Si reacted with the MCM surface OH
groups to afford the ligand functionalized materials MCM–L1Si and
MCM–L2Si. Then, the reaction between these materials and the pre-
cursor complexes [MX2(CO)3(CH3CN)2 (M = W, Mo; X = Br, I) led to the
formation of new containing materials of W and Mo: MCM–WBrL1,
MCM–WIL1, MCM–WBrL2, MCM–WIL2, MCM–MoBrL2, and
MCM–MoIL2, which are similar to MCM–MoXL1 (X = Br, I) described
previously [12]. For the tungsten materials MCM–WBrL2 and
MCM–WIL2, the metal content was found by elemental analysis to be
3.4 wt% (0.19 mmol gꢀ 1) and 4.47 wt% (0.26 mmol gꢀ 1). Mo loads of
5.11 wt% (0.53 mmol gꢀ 1) and 5.97 wt% (0.62 mmol gꢀ 1) were deter-
mined for MCM–MoBrL2 and MCM–MoIL2, respectively.
SBET and Vp calculated for MCM–L2 display a decrease relative to
MCM of 35% and 57%, respectively, in accordance with the decrease in
p/p0 coordinates of the inflection points of the isotherms [23]. The size
of the capillary condensation step, which is related to the volume of pore
space spanned by the pore walls, becomes even smaller for the modified
materials MCM-MoBrL2, MCM-MoIL2, MCM-WBrL2 and MCM–WIL2.
The maximum of the pore size distribution (PSD) curve (Fig. 3), deter-
mined by the BJH method (dBJH), drops from 3.60 to 2.87 nm when
going from MCM to MCM-L2. When the metal fragments are introduced
in MCM-L2, they produce a relative decrease in SBET which amounts to
44%, 56%, 39% and 41%, for MCM–WBrL2, MCM–WIL2,
MCM–MoBrL2 and MCM–MoIL2, respectively, when compared to
MCM. The values of Vp for materials MCM–WBrL23, MCM–WIL2,
MCM–MoBrL2 and MCM–MoIL2, also reveal a more pronounced
decrease of 60%, 72%, 60% and 55%, respectively.
It is possible to conclude that all the metal fragments reacted with the
ligand L2 at the surface of the material, as the load of ligand was much
higher. Indeed, based on the nitrogen analysis for each material, loads of
the L2 ligand of 0.80 and 0.39 mmol gꢀ 1, for MCM–WBrL2 and
MCM–WIL2, and 0.77 and 0.82 mmol gꢀ 1, for MCM–MoBrL2 and
MCM–MoIL2, were found.
All materials were also characterized by powder XRD, DRIFT, and
29Si MAS and 13C CP MAS NMR. Sorption/desorption N2 isotherms were
also determined for estimation of textural parameters. All character-
ization features are consistent with those reported in related mesoporous
materials [12,13,18].
The distribution of pore size, obtained by the BJH method, decreases
after the functionalization step. The size of the pore starts at 3.6 nm for
the parent MCM, dropping to 2.87 nm after the introduction of the L2Si
ligand, and continues to decrease to 2.67, 2.86, 2.86 and 2.65 nm,
respectively, for materials MCM–MoBrL2, MCM–MoIL2, MCM–WBrL2
and MCM–WIL2. This trend in the pore size shows that the ligand and
the metal fragments have functionalized the walls inside the pores.
The DRIFT spectrum of the MCM host (Figure S2) evidences a broad
band in the 3600–2600 cmꢀ 1 range, typical of silicates, assigned to
hydrogen bonded silanol groups. The band at ca. 1600 cmꢀ 1 due to OH
bending modes and the intense broad band at 1230–900 cmꢀ 1 assigned
to the asymmetric stretching vibration modes of the mesoporous
The X-ray diffraction powder patterns of the initial MCM (Figure S1,
SI) show peaks in the 2θ range between 2 and 10◦, which can be indexed
to the (100), (110), (200) and (210) reflections. They display a
diffraction pattern with a hexagonal symmetry revealing the existence of
a periodic arrangement of hexagonal channels in the materials. From the
d
100 interplanar value of the 100 reflection (35.7 Å), the lattice constant
√̅̅̅
a = 2d100
/
3 is estimated as 41.2 Å.
After functionalization of MCM with the silylated ligand L2Si, the
new MCM–L2Si material still shows the three peaks corresponding to
three reflections, but they are shifted slightly towards higher 2θ values
as compared with MCM-41. The d distance for the 100 reflection in-
creases to 36.2 Å, corresponding to a constant lattice of a = 41.8 Å. The
same periodic hexagonal arrangement was also found for materials
MCM–MoBrL2, MCM–MoIL2, MCM–WBrL2 and MCM–WIL2, obtained
from L2Si by reaction with the metal precursors. Their d100 distances of
35.9 Å, 36.4 Å, 36.0 Å, and 35.9 Å correspond to lattice constants a =
41.5 Å, 42.0 Å, 41.6 Å, and 41.5 Å, respectively, as collected in Table 1.
Similar results were observed for the tungsten materials MCM–WBrL1
framework (νSi–O–Si) are other relevant features. In MCM–L2 contain-
ing the grafted organic ligand L2Si, several changes and new bands are
detected. The most remarkable feature is the new band at 1640 cmꢀ 1
–
corresponding to the C N stretching vibrations from the imine part of
–
the ligand L2Si.
The introduction of the M(CO)3X2 fragments to yield MCM-MoBrL2,
MCM-MoIL2, MCM-WBrL2 and MCM-WIL2 also modifies the spectral
–
–
features, owing to the three new bands assigned to the νC O (Figure S2)
–
modes. These bands, slightly shifted relatively to those of the precursor
4

M. Vasconcellos-Dias et al.
Inorganica Chimica Acta 519 (2021) 120263
Fig. 3. Nitrogen adsorption studies of MCM and functionalized materials MCM-L2, MCM–MoBrL2, MCM-MoIL2, MCM-WBrL2 and MCM-WIL2 at 77 K: isotherms
(left) and pore size distribution curves (right).
complexes, are observed for MCM–WBrL2 at 2015, 1955 and 1865
cmꢀ 1. The
νC=N stretching vibrations of MCM–WBrL2 are detected at
1632 and 1591 cmꢀ 1 (Fig. 5), showing the integrity of the ligand and its
coordination to the core. The values are similar for the other materials.
All materials were also characterized by 13C and 29Si MAS solid-state
NMR (Fig. 4). The 29Si MAS NMR spectrum of pristine MCM displays two
broad resonances at ꢀ 109.3 and ꢀ 100.4 ppm, assigned to Q4 and Q3
species, and a weak shoulder at ꢀ 92.1 ppm assigned to Q2 species. Qn
species arise from the silica framework [Qn = Si(OSi)n(OH)4ꢀ n], the Q3
sites being associated with single silanols Si–OH, often involved in
hydrogen bonding, and the Q2 sites with Si–(OH)2 groups (Fig. 4).
MCM–L2 shows signals at ꢀ 109.6, ꢀ 101.2 and ꢀ 91.7 ppm for the Q4, Q3
and Q2 species, respectively, shifted from those of the MCM precursor.
The L2Si ligand present in MCM–L2 has a different silicon atom,
responsible for the signals associated to Tn species. Resonances at ꢀ 66.4,
ꢀ 57.5 and ꢀ 47.9 ppm assigned to the T3, T2 and T1 species are thus also
observed. The Qn and Tn species resonances are observed in all the
derivatized materials and reflect the immobilization of the ligands and
complexes in the MCM.
Fig. 5. 13C CP MAS NMR spectra of MCM functionalized materials MCM-L2,
MCM–MoBrL2, MCM-MoIL2, MCM-WBrL2 and MCM-WIL2.
The 29Si MAS spectra of MCM-MoBrL2, MCM-MoIL2, MCM-WBrL2
and MCM-WIL2 do not differ significantly from that of MCM–L2, indi-
cating that the reaction of the precursors MX2(MeCN)2(CO)3 (M = Mo or
W; X = Br or I) occurs with the nitrogen donor atoms in MCM–L2 ma-
terial and not with the material walls, as in this case the Si signals Qn
from the surface and Tn from the ligands would change more. For MCM-
WBrL2 the resonances appear at ꢀ 109.8 (Q4) and ꢀ 102.4 (Q3) ppm, and
at ꢀ 67.6 (T3) and ꢀ 59.0 (T2) ppm.
The 13C CP MAS NMR spectrum (not shown) of material MCM–L2 is
very similar to that of the L2 ligand, suggesting that the ligand is bound
to the surface of material and was not affected by the immobilization
procedure. It shows two broad signals corresponding to the aromatic
carbons of the pyridine ring at 123.2 and 127.5 ppm. The protons of the
alkyl chain of the ligand (1, 2, 3 in Scheme 2) show signals at 9.5, 21.0,
and 43.0 ppm, which were assigned to SiCH2CH2CH2N, SiCH2CH2CH2N
and SiCH2CH2CH2N. The 13C NMR spectra of the metal containing
materials MCM–MoBrL2, MCM–MoIL2, MCM–WBrL2 and MCM–WIL2
(Fig. 5) are not very different, suggesting that the presence of the Mo
nuclei practically does not modify the chemical environment experi-
enced by the carbon nuclei of the ligand. The spectra display signals at
around 9.0–10.0 ppm assigned to the SiCH2CH2CH2N atom, at around
22 ppm to SiCH2CH2CH2N, and at around 43 ppm corresponding to the
carbon adjacent to the nitrogen atom, SiCH2CH2CH2N. The Si atom is
still bound to one OEt group (Scheme 3), with signals at 18 and 59.5
ppm, belonging to CH3 and CH2, respectively. The carbon atom 4
(Scheme 2) is observed at 67 ppm and the CH3 (Y substituent, Scheme 2)
at ~ 60 ppm.
2.3. Catalytic studies. Complexes
Fig. 4. 29Si MAS NMR spectra of the MCM parent and corresponding func-
tionalized materials MCM-L2, MCM-MoBrL2, MCM-MoIL2, MCM-WBrL2 and
MCM-WIL2.
All complexes, WXLn (Ln = L1, L2, L3; X = Br, I), and MoXL2 (X = Br,
I), MoXL3 (X = Br, I), were tested as catalyst precursors in the epoxi-
dation catalysis of the set of multifunctional olefins cis–hex-3-en-1-ol
5

M. Vasconcellos-Dias et al.
Inorganica Chimica Acta 519 (2021) 120263
(cis), trans-hex-3-en-1-ol (trans), geraniol (ger), S-limonene (lim) and
1–octene (1-oct), with tert-butyl hydroperoxide (TBHP) as oxygen
donor, using dichloromethane as solvent, at 328 K. The results obtained
are summarized in Table 2 (see details in Experimental section). No
reaction occurs in the absence of a catalyst.
conversions, the tungsten catalysts being the ones with lowest conver-
sion. However, catalysts MoXL3, WXL2, and WXL3 (X = I, Br) showed
very high conversions above 92%.
The epoxidation of trans-hex-3-en-1-ol is promoted by all catalysts
with good substrate conversion (61–99%, Table 2, entries 11–20).
Epoxidation of the trans-hex-3-en-1-ol olefin led to the formation of
hexan-1-ol-3,4-epoxide (100%) in all cases, except with MoBrL2 and
WXL1 (X = I, Br), which present high conversions (67, 90, and 89%) but
a low or very low selectivity toward hexan-1-ol-3,4-epoxide (30, 20, and
10%).
With very few exceptions, the complexes containing the iodide
ligand were more active and selective than the complexes with the
bromide ligand.
All the tested catalysts oxidized cis-hex-3-en-1-ol with 100% selec-
tivity to the respective epoxide (entries 1–10). In general, those con-
taining iodide show higher conversions than those with bromide.
Tungsten complexes containing the ligand L1, WXL1 (X = I, Br), and
molybdenum complexes MoXL2 (X = I, Br) showed fair to lower
–
The two non-equivalent C C bonds of S-limonene and geraniol may
–
give rise to two different epoxides upon oxidation. Those favored by the
catalysts are sketched in Table 2 and correspond to epoxidation in the
Table 2
Olefin epoxidation catalyzed with complexes MXLy (M ¼ Mo, W, Ln ¼ L1, L2, L3 X ¼ Br, I) of cis-hex-3-en-1-ol (cis), trans-hex-3-en-1-ol (trans), geraniol (ger), S-
limonene and 1-octene (oct).
Entry
Substratea
Catalyst
Conv.b
%
Select.c
%
Yield %
1
2
WBrL1
WIL1
3
100
100
100
100
100
100
100
100
100
100
3
14
97
100
99
92
18
41
95
98
14
97
100
99
92
18
41
95
98
3
WBrL2
WIL2
cis-hex-3-en-1-ol
4
5
WBrL3
WIL3
6
7
MoBrL2
MoIL2
MoBrL3
MoIL3
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
WBrL1
WIL1
89
10
9
90
22
20
trans-hex-3-en-1-ol
WBrL2
WIL2
61
100
100
100
100
30
61
96
96
WBrL3
WIL3
75
75
82
82
MoBrL2
MoIL2
MoBrL3
MoIL3
WBrL1
WIL1
67
20
67
100
100
100
82
67
99
99
89
89
89
73d
86
80
59/10d
60d
WBrL2
WIL2
100
100
85
60
Geraniol
86
43/43d
45/25d
46/46d
61/6d
80/2d
97d
WBrL3
WIL3
82
92
92
MoBrL2
MoIL2
MoBrL3
MoIL3
WBrL1
WIL1
100
100
97
77
86
100
100
62
98
89/9d
3/0/0e
9/9/0e
8/0/78e
2/2/96e
12/14/76e
2/4/19e
64/14/0e
57/26/0e
5/4/90 e
8/12/80 e
5
24
76
S-limonene
WBrL2
WIL2
100
100
97
8
4
WBrL3
WIL3
26
31
18
MoBrL2
MoIL2
MoBrL3
MoIL3
100
100
100
100
78
83
9
20
41
42
43
44
45
46
47
48
49
50
WBrL1
WIL1
42
2
27
2
100
94
1
WBrL2
WIL2
30
34
5
26
31
4
1-octene
91
WBrL3
WIL3
77
39
1
93
36
1
MoBrL2
MoIL2
MoBrL3
MoIL3
100
67
9
6
33
68
96
31
64
94
a All reactions were carried out in dichloromethane in the presence of 2 eq. of oxidant (TBHP) and 1 mol % of M catalyst (M = Mo or W) at 328 K, preferred epoxide
shown; b Calculated after 24 h; c Calculated as ‘’Yield of epoxide’’/’’Conversion’’*100%; d Yield of 2,3-/6,7-oxyrane epoxides; e Yield of Z-epoxide, E-epoxide, and di-
epoxide.
6

M. Vasconcellos-Dias et al.
Inorganica Chimica Acta 519 (2021) 120263
2,3 positions, for geraniol, and in the ring, for S-limonene.
2.4. Catalytic studies. Materials
All complexes led to very high conversion (85–100%) of geraniol
with high selectivity (60–100%). The 2,3-oxyrane epoxide (Table 2) was
preferred, but variable amounts of the 6,7-oxyrane epoxide were also
obtained. The tungsten complexes WIL2 and WIL3 showed similar
selectivity for the two epoxides (43/43 and 46/46% for 2,3-/6,7-oxy-
rane epoxides, respectively), and WBrL3 behaved similarly with a 45/
25% ratio (entries 24–26). The highest conversion and selectivity to-
ward the 2,3-oxyrane epoxide was obtained with MoBrL3 (97%, entry
29). WBrL1 and the WIL2 had 100% conversions, but selectivity was
lower (entries 23 and 24). The remaining catalysts showed a preference
for the 2,3 product, but the 6,7 was also formed.
All materials MCM-WBrL1, MCM-WIL1, MCM-MoBrL2, MCM-
MoIL2, MCM-WBrL2 and MCM-WIL2 were tested as catalyst precursors
for olefin epoxidation using the same multifunctional olefins as above:
cis-hex-3-en-1-ol (cis), trans-hex-3-en-1-ol (trans), geraniol (ger), S-
limonene (lim) and 1-octene (oct), using tert-butylhydroperoxide (TBHP
in decane) as oxygen donor, in dichloromethane, at 328 K (see details in
Experimental section). No reaction occurred in the absence of a catalyst.
All the tested catalysts are 100% selective towards the epoxide in the
oxidation of cis-hex-3-en-1-ol (cis), as seen in Table 3, entries 1–6. While
the molybdenum containing materials lead to 100% yield for the desired
epoxide, the tungsten catalysts display modest conversions in the range
21–47%, the lowest being obtained with the Br containing materials.
MCM-MoBrL2 and MCM-MoIL2, as well as MCM-WBrL1 and MCM-
WIL1, are more active catalysts than the complexes, while MCM-WBrL2
and MCM-WIL2 become less active.
When the substrate is S-limonene (Table 2, entries 31–40), conver-
sions of 100% are reached with the molybdenum catalysts and WXL2,
WXL1 are almost inactive (5 and 24%, for X = Br and I, respectively),
while WXL3 present 97 and 31% conversions, for X = Br and I. The Z-
and E-epoxides are always observed, except with complexes WBrL1 and
WBrL2 (only the E-epoxide), and the yield is never very good. The
highest one for the Z-epoxide is 64% with MoBrL2 (entry 37), and the
values for the E-epoxide are much lower. High yield is observed for the
double epoxide with catalysts MoBrL3 (90%), MoIL3 (80%), WBrL2
(78%), WIL2 (96%), and WBrL3 (76%), in entries 39, 40, and 33–35.
The oxidation of 1-octene (Table 2, entries 41–50), an unbranched
(not activated) terminal olefin, was achieved with medium to fair con-
versions by all complexes, reaching 68% with MoIL3. The highest yield
of the epoxide was observed with the complexes WBrL2, WIL2, WIL3,
MoBrL3, and MoIL3, catalyst (entries 43, 44, 46, 49, and 50), with yields
from 26 to 64%.
The same trends are observed for all catalysts in the oxidation of
trans (Table 3, entries 7–12), with 100% selectivity for all of them, and
with the highest conversions for the molybdenum catalysts (MCM-
MoBrL2 100% and MCM-MoIL2 93%), while the tungsten ones exhibit
conversions ranging from 19 to 56%. Again the molybdenum materials
are more active than the respective complexes, while all the tungsten
complexes become less active and less selective when grafted in MCM.
MCM-MoBrL2 and MCM-MoIL2 are again the most active catalysts
in the oxidation of geraniol, with conversions of 98 and 100%, respec-
tively, similar to those afforded by the complexes (Table 3, entries
13–18). The tungsten catalysts lead to lower conversions than the mo-
lybdenum materials and their corresponding complexes. In general, the
heterogeneous catalysts are less selective toward the 2,3-oxyrane
product than the homogeneous ones, the 6,7-oxyrane being formed in
higher amounts.
We can compare the activity of the previous catalysts, especially
those containing molybdenum, with that of the complexes [Mo
(
η3–C3H5)X(CO)2(LY)] (LY = L1, L2, and L3, respectively 1, 2, and 3)
[13]. These three complexes oxidized the cis substrate with 99% con-
version and 100% selectivity, which is much better than that obtained
with MoBrL2 and WBrL1, slightly better than with MoBrL3 and WBrL2,
and the same as WBrL3. Complex 2 was more selective than MoBrL2 for
the epoxidation of trans, but MoIL3, on the other hand, converted trans
with much higher selectivity and conversion than complex 3. The con-
versions in the oxidation of ger catalyzed by complexes 1, 2, 3 and
MoXL2,3 (X = Br, I) were very similar. However, the four molybdenum
complexes MoXL2,3 (X = Br, I) were much more selective than 1, 2, and
3, with high yields of the 2,3-oxyrane product, while complexes 1–3 led
to similar amounts of 2,3- and 6,7-oxyrane. None of the two groups of
complexes exhibited high conversions for the epoxidation of 1-octene,
and selectivity was similar. MoBrL1 and MoIL1 were tested in a
different group of substrates and cannot be easily compared. In general,
no clear effect of the substitution pattern in L1, L2 and L3 is observed.
There are not extremely large differences between the W and the Mo
catalysts. However, when the complexes [MX2(CO)3(2-amino-1,3,4-
thiadiazole)2] (M = Mo, W) were tested, the W(II) ones were distinctly
more active from those of Mo(II) [16]. We can compare the activity of
these complexes with those in the present work in the oxidation of the
common substrate S-limonene. The Mo(II) 2-amino-1,3,4-thiadiazole
complexes have 0% conversion and MoX2L2/L3 (X = Br, I) 100%,
while the differences for W are small. This can be explained by the
stronger W-L bonds, suggesting that the binuclear W(II) active species
with the monodentate species is robust enough to undergo catalysis,
The two molybdenum heterogeneous catalysts MCM-MoXL2 (X = Br,
I) lead to 100% conversions in the oxidation of S-limonene (Table 3
entries 23, 24), but the other materials are much less active (17–35%,
Table 3, entries 19–22). However, MCM-MoBrL2 is very selective to-
ward the E-epoxide (88%), while MCM-MoIL2 produces the same
amount of the Z- and E-epoxides. The di–epoxide, formed by some ho-
mogeneous catalysts, is not observed. The activity of all the tungsten
materials was lower than that of the complexes.
The oxidation of 1-octene was hard to achieve by the MCM materials
(Table 3, entries 25–30), and both the conversions and the selectivity
were in general very low.
In conclusion it is possible to say that the immobilization of the
molybdenum complexes in MCM improved their capability as catalysts,
making them more active and selective for most of the substrates tested.
The tungsten complexes did not become better catalysts after their
immobilization. The results of conversions and selectivity for the cata-
lysts supported in MCM were lower or, in a few cases, similar to those
obtained for the complexes.
It is interesting to note that the [Mo(η3–C3H5)X(CO)2(L)] complexes
(L = L1, L2, and L3, respectively 1, 2, and 3) [13] became less active after
immobilization. However, the activity of molybdenum materials
MoBrL1 and MoIL1, described previously, increased after being sup-
ported in MCM.
Although tungsten complexes are good catalysts, they do not
improve when they are supported and act as heterogeneous catalysts,
and the same happens with the molybdenum complexes of the type [Mo
while the Mo(II) is not. The chelation effect introduced by the α-diimine
ligands stabilizes the active species, making the Mo(II) catalysts more
stable.
(
η3–C3H5)X(CO)2(LY)] [13]. On the contrary, complexes [Mo
On the other hand, a comparative study of MCl2(O2) (M = Mo, W)
derivatives showed that the performance of W catalysts in cis-cyclo-
octene epoxidation was higher in the presence of H2O2 as oxidant, and
lower when TBHP was used, emphasizing the subtle balance [24].
(X)2(CO)3(L1)] [12] and [MoX2(CO)3(2-amino-1,3,4-thiadiazole)2] [16]
become more active in MCM, while [WX2(CO)3(2-amino-1,3,4-thiadia-
zole)2] [16] increase their activity for X = I and decrease it for X = Br.
A quality of this kind of immobilized complexes is their recyclability
and absence of leaching, which was studied previously in a series of
functionalized materials. It was found that the conversions decreased by
small amounts in five successive runs of the epoxidation of some of trans-
7

M. Vasconcellos-Dias et al.
Inorganica Chimica Acta 519 (2021) 120263
Table 3
Olefin epoxidation catalyzed with materials MCM-WBrL1, MCM-WIL1, MCM-MoBrL2, MCM-MoIL2, MCM-WBrL2 and MCM-WIL2, of cis-hex-3-en-1-ol (cis), trans-
hex-3-en-1-ol (trans), geraniol (ger), S-limonene (lim) and 1-octene (1-oct).
Entry
Substrates
Catalyst
Conv.b
%
Select.c
%
Yield %
1
2
3
4
5
6
MCM-WBrL1
MCM-WIL1
30
100
100
100
100
100
100
30
47
47
MCM-WBrL2
MCM-WIL2
21
21
cis-hex-3-en-1-ol
40
40
MCM-MoBrL2
MCM-MoIL2
100
100
100
100
7
MCM-WBrL1
MCM-WIL1
31
43
19
56
100
93
100
100
100
100
100
100
31
43
19
56
100
93
8
trans-hex-3-en-1-ol
9
MCM-WBrL2
MCM-WIL2
10
11
12
MCM-MoBrL2
MCM-MoIL2
13
14
15
16
17
18
MCM-WBrL1
MCM-WIL1
75
89
85
58
98
100
74
45
77
98
98
92
30/25d
39/39d
50/15d
37/20d
70/28d
56/36d
geraniol
MCM-WBrL2
MCM-WIL2
MCM-MoBrL2
MCM-MoIL2
19
20
21
22
23
24
MCM-WBrL1
MCM-WIL1
25
86
61
84
86
97
98
10/12e
5/5e
17
S-limonene
MCM-WBrL2
MCM-WIL2
23
11/18e
15/15e
7/88e
35
MCM-MoBrL2
MCM-MoIL2
100
100
45/46e
25
26
27
28
29
30
MCM-WBrL1
MCM-WIL1
16
9
11
45
25
23
24
22
2
4
1-octene
MCM-WBrL2
MCM-WIL2
1
0.3
3
10
41
25
MCM-MoBrL2
MCM-MoIL2
10
6
a All reactions were carried out in dichloromethane in the presence of 2 eq. of oxidant (TBHP) and 175 mg of catalyst at 328 K; b Calculated after 24 h; c Calculated as
‘’Yield of epoxide’’/’’Conversion’’*100%; d Yield of 2,3-/ 6,7-oxyrane products; e Yield of Z-epoxide, E-epoxide.
hex-2-en-1-ol catalyzed by MoI2(CO)3(IMP) or MoBr(
η
3-C3H5)
2.5. Catalytic studies. Mechanisms
(CO)2(IMP) complexes immobilized in MCM in the presence of TBHP
[25], and also by MoI2(CO)3 complexes anchored in a helical MCM
material by an asymmetric bidentate nitrogen ligand derived from 2-
aminopyridine glycidyloxypropyl [26]. The Mo(CO3)I2 complex of 2-
(2′-hydroxyphenyl)imidazoline underwent a 20% deactivation in the
epoxidation of R(+)-limonene, but after grafting a similar deactivation
was only observed in the 3rd run [27]. These examples indicate that the
immobilized Mo(II) precursors provide recyclable heterogeneous
catalysts.
The reaction mechanism should help to understand the catalysis
results. Since the M(II) complexes and materials are catalyst precursors,
the TBHP oxidant starts by oxidizing them to M(VI) species. This reac-
tion has been well studied for the complex [Mo(II)(η5–C5H5)(CO)3(X)]
(X = Cl or CH3), which is converted to [Mo(VI)(η5–C5H5)O2(X)], the
catalyst. The latter complexes have been isolated and characterized,
[14,28,29] and a complete computational study led to the identification
of the active species [30,31]. The reaction mechanism of the alkene
epoxidation was very similar to that proposed for the [Mo(VI)
Leaching was also examined previously. For instance, the Mo(CO3)I2
complex of 2-(2′-hydroxyphenyl)imidazoline grafted in MCM was
applied to the epoxidation of R(+)-limonene. The solution was filtered
after 2 h and the remaining solution showed no activity during the
following 22 h, while without filtering, the heterogeneous catalyst
remained active for a few hours [25]. The suspension containing the [Mo
–
–
O2X2(N N)] complex, with X = Cl, Br, CH3, and N N = bidentate
–
nitrogen ligand (see Figure S3) [32,33]. The complex [MoO2X2(N N)]
is the active Mo(VI) catalyst (A1). In the first step, the HOOR oxidant
– –
O X face
approaches the metal over the center of a triangular O
(Figure S3, top) to form a heptacoordinate intermediate [MoO(OH)
η5–C3H5)Cl(CO)2(L1)] complex immobilized in MCM lost its activity in
(OOR)X2(N N)] (B1), where the OH and OOR ligands are bound by a
–
(
–
the epoxidation of cis-cyclooctene after removing the catalyst by filtra-
OH⋅⋅⋅H hydrogen bond (B ) [32]. The C C bond adds to the Mo-O bond,
–
1
tion [12].
forming a large metallacycle including the OH⋅⋅⋅H hydrogen bond (C1),
8

M. Vasconcellos-Dias et al.
Inorganica Chimica Acta 519 (2021) 120263
which breaks down into the epoxide, the ROH alcohol, and the regen-
erated catalyst [33]. In Figure S4, we show a similar catalytic cycle for
the [Mo(η5–C5H5)(CO)3(X)] system in a simplified version. The first step
is the oxidation of this Mo(II) complex to the active Mo(VI) species [Mo
oxidation of the allyl complexes is much less exergonic than that of the
other ones, with Gibbs energies ranging between ~ꢀ 74 and ꢀ 117 kcal
molꢀ 1, and ~ꢀ 259 and ꢀ 292 kcal molꢀ 1, respectively, but this results
from the different nature of the complexes and their products (the en-
ergy values are collected in Table S3, SI). On the other hand, the tung-
sten complexes are easier to oxidize than the molybdenum ones, by
~20–25 kcal molꢀ 1, reflecting the lower ionization energy of W
compared to Mo, and the differences between Br and I derivatives are
smaller, though reaction of iodides are more exergonic. However, the
catalytic cycles must also be considered.
(
η5–C5H5)O2(X)] (A2), from which the catalysis proceeds, over in-
termediates B2 (OH activation of the HOOR oxidant) and C2 (alkene
addition) [30,31,34]. These mechanisms have been reviewed [35,36].
The complexes [Mo(η5-C3H5)Cl(CO)2(N N)] (N
= 1,4-(4-
–
–
N
chloro)phenyl-2,3-naphthalenediazabutadiene) behave differently in
the presence of TBHP. IR spectroscopic analysis and elemental analysis
suggested that the most likely Mo(VI) intermediate was the dimer
The catalytic cycles are shown in Fig. 6, and S3-S5. Only the three
structures of intermediates A, B, and C were calculated for all systems. A
set of structures of all the intermediates is shown in Fig. 7 for
[MI2(CO)3(L1)] and in Figure S6 for [Mo(η5–C3H5)Br(CO)2(L1)] and all
their derivatives. The others are extremely similar. The determination of
transition states was not attempted, owing the size of the binuclear
complexes.
–
[{MoO2Cl(N N)}2(μ-O)] (A3) [15]. The determination of the single
crystal X–ray diffraction structure of the analogue [{MoO2Cl(bpy*)}2(μ-
O) ] (bpy* = 4,4′-di-tert-Bu-2,2′-bipyridine) [37] and a computational
study [15] confirmed the nature of the catalyst. The M(II) precursor in
the present study, [MX2(CO)3(L1,2,3)] (M = W, Mo), where the allyl
ligand is replaced by CO and X, should by analogy, be converted in the
new catalyst [{MX2(L1,2,3)}2(
μ
-O) ]2+(A4), after loss of the carbonyl li-
The more deeply studied reactions of the Mo(VI) mononuclear cat-
alysts show very clearly that the step with the largest barrier corre-
gands. The proposed catalytic cycles, including the three more relevant
intermediates (A, B, C), are shown in Figures S5 for the [Mo(η5–C3H5)X
sponds to the B → C conversion. It involves the entropy unfavored
–
–
(CO)2(N N)] catalysts (A3, B3, C3) and in Fig. 6 for the complexes
addition of the C C bond of the alkene to the Mo-O(OR) bond of B (see
–
[MX2(CO)3(L1,2,3)] addressed in the present work (A4, B4, C4).
Figures S3 and S4). C is the intermediate with the highest energy and
gives rise to the products and the regenerated catalyst. The relative
energies of the three tungsten and molybdenum intermediates addressed
in this work are collected in Table 4. The intermediate C always has the
highest energy and by comparison with previous systems [30,31,33] it
may be inferred that the activation barriers associated with the forma-
tion of C and its decomposition will be slightly higher than the energy of
The activity of the catalysts was theoretically ascertained through
DFT calculations [38] (ADF program [39–41], see Computational de-
tails). We started by analyzing the oxidation step, and calculated the
electronic and Gibbs energy of the oxidation reaction for the four
[MX2(CO)3(L1,2,3)] systems (W, Mo; Br, I) and compared them with the
related [Mo(η5–C3H5)Cl(CO)2(N N)] systems shown in Fig. S5. The
–
Fig. 6. Simplified mechanism of alkene epoxidation with HOOR by the [{MoX2(L1)}2(
μ
–O)]2+ catalyst (A4). The oxidation of [MX2(CO)3(L1,2,3)] (M = W, Mo; X = Br,
I) to A4 by HOOR is shown in the top rectangle.
9

M. Vasconcellos-Dias et al.
Inorganica Chimica Acta 519 (2021) 120263
complexes of this type and also of their immobilization in MCM
materials.
We studied the catalytic activity of these complexes and materials in
the oxidation of cis–hex-3-en-1-ol (cis), trans-hex-3-en-1-ol (trans), ge-
raniol (ger), S-limonene (lim) and 1–octene (1-oct), with TBHP. The
complexes studied here were in general less active in the oxidation of
trans than [Mo(η3–C3H5)X(CO)2(LY)] (L = L1, L2, and L3, respectively 1,
2, and 3) [13], which oxidized this substrate with 99% conversion and
100% selectivity. Complex 2 was more selective than MoBrL2 for the
epoxidation of trans, but MoIL3, on the other hand, converted trans with
much higher selectivity and conversion than complex 3. The two fam-
ilies behaved similarly in the oxidation of ger (very active) and 1-octene
(poorly).
DFT calculations suggested that the iodide complexes were more
easily oxidized by ROOH, but the catalytic reaction was less favored
than for bromides. This and the very close energies found for the cata-
lytic reaction prevent us from drawing reliable conclusions.
The immobilization of the molybdenum complexes MXLY in MCM
made them more active and selective for most of the substrates tested
(also MXL1 with other substrates). The opposite, however, happened
with the tungsten complexes, as the conversions and selectivity for the
catalysts supported in MCM were lower or, in a few cases, similar to
those obtained for the complexes.
Although tungsten complexes are good catalysts, they do not become
better when supported in MCM. Since their synthesis is more difficult
than the synthesis of the molybdenum analogues, they do not prove to
be very promising. Some of the molybdenum heterogeneous catalysts
are more active than the complexes and share the advantages of het-
erogeneous catalysts.
Fig. 7. Three dimensional structures of complex [WI2(CO)3(L1)], the dimer
formed upon oxidation by HOOR (A), the binuclear species resulting from OH
activation of HOOR (B’), and the binuclear complex formed after addition of
ethylene to B.
C. These energies of C are higher for the iodide complexes than the
bromide ones, so that their higher activity must be associated with the
previous oxidation reaction. Two different intermediates of type B in-
4. Experimental
–
termediate (B and B’) with a O H⋅⋅O hydrogen bond, differing by their
orientation relative to the Mꢀ Oꢀ M line, were found for the allyl system
with Br (Fig S6). B, with a Br-Mo-O-Mo-Br axis, is more stable than B’
with a Br-Mo-O-Mo-OH one. However, a type B intermediate could not
be obtained for the iodides. In all the iodides, the second intermediate is
B’ as shown in Fig. 7 for the [WI2(CO)3(L1)] system (notice the I-W-O-W-
OH line).
4.1. General considerations
All reagents were obtained from Aldrich and used as received. All the
work involving sensitive compounds was carried out using standard
Schlenk techniques. Commercial-grade solvents were dried and deoxy-
genated by standard procedures (i.e., toluene over Na/benzophenone
ketyl, dichloromethane and hexane over CaH2), distilled under nitrogen,
and kept over 4 Å molecular sieves.
The Gibbs energies are much higher than the electronic energies,
because the two reactions, the OH activation of the ROOH oxidant and
the addition of ethylene to the Mo-O bond, are both associative. The
values are extremely similar for all complexes, even more than the en-
ergies, in this model system where the substrate was ethylene and the
oxidant MeOOH. This helps to explain the apparent random results and
the absence of a well defined pattern.
The organometallic complexes [MX2(CH3CN)2(CO)3] (M = Mo, X =
I, Br; M = W, X = I, X = Br) were prepared as described before [42,43].
Ligands C5H4NCY = N(CH2)2CH3 (Y = H, L1; Y = CH3, L2; Y = Ph, L3)
and C5H4NCY = N(CH2)3Si(OEt3) (Y = H, L1Si Y = CH3, L2Si) were also
prepared as reported [12]. MCM-41 and post synthetic derivatized
materials were obtained as described previously, using [(C16H33)N
(CH3)3]Br (CTAB) as the templating agent [13,18]. Physisorbed water
3. Conclusions
◦
was removed (540 C for six hours under air) from calcined MCM-41
A family of [MX2(CO)3(LY)] complexes (MXLy: M = W, Mo; X = Br, I;
LY = L1-3, Scheme 2) was prepared and some of them immobilized in
MCM. Complexes [WX2(CO)3(L1)] (X = Br, WBrL1 and X = I, WL2I)
were structurally characterized by single crystal X-ray diffraction. To the
best of our knowledge, these are the first examples of tungsten
before the grafting experiment, by heating at 180 ◦C in vacuum (10-2 Pa)
during two hours. Hybrid materials were prepared using ligands L1Si
and L2Si as reported, with an organic material load of ≈ 3% [12,44,45].
FTIR spectra were obtained as KBr pellets and diffuse reflectance
measurements (DRIFT) on a Nicolet 6700 in the 400–4000 cmꢀ 1 range
Table 4
Calculated relative energies and Gibbs energies (kcal molꢀ 1) for the A, B, and C intermediates in the catalytic epoxidation of ethylene with MeOOH.
ΔE
ΔG
Initial complex
A
B
B’
C
A
B
B’
C
[WBr2(CO)3(L1)]
0
0
0
0
0
0
0
0
9.94
20.82
24.90
18.92
23.79
23.99
25.45
24.07
25.14
0
0
0
0
0
0
0
0
25.14
50.59
54.95
47.79
54.49
53.99
55.4
[WI2(CO)3(L1)]
15.79
30.55
[MoBr2(CO)3(L1)]
[MoI2(CO)3(L1)]
11.9
26.86
16.17
9.79
30.98
25.81
[W(η3–C3H5)Br(CO)2(L1)]
[W(η3–C3H5)I(CO)2(L1)]
8.94
24.03
23.59
25.67
25.67
9.20
10.68
10.78
[Mo(
η
3–C3H5)Br(CO)2(L1)]
12.31
28.1
54.11
54.11
[Mo(η3–C3H5)I(CO)2(L1)]
10

M. Vasconcellos-Dias et al.
Inorganica Chimica Acta 519 (2021) 120263
(cmꢀ 1 resolution). Powder XRD measurements were taken on a Philips
Analytical PW 3050/60 X’Pert PRO (θ/2θ) equipped with X’Celerator
detector and with automatic data acquisition (X’Pert Data Collector
13C{1H} NMR (400.13 MHz, (CD3)2(CO), room temp., ppm): 10.2
(C1), 20.7 (C2), 41.9 (C3), 122.1 (C8), 128.6 (C7), 137.7 (C6), 150.3
(C9), 165.8 (C4), 184.7 (CO) 194.4 (CO).
(v2.0b) software), using a monochromatic Cu Kα radiation as the inci-
dent beam, operating at 40 kV–30 mA. XRD diffraction patterns were
obtained by continuous scanning in a 2θ range of 2 to 10◦ with 2θ step
size of 0.017◦ and a scan step time of 99.695 s. 1H and 13C solution NMR
spectra were obtained with a Bruker Avance 400 spectrometer. 29Si and
13C solid state NMR spectra were performed at University of Aveiro by
Dr Paula Ferreira, and were recorded at 79.49 MHz and 100.62 MHz,
respectively, on a (9.4 T) Bruker Avance 400P spectrometer. 29Si CP
4.3.1.2. [WI2(CO)3C5H4NCH = N(CH2)2CH3] (WIL1). Yield (η): 98%
(0.66 g) (red powder),
Elemental analysis WI2C12H12N2O3 (669.9) (%): Calc C 21.52, N
4.18, H 1.81; found: C 21.65, N 4.31, H 2.08.
IR (KBr pellets,
ν
cmꢀ 1): 3071 (w), 3016 (w), 2972 (m), 2969 (m),
2933 (w), 2885 (m), 2014 (vs), 1935 (vs), 1896 (vs), 1617 (m), 1595
(m), 1554 (w), 1480 (s), 1391 (m), 1310 (s), 1260 (m), 1239 (m), 1159
MAS NMR spectra were recorded with 5.5 μ
s 1H 90◦ pulse, 2 ms contact
(m), 770 (vs), 754 (s), 519 (s), 494 (s) cmꢀ 1
.
1H NMR (400.13 MHz, (CD3)2(CO), room temp., ppm): 1.05 (t, 3H,
H1), 2.05 (m, 2H, H2), 4.46 (t, 2H, H3), 7.88 (t, 1H, H8), 8.41 (t, 1H, H7),
8.51 (t, 1H, H6), 9.33 (s, 1H, H4), 9.81 (d, 1H, H9).
time, a spinning rate of 8 kHz, and 4 s recycle delays. Chemical shifts are
quoted in ppm from TMS. 13C spectra (solid state) were recorded at
125.76 MHz on a Bruker Avance 500 spectrometer. The N2 sorption
measurements were obtained in an automatic apparatus (ASAP 2010;
Micrometrics). BET specific surface areas (SBET, p/p0 from 0.03 to 0.13)
and specific total pore volume, Vp were estimated from N2 adsorption
isotherms measured at 77 K. The pore size distribution (PSD) was
calculated by the BJH method using the modified Kelvin equation, with
correction for the statistical film thickness on the pore walls [23]. The
statistical film thickness was calculated using the Harkins-Jura equation
in the p/p0 range from 0.1 to 0.95. Microanalyses (C, N, H) and ICP
analyses for determination of molybdenum and tungsten were per-
formed at the University of Vigo.
13C{1H} NMR (400.13 MHz, (CD3)2(CO), room temp., ppm): 10.3
(C1), 24.4 (C2), 67.4 (C3), 128.0 (C8), 130.3 (C7), 140.4 (C6), 155.4
(C9), 168.2 (C4).
4.3.2. [MX2(CO)3{C5H4NCCH3 = N(CH2)2CH3}] (M = W, Mo and X =
I, Br)
A solution of [MX2(CO)3(CH3CN)2] (M = W, Mo and X = I, Br) (0.5
mmol) in dichloromethane (10 mL) was added to a solution of the ligand
C5H4NCCH3 = N(CH2)2CH3 (L2, 1.2 mmol, 0.195 g) also in dichloro-
methane (10 mL). The resulting mixture was stirred for 4 h (tungsten) or
6 h (molybdenum) at room temperature and under a nitrogen atmo-
sphere. The solution was filtered and concentrated to a solid. After
filtration, the product was recrystallized from hexane.
4.2. Catalytic tests
The complexes and materials were tested in epoxidation of multi-
functional olefins, cis-hex-3-en-1-ol (cis), trans-hex-3-en-1-ol (trans),
geraniol (ger), S-limonene (lim) and 1-octene (1-oct), using tert-butyl-
hydroperoxide (TBHP) as oxidant (5.5 M in n-decane). The catalytic
oxidation tests were carried out at 328 K, using dichloromethane as
solvent. The reactions occurred under air in a reaction vessel equipped
with a magnetic stirrer and a condenser. In a typical experiment, the
vessel was loaded with catalyst (1 mol% for complexes and 175 mg for
materials), olefin (100 mol%), oxidant (200 mol%) and 3 mL of solvent.
The final volume of the reaction was ca. 6 mL. The addition of the
oxidant determined the initial time of the reaction. Conversion, product
yields and stereochemistry were monitored by sampling periodically
and analyzing the products on a Shimadzu QP2010-Plus GC/MS system
and a capillary column (Teknokroma TRB- 5MS, TRB-1MS) operating in
the linear velocity mode.
4.3.2.1. [WBr2(CO)3C5H4NCCH3 = N(CH2)2CH3] (WBrL2). Yield (η):
80% (0.47 g) (red powder)
Elemental analysis WBr2C13H14N2O3 (589.92) (%): Calc: C 26.47, N
4.75, H 2.39 found: C 26.13, N 4.59, H 2.03.
IR (KBr pellets,
ν
cmꢀ 1): 3411 (s), 3046 (m), 2960 (s), 2932 (s), 2874
(m), 2033 (vs), 1949 (vs), 1902 (vs), 1619 (s), 1603 (s), 1530 (m) 1435
(s), 1383 (m), 1260 (m), 1163 (m), 1093 (m), 1049 (m), 1026 (w), 951
(m), 779 (s), 750 (s).
1H NMR (400.13 MHz, CDCl3, room temp., ppm): 1.06 (t, 3H, H1),
2.13 (t, 2H, H2), 2.69 (s, 3H, H10) 3.84 (t, 2H, H3), 7.92 (t, 1H, H8), 8.20
(d, 1H, H6), 8.35 (t, 1H, H7), 8.76 (d, 1H, H9).
13C{1H} NMR (400.13 MHz, CDCl3, room temp., ppm): 12.1 (C1),
14.9 (C2), 23.7 (C10), 54.2 (C3), 124.5 (C8), 128.3 (C6), 141.8 (C7),
149.1 (C9), 149.9 (C5), 167.2 (C4).
4.3.2.2. [WI2(CO)3C5H4NCCH3 = N(CH2)2CH3] (WIL2). Yield (η): 88%
4.3. Synthesis of complexes
(0.60 g) (purple powder)
Elemental analysis WI2C13H14N2O3 (683.92) (%): Calc C 22.83, N
4.10, H 2.06 found: C 22.77, N 3.81, H 1.88.
4.3.1. [WX2(CO)3{C5H4NCH = N(CH2)2CH3}] (X = Br, I)
cmꢀ 1): 3421 (s), 3041 (m), 2966 (m), 2903 (m),
A solution of [WX2(CO)3(CH3CN)2] (X = I, Br) (1 mmol) in
dichloromethane (10 mL) was added to a solution of the ligand
C5H4NCH = N(CH2)2CH3 (L1, 1.2 mmol, 0.178 g) in dichloromethane
(10 mL). The resulting mixture was stirred for 6 h at room temperature
and under a nitrogen atmosphere. The solution was filtered and
concentrated by evaporation, and pentane was added. The precipitate
was filtered, and the product dried under vacuum.
IR (KBr pellets,
ν
2873 (w), 2065 (m), 2019 (s), 2003 (m), 1925 (vs), 1654 (m), 1605 (m),
1528 (m), 1488 (w), 1457 (s), 1384 (m), 1305 (m), 1227 (m), 1162 (m),
1092 (w), 1027 (w), 946 (w), 776 (s), 732 (s).
1H NMR (400.13 MHz, (CD3)2(CO), room temp., ppm): 1.08 (t, 3H,
H1), 2.10 (s, 3H, H10), 2.85 (t, 2H, H2) 3.78 (t, 2H, H3), 7.82 (t, 1H, H8),
8.40 (d, 1H, H6), 8.52 (t, 1H, H7), 9.56 (d, 1H, H9).
13C{1H} NMR (400.13 MHz, CDCl3, room temp., ppm): 13.2 (C1),
22.9 (C2), 27.7 (C10), 44.0 (C3), 123.6 (C8), 128.8 (C6), 138.8 (C7),
151.0 (C9), 203.2 (CO), 209.3 (CO).
4.3.1.1. [WBr2(CO)3C5H4NCH = N(CH2)2CH3] (WBrL1). Yield (η):
88% (0.51 g) (orange-brick powder)
Elemental analysis WBr2C12H12N2O3 (575.9) (%): Calc C 25.03, N
4.86, H 2.10; found C 25.64, N 4.40, H 2.52.
4.3.2.3. [MoBr2(CO)3{C5H4NCCH3 = N(CH2)2CH3}] (MoBrL2). Yield
IR (KBr pellets,
ν
cmꢀ 1): 3069 (m), 3022 (w), 2971 (m), 2942 (w),
(
η
): 93% (0.47 g) (red powder)
2921 (m), 2878 (w), 2036 (vs), 1956 (vs), 1904 (vs), 1621 (m), 1595
(m), 1477 (s), 1309 (m), 1261 (m), 1242 (m), 782 (s), 543 (s), 490 (s).
1H NMR (400.13 MHz, (CD3)2(CO), room temp., ppm): 1.03 (t, 3H,
H1), 1.87 (m, 2H, H2), 3.14 (t, 2H, H3), 7.91 (t, 1H, H8), 8.43 (t, 1H, H7),
8.52 (t, 1H, H6), 8.97 (d, 1H, H9), 9.37 (s, 1H, H4) ppm.
Elemental analysis MoBr2C13H14N2O3 (502.01) (%): Calc: C 31.10, N
5.58, H 2.81 found: C 29.98, N 5.23, H 2.51.
IR (KBr pellets,
ν
cmꢀ 1): 3440 (s) 3070 (m), 3028 (m), 2964 (m),
2930 (m), 2873 (w), 2044 (m), 1971 (vs), 1921 (s), 1901 (m), 1617 (s),
11

M. Vasconcellos-Dias et al.
Inorganica Chimica Acta 519 (2021) 120263
1531 (w), 1472 (m), 1477 (m), 1435 (m), 1305 (w), 1250 (w), 1164 (w),
1092 (w), 1023 (w), 945 (m), 776 (s), 748 (s).
(C1), 20.7 (C2), 54.1 (C3), 124.3 (C12), 126.5 (C14), 128.0 (C11 C15),
129.5 (C8), 130.8 (C13), 132.7 (C7), 137.4 (C6), 148.4 (C9).
1H NMR (400.13 MHz, (CD3)2(CO), room temp., ppm): 1.21 (t, 3H,
H1), 2.20 (s, 2H, H10), 2.67 (d, 3H, H2), 4.64 (t, 2H, H3), 8.21 (t, 1H, H8),
8.45 (t, 1H, H6), 8.57 (d, 1H, H7), 9.26 (d, 1H, H9).
4.3.3.3. [MoBr2(CO)3C5H4NCPh = N(CH2)2CH3] (MoBrL3). Yield (η):
76% (0.43 g) (green powder)
13C{1H} NMR (400.13 MHz, (CD3)2(CO), room temp., ppm): 13.7
(C1), 18.1 (C2), 24.6 (C10), 55.0 (C3), 121.3 (C8), 124.7 (C6), 136.1
(C7), 148.8 (C9), 158.3 (C5), 166.2 (C4).
Elemental analysis MoBr2C18H16N2O3 (564.08) (%): Calc: C 38.33, N
4.97, H 2.86; found: C 38.16, N 4.42, H 2.71.
IR (KBr pellets,
ν
cmꢀ 1): 3073 (w), 3019 (w), 2965 (m), 2919 (m),
2874 (m), 2040 (s), 1967 (vs), 1918 (s), 1660 (m), 1594 (s), 1522 (m),
1449 (s), 1281 (m), 1162 (m), 1093 (w), 1018 (w), 940 (s), 775 (s), 695
(s).
4.3.2.4. [MoI2(CO)3C5H4NCCH3 = N(CH2)2CH3] (MoIL2). Yield (η):
91% (0.54 g) (red-wine powder)
Elemental analysis MoI2C13H14N2O3 (596.01) (%): Calc: C 26.20, N
4.70, H 2.37. found: C 26.09, N 4.45, H 2.01.
1H NMR (400.13 MHz, (CD3)2(CO), room temp., ppm): 1.05 (t, 3H,
H1), 1.90 (t, 2H, H2), 3.01 (m, 2H, H3), 7.41 (dd, 2H, H12 H14), 7.56 (m,
2H, H11 H15), 7.68 (m, 1H, H8), 7.73 (d, 1H, H13), 8.14 (t, 1H, H7), 8.77
(m, 1H, H6), 9.53 (d, 1H, H9).
IR (KBr pellets,
ν
cmꢀ 1): 3415 (s), 3077 (m), 2990 (m), 2964 (m),
2905 (w), 2872 (w), 2012 (m), 1963 (vs), 1920 (vs), 1850 (m) 1616 (m)
1593 (m), 1433 (m), 1375 (m), 1302 (w), 1250 (m), 1162 (m), 1096 (w),
1021 (w), 994 (w)945 (w), 774 (s), 746 (s).
13C{1H} NMR (400.13 MHz, (CD3)2(CO), room temp., ppm): 11.4
(C1), 22.8 (C2), 39.6 (C3), 124.4 (C12), 126.9 (C14), 127.9 (C11), 129.5
(C15), 130.8 (C8), 132.6 (C13), 136.4 (C7), 138.2 (C6), 143.8 (C10),
148.5 (C9), 154.1 (C4), 154.6 (C5), 192.8 (CO).
1H NMR (400.13 MHz, (CD3)2(CO), room temp., ppm): 1.07 (t, 3H,
H1), 2.01 (s, 3H, H10), 2.73 (d, 2H, H2) 4.48 (t, 2H, H3), 7.93 (t, 1H, H8),
8.46 (t, 1H, H6), 8.52 (d, 1H, H7), 9.68 (d, 1H, H9).
13C{1H} NMR (400.13 MHz, (CD3)2(CO), room temp., ppm): 10.9
(C1), 14.7 (C2), 21.0 (C10), 53.9 (C3), 127.9 (C8), 130.8 (C6), 132.6
(C7), 137.4 (C9), 148.3 (C5).
4.3.3.4. [MoI2(CO)3C5H4NCPh = N(CH2)2CH3] (MoIL3). Yield (η):
87% (0.57 g) (dark-green powder)
Elemental analysis MoI2C18H16N2O3 (658.09) (%): Calc: C 32.85, N
4.26, H 2.45; found: C 32.49, N 4.07, H 2.19.
IR (KBr pellets,
ν
cmꢀ 1): 3462 (s), 3055 (m), 2964 (m), 2930 (w),
4.3.3. [MX2(CO)3{C5H4NC6H5 = N(CH2)2CH3}] (M = W, Mo and X =
I)
2867 (w), 2013 (s), 1966 (s), 1918 (s), 1843 (s), 1662 (m), 1593 (m),
1457 (m), 1445 (m), 1336 (m), 1319 (m), 1282 (w), 1255 (m), 1157 (m),
1088 (w), 1058 (w), 994 (w), 939 (m), 791 (w), 749 (s), 696 (s), 649 (s).
1H NMR (400.13 MHz, (CD3)2(CO), room temp., ppm): 1.05 (t, 3H,
H1), 1.90 (t, 2H, H2), 3.27 (m, 2H, H3), 7.30 (dd, 2H, H12 H14), 7.55 (m,
2H, H11 H15), 7.64 (m, 1H, H8), 7.68 (m, 1H, H13), 8.09 (t, 1H, H6), 8.16
(d, 1H, H7), 8.78 (d, 1H, H9).
A solution of [MX2(CO)3(CH3CN)2] (M = W, Mo and X = I, Br) (0.5
mmol) in dichloromethane (10 mL) was added to a solution of the ligand
C5H4NCCH3 = N(CH2)2CH3 (L3, 1.2 mmol, 0.270 g) also in dichloro-
methane (10 mL). The resulting mixture was allowed to stir for 4 h
(tungsten) or 6 h (molybdenum) at room temperature and under a ni-
trogen atmosphere. The solution was then filtered and concentrated by
evaporation. The product was filtered and recrystallized with hexane
and dried under vacuum.
13C{1H} NMR (400.13 MHz, (CD3)2(CO), room temp., ppm): 10.5
(C1), 20.5 (C2), 41.9 (C3), 124.3 (C12), 126.5 (C14), 128.0 (C11 C15),
129.7 (C8), 130.7 (C13), 137.6 (C7), 142.3 (C6), 148.3 (C9).
4.3.3.1. WBr2(CO)3C5H4NCPh = N(CH2)2CH3] (WBrL3). Yield (η):
77% (0.50 g) (black crystalline powder)
4.4. Synthesis of materials MCM-WBrL1, MCM-WIL1, MCM- WBrL2,
MCM-WIL2, MCM-MoBrL2, and MCM-MoIL2
Elemental analysis WBr2C18H16N2O3 (652.00) (%): Calc: C 33.16, N
4.30, H 2.47; found: C 33.07, N 4.19, H 2.34.
IR (KBr pellets,
ν
cmꢀ 1): 3092 (w), 3056 (s), 2966 (s), 2872 (m), 2083
A solution of each complex (0.65 mmol) in dry toluene (10 mL) was
added to a suspension of 1 g of the material MCM-L1Si or MCM-L2Si in
dry toluene (20 mL). The reaction mixture was stirred under a N2 at-
mosphere at room temperature. The resulting material was then filtered
off, washed with dichloromethane, and dried under vacuum for 12 h.
(w), 2036 (vs), 1956 (vs), 1910 (s), 1671 (s), 1607 (s), 1525 (m), 1449
(s), 1347 (m), 1286 (s), 1226 (w), 1163 (w), 1160 (w), 1054 (m), 949 (s),
812 (m), 770 (s), 698 (s), 646 (m).
1H NMR (400.13 MHz, (CD3)2(CO), room temp., ppm): 1.02 (t, 3H,
H1), 1.88 (t, 2H, H2), 3.03 (m, 2H, H3), 7.63 (dd, 2H, H12 H14), 7.79 (t,
1H, H13), 8.06 (dd, 2H, H11 H15), 8.33 (t, 1H, H8), 8.44 (d, 1H, H6), 8.77
(t, 1H, H7), 9.23 (d, 1H, H9).
4.4.1. MCM-[WBr2(CO)3C5H4NCH = N(CH2)2CH3] (MCM-WBrL1,
pink powder)
13C{1H} NMR (400.13 MHz, (CD3)2(CO), room temp., ppm): 10.5
(C1), 20.5 (C2), 40.8 (C3), 127.0 (C6), 128.8 (C12 C14), 129.2 (C8),
130.7 (C11 C15), 134.2 (C13), 143.9 (C7), 145.2 (C9), 148.6 (C4), 189.0
(CO).
Elemental analysis (%) found C 10.57, N 2.39, H 1.95, W 4.44.
IR (KBr
ν
cmꢀ 1) 2981 (m), 2952 (w), 2893 (w), 2044, 1931, 1853 (m;
–
ν
O), 1624 (s; νC=N), 1475 (w), 1460 (w), 1390 (m), 1244 (vs), 1087
–
C
–
(vs), 968 (vs), 795 (s).
13C CP MAS NMR (δ ppm) 9.22 (SiCH2), 17.30 (SiOCH2CH3), 21.06
(CH2CH2CH2), 42.76 (CH2N), 58.21 (SiOCH2CH3).
29Si MAS NMR (δ ppm) ꢀ 49.49 (T1), ꢀ 57.76 (T2), ꢀ 67.78 (T3),
ꢀ 104.39 (Q3), ꢀ 109.93 (Q4).
4.3.3.2. [WI2(CO)3C5H4NCPh = N(CH2)2CH3] (WIL3). Yield (η): 72%
(0.54 g) (dark-blue powder)
Elemental analysis WI2C18H16N2O3 (746.00) (%): Calc: C 28.98, N
3.76, H 2.16; found: C 28. 71, N 3.57, H 1.92.
IR (KBr pellets,
ν
cmꢀ 1): 3041 (w), 3028 (w) 2964 (w), 2928 (m),
4.4.2. MCM-[WI2(CO)3C5H4NCH = N(CH2)2CH3] (MCM-WIL1, pink
2852 (w), 2002 (m), 1949 (s), 1919 (s), 1663 (m), 1567 (s), 1444 (m),
1385 (w), 1281 (w), 1187 (w), 1054 (w), 1028 (m), 942 (m), 864 (m),
768 (s), 699 (s).
powder)
Elemental analysis (%) found C 9.86, N 2.16, H 1.78, W 3.3.
–
–
IR (KBr
ν
cmꢀ 1) 2976 (w), 2920 (w), 2033, 1944, (m;
ν
O), 1622 (s;
1H NMR (400.13 MHz, (CD3)2(CO), room temp., ppm): 1.05 (t, 3H,
H1), 1.91 (t, 2H, H2), 3.22 (m, 2H, H3), 7.27 (dd, 2H, H12), 7.38 (m, 1H,
–
C
νC=N), 1448 (w), 1392 (w), 1227 (vs), 1066 (vs), 949 (vs), 789 (s).
13C CP MAS NMR (δ ppm) 9.06 (SiCH2), 16.79 (SiOCH2CH3), 20.91
(CH2CH2CH2), 42.88 (CH2N), 58.12 (SiOCH2CH3).
H
14), 7.55 (m, 2H, H11 H15), 7.67 (m, 1H, H8), 7.76 (H13), 8.06 (t, 1H,
H7), 8.09 (d, 1H, H6), 8.74 (d, 1H, H9).
29Si MAS NMR (δ ppm) ꢀ 48.72 (T1), ꢀ 58.79 (T2), ꢀ 67.45 (T3),
ꢀ 101.61 (Q3), ꢀ 110.66 (Q4).
13C{1H} NMR (400.13 MHz, (CD3)2(CO), room temp., ppm): 10.3
12

M. Vasconcellos-Dias et al.
Inorganica Chimica Acta 519 (2021) 120263
4.4.3. MCM-[WBr2(CO)3C5H4NCCH3 = N(CH2)2CH3] (MCM-
WBrL2, orange powder)
4.6. Computational studies
Elemental analysis (%) found C 7.29, N 2.25, H 1.82, W 3.4.
Density Functional Theory calculations [38] were carried out using
the Amsterdam Density Functional program (ADF) [39–41]. Geometries
were optimized with gradient correction, without symmetry constraints,
using the Local Density Approximation of the correlation energy (Vosko-
Wilk-Nusair) [51] and the Generalized Gradient Approximation
(Becke’s [52]exchange and Perdew’s [53,54] correlation functionals).
The COSMO approach implemented in ADF was considered to take into
account solvent (dichloromethane) in all geometry optimizations and
single point calculations. Relativistic effects were treated with the ZORA
approximation [55]. Triple ζ Slater-type orbitals (STO) with a small
frozen core were used to describe all the valence electrons of H, O, C, N,
Br, I, Mo and W. A set of two polarization functions was added to all
atoms. Frequency calculations were carried out to check that in-
termediates were minima in the potential energy surface. Three-
dimensional representations of the structures and molecular orbitals
were obtained with Chemcraft [56].
IR (KBr
1955, 1865 (s;
ν
cmꢀ 1) 3053 (w), 2968 (m), 2945 (w), 2879 (w), 2015,
–
ν
O), 1632, 1591 (s; νC=N), 1539 (w), 1491 (w), 1460
–
C
–
(m), 1392 (m), 1230 (vs), 1147 (vs), 1061 (vs), 893 (s), 820 (vs), 582 (s).
13C CP MAS NMR (δ ppm) 9.8 (SiCH2), 17.7 (SiOCH2CH3), 21.2
(CH2CH2CH2), 42.6 (CH2N), 58.3 (SiOCH2CH3), 59.9 (SiOCH2CH3),
67.9 (CH3-ring), 125.5, 128.9, 137.6, 144.4 (Ph ring), 198.7.
29Si MAS NMR (δ ppm) ꢀ 59.0 (T2), ꢀ 67.6 (T3), ꢀ 102.4 (Q3), ꢀ 109.8
(Q4).
4.4.4. MCM-[WI2(CO)3C5H4NCCH3 = N(CH2)2CH3] (MCM-WIL2)
(orange powder)
Elemental analysis (%) found C 6.91, N 1.09, H 1.55, W 4.47.
IR (KBr
1952, 1876 (s;
ν
cmꢀ 1) 3055 (w), 2987 (m), 2935 (w), 2879 (w), 2023,
–
ν
O), 1707 (m), 1648, 1616 (vs; νC=N), 1529 (w), 1473
–
C
–
(w), 1458 (w), 1387 (m), 1232 (vs), 1113 (vs), 1063 (vs), 906 (s), 814
(vs), 590 (s).
The geometries were modelled from the structures described above
and from that of [Mo(η3–C3H5)Br(CO)2(L1)]. In the search for the in-
termediates, several isomers of A and B were tested. For the B type
species, B and B’ described above have close energies and can be
interconverted. Although B was the more stable, it could not be found
for the iodide complexes.
13CP MAS NMR (δ ppm) 9.7 (SiCH2), 17.6 (SiOCH2CH3), 21.0
(CH2CH2CH2), 42.6 (CH2N), 58.2 (SiOCH2CH3), 60.0 (SiOCH2CH3),
125.3, 129.0, 137.6, 148.7 (Ph ring), 197.8.
29Si MAS (δ ppm) ꢀ 58.4 (T2), ꢀ 66.8 (T3), ꢀ 102.2 (Q3), ꢀ 109.4 (Q4).
4.4.5. MCM-[MoBr2(CO)3C5H4NCCH3 = N(CH2)2CH3] (MCM-
MoBrL2, pink powder)
Declaration of Competing Interest
Elemental analysis (%) found C 6.83, N 2.16, H 1.91, Mo 5.11.
IR (KBr
1934, 1853 (s;
ν
cmꢀ 1) 3033 (w), 2976 (m), 2922 (w), 2883 (w), 2003,
–
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
ν
O), 1707 (m), 1677, 1610 (s; νC=N), 1493 (w), 1448
–
C
–
(m), 1396 (m), 1240 (vs), 1117 (vs), 796 (vs), 721 (vs), 582 (s).
13C CP MAS NMR (δ ppm) 9.0 (SiCH2), 18.0 (SiOCH2CH3), 21.5
(CH2CH2CH2), 25.8 (CH2CH2CH2), 42.9 (CH2N), 60.0 (SiOCH2CH3),
69.2 (CH3-ring), 125.8, 129.4 (Ph ring).
Acknowledgements
29Si MAS NMR (δ ppm) ꢀ 58.3 (T2), ꢀ 66.7 (T3), ꢀ 102.1 (Q3), ꢀ 109.6
(Q4).
˜
ˆ
We thank the Fundaçao para a Ciencia e a Tecnologia (FCT),
Portugal, for financial support to Centro de Química e Bioquímica (UID/
MULTI/00612/2019), Centro de Química Estrutural (UIDB/00100/
2020 and UIDP/00100/2020), and BioISI – Biosystems & Integrative
Sciences Institute UIDB/04046/2020 and UIDP/04046/2020). M.
Vasconcellos-Dias thanks FCT for a PhD grant (SFRH/BD/37690/2007).
The crystallographic studies were supported by CICECO-Aveiro Institute
of Materials (UIDB/50011/2020 & UIDP/50011/2020), financed by
National Funds through the FCT/MEC, and co-financed by QREN-FEDER
through COMPETE under the PT2020 Partnership Agreement. We
acknowledge M. S. Saraiva and P. Ferreira for help with some catalytic
reactions and solid-state NMR, respectively.
4.4.6. MCM-[MoI2(CO)3C5H4NCCH3 = N(CH2)2CH3] (MCM-MoIL2,
pink powder).
Elemental analysis (%) found C 7.19, N 2.31, H 1.84, Mo 5.97.
IR (KBr
1946, 1871 (s;
ν
cmꢀ 1) 3049 (w), 2979 (m), 2937 (w), 2893 (w), 2023,
–
ν
O), 1671, 1606 (s; νC=N), 1469 (w), 1444 (m), 1387
–
C
–
(m), 1246 (vs), 1142 (vs), 1051 (vs), 914 (s), 810 (s), 579 (s).
13C CP MAS NMR (δ ppm) 10.0 (SiCH2), 17.4 (SiOCH2CH3), 21.1
(CH2CH2CH2), 25.5 (CH2CH2CH2), 42.9 (CH2N), 59.9 (SiOCH2CH3),
68.0 (CH3-ring), 125.6, 128.6, 129.2 (Ph ring).
29Si MAS NMR (δ ppm) ꢀ 57.9 (T2), ꢀ 66.7 (T3), ꢀ 102.4 (Q3), ꢀ 109.8
(Q4).
Appendix A. Supplementary data
4.5. Crystallography
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.ica.2021.120263.
X-ray diffraction data were collected at low temperature with mon-
ochromated Mo-Kα radiation (MoKα (λ = 0.71073 Å) on a Bruker Kappa
References
APEX II diffractometer equipped with a CCD area detector. Absorption
corrections were applied using SADABS [46]. The structures were solved
by intrinsic phasing with SHELXT [47] and refined by least-squares
methods with SHELXL [48], using the Olex2 software [49]. All non-
hydrogen atoms were refined with anisotropic thermal parameters.
The hydrogen atoms were positioned in geometrically idealized posi-
tions and refined using the riding model. All molecular diagrams were
made with Mercury [50].
[1] S.S. Balula, S.M. Bruno, A.C. Gomes, A.A. Valente, M. Pillinger, I.S. Gonçalves, D.J.
Macquarrie, J.H. Clark, Inorg. Chim. Acta 387 (2012) 234–239.
[2] A. Mentes, M.E. Hanhan, Transition Met. Chem. 33 (2008) 91–97.
[3] S.R. Gilani, Z. Mahmood, J. Chem. Soc. Pak. 25 (2003) 41–43.
[4] A.C. Gomes, S.M. Bruno, S. Gago, R.P. Lopes, D.A. Machado, A.P. Carminatti, A.
A. Valente, M. Pillinger, I.S. Gonçalves, J. Organomet. Chem. 696 (2011)
3543–3550.
[5] A. Mentes, R.D.W. Kemmitt, J. Fawcett, D.R. Russell, J. Organomet. Chem. 660
(2002) 91–97.
The Crystal data and selected refinement details are listed in Table
S4. The structures were deposited with the Cambridge Crystallographic
Data Centre with the numbers 1,975,301 (WBrL1) and 1,975,301
(WIL1).
[6] A. Mentes, Transition Met. Chem. 24 (1999) 77–80.
[7] G.-E. Matsubayashi, T. Ohara, T. Tanaka, Inorg. Chim. Acta 161 (1989) 67–72.
[8] D.M. Haddleton, D.J. Duncalf, D. Kukulj, M.C. Crossman, S.G. Jackson, S.A.F. Bon,
A.J. Clark, A.J. Shooter, Eur. J. Inorg. Chem. (1998) 1799–1806.
[9] S. Shabbir, S. Lee, M. Lim, H. Lee, H. Ko, Y. Lee, H. Rhee, J. Organomet. Chem. 846
(2017) 296–304.
[10] H. Kotze, S. Mapolie, Appl. Organomet. Chem. 31 (2016) 1–13.
13

M. Vasconcellos-Dias et al.
Inorganica Chimica Acta 519 (2021) 120263
[11] F. Hartl, M.J. Bakker, V.F. Santos, P.J. Costa, M.J. Calhorda, Inorg. Chem. 57
[32] F.E. Kühn, M. Groarke, E. Bencze, E. Herdtweck, A. Prazeres, A.M. Santos, M.
˜
(2018) 11704–11716.
J. Calhorda, C.C. Romao, I.S. Gonçalves, A.D. Lopes, M. Pillinger, Chem. Eur. J. 8
´
˜
[12] M. Vasconcellos-Dias, J. Marreiros, R. Sales, V. Felix, P. Brandao, C.D. Nunes, M.
(2002) 2370–2383.
ˆ
˜
J. Calhorda, Molecules 24 (2019) 578.
[33] L.F. Veiros, A. Prazeres, P.J. Costa, C.C. Romao, F.E. Kühn, M.J. Calhorda, Dalton
˜
´
[13] M. Vasconcellos-Dias, C.D. Nunes, P.D. Vaz, P. Ferreira, P. Brandao, V. Felix, M.
Trans. (2006) 1383–1389.
´
J. Calhorda, J. Catal. 256 (2008) 301–311.
[34] A. Comas-Vives, A. Lledos, R. Poli, Chem. Eur. J. 16 (2010) 2147–2158.
˜
[14] M. Abrantes, A.M. Santos, J. Mink, F.E. Kühn, C.C. Romao, Organometallics 22
[35] M.J. Calhorda, P.J. Costa, Curr. Org. Chem. 16 (2012) 65–72.
[36] C.D. Nunes, M.J. Calhorda, Inorg. Chim. Acta 431 (2015) 122–131.
[37] A.C. Gomes, J.A. Fernandes, C.A. Gamelas, I.S. Gonçalves, F.A.A. Paz, Acta
Crystallogr., Sect. E 67 (2011), m1738.
(2003) 2112–2118.
[15] 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. Felix, M.G.B. Drew, Organometallics 26
(2007) 5548–5556.
[38] R.G. Parr, W. Yang, Density-Functional Theory of Atoms and Molecules, Oxford
University Press, 1989.
[16] M. Vasconcellos-Dias, M.S. Saraiva, P. Ferreira, M.J. Calhorda, Organometallics 34
(2015) 1465–1478.
[39] G. te Velde, F.M. Bickelhaupt, E.J. Baerends, C. Fonseca Guerra, S.J.A. van
Gisbergen, J.G. Snijders, T. Ziegler, Chemistry with ADF. J. Comput. Chem. 22
(2001) 931–967.
[17] M. Vasconcellos-Dias, C.D. Nunes, P.D. Vaz, P. Ferreira, M.J. Calhorda, Eur. J.
Inorg. Chem. (2007) 2917–2925.
˜
[18] C.D. Nunes, A.A. Valente, M. Pillinger, A.C. Fernandes, C.C. Romao, J. Rocha, I.
[40] C. Fonseca Guerra, J.G. Snijders, G. te Velde, E.J. Baerends, Towards an Order-N
DFT Method, Theor. Chem. Acc. 99 (1998) 391–403.
S. Gonçalves, J. Mat. Chem. 12 (2002) 1735–1742.
[19] B. Marler, U. Oberhagemann, S. Voltmann, H. Gies, Micropor. Mesopor. Mat. 6
(1996) 375–383.
[41] Theoretical Chemistry, Vrije Universiteit, Amsterdam, T. N. ADF2013 SCM http://
www.scm.com (accessed September, 2020).
[20] W. Hammond, E. Prouzct, S.D. Mahanti, T.J. Pinnavia, Micropor. Mesopor. Mat. 27
(1999) 19–25.
[42] P.K. Baker, S.G. Fraser, E.M. Keys, J. Organomet. Chem. 309 (1986) 319–321.
[43] R.G. Hayter, Synthesis 13 (1968) 4–6.
[21] M.D. Alba, A. Becerro, J. Klinowski, J. Chem. Soc., Faraday Trans. 92 (1996)
849–854.
[44] J.M. Maher, N.J. Cooper, J. Am. Chem. Soc. 192 (1980) 7606–7607.
[45] M.S. Saraiva, N.L. Dias Filho, C.D. Nunes, P.D. Vaz, T.G. Nunes, M.J. Calhorda,
Micropor. Mesopor. Mat. 117 (2009) 670–677.
[22] A.A. Romero, M.D. Alba, W. Zhou, J. Klinowski, J. Phys. Chem. B 101 (1997)
5294–5300.
[46] L. Krause, R. Herbst-Irmer, G.M. Sheldrick, D. Stalke, J. Appl. Crystallogr. 48
(2015) 3–10.
[23] M. Kruk, M. Jaroniec, Langmuir 15 (1999) 5410–5413.
[24] A. Jimtaisong, R.L. Luck, Inorg. Chem. 45 (2006) 10391–10402.
[25] M.S. Saraiva, C.I. Fernandes, T.G. Nunes, M.J. Calhorda, C.D. Nunes, Appl. Catal. A
504 (2015) 328–337.
[47] G.M. Sheldrick, Acta Cryst., A 71 (2015) 3–8.
[48] G.M. Sheldrick, Acta Cryst., C 71 (2015) 3–8.
[49] L.J. Bourhis, O.V. Dolomanov, R.J. Gildea, J.A.K. Howard, H. Puschmann, Acta
Cryst., A 71 (2015) 59–75.
[26] M.S. Saraiva, C.I. Fernandes, T.G. Nunes, C.D. Nunes, M.J. Calhorda, Top. Catal. 59
(2016) 1237–1248.
[50] C.F. Macrae, I.J. Bruno, J.A. Chisholm, P.R. Edgington, P. McCabe, E. Pidcock,
L. Rodriguez-Monge, R. Taylor, J. van de Streek, P.A. Wood, J. Appl. Crystallogr.
41 (2008) 466–470.
[27] M.S. Saraiva, C.I. Fernandes, T.G. Nunes, C.D. Nunes, M.J. Calhorda, J. Organomet.
Chem. 751 (2014) 443–452.
˜
[28] A.A. Valente, J.D. Seixas, I.S. Gonçalves, M. Abrantes, M. Pillinger, C.C. Romao,
[51] S.H. Vosko, L. Wilk, M. Nusair, Can. J. Phys. 58 (1980) 1200–1211.
[52] A.D. Becke, J. Chem. Phys. 109 (1998) 2092–2098.
[53] J.P. Perdew, Phys. Rev. B 33 (1986) 8822–8824.
Catal. Lett. 101 (2005) 127–130.
[29] M. Pratt, J.B. Harper, S.B. Colbran, Dalton Trans. (2007) 2746-.2748.
[30] P.J. Costa, M.J. Calhorda, F.E. Kühn, Organometallics 29 (2010) 303–311.
[54] J.P. Perdew, Phys. Rev. B 34 (1986) 7406–7406.
´
[31] A.M. Al-Ajlouni, D. Veljanovski, A. Capape, J. Zhao, E. Herdtweck, M.J. Calhorda,
[55] E. van Lenthe, A. Ehlers, E.-J. Baerends, J. Chem. Phys. 110 (1999) 8943–8953.
[56] Chemcraft Program http://www.chemcraftprog.com/index.html (accessed
September, 2020).
F.E. Kühn, Organometallics 28 (2009) 639–645.
14