A Recoverable Ruthenium Aqua Complex Supported on Silica Particles: An Efficient Epoxidation Catalyst

Article abstract of DOI:10.1002/chem.201604463

The preparation and characterization of complexes with a phosphonated terpyridine (trpy) ligand (trpy-P-Et) and a bidentate pyridylpyrazole (pypz-Me) ligand, with formula [Ru^II(trpy-P-Et)(pypz-Me)X]ⁿ⁺ (2: X=Cl, n=1; 3: X=H₂O, n=2), is described, together with the anchoring of 3 on two types of supports: mesoporous silica particles (SP) and silica-coated magnetic particles (MSP). Aqua complex 3 is easily obtained by heating 2 in refluxing water and exhibits a two-electron Ru^IV/II redox process. It was anchored on SP and MSP supports by two different synthetic strategies, yielding the heterogeneous systems SP@3 and MSP@3, which were fully characterized by IR and UV/Vis spectroscopy, SEM, cyclic voltammetry, and differential pulse voltammetry. Catalytic olefin epoxidation was tested with molecular complex 3 and its SP@3 and MSP@3 heterogeneous counterparts, including reuse of the heterogeneous systems. The MSP@3 material can be easily recovered by a magnet, which facilitates its reusability.

Full text of DOI:10.1002/chem.201604463

A Journal of
Accepted Article
Title: A recoverable ruthenium aqua complex supported onto silica
particles: an efficient epoxidation catalyst
Authors: M.Isabel Romero, Ingrid Ferrer, Xavier Fontrodona, Anna
Roig, and Montserrat Rodríguez
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201604463
Link to VoR: http://dx.doi.org/10.1002/chem.201604463
Supported by

10.1002/chem.201604463
Chemistry - A European Journal
A recoverable ruthenium aqua complex supported onto silica
particles: an efficient epoxidation catalyst
Íngrid Ferrer,[a] Xavier Fontrodona,[a] Anna Roig,[b] Montserrat Rodríguez*[a] and Isabel
Romero*[a]
[a] Departament de Química and Serveis Tècnics de Recerca, Universitat de Girona, Campus de Montilivi, E-
17003 Girona, Spain.
[b] Institut de Ciència de Materials de Barcelona, ICMAB (CSIC), Campus UAB, 08193 Bellaterra, Spain
1
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
Abstract
The preparation and characterization of new complexes with a phosphonated trpy ligand (trpy-
P-Et) and a bidentate pyridylpyrazole (pypz-Me) ligand, with formula [RuII(trpy-P-Et)(pypz-
Me)X]n+ (X = Cl, n= 1, 2; X=H2O, n=2, 3) is described, together with the anchoring of 3 onto two
types of supports: mesoporous silica particles (SP) and silica coated magnetic particles (MSP).
The aqua complex 3 is easily obtained through reflux of 2 in water and displays a bielectronic
Ru(IV/II) redox process. It has been anchored onto SP and MSP supports through two different
synthetic strategies, yielding the heterogeneous systems SP@3 and MSP@3 that have been
fully characterized by IR, UV-vis, SEM, CV and DPV. Catalytic olefin epoxidation has been
tested with the molecular complex 3 and the SP@3 and MSP@3 heterogeneous counterparts,
including the reuse of the heterogeneous systems. The MSP@3 material can be easily
recovered by a magnet facilitating their reusability.
KEYWORDS: ruthenium / aqua complexes / surface functionalization / epoxidation / heterogeneous
catalysis.
2
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
Introduction
Olefin epoxidation has received considerable interest from both academics and industry since
epoxides play an important role as intermediates and building blocks in organic chemistry and
materials science.1 During the last decades, a series of homogeneous catalysts were identified
to be active in this reaction and, among them, we have developed efficient polypyridyl Ru-OH2
catalysts for epoxidation reactions.2 These Ru-aqua complexes can easily lose protons and
electrons and reach reactive, higher oxidation state RuIV=O species, and as a consequence
they have been widely used as redox catalysts for the oxidation of both organic and inorganic
species. In addition, the reaction mechanisms for the oxidation of several substrates by RuIV=O
have been thoroughly described together with the establishment and optimization of catalytic
processes.3 The redox potentials, and therefore the performance of the polypyridyl Ru-OH2
catalysts can be tuned through the electronic and steric effects of the auxiliary polypyridyl
ligands coordinated to the Ru metal center. The main advantages of homogenous catalysis are
selectivity and easy tuning of properties of catalysts through the modification of ligands;
however, from the perspective of sustainability and industrial applicability, homogeneous
catalysts are often expensive, create additional waste and their regeneration is difficult. In this
context, covalent grafting of homogeneous catalysts onto an insoluble porous solid support with
high surface, as mesoporous silica, 4 is a good approach where the catalyst can be recovered
by simple filtration.5 Also, the combination of different materials to create multifunctional
systems is a good alternative to obtain recyclable and efficient catalysts. This is the case of
nano o micro core-shell magnetic particles where the active sites are mainly distributed on the
“outer” surface of the support. These particles have been recently subject of extensive research,
since such materials combine the unique magnetic properties of the core together with the
possibility to further functionalize the surface; in addition the application of an external magnetic
field enables the removal of the particles in a simple way.6
Few examples of well-defined molecular epoxidation catalysts supported onto the previously
mentioned supports have been described in the literature.7 In this context, the design of new
catalysts anchored to different supports for alkene oxidation is an important challenge in the
development of sustainable methodologies.
In the present paper we describe the synthesis and characterization of new ruthenium
compounds with general formula [RuII(trpy-P-Et)(pypz-Me)X]n+,where trpy-P-Et and pypz-Me are
N-donor ligands (see Scheme 1), and X = Cl, H2O. The ruthenium aqua complex has been
anchored onto two different supports via covalent bonds and the anchored Ru-aqua compound
does not modify its coordination and electronic properties upon heterogenization. We also report
the good performance of the Ru complex as catalyst for olefin epoxidation reaction in
homogeneous and heterogeneous phase, along with the reutilization of the corresponding
supported catalysts and their recovery by a magnet when magnetic particles are used as
anchoring support. The stereospecific epoxidation of cis-epoxide observed in homogeneous
phase is consistent with the occurrence of a bielectronic transfer process in the catalyst.
3
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
Scheme 1. Ligands structures and synthetic strategy for the preparation of complexes 2 and 3.
trpy-P-Et, L1
pypz-Me, L3
Results and discussion
Synthesis, Structure and Redox Characterization
Molecular Complexes
The synthetic strategy followed for the preparation of RuII complexes 2 and 3, containing the
trpy-P-Et L1, and pypz-Me L3, ligands, is outlined in Scheme 1. The trichlorido Ru complex
[RuIIICl3(trpy-P-Et)], 1 (previously obtained by reaction of equimolar amounts of RuCl3·2.38H2O
and the trpy-P-Et ligand, L1, in dry methanol at reflux for 3 h7a), was treated with the didentate
ligand pypz-Me, L3, in presence of NEt3, producing the substitution of two chlorido ligands on 1.
After addition of a saturated NH4PF6 aqueous solution the corresponding chlorido complex 2
was obtained as a 1:1 mixture of trans and cis isomers, due to the no symmetric nature of the
pypz-Me ligand (the nomenclature cis or trans refers to the relative position of the monodentate
ligand, Cl or H2O, with regard to the pyrazole ring of the pypz-Me ligand, see Figure 1). Further
addition of saturated aqueous solution of NH4PF6 to the mother liquor lead to the precipitation of
a little fraction of the almost isomerically pure cis-[RuIICl(trpy-P-Et)(pypz-Me)](PF6) complex. All
attempts to separate the isomers were unsuccessful. The aqua complex [RuII(trpy-P-Et)(pypz-
Me)(OH2)](PF6)2, 3, is easily obtained by dissolving complex 2 in water under reflux and the
complex is also isolated as a mixture of cis-3 and trans-3 isomers. It is worth mentioning here
4
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
that this behaviour is different to that shown by the analogous complex [Ru(trpy)(pypz-
Me)(OH2)]2+ 8 where addition of AgNO3 to the chlorido complex precursor is necessary to
produce the exchange of Cl by aqua ligand. Probably, the presence of the phosphonate group
on the trpy-P-Et ligand in complex 2 produces different electronic effects that make the Cl ligand
more labile, as can be found in the literature for other substituted trpy ligands.9
Crystal structures of trans- and cis-2 have been solved jointly by X-ray diffraction analysis from
a single crystal where the two isomers crystallize together in an equimolar ratio. Figure 1
displays their molecular structure whereas the main crystallographic data and selected bond
distances and angles can be found in the Supplementary Information section (Tables S1 and
S2).
Figure 1. Ortep plots and labelling schemes for a) trans-2 and b) cis-2.
a)
Cl1
C17
O2
C13
N3
C16
C12
O1
C14
C9
C6
C10
N2
C20
C21
C11
C15
N4
C8
C3
P1
C22
C7
C4
O3
C5
C1
N1
C18
N6
C24
C23
C25
C2
C28
N5
C19
C26
C27
b)
O1
C16
P1
C17
O2
Cl1
C13
C9
C14
C10
C12
N3
C8
C11
C6
C5
C15
N4
C28
N5
O3
C7
C4
N2
C18
N1
C1
N6’
C25
C27
C19
C20
C3
C2
C24
C23
C21
C26
C22
In both cases, the Ru metal center adopts an octahedral distorted type of coordination where
the trpy-P-Et ligand is bonded in a meridional manner and the pypz-Me ligand acts in a
didentate fashion. The sixth coordination site is occupied by the chlorido ligand. All bond
distances and angles are within the expected values for this type of complexes,7a,8,10 although
5
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
the diffraction data does not allow a precise discussion of structural parameters, in particular
those corresponding to the pypz-Me ligand; the data only allows its refinement as a single
molecule with disordered positions at the area of atoms C20-C28. The N(1)-Ru-N(2) and N(2)-
Ru-N(3) angles are 80º and 81.1º respectively, showing the geometrical restrictions imposed by
the tridentate trpy-P-Et ligand, which is considered to define the equatorial plane of the
structure; as a consequence of this, the other two equatorial angles, N(1)-Ru-N(4) and N(3)-Ru-
N(4), are larger than the 90º expected for an ideal octahedral geometry.
The IR spectra for both complexes 2 and 3 show vibrations around 1200, 1000 and 800 cm-1
that can be assigned to P=O, P-O and P-C stretching modes, respectively. In the case of 3, it
can be observed a band around 3000 cm-1 that can be assigned to O-H stretching of the aqua
ligand. The one-dimensional (1D) and two-dimensional (2D) NMR spectra of both complexes
were registered in d6-acetone and are presented in Supplementary Information (Figures S1-S3).
The resonances found are consistent with the structures obtained in the solid state for 2. 1H-
NMR shows two sets of signals, one in the aromatic region corresponding to the nitrogen
ligands and the other in the aliphatic region assigned to the ethyl groups bound to phosphonate
and the methyl group of de pypz-Me ligand. It is worth mentioning the difference observed in the
chemical shift of H20 for the two isomers of compound 2; in the case of trans isomer this signal
is found at 10.16 ppm whereas for the cis isomer it appears at 8.17 ppm. This difference is due
to the deshielding effect of the spatially close Cl ligand in the case of the trans isomer.
The UV-Vis spectra for both complexes 2 and 3 registered in CH2Cl2 are displayed in Figure S4.
The complexes exhibit ligand based -* bands below 350 nm and relatively intense bands
above 350 nm mainly due to d(Ru)-*(L) MLCT transitions.11 In the case of the Ru-Cl complex
the MLCT bands are shifted to longer wavelengths with regard to the corresponding Ru-OH2
complex because of the relative destabilization of the d (Ru) levels provoked by the anionic
chlorido ligand.12-14
The redox properties of complexes 2 and 3 have been determined by cyclic voltammetry (CV)
and differential pulse voltammetry (DPV) experiments (see Figures S5 and S6). Chlorido
complex 2 (Figure S5a) exhibits a quasireversible monoelectronic Ru(III/II) redox wave at E1/2
around 0.93 V versus SCE; the analogous complex [RuCl(trpy)(pypz-Me)]+ 8 reaches the Ru(III)
oxidation state at lower potential (0.80 V), evidencing the higher electron-withdrawing capacity
of the trpy-P-Et ligand with regard to trpy, that provokes a destabilization of the Ru(III) oxidation
state in complex 2. On the other hand, the CV of the aqua complex 3 (Figure S5b) exhibits a
quasireversible wave at E1/2 around 0.5 V vs SCE in aqueous phosphate buffer at pH= 7
corresponding, as we could confirm through Pourbaix diagrams (see below), to the bielectronic
Ru(IV/II) redox couple of the two cis and trans-3 isomers, which display closely similar potential
values. Differential pulse voltammetry (DPV, Figure S6) obtained at pH = 5.5 allows
differentiating these two redox processes at E1/2 = 0.51 and 0.61 V that, by comparison with the
analogous complex [Ru(trpy)(pypz-Me)(OH2)]2+,8 can be assigned respectively to the trans-3
and cis-3 isomers. Figure 2 shows the Pourbaix diagrams for both trans- and cis-3 complexes,
with a unique pH-dependent redox process throughout the whole pH range with a change in the
slope value (from approximately 58 to 30 mV/pH unit) at pH around 7.6 for trans-3 and 9.6 for
cis-3, a behavior that is generally indicative of the occurrence of a two-electron (IV/II) redox
process.2b,c The changes in the slope throughout the whole pH range correspond to the pKa
values of Ru(II) or Ru(III) species and are indicated by vertical lines in the Pourbaix diagrams.
Electrochemical and thermodynamic (pKa values) data corresponding to trans- and cis-3 are
gathered in Table 1 together with analogous information for similar complexes described in the
literature.
6
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
Table 1. pKa and electrochemical data (pH = 7, E1/2 in V vs SCE) for aqua complexes described in this
work and others for purposes of comparison.
E1/2
(III/II) (IV/III)
E1/2
E
E
Entry
Compound
pKa(II) pKa(III)
ref.
(IV/II)b
(IV/II)a
1
2
3
4
5
6
7
trans-3
cis-3
0.42c
0.52c
<0
<0
0.42
0.52
0.48
0.495
0.55
0.66
0.65
7.6
9.6
1.1
1.5
this work
this work
trans-[Ru(trpy)(pypz-Me)(OH2)]2+ 0.39
0.57
0.52
180
50
10.1
10.75
9.75
0.95
1.65
1.21
8
8
cis-[Ru(trpy)(pypz-Me)(OH2)]2+
trans-[Ru(trpy)(pyrpy-O)(OH2)]+
[Ru(trpy-P)(bpm)(OH2)]2+
[Ru(trpy)(bpy)(OH2)]2+
0.47
0.55c
0.66c
0.57
<0
8
<0
7a
15
0.74
0.17
9.7
1.7
a E = E1/2(IV/III) - E1/2(III/II) in mV. b Average value calculated according to: E(IV/II) = [E1/2(IV/III) + E1/2(III/II)]/2, in V. c
E1/2(IV/II) in V.
A first look at Table 1 shows that both isomers of complex 3 (entries 1 and 2) present Ru(III/II)
potential values slightly higher than the analogous cis- and trans-[Ru(trpy)(pypz-Me)(OH2)]2+
complexes (entries 3 and 4), which is in accordance with the electron-withdrawing character of
the phosphonate group of the trpy-P-Et ligand, as we have noted above. Another complex also
containing this modified trpy ligand (entry 6) displays the highest Ru(III/II) potential value for the
set of complexes showed in Table 1. Nevertheless, the E1/2 values are not only governed by the
electronic properties of the ligands but also by the geometry of the complex. Thus, when
comparing for instance entries 1 and 2 or 3 and 4, we can see that coordination of a relatively
good σ-donor (or less π-acceptor) ligand in a trans fashion with respect to the aqua ligand, as in
the case for the trans isomers, lowers E1/2(III/II) with respect to the analogous cis isomers hence
stabilizing the high oxidation state species.16 However, in complexes displaying two
monoelectronic processes (entries 3 and 4), the effect on the Ru(IV/III) wave seems to be
reversed and, consequently, the difference between Ru(IV/III) and Ru(III/II) E values (E, see
Table 1) is increased for the trans geometry. Thus, the occurrence of a bielectronic process
seems to arise from a specific balance between the geometry and the -donor and -acceptor
properties of the ancillary ligands.17
7
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
Figure 2. Pourbaix diagram of a) trans-3 and b) cis-3. The pH-potential regions of stability for the different
oxidation states and their dominant proton compositions are indicated.
a)
1
pKaRu(III)
0.9
RuIII-OH2
0.8
RuIV=O
0.7
0.6
0.5
pKaRu(II)
0.4
RuII-OH2
0.3
RuII-OH
0.2
0.1
0
0
2
4
6
8
10
12
pH
b)
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
pKaRu(III)
RuIII-OH2
RuIV=O
pKaRu(II)
RuII-OH2
RuII-OH
0
2
4
6
8
10
12
pH
8
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
Finally, regarding the pKa values for the Ru(II) species, the trans- and cis-3 isomers display
lower pKa values than the corresponding [Ru(trpy)(pypz-Me)(OH2)]2+ isomers (7.6 and 9.6 vs.
10.1 and 10.75 for trans and cis, see entries 1-4). This higher acidic character for 3 is again in
accordance with the enhanced electron-withdrawing nature of the trpy-P-Et ligand when
compared to trpy.
Heterogeneous Complexes
The two synthetic strategies followed for the immobilization of [RuII(trpy-P)(pypz-Me)(OH2)]2+, 3,
are shown in Scheme 2. Both strategies are based in the formation of hybrid materials in which
the organic and inorganic parts are linked by a chemical bond. A covalent binding is formed
between the phosphonate-modified terpyridine ligands (trpy-P-Et, L1) and (trpy-P-H, L2) and
silica particles, leading to the modification of the surface of silica supports. Two different SiO2
supports were used: commercial silica mesoporous particles (SP) and silica-coated -Fe2O3
magnetic particles (MSP), with mean diameters of aprox. 200 nm in both cases, and surface
area of 886 m2/g and 25 m2/g, respectively. Mass fraction of -Fe2O3 in MSP is 7.2% as
determined by ICP-MS; these values are summarized in Table S3.
Scheme 2. Synthetic strategies for the immobilization of Ru complex 3. L3 represents the bidentate pypz-
Me ligand and grey spheres represent the silica particles.
9
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
Following the strategy I, (Scheme 2, 1a/b) terpyridine phosphonate-coated silica particles were
synthesized by grafting (trpy-P-Et, L1) or (trpy-P-H, L2) ligands onto silica materials.
Diethylphosphonate L1 was mixed with SP or MSP in toluene at reflux overnight, in order to
achieve the hydrolysis of the ethyl groups, whereas L2 was mixed with SP or MSP in water
overnight at room temperature. Removal of the ethyl groups upon grafting in the case of ligand
L1 is confirmed through NMR spectroscopy (see Figure S8). In both cases the immobilized
terpyridine ligands were obtained after centrifugation. The corresponding ligand-coated silica
particles were then mixed with RuCl3 in methanol overnight to afford [RuCl3(trpy-P)]-coated,
deep brown silica particles that were washed, centrifuged and dried in a hot air oven. In parallel,
a sample of naked silica particles (SP) was treated with RuCl3 following the same procedure but
a whitish silica was recovered and thus a significant degree of direct deposition of
uncoordinated Ru at the surface can be ruled out. Finally, the bidentate pypz-Me ligand, L3,
was coordinated to [RuCl3(trpy-P)]-coated silica particles by reflux in methanol using NEt3 as
reducing agent, leading to the final Ru(II) aqua complex 3 attached to the corresponding
support. This heterogeneous system was centrifuged, washed and dried in a hot air oven. It is
important to note here that the final immobilized compound was the aqua complex 3 instead of
the expected chlorido complex 2. This is also the case for strategy II (see below) and is
10
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
probably due to the lability of chlorido ligand that, in presence of non-anhydrous solvents, leads
to the exchange of Cl by the aqua ligand.
Strategy II consists in the reaction of molecular chlorido complex 2 with SP particles as is shown
in Scheme 2(2). A mixture of trans- and cis-[RuCl(trpy-P-Et)(pypz-Me)](PF6), 2, was dissolved in
toluene under reflux (with addition of acetone after 4 hours of reflux to further solubilize complex
2, see experimental section), in presence of SP support. After 20 hours, the hydrolysis of the
ester groups was produced and these conditions also lead to the hydrolysis of the chlorido
ligand, generating the corresponding aqua complex 3 anchored to the support. An alternative
anchoring protocol using a mixture of toluene/acetone as refluxing solvent from the beginning of
the process was tested; however, under these conditions the exchange of chlorido ligand by
aqua ligand took place before the hydrolysis of the ester groups causing the anchorage of the
complex through the aqua ligand position preferentially, leading to a blockage of the labile
position needed for the catalytic activity. To the best of our knowledge, this is the first example
of an aqua ruthenium complex immobilized and fully characterized (see below) onto silica
supports including magnetic particles since in previous works an aqua complex is generated in
situ by dissolving in water the Ru-Cl precursor supported onto Fe3O4 nanoparticles.7a
The ruthenium loading in the corresponding supports were measured by inductively coupled
plasma-atomic emission spectrometry (ICP-AES) and the values are shown in Table S4. As it
can be observed strategy 1a and 1b (starting from the ethyl or the acid trpy-P ligand
respectively, see scheme 2) were tested using MSP support, being the strategy 1a the most
effective probably due to the harsher reaction conditions used, necessary to hydrolyze the ethyl
groups of the phosphonate ligand. Strategy 1b was used with the SP support yielding a high
amount of Ru anchored, and this support was also tested using strategy II (direct anchoring of
the molecular chlorido complex). A lower amount of Ru was anchored through this strategy
when compared to strategy 1b, probably due to the incomplete hydrolysis of the ester groups.
Scanning electron microscopy (SEM) was performed on the two supports after functionalization,
and the images obtained are displayed in Figure 3.
Figure 3. SEM images of a) SP and b) MSP after the anchoring of the Ru complex.
MSP@3
SP@3
1
1m
1 m
As can be observed, all supports remain practically intact even after harsh synthetic conditions
in some cases, maintaining their particle size and morphology after the grafting process. Figure
11
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
S7 shows the SEM images obtained for supports SP and MSP before anchoring the complex,
and full characterization of MSP particles can be found elsewhere18. Thermogravimetric profiles
have been registered for the naked SP support and for the two (SP and MSP) supports after
anchoring the Ru complex, and the results obtained are shown in Figure 4. In all cases, the
supports present an initial weight loss at temperatures below 150˚C, corresponding to the
evaporation of adsorbed solvent. At temperatures above 300˚C, the weight loss for the three
functionalized supports corresponds to the release of the Ru complex. For SP the weight loss
was 7.7%, whereas it was 3.9% for MSP (both prepared through strategy 1b).
Figure 4. Thermogravimetric profiles of SP (blue), SP@3 (green), MSP@3 (grey).
The spectroscopic (IR and UV-vis) and electrochemical characterization of the anchored
catalyst 3 has been performed on SP functionalized systems. Figure 5a displays the IR spectra
for SP and SP@3 and the inset shows the detail of the bands found around 2900-3000 cm-1 that
can be assigned to N-H and C-H stretching modes of the supported trpy-P ligand in SP@3. On
the other hand, difference IR spectra are provided in the supporting information (Figure S9)
together with the IR spectrum of the molecular complex 3. As can be observed in Figure S9a, a
shift of the vibrations around 800 cm-1 to higher wavenumber values take place upon grafting of
the complex onto SP particles, presumably indicating a modification on the phosphonate group.
The UV-Vis spectra for SP@3 obtained through strategies 1b and 2 have been registered on a
suspension of the heterogeneous support in dichloromethane and are displayed in Figure 5b
together with the UV-vis spectrum corresponding to the molecular aqua complex 3. The
anchored complexes exhibit ligand based -* bands below 350 nm and relatively intense
bands above 350 nm mainly due to d(Ru)-*(L) MLCT transitions11 that are similar to those
obtained for the homogeneous complex 3; therefore, we can assert that the anchored 3
complex is in both cases analogous to the molecular species, and its electronic properties have
not been substantially modified.
12
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
Figure 5. a) IR spectra of SP (orange) and SP@3 (blue); b) UV-vis spectra for SP@3 obtained through
strategy 1b (red), 2 (blue) and homogeneous complex 3 (dotted grey).
a)
100
90
100
99
80
98
70
97
96
60
95
3200 3100 3000 2900 2800 2700
50
wavenumber (cm‐1)
40
4000 3500 3000 2500 2000 1500 1000
500
wavenumber (cm‐1)
b)
1,8
1,6
1,4
1,2
1,0
0,8
0,6
0,4
0,2
0,0
250
300
350
400
450
500
550
600
λ(nm)
The redox properties of the functionalized particles have been investigated using cyclic
voltammetry (CV) and differential pulse voltammetry (DPV) experiments by settling a layer of
the heterogeneous functionalized particles on the working electrode surface. Figure 6 displays
13
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
the differential pulse voltammetry obtained for SP@3 together with that of the homogeneous
complex 3 for purposes of comparison, and the cyclic voltammetries of SP@3 registered in
MeOH and in aqueous phosphate buffer are shown in the supporting information (Figure S10).
The value of E1/2=0.56V obtained for SP@3 is similar to the average of the potential values
obtained for the two isomers cis and trans of complex 3 in solution. In the case of MSP@3 this
value was E1/2=0.60V. The DPV experiment performed on the 3-modified support does not allow
the distinction of the two waves for the cis- and trans- isomers of the complex though an
isomeric mixture is expected to be formed in the heterogeneous supports as suggested by the
UV-vis spectrum obtained.
Figure 6. Differential Pulse Voltammetry of SP@3 (red) and homogeneous complex 3 (dotted grey) in
aqueous phosphate buffer at pH=5.5.
4,0E‐06
3,0E‐06
I(A)
2,0E‐06
1,0E‐06
0,0E+00
0,2
0,4
0,6
E(V)
0,8
1
The redox properties of Ru-aqua complexes are pH dependent due to the simultaneous
exchange of protons and electrons during the redox processes, and this dependence can be
plotted as E vs. pH, or Pourbaix, diagrams. The Pourbaix diagram obtained for a SP@3 sample
prepared through strategy I (1b) is displayed in Figure 7 and it is similar to that obtained when
the functionalized support is prepared by following strategy II (Figure S11). In both cases, the
change of E1/2 value with pH displayed by the heterogeneous SP@3 systems confirm that the
species anchored is indeed the aqua complex since a parallel behaviour to the homogeneous
phase is observed (see Figure 2); a single wave was observed throughout the whole the pH
range, which can tentatively be assigned to a Ru(IV/II) 2-electron transition of the supported
complex. The values shown correspond to an average of the data for cis- and trans- isomers.
14
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
Figure 7. Pourbaix diagram of SP@3 obtained through strategy 1b. The pH-potential regions of
stability for the various oxidation states and their dominant proton compositions are indicated.
1
pKaRu(III)
0.9
RuIII-OH2
0.8
RuIV=O
0.7
0.6
0.5
pKaRu(II)
0.4
RuII-OH2
0.3
RuII-OH
0.2
0.1
0
0
2
4
6
8
10
12
pH
Catalytic epoxidation of alkenes
The catalytic activity of an isomeric trans- and cis- mixture of aqua complex 3 was checked in
the epoxidation of cis--methylstyrene in dichloromethane, using PhI(OAc)2 as oxidant. No
epoxidation occurred in the absence of catalyst or in the presence of naked silica supports. The
catalytic results are shown in Table 2 together with those corresponding to previously reported
complexes for comparison purposes.
Table 2. Catalytic epoxidation of cis--methylstyrene by Ru aqua complexes using PhI(OAc)2 as
oxidant.a
Conversion Selectivity
% of
cis-epoxide
ref.
Entry
Compound
(%)
(%)b
1
2
3
4
3
>99
70
>99
98
This work
trans-[Ru(trpy)(pypz-Me)(OH2)]2+
cis-[Ru(trpy)(pypz-Me)(OH2)]2+
[Ru(trpy-P)(bpm)(OH2)]2+
90
99
86
68
8
8
>99
>99
97
>99
7a
a Conditions: complex (1.25 mol), substrate (125 mol), PhI(OAc)2 (250 mol), CH2Cl2 (2.5 mL), 25ºC, 24 h,
biphenyl (125 mol) as internal standard.
b Selectivity for epoxide, (Yield/Conversion)x100.
15
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
An interesting reaction in epoxidation catalysis is the stereoselective epoxidation of cis-
alkenes,19-24 in which the well-known and efficient Mn-salen complexes 25 and others26 generally
give mixtures of cis and trans-epoxides. The cis→trans isomerization in epoxidation processes
involving cis alkenes is a common phenomenon due to the higher thermodynamic stability of the
trans epoxides. To undergo such isomerization, a rotation around the C-C bond is required
involving the presence of a relatively long-lived free substrate radical during the oxygen transfer
process.27
As we can observe in Table 2, the mixture of isomers 3 presents high levels of conversion and
selectivity for the epoxide (with benzaldehyde identified as main by-product), comparable to
other complexes described in the literature.7a,8 The corresponding cis-epoxide was detected as
the sole product, with no formation of the trans-epoxide. This result indicates that, if an alkene-
localized radical species was formed as intermediate, the closure of the ring would be much
faster than the C-C rotation at the radical intermediate. However, a potential O-atom transfer
concerted mechanism cannot be ruled out, and this would be in turn consistent with the low
stability of the Ru(III) oxidation state that favors 2e- versus 1e- transfer process.28
We have also evaluated the heterogeneous catalytic activity of the SP@3 anchored
aquacomplex in the epoxidation of a diversity of alkenes in dichloromethane and using
PhI(OAc)2 as oxidant (peracetic acid was also tested but the conversion values were low). All
substrates were tested under analogous conditions (catalyst:substrate:oxidant, 1:100:200). No
epoxidation occurred in the absence of catalyst.
A preliminary essay of the catalytic activity was performed using two different samples of the
catalyst SP@3, obtained through strategies 1b and 2 respectively, to check the influence of the
synthetic strategy on the catalytic performance. The substrate tested was styrene and in both
cases the conversion and selectivity values obtained were similar. This result seems to indicate
that the two heterogeneous catalytic materials are similar, in consistency with the UV-vis and
electrochemical characterization of the systems obtained through both strategies (see above).
Then, SP@3 prepared through 1b was used for the evaluation of the catalytic oxidation of
further substrates.
First of all, the reaction time for the epoxidation of different substrates was optimized and the
conversion values registered at different times are collected in Table S5. Almost total
conversion was attained in 6 h reaction for aromatic substrates (styrene and cis--
methylstyrene) whereas for the aliphatic 4-vinylcyclohexene substrate longer reaction times
were needed. Thus, 6 hours was chosen as the reaction time for all the essays. The conversion
and selectivity values obtained for the epoxidation of four different alkenes using SP@3 as
catalyst are gathered in table 3. In most cases, high conversions and moderate selectivity
values are obtained except for styrene which displays a high selectivity value. It is also
remarkable that the percentage of cis epoxide in the epoxidation of cis--methylstyrene (91%) is
still high but the process is not stereospecific in contrast with the quantitative generation of the
cis epoxide in the case of the homogeneous counterpart. This indicates that the silica support
plays a certain role in the catalytic pathway, perhaps inducing a more significant occurrence of a
radical pathway mechanism in parallel to the oxo-transfer mechanism. The support could also
lead to slight modifications of the electronic and/or structural parameters of the active species
that could slower the ring closure step in a putative Ru-coordinated radical intermediate, thus
allowing the cistrans C-C bond rotation to a higher extent.27
16
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
Table 3. Catalytic epoxidation of selected alkenes by SP@3. Conditions: SP@3 (1 mol Ru),
substrate (100 mol), PhI(OAc)2 (200 mol), CH2Cl2 (2.5 mL), 25ºC, 6h, biphenyl (100 mol) as
internal standard.
Substrate
Conversion (%)
Selectivity (%)a
92(58)b
79(40)b
94
58c
95d
43d
34(100/0)e
25
a Selectivity for epoxide: (Yield/Conversion)x100. b In parentheses, conversion and selectivity values obtained using
MSP@3.c 91% of cis-epoxide was obtained. d Analysis performed after 15h. e In parentheses, ratio [ring epoxide/vinyl
epoxide].
The performance of the heterogeneous catalyst was also tested with aliphatic substrates; with
4-vinylcyclohexene the process was regiospecific towards the ring epoxidation, and with 1-
octene modest performance was obtained in agreement with the lower reactivity towards non-
activated monosubstituted alkenes. The MSP@3 catalyst was also evaluated in the epoxidation
of styrene showing moderate conversion and selectivity values (Table 3). The performance of
our heterogeneous catalysts are comparable to that displayed by other supported Ru
complexes7a,29 though the SP@3 system reaches almost complete conversion for most
substrates in shorter reaction times. The lower performance in the case of aliphatic substrates
such as 1-octene is commonly described also for heterogeneous catalysts based on other
metals,7c although in the case of some heterogeneous molybdenum catalysts the conversion
and selectivity values are remarkable in relatively short reaction times.7b
Based in the above results, we have also investigated the recyclability of the SP@3 and
MSP@3 heterogeneous catalytic systems in CH2Cl2. In most cases, the catalyst maintains a
good performance through up to four runs. In the epoxidation of styrene by SP@3, a reaction
time of 6 h was initially set for each run as in the previous essays. However, the conversion
value decreased dramatically after the first run (Figure 8a). Then, a similar experiment (Figure
8b) was performed keeping each of the catalytic runs throughout 15 h and the conversion
values were improved, indicating that the low conversion values obtained in the first case (6h
per run) were in part a consequence of slow kinetics in heterogeneous phase. However, a
gradual decrease in the conversion values was observed throughout successive runs when kept
for either 6 or 15 hours and thus a progressive deactivation of the catalyst is presumably taking
place. We have also considered the possibility of catalyst leaching, which can similarly lead to a
decrease of heterogeneous catalytic activity throughout the successive runs and can also cause
difficulty to unravel whether the activity arises from the heterogeneous or homogeneous
(leached) catalyst. Thus, the amount of Ru complex was measured through ICP-AES in the
reaction solvent for a whole set of reuses in the case of styrene and the rest of substrates tested
(see below). In all cases, the amount of Ru present in the final solution was under the detection
limit of the technique (1.8-2% of the initial Ru anchored). We can then assert that leaching of Ru
17
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
is negligible for the heterogeneous systems described and then the decrease in activity is
seemingly due to catalyst deactivation. On the other hand, as can be observed in Figure 8b,
selectivity values above 75% were achieved for the styrene oxide product in all runs.
Figure 8. Conversion and selectivity values obtained throughout a number of consecutive reuses of
catalyst SP@3 in the epoxidation of styrene: a) 6 h per run; b) 15 h per run.
a)
b)
The recyclability of the catalytic SP@3 system has been also investigated on other substrates
as cis--methylstyrene, 4-vinylcyclohexene and 1-octene and the results are displayed in
Figures S12-S14. These results evidence that the catalyst maintains good conversion values for
cis--methylstyrene and 4-vinylcyclohexene substrates for 3-4 runs whereas the less activated
1-octene displays much lower activity. The overall turnover numbers are 172 for cis--
18
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
methylstyrene, 108 for 4-vinylcyclohexene and 12 for 1-octene. The selectivity for the epoxide is
also well maintained except for this latest substrate, with a dramatic drop of selectivity at the
second run and practically null conversion at the fourth run. In the case of cis--methylstyrene
the stereoselectivity for the cis-epoxide is maintained around 80-90% through all the runs For 4-
vinylcycloxexene the epoxidation takes place exclusively at the ring position in all runs.
The reusability of the MSP@3 system has been tested in the epoxidation of styrene and the
results obtained are displayed in Figure 9a, which shows that the overall values of conversion
and selectivity are moderate and are clearly lower than the ones obtained for this substrate
using catalyst SP@3 under the same reaction conditions. The origin of this difference could
arise from the large difference in the surface area values between the two supports (886 m2/g
for SP and only 25 m2/g for MSP, see Table S1), and the amount of ruthenium anchored onto
these supports (1·10-4 mmol/gsilica for SP@3 and 2,96·10-4 mmol/gsilica for MSP@3), that would
entail that molecules of anchored catalyst were placed much closer to each other at the MSP
support, due to the higher concentration of complex 3 onto the surface of the MSP when
compared to SP support (4·10-6 mmol/m2 for MSP vs 0,33·10-6 mmol/m2 for SP). This fact could
hinder the approach of the substrate to the active site or alternatively favor the catalyst self-
deactivation in the MSP@3. However, their magnetic core allows us to attract and remove them
from the reaction medium by a magnet, facilitating the reusability of the material as we can see
in Figure 9b.
19
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
Figure 9. a) Conversion and selectivity values obtained throughout a number of consecutive reuses of
catalyst MSP@3 in the epoxidation of styrene. b) Photographs of the CH2Cl2 dispersion of MSP@3
without and with a magnet.
a)
b)
Conclusions
We have synthesized and fully characterized two new complexes, [RuIICl(trpy-P-Et)(pypz-
Me)](PF6), 2, and [RuII(trpy-P-Et)(pypz-Me)(OH2)](PF6)2, 3, both containing the trpy-P-Et ligand
which is used as linker to anchor the molecular aqua complex 3 to silica supports. Unlike other
complexes found in the literature, the molecular aqua complex was easily obtained after reflux
of chlorido complex in water, evidencing the influence of electronic factors arising from the
phosphonate group of the electron-withdrawing trpy-P-Et ligand. The aqua complex 3 was
tested in the epoxidation of cis--methylstyrene showing high levels of conversion and
specificity for the cis-epoxide. This behavior is in accordance with the occurrence of a
bielectronic (IV/II) redox process displayed by the catalyst 3 that could allow the occurrence of a
concerted oxo-transfer mechanism. The aqua complex 3 was easily and successfully
immobilized onto two silica supports (including magnetic particles) using two different synthetic
strategies that allowed either i) the anchoring of the trpy-P-Et ligand onto the surface of the
supports and the subsequent preparation of compounds in situ or ii) the immobilization of the
previously synthesized molecular ruthenium complex onto the supports. The characterization of
these hybrid materials corroborated that the aqua ruthenium compound, anchored through the
different strategies, doesn’t modify significantly its coordination and electronic properties with
20
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
respect to the molecular counterpart. To the best our knowledge, our aqua complex is the first
one anchored onto silica supports that has been isolated and fully characterized. The
heterogeneous SP@3 and MSP@3 systems were effective in the epoxidation of a variety of
olefin substrates being selective in the epoxidation of cis--methylstyrene with high levels of
conversion and stereoselectivity for the cis epoxide. System MSP@3 showed moderate
conversion and selectivity values in the epoxidation of styrene with regard to the values
displayed by the analogous SP@3 catalyst, manifesting a different effect of the SP and MSP
supports on the catalytic performance, probably as consequence of the higher grafting density
of complex on the MSP supports, that could hinder the approach of the substrate to the active
site or alternatively favour the catalyst self-deactivation. However, they can be recovered
through a simple magnetic separation avoiding the centrifugation process.
The reutilization of SP@3 system in the epoxidation of alkenes has been evaluated
demonstrating the effective recyclability of the catalytic system up to 3 or 4 successive runs in
all cases, keeping high conversion and selectivity values in the epoxidation of styrene through
all the runs. Moderate values were obtained for the rest of the alkenes tested. The epoxidation
of cis--methylstyrene is not stereospecific in the heterogeneous catalytic processes but high
steroeselectivity values were obtained that were maintained through the different reuses.
Experimental Section
Materials
All reagents used in the present work and silica mesoporous particles (SP) were obtained from Sigma-
Aldrich and were used without further purification. Reagent grade organic solvents were obtained from
SDS and high purity de-ionized water was obtained by passing distilled water through a nano-pure Mili-Q
water purification system. Diethyl 2,2’:6’,2’’-terpyridine-4’-phosphonate (trpy-P-Et, L1) and diethyl
2,2’:6’,2’’-terpyridine-4’-phosphonic acid (trpy-P-H, L2) ligands were supplied by HetCat and were directly
used without further purification. RuCl3·2.38H2O was supplied by Johnson and Matthey Ltd. and was
used as received.
Instrumentation and Measurements
IR spectra were recorded on an ATR MK-II Golden Gate Single Reflection. UV-Vis spectroscopy was
performed on a Cary 50 Scan (Varian) UV-Vis spectrophotometer with 1 cm quartz cells. Cyclic
voltammetric (CV) and Differential Pulse Voltammetry (DPV) experiments were performed in an IJ-
Cambria IH-660 potentiostat using a three electrode cell. Glassy carbon electrode (3 mm diameter) from
BAS was used as working electrode, platinum wire as auxiliary and Ag/AgCl as the reference electrode.
All cyclic voltammograms presented in this work were recorded under nitrogen atmosphere. The
complexes were dissolved in solvents containing the necessary amount of n-Bu4NH+PF6 (TBAH) as
-
supporting electrolyte to yield a 0.1 M ionic strength solution. All E1/2 values reported in this work were
estimated from cyclic voltammetric experiments as the average of the oxidative and reductive peak
potentials (Epa+Epc)/2, or directly from DPV. Unless explicitly mentioned the concentration of the
complexes was approximately 1mM. The NMR spectroscopy was performed on a Bruker DPX 300 and
400 MHz. Samples were run in CD2Cl2 or d3-acetonitrile with internal references (residual protons and/or
tetramethylsilane). Elemental analyses were performed using a CHNS-O Elemental Analyser EA-1108
from Fisons. ESI-MS experiments were performed on a Navigator LC/MS chromatograph from Thermo
Quest Finnigan, using acetonitrile as a mobile phase. SEM images were recorded using a SEM QUANTA
FEI 200FEG-ESEM. Gas chromatography experiments were performed by capillary GC, using a GC-2010
Gas Chromatograph from Shimadzu, equipped with an Astec CHIRALDEX G-TA Column (30 m x 0.25
mm diameter) incorporating a FID detector. All the product analyses in the catalytic experiments were
21
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
performed by means of GC using biphenyl as internal standard. For metal content determination a
sequential inductively coupled plasma atomic emission spectrometer (ICP-AES, Liberty Series II, Varian,
Australia) in radial configuration and a quadrupole-based inductively coupled plasma mass spectrometer
system (ICP-MS, Agilent 7500c, Agilent Technologies, Tokyo, Japan) were used. This latest instrument is
equipped with an octapole collision reaction cell. However, in this work, the collision reaction cell acts only
as an ion focusing lens because it was not filled with any pressurized gas. Prior to measurements,
samples were digested with HCl:HNO3. Thermogravimetric analysis (TGA) was performed under N2
atmosphere with a 10ºC min-1 heating rate from 30ºC to 700ºC. The BET specific surface area (Sa) was
determined by N2 adsorption-desorption measurements at 77 K using an ASAP 2000 (Micrometrics
Instrument Corporation, USA) after degasification at 120ºC under vacuum for 24h.
Crystallographic Data Collection and Structure Determination
Measurement of the crystals were performed on a Bruker Smart Apex CCD diffractometer using
graphite-monochromated Mo Kα radiation (λ = 0.71073Å) from an X-Ray tube. Data collection, Smart V.
5.631 (BrukerAXS 1997-02); data reduction, Saint+ Version 6.36A (Bruker AXS 2001); absorption
correction, SADABS version 2.10 (Bruker AXS 2001) and structure solution and refinement, structure
solution and refinement, SHELXL-2013 (Sheldrick, 2013). The crystallographic data as well as details of
the structure solution and refinement procedures are reported in supplementary information. CCDC
1505432 contains the supplementary crystallographic data for this paper. These data can be obtained
free of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/products/csd/request/
Preparations
[RuCl3(trpy-P-Et)] 1 complex7a and 2-(1-methyl-3-pyrazolyl)pyridine30 (pypz-Me, L3) ligand were prepared
according to literature procedures. Maghemite-silica composite nanoparticles (MSP) with mean diameters
of aprox. 2 m and 200 nm, respectively, were synthesized in our laboratory by hydrolysis and
polycondensation of tetramethylorthosilicate (TMOS, 98%, Sigma-Aldrich) and high temperature
supercritical drying, according to the procedures described elsewhere (Sol-gel route to direct formation of
silica aerogel microparticles using supercritical solvents).31 Supercritical fluid assisted one-pot synthesis
of biocompatible core(-Fe2O3)@shell(SiO2) nanoparticles as high relaxivity T2-contrast agents for
Magnetic Resonance Imaging.18
Electrochemical experiments were performed under N2 atmosphere with degassed solvents. All
spectroscopic, electrochemical and synthetic experiments were performed in the absence of light unless
explicitly mentioned.
Synthesis of the complexes
Trans- and cis-[RuCl(trpy-P-Et)(pypz-Me)](PF6), trans- and cis-2. A sample of 1 (0.15 g, 0.26 mmol)
was added to a 100 mL round bottomed flask containing a solution of LiCl (0.022 g, 0.52 mmol) dissolved
in 40 mL of EtOH/H2O (3:1) under magnetic stirring. Then, NEt3 (0.06 mL, 0.52 mmol) was added and the
reaction mixture was stirred at room temperature for 30 min. Afterwards, pypz-Me, L3, (0.041 g, 0.26
mmol) was added and the mixture was heated at reflux for 3h. The hot solution was then filtered off in a
frit and the volume was reduced in a rotary evaporator. After addition of a saturated aqueous solution of
NH4PF6 a precipitate formed which was filtered off and washed with water. The solid obtained in this
manner was a mixture approximately 1:1 of complexes trans-2 and cis-2. Yield: 111 mg (61%). Anal.
Found (Calc.) for 2·H2O: C, 40.5 (40.6); H, 3.32 (3.77); N, 9.81 (10.1). IR (ν, cm-1): 3050 (ν (=C-H)), 1598
(ν (C=C)), 1439 (ν (C=N)), 1243 (ν (P=O)), 1015 (ν (P-O-C)), 840-752 (ν (P-C)). E1/2 (CH2Cl2 + 0.1M
TBAH) = 0.93 V vs. SCE. UV-Vis (CH2Cl2) [λmax, nm (ε, M-1 cm-1)]: 282 (28207), 324 (23103), 394 (7658),
512 (8615).
For trans-2, 1H-NMR (acetone- d6, 400 MHz): δ 1.42 (t, 3H, H17, J17-16 = 7 Hz), 1.44 (t, 3H, H19, J19-18 = 7
Hz), 3.02 (s, 3H, H28), 4.34 (m, 4H, H16, H18), 7.24 (d, 1H, H26, J26-27 = 2.8 Hz), 7.51 (ddd, 2H, H2, H14,
J2-1
=
J14-15 = 5.6 Hz; J2-3
=
J14-13 = 7.5 Hz; J2-4
=
J14-12 = 1.1 Hz), 7.58 (d, 1H, H27, J27-26 = 2.8 Hz), 7.85 (d,
22
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
2H, H1, H15, J1-2
8.06 (ddd, 2H, H3, H13, J3-2
=
J15-14 = 5.6 Hz), 7.91 (ddd, 1H, H21, J21-20 = 5.6 Hz; J21-22 = 7.2 Hz; J21-23 = 1.3Hz),
J13-14 = 7.7 Hz; J3-4 J13-12 = 5.2 Hz; J3-1 = 1.4 Hz), 8.35 (dt, 1H, H22, J22-21
=
=
=
J22-23 = 7.2 Hz; J22-20 = 1.5 Hz), 8.53 (d, 1H, H23, J23-22 = 7.2 Hz), 8.84 (d, 1H, H4, J4-3 = 5.2 Hz), 8.86 (d,
1H, H12, J12-13 = 5.2 Hz), 8.87 (s, 1H, H7), 8.90 (s, 1H, H9), 10.16 (d, 1H, H20, J20-21= 5.6 Hz). 13C-NMR
(acetone- d6): δ 16.8 (C17, C19), 38.3 (C28), 63.9 (C16, C18), 105.6 (C26), 123.1 (C23), 124.5 (C4, C7),
124.6 (C9), 125.1 (C12), 125.6 (C21), 128.8 (C2, C14), 137.2 (C27), 137.9 (C3, C13), 138.3 (C8, C22),
153.4 (C1, C15), 153.8 (C24, C25), 153.9 (C20), 159.7 (C5), 160.6 (C11), 160.8 (C6), 161.6 (C10).
1
For cis-2, H-NMR (acetone-d6, 400 MHz): δ 1.42 (t, 3H, H17, J17-16 = 7.1 Hz), 1.44 (t, 3H, H19, J19-18
=
7.1 Hz), 4.34 (m, 4H, H16, H18), 4.73 (s, 3H, H28), 6.91 (ddd, 1H, H22, J22-21 = 7.8 Hz; J22-23 = 5.8 Hz; J22-
20 = 1.3 Hz), 7.37 (d, 1H, H23, 23-22 = 5.8 Hz), 7.53 (ddd, 2H, H2, H14, J2-1
8 Hz; J2-4 J14-12 = 1.1 Hz), 7.61 (d, 1H, H26, J26-27 = 2.8 Hz), 7.72 (dt, 1H, H21, J21-20
J21-23 = 1.3Hz), 8.03 (ddd, 2H, H3, H13, J3-2 J13-14 = 8 Hz; J3-4 J13-12 = 5.2 Hz; J3-1 = 1.4 Hz), 8.10 (d,
2H, H1, H15, J1-2
=
J14-15 = 6.1 Hz; J2-3
= J14-13 =
=
=
J21-22 = 7.8 Hz;
=
=
=
J15-14 = 6.1 Hz), 8.17 (d, 1H, H20, J20-21= 7.8 Hz), 8.47 (d, 1H, H27) J27-26 = 2.8 Hz,
8.81 (s, 1H, H7), 8.82 (d, 1H, H4, J4-3 = 5.2 Hz), 8.83 (d, 1H, H12, J12-13 = 5.2 Hz), 8.85 (s, 1H, H9). 13C-
NMR (acetone- d6): δ 16.8 (C17, C19), 42.0 (C28), 63.9 (C16, C18), 106.2 (C26), 122.6 (C20), 124.4
(C7), 124.5 (C4), 125.0 (C12, C22), 125.1 (C9), 128.8 (C2, C14), 137.0 (C21), 138.0 (C3, C13), 138.1
(C8, C27), 151.2 (C24), 152.7 (C23, C25), 154.3 (C1, C15), 159.5 (C5), 159.5 (C11), 161.4 (C6), 161.5
(C10). For the NMR assignments we use the same labeling scheme as for the X-ray structures (Figure 1).
trans and cis-[Ru(trpy-P-Et)(pypz-Me)(OH2)](PF6)2, trans- and cis-3. A 0.03 g (0.036 mmol) sample of
trans and cis-2 was dissolved in 40 ml of water, the resulting solution was heated at reflux for 18h. After
reduction of the volume in a rotary evaporator a saturated aqueous solution of NH4PF6 was added. The
precipitate formed was filtered off and washed several times with cold water. The solid obtained in this
manner was a mixture of complexes trans-3 and cis-3. Yield: 16 mg (42%). Anal. Found (Calc.) for
3·1.5(C2H5)2O: C, 39.1 (38.9); H, 4.1 (4.4); N, 8.0 (8.0). IR (ν, cm-1): 3081 (ν (=C-H)), 2918 (ν (O-H)),
1602 (ν (C=C)), 1441 (ν (C=N)), 1227 (ν (P=O)), 1015 (ν (P-O-C)), 840-753 (ν (P-C)). UV-Vis (CH2Cl2)
[λmax,nm (ε, M-1 cm-1)]: 278 (45478), 320 (36314), 382 (7540), 480 (10418).
1
For trans-3, H-NMR (acetone-, 400 MHz): δ 1.45 (m, 6H, H17, H19), 3.08 (s, 3H, H28), 4.39 (m, 4H,
H16, H18), 7.28 (d, 1H, H26), 7.65 (m, 3H, H2, H14, H27), 8.03 (ddd, 1H, H21), 8.07 (d, 2H, H1, H15),
8.19 (m, 2H, H3, H13), 8.46 (dt, 1H, H22), 8.64 (d, 1H, H23), 8.96 (m, H4), 8.99 (dd, 1H, H12), 9.03 (s,
1H, H7), 9.06 (s, 1H, H9), 9.68 (d, 1H, H20). E1/2 (IV/II), phosphate buffer pH = 7.12: 0.42 V vs. SCE.
1
For cis-3, H-NMR (acetone- d6, 400 MHz): δ 1.45 (m, 6H, H17, H19), 4.39 (m, 4H, H16, H18), 4.64 (s,
3H, H28), 6.94 (ddd, 1H, H22), 7.40 (d, 1H, H23), 7.65 (m, 3H, H2, H14, H26), 7.76 (dt, 1H, H21), 8.19
(m, 4H, H1, H3, H13, H15), 8.29 (ddd, 1H, H20), 8.56 (d, 1H, H27), 8.95 (dd, 1H, H4), 8.96 (m, H12), 8.97
(s, 1H, H7), 9.00 (s, 1H, H9). E1/2 (IV/II), phosphate buffer pH = 7.12: 0.52 V vs. SCE.
Preparation of the heterogeneous systems
Two synthetic strategies have been followed:
Strategy I
I.1) Preparation of ligand-functionalized silica
.
For ligand L1: a sample of MSP (0.1 g) was added to a solution of trpy-P-Et, L1, (0.05 g, 0.14 mmol) in
toluene (10 ml) and the suspension was refluxed overnight. The resulting support, (trpy-P)-modified silica,
was centrifuged, washed with acetone (2 x 10 ml) and dried in a hot air oven at 110°.
For ligand L2: A sample of SP or MSP (0.1 g) was added to a solution of trpy-P-H, L2, (0.05 g, 0.16
mmol) in water (10 ml). The suspension was stirred overnight. The resulting support, (trpy-P)-modified
silica, was centrifuged, washed with water (2 x 10 ml) and dried in a hot air oven at 110°.
I.2) Preparation of complex-functionalized silica.
A 0.1 g sample of organically modified silica (SP@(trpy-P) or MSP@(trpy-P) was added to a solution of
RuCl3·2.38H2O (0.040 g, 0.16 mmol) in methanol. The mixture was stirred overnight. The resulting solid
(SP@[RuCl3(trpy-P)] or MSP@[RuCl3(trpy-P)]) was centrifuged, washed with methanol (2 x 10 ml) and
dried in a hot air oven at 110°C.
A 0.1 g sample of ruthenium-functionalized silica (SP@[RuCl3(trpy-P)] or MSP@[RuCl3(trpy-P)]) was
placed in a round-bottomed flask together with 6 ml of methanol. Then, LiCl (0.014 g, 0.32 mmol) and
23
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
triethylamine (44.6 μl, 0.32 mmol) were added under nitrogen atmosphere and the mixture was stirred at
room temperature for 30 min. After this time, pypz-Me (L3) ligand (0.026 g, 0.16 mmol) dissolved in 4 ml
of methanol was added and the mixture was heated at reflux for 2 h. The resulting modified silica
(SP@[Ru(trpy-P)(pypz-Me)(OH2)]2+, SP@3, or MSP@[Ru(trpy-P)(pypz-Me)(OH2)]2+, MSP@3) was
centrifuged, washed with methanol (2 x 10 ml) and dried in a hot air oven at 110°C. For SP@3: E1/2 (IV/II),
phosphate buffer pH = 5.5: 0.56 V vs SCE. UV-Vis (CH2Cl2) [λmax, nm]: 276, 316, 378, 484. For MSP@3:
E1/2 (IV/II), phosphate buffer pH = 5.5: 0.60 V vs SCE.
Strategy II
SP (0.1 g) was added to a solution of 2 (0.05 g, 0.06 mmol) in 9 ml of toluene. The mixture was stirred
and heated at reflux for 4 h. afterwards, 1 ml of acetone was added and the solution was refluxed for 20h.
The resulting product was centrifuged, washed with acetone (2 x 10 ml) and dried in a hot air oven at
110°C. UV-Vis (CH2Cl2) [λmax, nm]: 280, 328, 386, 490. E1/2(IV/II), phosphate buffer pH = 7.1: 0.55 V vs.
SCE.
Catalytic studies
Alkene epoxidation
Homogeneous System
The ruthenium catalyst (1.25 μmol), alkene (125 μmol) and PhI(OAc)2 (250 μmol) were stirred at room
temperature in CH2Cl2 (2.5 ml) for 24 h. After addition of an internal standard, the sample was filtered
through a basic alumina plug and quantified by gas chromatographic (GC) analysis based on calibration
curves for each substrate and epoxide.
Heterogeneous System
Alkene (100 μmol) and PhI(OAc)2 (200 μmol) were dissolved in 2.5 ml of CH2Cl2 together with 1 μmol of
SP or MSP-supported ruthenium catalyst. The amount of heterogenized catalyst was calculated taking
into account the functionalization of SPs and MSPs (mmol Ru·g-1) in each case. For experiments using
SP-based system, the resulting solution was centrifuged while for MSP-based system a magnet was used
for the separation of the catalytic material from the reaction medium. The catalytic heterogeneous
systems were washed twice with CH2Cl2 and were used in a subsequent catalytic run. For leaching
calculations, the resulting solution after several reuses was centrifuged and filtered through celite in order
to avoid particles in the ICP analysis.
Acknowledgements
This research has been financed by AGAUR (Generalitat de Catalunya, projects 2014-SGR-149 and
2014-SGR-213), by MINECO (project CTQ2015-66143-P) and by Universitat de Girona (project
MPCUdG2016/048). IF thanks UdG for a predoctoral grant. Dr. M. Iglesias from the Chemistry
Department of UdG is gratefully acknowledged for her help with the ICP quantification measurements.
24
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
References
[1] a) S. Y. Ko, A. W. M. Lee, S. Masamune, L. A. Reed, K. B. Sharpless, F. J. Walker, Science 1983, 220, 949–951; b)
K. C. Nicolaou, N. Winssinger, J. Pastor, S. Ninkovic, F. Sarabia, Y. He, D. Vourloumis, Z. Yang, T. Li, P.
Giannakakou, E. Hamel, Nature 1997, 387, 268–272; c) S. D. Gagnon in Encyclopedia of Polymer Science,
Engineering, 2nd ed. (Eds: H. F. Mark, N. M. Bikales, C. G. Overberger, G. Menges, J. I. Kroschwitz), Wiley-VCH:
Weinheim, 1994, pp. 275–307; d) D. J. Darensbourg, R. M. Mackiewicz, A. L. Phelps, D. R. Billodeaux, Acc.
Chem. Res. 2004, 37, 836–844; e) E. N. Jacobsen, Catalytic Asymmetric Synthesis (Ed: I. Ojima), VCH: New York,
1993, p. 229; f) G. De Faveri, G. Ilyashenko, M. Watkinson, Chem. Soc. Rev. 2011, 40, 1722–1760.
[2] a) I. Serrano, M. I. López, Í. Ferrer, A. Poater, T. Parella, X. Fontrodona, M. Solà, A. Llobet, M. Rodríguez, I.
Romero, Inorg. Chem. 2011, 50, 6044–6054; b) E. Masllorens, M. Rodríguez, I. Romero, A. Roglans, T. Parella, J.
Benet-Buchholz, M. Poyatos, A. Llobet, J. Am. Chem. Soc. 2006, 128, 5306–5307; c) M. Dakkach, X. Fontrodona,
T. Parella, A. Atlamsani, I. Romero, M. Rodríguez, Adv. Synth. Catal. 2011, 353, 231–238; d) J. Mola, C. Dinoi, X.
Sala, M. Rodríguez, I. Romero, T. Parella, X. Fontrodona, A. Llobet, Dalton Trans. 2011, 40, 3640–3646; e) M.
Dakkach, A. Atlamsani, T. Parella, X. Fontrodona, I. Romero, M. Rodríguez, Inorg. Chem. 2013, 52, 5077–5087.
[3] a) A. S. Hazari, A. Das, R. Ray, H. Agarwala, S. Maji, S. M. Mobin, G. K. Lahiri, Inorg. Chem. 2015, 54, 4998–
5012; b) J. R. Bryant, T. Matsuo, J. M. Mayer, Inorg. Chem. 2004, 43, 1587–1592; c) X. Sala, A. Poater, I. Romero,
M. Rodríguez, A. Llobet, X. Solans, T. Parella, T. M. Santos, Eur. J. Inorg. Chem. 2004, 2004, 612–618; d) W.-P.
Yip, W.-Y. Yu, N. Zhu, C.-M. Che, J. Am. Chem. Soc. 2005, 127, 14239–14249; e) R. Zong, R. P. Thummel, J. Am.
Chem. Soc. 2005, 127, 12802–12803.[
[4] a) C. D. Nunes, A. A. Valente, M. Pillinger, A. C. Fernandes, C. C. Romão, J. Rocha, I. S. Gonçalves, J. Mater.
Chem. 2002, 12, 1735–1742; b) M. Jia, A. Seifert, W. R. Thiel, Chem. Mater. 2003, 15, 2174–2180; c) A. Sakthivel,
J. Zhao, G. Raudaschl-Sieber, M. Hanzlik, A. S. T. Chiang, F. E. Kühn, Appl. Catal. A Gen. 2005, 281, 267–273.
[5] a) L. Yin, J. Liebscher, Chem. Rev. 2007, 107, 133–173.
[6] a) M. Shokouhimehr, Y. Piao, J. Kim, Y. Jang, T. Hyeon, Angew. Chemie 2007, 119, 7169–7173; b) M.
Shokouhimehr, Y. Piao, J. Kim, Y. Jang, T. Hyeon, Angew. Chem. Int. Ed. Engl. 2007, 46, 7039–7043; c) E. Ertem,
N. Murillo-Cremaes, R. P. Carney, A. Laromaine, E.-R. Janeček, A. Roig, F. Stellacci, Chem. Commun. 2016, 52,
5573–5576; d) S. Shylesh, V. Schünemann, W. Thiel, Angew. Chem. Int. Ed. 2010, 49, 3428 –3459; e) S. Shylesh,
L. Wang, W. R. Thiel, Adv. Synth. Catal. 2010, 352, 425- 432; f) K. S. M. Salih, P. Mamone, G. Dörr, T. O. Bauer,
A. Brodyanski, C. Wagner, M. Kopnarski, R. N. Klupp-Taylor, S. Demeshko, F. Meyer, V. Schünemann, S. Ernst,
L. J. Gooßen, W. R. Thiel, Chem. Mater. 2013, 25, 1430-1435; g) D. Dehe, L. Wang, M. K. Müller, G. Dörr, Z.
Zhou, R. N. Klupp-Taylor, Y. Sun, S. Ernst, M. Hartmann, M. Bauer, W. R. Thiel, ChemCatChem 2015, 7, 127-
136.
[7] a) L. Vaquer, P. Riente, X. Sala, S. Jansat, J. Benet-Buchholz, A. Llobet, M. A. Pericàs, Catal. Sci. Technol. 2013,
3, 706–714; b) S. Shylesh, J. Schweizer, S. Demeshko, V. Schünemann, S. Ernst, W. R. Thiel, Adv. Synth. Catal.
2009, 351, 1789–1795; c) S. Shylesh, M. Jia, W. R. Thiel, Eur. J. Inorg. Chem. 2010, 4395-4410.
[8] M. Dakkach, M. I. López, I. Romero, M. Rodríguez, A. Atlamsani, T. Parella, X. Fontrodona, A. Llobet, Inorg.
Chem. 2010, 49, 7072–7079.
[9] A. Rilak, I. Bratsos, E. Zangrando, J. Kljun, I. Turel, Ž. D. Bugarčić, E. Alessio, E. Inorg. Chem. 2014, 53, 6113–
6126.
[10] C. Sens, M. Rodríguez, I. Romero, A. Llobet, T. Parella, J. Benet-Buchholz. Inorg. Chem. 2003, 42, 8385–8394.
[11] V. Balzani, A. Juris, M. Venturi, S. Campagna, S. Serroni, Chem. Rev. 1996, 96, 759–834.
[12] E. T. Bell-Loncella, C. A. Bessel, Inorganica Chim. Acta 2000, 303, 199–205.
[13] R. Llanguri, J. J. Morris, W. C. Stanley, E. T. Bell-Loncella, M. Turner, W. J. Boyko, C. A. Bessel, Inorganica
Chim. Acta 2001, 315, 53–65.
[14] P. Passaniti, W. R. Browne, F. C. Lynch, D. Hughes, M. Nieuwenhuyzen, P. James, M. Maestri, J. G. Vos, J.
Chem. Soc. Dalton Trans. 2002, 1740.
[15] J. A. Gilbert, S. W. Gersten, T. J. Meyer, J. Am. Chem. Soc. 1982, 104, 6872–6873.
[16] R. A. Leising, K. J. Takeuchi, Inorg. Chem. 1987, 26, 4391–4393.
[17] A. B. P. Lever, Inorg. Chem. 1990, 29, 1271–1285.
[18] E. Taboada, R. Solanas, E. Rodríguez, R. Weissleder, A. Roig, Adv. Funct. Mater. 2009, 19, 2319–2324
[19] W. Mägerlein, C. Dreisbach, H. Hugl, M. K. Tse, M. Klawonn, S. Bhor, M. Beller, Catal. Today 2007, 121, 140–
150.
25
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
[20] M. K. Tse, S. Bhor, M. Klawonn, G. Anilkumar, H. Jiao, A. Spannenberg, C. Döbler, W. Mägerlein, H. Hugl, M.
Beller, Chem. Eur. J. 2006, 12, 1875–1888.
[21] J. Benet-Buchholz, P. Comba, A. Llobet, S. Roeser, P. Vadivelu, H. Wadepohl, S. Wiesner, Dalton Trans. 2009,
5910–5923.[
[22] G. A. Barf, R. A. Sheldon, J. Mol. Catal. A Chem. 1995, 102, 23–39.
[23] A. D. Chowdhury, A. Das, I. K, S. M. Mobin, G. K. Lahiri, Inorg. Chem. 2011, 50, 1775–1785.
[24] M. K. Tse, C. Döbler, S. Bhor, M. Klawonn, W. Mägerlein, H. Hugl, M. Beller, Angew. Chem. Int. Ed. Engl. 2004,
43, 5255–5260.
[25] a) W. Adam, K. J. Roschmann, C. R. Saha-Möller, D. Seebach, J. Am. Chem. Soc. 2002, 124, 5068–5073; b) S.-E.
Park, W. J. Song, Y. O. Ryu, M. H. Lim, R. Song, K. M. Kim, W. Nam, J. Inorg. Biochem. 2005, 99, 424–431.
[26] E. Manrique, A. Poater, X. Fontrodona, M. Solà, M. Rodríguez, I. Romero, Dalton Trans. 2015, 44, 17529–17543.
[27] M. Palucki, N. S. Finney, P. J. Pospisil, M. L. Güler, T. Ishida, E. N. Jacobsen, J. Am. Chem. Soc. 1998, 120, 948–
954.
[28] X.-Q. Yu, J.-S. Huang, W.-Y. Yu, C.-M. Che, J. Am. Chem. Soc. 2000, 122, 5337–5342.
[29] a) M. Dakkach, X. Fontrodona, T. Parella, A. Atlamsani, I. Romero, M. Rodríguez, Dalton Trans. 2014, 43, 9916-
9923; b) M. Tada, S. Muratsugu, M. Kinoshita, T. Sasaki, Y. Iwasawa, J. Am. Chem. Soc. 2010, 132, 713-724; c) M.
Tada, R. Coquet, J. Yoshida, M. Kinoshita, Y. Iwasawa, Angew. Chem. Int. Ed. 2007, 46, 7720-7723; d) X-Q. Yu, J-
S. Huang, W-Y- Yu, C-M. Che, J. Am. Chem. Soc. 2000, 122, 5537-5342.
[30] W. R. Thiel, M. Angstl, T. Priermeier, T. Chem. Ber. 1994, 127, 2373–2379.
[31] M. Moner-Girona, A. Roig, E. Molins, J. Llibre, J. Sol-Gel Sci. Technol. 2003, 26, 645–649.
26
This article is protected by copyright. All rights reserved.

10.1002/chem.201604463
Chemistry - A European Journal
FULL PAPER
A molecular ruthenium aqua complex
has been anchored onto silica
mesoporous and silica coated
magnetic particles, through two
synthetic strategies without a
Author(s), Corresponding Author(s)*
Page No. – Page No.
Title
significant modification of the
coordination and electronic properties
with respect to the molecular
counterpart. The catalytic results and
the reutilization of these hybrid
materials highlight their performance
in the epoxidation of alkenes.
27
This article is protected by copyright. All rights reserved.