Synthesis of Co, Rh and Ir nanoparticles from metal carbonyls in ionic liquids and their use as biphasic liquid-liquid hydrogenation nanocatalysts for cyclohexene

Article abstract of DOI:10.1016/j.jorganchem.2008.09.050

Stable cobalt, rhodium and iridium nanoparticles are obtained reproducibly by thermal decomposition under argon from Co₂(CO)₈, Rh₆(CO)₁₆ and Ir₄(CO)₁₂ dissolved in the ionic liquids BMim

Full text of DOI:10.1016/j.jorganchem.2008.09.050

Journal of Organometallic Chemistry 694 (2009) 1069–1075
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Synthesis of Co, Rh and Ir nanoparticles from metal carbonyls in ionic liquids
and their use as biphasic liquid–liquid hydrogenation nanocatalysts for cyclohexene
Engelbert Redel a, Jérôme Krämer a, Ralf Thomann b,1, Christoph Janiak a,
*
a Institut für Anorganische und Analytische Chemie, Universität Freiburg, Albertstrasse 21, D-79104 Freiburg, Germany
b FMF (Freiburger Material Forschungszentrum), Universität Freiburg, Stefan-Meier-Strasse 21, D-79104 Freiburg, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Stable cobalt, rhodium and iridium nanoparticles are obtained reproducibly by thermþal decomposition
under argon from Co2(CO)8, Rh6(CO)16 and Ir4(CO)12 dissolved in the ionic liquids BMim BFꢀ4 , BMim+OTfꢀ
and BtMAþNTf2ꢀ ꢀ [BMim+ = n-butyl-methyl-imidazolium, BtMA+ = n-butyl-tri-methyl-ammonium,
in BMimþBFꢀ4 increases with the molecular volume of the ionic liquid anion in BMim+OTfꢀ and
BtMAþNTfꢀ2 . Characterization of the nanoparticles was done by transmission electron microscopy
(TEM), transmission electron diffraction (TED), X-ray powder diffraction (XRPD) and dynamic light scat-
tering (DLS). The rhodium or iridium nanoparticle/IL systems function as highly effective and recyclable
catalysts in the biphasic liquid–liquid hydrogenation of cyclohexene to cyclohexane with activities of up
to 1900 molproduct/(molmetal h) and quantitative conversion within 1 h at 4 bar H2 pressure and 75 °C.
Ó 2008 Elsevier B.V. All rights reserved.
Received 22 July 2008
Received in revised form 15 September
2008
Accepted 29 September 2008
Available online 2 October 2008
OTfꢀ
=
ꢀO3SCF3, NTf2
¼
ꢀNðO2SCF3Þ2]. The very small and uniform nanoparticle size of about 1–3 nm
Dedicated to Prof. Christoph Elschenbroich
on the occasion of his 70th birthday
Keywords:
Metal nanoparticles
Nanocatalysts
Catalytic olefin hydrogenation
Ionic liquids
Metal carbonyl
1. Introduction
ideal point charges. Real-life anions with a molecular volume
would be classified as ‘‘electrosteric stabilizers”. However, the term
Transition-metal nanoparticles are very important for techno-
logical applications in several areas of science and industry, includ-
ing catalysis or chemical sensors [1–3]. In particular, Ir and Rh
nanoparticles can be used for olefin hydrogenation reactions [4].
The controlled and reproducible synthesis of defined and stable
metal nanoparticles (MNPs) is of very high importance for different
fields of applications [5–10]. Metal nanoparticle syntheses in ionic
liquids (ILs) are carried out through the reduction [11–14] or
decomposition [15] of metal salts or metal complexes, photochem-
ical [16,17] or electro-reduction [18–20], and recently also by ther-
mal and photochemical decomposition of metal carbonyls [21–23].
Ionic liquids stabilize metal nanoparticles on the basis of their
high ionic charge, high polarity, high dielectric constant and supra-
molecular network [24–27]. According to DLVO (Derjaugin–
Landau–Verwey–Overbeek) theory [28,29], ILs should provide an
electrosteric protection in the form of a ‘‘protective shell” for MNPs,
so that no extra stabilizing molecules or organic solvents are
needed [30–32]. Here we note that DLVO theory treats anions as
‘‘electrosteric” is contentious and ill-defined [33]. The stabilization
of metal nanoclusters in ionic liquids could, thus, be attributed to
‘‘extra-DLVO” forces [33] including effects from the network prop-
erties of ionic liquids like hydrogen bonding, hydrophobic effects
and steric interactions. However, the DLVO theory predicts that
the anionic charges should be the primary source of stabilization
for the electrophilic metal nanocluster. Experimentally PFꢀ6 anions
from BMimþPFꢀ6 were found on a Pd nanocluster surface by XPS.
This supports the hypothesis that weakly coordinating anions can
contribute to the stability of transition-metal nanoclusters in or-
ganic solutions or ILs [34]. In the case of Ir nanoclusters stabilized
by imidazolium ionic liquids, it was found that in addition to the
BF4 anion N-heterocyclic carbenes are another source of potential
nanocluster stabilization in ILs [35]. Also, a correlation was found
between the molecular volume of the anion in the ionic liquid
and the size of the synthesized metal nanoparticles [11,21].
However, transition-metal nanoparticles in ionic liquids are only
kinetically stable because the formation of bulk metal is thermody-
namically favored. Yet, in the absence of strongly coordinating pro-
tective ligand layers, MNPs in ILs should be effective catalysts. The
IL network contains only weakly coordinating cations and anions
as stabilizers that bind less strongly to the metal surface than other
anions or capping ligands. Generally, the catalytic properties
* Corresponding author. Tel.: +49 761 2036147; fax: +49 761 2036127.
E-mail addresses: ralf.thomann@fmf.uni-freiburg.de (R. Thomann), janiak@uni-
freiburg.de (C. Janiak).
1
Tel.: +49 761 2035379.
0022-328X/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2008.09.050

1070
E. Redel et al. / Journal of Organometallic Chemistry 694 (2009) 1069–1075
(activity and selectivity) of dispersed metal nanoparticles indicate
that they possess pronounced surface-like (multi-site) rather than
single-site-like character [1,36,37].
hours in the ionic liquids n-butyl-tri-metꢀhylammonium
methyl-imidazolium tetrafluoroborate (BMimþBF4ꢀ) or trifluoro-
methanesulfonate (BMim+OTfꢀ) (Scheme 1). Co2(CO)8 was decom-
posed at 180 °C, Rh6(CO)16 and Ir4(CO)12 at 230 °C. Literature
N-bis(trifluoromethylsulfonyl)imide
(BtMAþNTf2 ),
n-butyl-
The hydrogenation of benzene to cyclohexane is a multi-million
ton process for the subsequent oxidation of the cycloalkane to adi-
pic acid and caprolactam as building blocks for Nylon 6.6 and Ny-
lon 6 [38,39]. Cyclohexene is often taken as a model substrate for
hydrogenation activity studies of biphasic liquid–liquid systems
involving ILs [40–42], also because its solubility is limited in the
polar ionic liquid BMimþBF4ꢀ [43]. ILs are also interesting under
green catalysis aspects which require that catalysts be designed
for easy product separation from the reaction media and multi-
time catalyst recycling with very high efficiency [44]. Ionic liquids
as catalytic reaction media fulfil these requirements as no surplus
organic solvent is needed in biphasic catalytic transformations of
liquid substrates. A low miscibility of hydrophobic substrates and
products with the polar ionic liquid phase, e.g., BMimþBF4ꢀ allows
for easy separation by simple decantation of the hydrophobic
phase. Also, low-volatile products can simply be distilled off or re-
moved under vacuum from ILs with their very low vapor pressure
to allow reuse of the catalyst/IL system.
reported decomposition temperatures are 220 °C for Rh6(CO)16
210 °C for Ir4(CO)12 and above 100 °C for Co2(CO)8 [45].
,
Dark-brown to black Co, Rh or Ir nanoparticle dispersions
(Fig. S27, Supplementary material) are reproducibly obtained
through decomposition from their metal carbonyl in ionic liquids
(Figs. 1–4). The resulting nanoparticles were analyzed by transmis-
sion electron microscopy (TEM), transmission electron diffraction
(TED), X-ray powder diffractometry (XRPD) and dynamic light
scattering (DLS) (Table 1). The dispersions are found stable for
more than six months under argon atmosphere. Under air the
black, magnetic Co dispersion (Fig. S28) turns violet and loses its
magnetic properties. The Rh and Ir dispersions are stable, and the
particle size does not change, after more than 70 h with air/oxygen
contact (Fig. S7 DLS). TED and XRPD studies show that the resulting
Co, Rh and Ir nanoparticles produced under argon are not oxidized.
The diffraction patterns correspond to their metal lattices (Fig. 5,
Figs. S21–S26, Tables S2 and S3) [46], thereby proving their metal
character and the absence of significant oxidation.
The median metal nanoparticle size of 3 nm for Rh and 1 nm for
Ir is extremely small with a narrow or uniform, albeit not monodis-
perse, size distribution (Figs. 2 and 3). It is, at present, not trivial to
routinely and easily prepare uniform nanoparticles of such small
1–3 nm size. No extra stabilizers or capping molecules are needed
to achieve this small particle size. The size of the Co, Rh and Ir
nanoparticles in BMimþBF4ꢀ could also be estimated with the
Scherrer equation [47] from the half-width of the diffraction peaks
(Figs. S24–S26). Ir nanoparticles obtained in BMim+OTfꢀ and
BtMAþNTfꢀ2 range from about ꢁ4 nm in the former to ꢁ80 nm
the later (ableit this value corresponds to large aggregates of smal-
ler Irþ nanoparticles) (Fig. 4). Co nanoparticles obtained in
BMim BFꢀ4 with a size of ꢁ14 nm are magnetic and agglomerate
as a result of their superparamagnetic properties (Fig. 1 and
Fig. S28) [5]. The same observation of agglomerated Co nanoparti-
cles was reported from a recent Co2(CO)8 decomposition in differ-
ent imidazolium ionic liquid/hexane mixture, e.g., BMimþNTf2ꢀ/
hexane [23].
Here we report the preparation of Co, Rh and Ir-metal nanopar-
ticles by thermal decomposition of Co2(CO)8, Rh6(CO)16 and
Ir4(CO)12, respectively, in different ILs. Moreover, the synthesized
nanoparticles in ILs were used as highly effective and recyclable
catalysts for the biphasic liquid–liquid hydrogenation of cyclohex-
ene to cyclohexane.
2. MNP syntheses and characterization
In a typical experiment the metal carbonyl was dissolved under
argon atmosphere in the dried and deoxygenated ionic liquid (IL).
The solution was heated under argon up to 230 °C (above the
decomposition temperature of the metal carbonyl) for several
A correlation exists between the molecular volume of the anion
in the ionic liquid and the synthesized metal nanoparticles
[11,21,22], indicated here for, but not limited to, Rh and Ir
(Fig. 6). The crystallinity and the size of the Ir nanoparticles depend
on the decomposition time. The longer the dispersions were
Scheme 1. Formation of Co, Rh and Ir nanoparticles by thermal decomposition of
their metal carbonyls in different ionic liquids (ILs).
Fig. 1. Co nanoparticles from thermal decomposition of Co2(CO)8 in BMimþBFꢀ (entry 1 in Table 1).
4

E. Redel et al. / Journal of Organometallic Chemistry 694 (2009) 1069–1075
1071
Fig. 2. Rh nanoparticles from thermal decomposition of Rh6(CO)16 in BMimþBF4ꢀ (left) and Rh nanoparticles after 6 hydrogenation cycles (right) (entry 2 and 5, respectively, in
Table 1).
Fig. 3. Ir nanoparticles from 3.5 h (left) and 18 h (right) thermal decomposition of Ir4(CO)12 in BMimþBFꢀ (entry 6 and 7, respectively, in Table 1).
4
Fig. 4. Ir nanoparticles from thermal decomposition of Ir4(CO)12 in BMim+OTfꢀ (left) and BtMAþNTfꢀ2 (right) (entry 9 and 10, respectively, in Table 1).
heated, the more crystalline are the particles (Fig. 5). The diffrac-
tion patterns of Ir samples which have been thermolyzed for
3.5 h show only diffuse halos. Ir nanoparticles from 18 h thermol-
ysis show narrow and well defined reflections (Fig. 5 and Fig. S23).
Nanoparticles must be stabilized in order to prevent their
agglomeration or aggregation which eventually leads to the forma-
tion of small metal particles. MNPs (core) are considered stabilized
in ILs by the formation of ‘‘protective” anionic and cationic layers

1072
E. Redel et al. / Journal of Organometallic Chemistry 694 (2009) 1069–1075
Table 1
MNP (M = Co, Rh and Ir) size and size distribution in different ionic liquids (ILs)
Ionic
VIL-Anion/nm3
[56]]
Metal carbonyl/metal weight
percent in IL/decomposition time
TEM median diameter/nm
Dynamic light scattering median
diameter/nm (standard deviation r)
b
c
liquida
(standard deviation
r)
1
2
3
4
5
6
7
8
9
10
BMimþBFꢀ4
BMimþBFꢀ4
BMim+OTfꢀꢀ
BtMAþNTf2
BMimþBFꢀ4
BMimþBFꢀ4
BMimþBFꢀ4
BMimþBFꢀ4
BMim+OTfꢀꢀ
BtMAþNTf2
0.073 0.009
0.073 0.009
0.131 0.015
0.232 0.015
0.073 0.009
0.073 0.009
0.073 0.009
0.073 0.009
0.131 0.015
0.232 0.015
Co2(CO)8/0.16/18 h
Rh6(CO)16/0.2/18 h
Rh6(CO)16/0.2/18 h
Rh6(CO)16/0.5/18 h
Rh6(CO)16/1/18 hd
Ir4(CO)12/0.2/3.5 h
Ir4(CO)12/0.2/18 h
Ir4(CO)12/0.5/18 h
Ir4(CO)12/0.5/18 h
Ir4(CO)12/0.5/18 h
14 ( 8)
3.0 ( 0.6)
4.4 ( 1.1)
14 ( 7)
3.5 ( 0.8)
1.1 ( 0.2)
1.7 ( 0.3)
1.3 ( 0.2)
3.6 ( 0.6)
81 ( 23)e
6.0 ( 1.6)
9.5 ( 1.7)
–
7.0 ( 1.2)
4.1 ( 0.7)
3.6 ( 0.7)
3.4 ( 1.0)
7.1 ( 1.1)
–
Solubility of metal carbonyl precursors in BMimþBF4ꢀ is limited to a maximum value of about 1 wt.%.
Statistical evaluation of the total sample pictures (see also Supplementary material).
Hydrodynamic radius, median diameter form the first 3 measurements at 633 nm. The hydrodynamic radius is roughly 2–3 times the size of the pure kernel cluster. For
a
b
c
very small MNPs (ꢁ1 nm) the size of the hydrodynamic radius can even increase to more than 3 times the MNP radius. The resolution of the DLS instrument is 0.6 nm.
d
After six catalytic cycles.
This value corresponds to large aggregates of smaller Ir nanoparticles.
e
Fig. 5. Transmission electron diffraction (TED) pattern of Ir nanoparticles from Ir4(CO)12 in BMimþBFꢀ by thermal decomposition after 3.5 h (left) and 18 h (right) (entry 6 and
4
0
7, respectively, in Table 1). The black bar is the beam stopper. The diffraction rings at (ÅA) 2.2 (very strong), 1.9, 1.4, 1.2 (all strong) match with the D-spacing of the Ir-metal
diffraction pattern (see Fig. S23 and Table S3).
(shells) around them in a ‘‘core-shell system” [30,48–51]. According
to DLVO theory the first inner shell must be anionic [28]: Then the
IL anion will have the highest influence on the size and electro-
static stabilization of the Co, Rh and Ir nanoparticle. The anion
molecular volume determines the region of the nanoparticle size
(Fig. 6), although the supramolecular imidazolium-anion clusters
of ILs should be taken into account [11,21]. Pure ILs can be re-
garded as supramolecular polymeric structures with a high degree
of self-organisation and weak interactions of an extended network
of cations and anions connected together by hydrogen bonds [6].
When mixed with other molecules or MNPs, ionic liquids become
nano-structured materials with polar and nonpolar regions
[6,52–55]. The ionic liquid represents an excellent medium for
the formation of MNPs with, in most cases, a small size and size
distribution under mild conditions.
3. Cyclohexene hydrogenation catalysis with MNP/IL
The obtained Co, Rh and Ir nanoparticle/IL dispersions were
tested for their catalytic activity in the biphasic liquid–liquid
hydrogenation of cyclohexene to cyclohexane (Scheme 2 and
Table 2).
Fig. 6. Correlation between the molecular volume of the ionic liquid anion (VIL-anion
and the observed Rh nanoparticle size with standard deviations as error bars (from
TEM measurements, entry 2–4 in Table 1).
)

E. Redel et al. / Journal of Organometallic Chemistry 694 (2009) 1069–1075
1073
ners, steps) rather than the sole amount of surface atoms strongly
influence the catalytic activity [57].
In the context of green catalysis aspects the recyclability of the
MNP/IL system was investigated. The biphasic liquid–liquid cata-
lytic hydrogenation of cyclohexene could utilize the low miscibility
of the hydrophobic substrate and product (cyclohexene and cyclo-
hexane [43]) with the polar ionic liquid phase BMimþBF4ꢀ for phase
separation by decantation. With our small volumes this procedure
proved not very practical, however, and complete organic phase
separation could not be ensured this way. Hence, we separated
the cyclohexene/cyclohexane mixture by distillation from the
low-volatile IL phase. Based on the conditions for entry 4 in Table
2 the catalysis was started with 0.08 ml RhNP/BMimþBF4ꢀ disper-
sion (c = 0.025 mol/l; and 0.2 ml (2 mmol) cyclohexene (molar
cyclohexene/metal ratio = 1000)) at 75 °C and 4 bar H2 pressure.
After 2.5 h the organic phase was distilled off and fresh cyclohex-
ene substrate (0.2 ml) was added. It is evident that the RhNP/
BMimþBFꢀ4 dispersion can be reused more than six times without
any significant loss in catalytic activity (see Scheme 3), contrary
to reports in the literature where RhNP was prepared from
RhCl3 ꢂ 3H2O through H2 reduction in BMimþPF6ꢀ [13]. In the ab-
sence of a scavenger the reduction route will generate HCl. Also,
the PFꢀ6 anion is known to more easily decompose to HF than
BFꢀ4 in the presence of water and/or metal [58–61]. The formation
of acids in ILs is disadavantages for the NP/IL network stability be-
cause of proton þincorporation in the IL dynamic matrix [11]. With
our RhNP/BMim BFꢀ4 dispersion there may even be a small activity
increase through the first cycles (see Scheme 3). This should not be
over interpreted for the moment as the activity increase from 350
(1st cycle, 88% conversion) to 400 mol product (mol metal)ꢀ1 hꢀ1
(6th cycle, quantitative conversion) can be seen to fluctuate around
375 mol product (mol metal)ꢀ1 hꢀ1 with some measurement errors
(from GC). Still, some activity increase could be reasoned by the
formation of Rh-hydride species and their immediate presence in
subsequent cycles which prevents an induction time and, thereby
speeds up the reaction. Furthermore, a TEM and DLS investigation
of the Rh nanoparticles after six catalytic cycles, albeit based on a
1 wt.% RhNP dispersion (entry 5 in Table 1 and Fig. 2, right, see also
Figs. S4, S14 and S15) indicates particle agglomeration regions of
smaller particles, yet the individual small particles can still be seen.
With the 0.2 wt.% RhNP/IL solution used in catalysis the agglomer-
ation is likely to be less and the expected concomitant activity loss
apparently outweighted by the formation of more active (Rh-hy-
dride) surface species.
Scheme 2. Hydrogenation of cyclohexene catalyzed by Co, Rh and Ir nanoparticles
in BMimþBFꢀ4
.
Table 2
Catalytic biphasic liquid–liquid hydrogenation of cyclohexene to cyclohexane with
MNP (M = Co, Rh or Ir) in BMimþBF4ꢀ ionic liquid
Dispersion Molar ratio
Reaction
Time Conversion Activity mol
(metal
wt.%)
cyclohexene/ temperature (h)
(%)
product (mol
metal
(°C)
metal)ꢀ1 hꢀ1
1a
2a
3b
4c
5b
6b
Co 0.8
Co 0.8
Rh 0.2
Rh 1
Rh 0.2
Rh 0.2
Ir 0.2
Ir 0.2
Ir 0.2
60
60
1000
80
1000
1000
2000
2000
2000
2000
80
90
75
75
75
75
75
75
75
75
3
0.8
1.1
95
98
75
93
97
100
90
80
0.16
0.04
380
32
300
370
1940
800
1800
1600
18
2.5
2.5
2.5
2.5
1
2.5
1
1
b
7
8b
9b
10b Ir 0.2
a
0.5 ml CoNP/IL dispersion (c = 0.166 mol/l) and 0.5 ml cyclohexene.
0.08 ml MNP/IL dispersion (M = Rh, c = 0.025 mol/l; M = Ir, c = 0.0128 mol/l) and
b
0.2 ml cyclohexene.
c
1.0 ml RhNP/IL dispersion and 1.0 ml cyclohexene. Cyclohexene den-
sity = 0.811 g/ml, M = 82.14 g/mol, 1.0 ml corresponds to 10 mmol. BMimþBF4ꢀ
density = 1.21 g/ml.
While the Co nanoparticles showed very little to no activity
(Table 2), Ir and Rh nanoparticles were found to be highly active
nanocatalysts under very mild reaction conditions (75 °C, 4 bar
H2 pressure). Especially Ir nanoparticles exhibited very high cata-
lytic activities of up to 1900 mol product (mol metal)ꢀ1 hꢀ1. Rh
nanoparticles showed good catalytic activities in the range of
300–380 mol product (mol metal)ꢀ1 hꢀ1 through different catalytic
runs (Table 2). The higher catalytic activity of Ir in comparison to
Rh nanoparticles may be explained by their somewhat smaller size
and concomitant lager surface-to-volume ratio (cf. Table 1). During
our studies it became evident that the glass vessels need to be
carefully cleaned with conc. HCl/HNO3 mixture (aqua regia) fol-
lowed by surface passification with Me3SiCl to avoid both catalyst
transfer by glass surface adsorption and partial catalyst deactiva-
tion, respectively. Cyclohexene conversions of over 50% could still
be observed without any new catalyst addition or with normally
low active CoNP/IL when the glass vessel was only washed with
acetone, water and isopropanol but no acid treatment following a
Rh- or IrNP/IL run. On the other hand, the high activity of IrNP/IL
was significantly lowered without the Me3SiCl glass surface passi-
fication after the acid treatment.
Activity calculations with metal nanoparticles are typically
based on the total amount of metal and not only the surface metal
part. Different studies show that the catalytic activity results not
only from the exposed surface metal atoms. During heterogeneous
catalysis the surface can restructure and atoms can shift positions.
These changes in the surface lead to a difficult to determine num-
ber of surface atoms. Partial aggregation is also possible during
catalysis which would have a strong influence on the fraction of
the surface atoms. Heterogeneous catalytic reactions are extremely
complex and it is also possible that atoms under the exposed sur-
face atoms play a significant role in the catalytic activity. Moreover,
defects, the surface topology and surface atom sites (edges, cor-
The catalytic activity of RhNPs prepared from Rh6(CO)16 in
BMimþBFꢀ4 can be compared with other cyclohexene hydrogena-
tion catalysts and their precursors in Table 3. It is evident from
420
400
380
360
340
320
1
2
3
4
5
6
7
catalytic cycle
Scheme 3. Activity over seven catalytic cycles for the hydrogenation of cyclohex-
ene (0.2 ml, 2 mmol) with the same 0.2 metal wt.% RhNP/BMimþBF4ꢀ catalyst
(0.08 ml, 0.002 mmol) and molar cyclohexene/metal ratios of 1000 at 75 °C, 4 bar
H2 pressure and 2.5 h reaction time. An activity of 350 mol product (mol
metal)ꢀ1 hꢀ1 corresponds to 88% and an activity of 400 to quantitative (100%)
conversion.

1074
E. Redel et al. / Journal of Organometallic Chemistry 694 (2009) 1069–1075
Table 3
Mo K
a radiation (k = 0.7093 Å), a Ge(111) monochromator and
Catalyst activities of MNPs in the biphasic liquid–liquid cyclohexene hydrogenation
the samples in glass capillaries on a rotating probe head. Simulated
powder patterns were based on single-crystal data and calculated
using the STOE WINXPOW software package [46].
A Malvern Zetasizer Nano-ZS was used for the dynamic light
scattering measurements working at 633 nm wavelength. Care
was taken for choosing the right parameters, such as the index of
refraction of Co (2.8), Rh (1.7) and Ir (2.3) each at 0.1 absorption
at their wavelength. Samples were prepared by dilution of 50
(0.2 wt.% of M); 10 L (1 wt.% of M) of the metal/IL dispersion in
2 mL n-butyl-imidazole (99% p.a.; particle free). Hundred and fifty
microliters of this M/IL/n-butyl-imidazole solution were diluted
further with 1.5 mL n-butyl-imidazole in a glass cuvette before
measurement.
with BMim+ containing ILs
Precursor/MNP
Ionic liquid
anion
Activity mol product
(mol metal)ꢀ1 hꢀ1
Reference
[IrCl(cod)]2/Ir
[Ru(cod)(cot)]/Ru
PFꢀ6
BFꢀ4
PFꢀ6
BFꢀ4
PFꢀ6
PFꢀ6
PFꢀ6
343
100
62
388
943
147
143
[40]
[43]
RuO2/Ru
[41]
[42]
lL
l
Pd(acac)2/Pd
the data in Table 3 that the catalytic properties of MNP/ILs are
influenced by the nature of the IL anion [31]. The catalytic activity
of IrNPs from Ir4(CO)12 is remarkably higher than that of Ir nano-
particles made by hydrogen reduction of [IrCl(cod)]2 [40]. We sug-
gest that the reduction route leads to, e.g., HCl impurities in the
ionic liquid which lower its stabilization effect and result in cata-
lyst deactivation. The carbonyl decomposition route helps to avoid
any alteration of the IL phase. We note that in solvents other than
ILs, acids can have a positive effect on the catalyst activity. This
was shown, for example, in the acetone hydrogenation carried
out in neat acetone with IrNP (generated from IrCl(1,5-cod)/H2,
thereby yielding one equivalent of HCl) [62].
5.1. Metal nanoparticle syntheses
Thermal decompositions were carried out under argon in a glass
vessel which was connected to an oil bubbler. In a typical experi-
ment (see Tables 1 and 4) the metal carbonyl was dissolved/dis-
persed (during ꢁ1 h) under argon at room temperature in 3.0 g
of the ionic liquid to give a solution/suspension with a defined
weight percentage (wt.%) in metal. The solution was slowly heated
to 180 °C (Co) or 230 °C (Rh, Ir) for 18 h under stirring. After cooling
to room temperature under argon an aliquot of the ionic liquid was
collected under argon atmosphere for in situ TEM, TED and dy-
namic light scattering characterization.
4. Conclusions
We describe here a simple and reproducible metal carbonyl
thermolysis for the synthesis of stable Co, Rh and Ir-metal nano-
particles in ionic liquids, in particular in BMimþBF4ꢀ, with rather
small and uniform MNP sizes of about 1–3 nm for Rh and Ir nano-
particles. No extra stabilizers or capping molecules are needed to
achieve this small particle size. The synthesis uses easily commer-
cially available metal carbonyl precursors and can readily be ex-
panded to the broad range of other metal carbonyl complexes.
ILs are once again demonstrated to present a favorable template
for the preparation of predictable chemical nanostructures. Fur-
thermore, the obtained RhNP and IrNP/IL dispersions were em-
ployed as highly active and reusable ‘‘green catalysts” without the
need of other organic solvents in the biphasic liquid–liquid hydro-
genation of cyclohexene.
5.2. Catalysis
Catalytic reactions were carried out either in a 120 ml Büchi
glass autoclave or in a 30 ml glass inlay within a steel autoclave
at 4 bar H2 pressure (constant pressure setting for Rh and Ir, initial
pressure for Co) and 75 °C. Cylohexene was added to the metal-dis-
persions under argon. The ionic liquid containing the particles and
the substrate formed a biphasic system. The ionic liquid is the low-
er, more dense phase with the Co, Rh or Ir nanoparticles, the upper
phase is the cyclohexene/cyclohexane. During the reaction at 75 °C
a homogenous, single liquid phase was obtained. After cooling to
room temperature phase separation occured again. The cyclohex-
ene to cyclohexane conversion was analyzed by putting a drop of
the mixture into a GC sample vial with 1 ml of water. The addition
of water as a non-electrolyte can enlarge the activity coefficient of
organic components, thereby increase their detection sensitivity
through the increase in peak area. The FID does not detect the
water itself [63]. The vial was closed with aluminum crimp caps
(with butyl rubber septum), placed into the headspace sampler
(thermostated for 20 min to 50 °C) of a GC-headspace Perkin–El-
mer 8500 HSB 6, equipped with a DB-5 film capillary column
5. Experimental
Materials and Instrumentation: Co2(CO)8, Rh6(CO)16
, and
Ir4(CO)12 were obtained from Strem, ionic liquids n-butyl-
tri-methylammonium bis(trifluoromethylsulfonyl)imide (BtMAþ
NTfꢀ2 ), n-butyl-methyl-imidazolium tetrafluoroborate (BMimþ
BFꢀ4 ) and trifluoromethanesulfonate (BMim+OTfꢀ) from IoLiTec
(H2O content ꢃ 100 ppm; Clꢀ content ꢃ 50 ppm). Chloride anion
impurities could be another possible source of the stabilization of
metal nanoclusters in ionic liquids [33]. Because of the certified
small chloride content in our commercial ILs, the Cl anions will
only have a minor role, if any, for such MNP stabilization. All
manipulations were done using Schlenk techniques under argon
since the metal carbonyls are hygroscopic and air sensitive. The io-
nic liquids were dried at high vacuum (1þ0ꢀ3 mbar) for several days
to avoid especially in the case of BMim BFꢀ4 the hydrolysis to HF
[58–61].
Transmission electron microscopy (TEM) photographs were ta-
ken at room temperature from a carbon-coated copper grid on a
Zeiss LEO 912 transmission electron microscope operating at an
accelerating voltage of 120 kV.
Powder X-ray diffractograms were measured at ambient tem-
perature using a STOE STADI-P with Debye–Scherrer geometry,
Table 4
Decomposition conditions and experimental data for MNP synthesis
Metal
carbonyl
Ionic
Metal
(wt.%)
Mass metal
carbonyl (mg)
Decomposition
temperature (°C)
liquida
1a Co2(CO)8
1b Co2(CO)8
BMimþBFꢀ4
BMimþBFꢀ4
0.16
0.8
0.2
14
71
11
11
26
52
9
43
21
21
21
180
180
230
230
230
230
230
230
230
230
230
2
3
4
5
6
7
8
9
Rh6(CO)16 BMimþBF4ꢀ
Rh6(CO)16 BMim+OTfꢀꢀ 0.2
Rh6(CO)16 BtMAþNTf2
Rh6(CO)16 BMimþBF4ꢀ
Ir4(CO)12
Ir4(CO)12
Ir4(CO)12
Ir4(CO)12
0.5
1
0.2
1
BMimþBFꢀ4
BMimþBFꢀ4
BMimþBFꢀ4
0.5
BMim+OTfꢀꢀ 0.5
10 Ir4(CO)12
BtMAþNTf2
0.5
a
3.0 g ionic liquid.

E. Redel et al. / Journal of Organometallic Chemistry 694 (2009) 1069–1075
1075
[19] A. Safavi, N. Maleki, F. Tajabadi, E. Farjami, Electrochem. Commun. 9 (2007)
1963.
(60 m ꢄ 0.32 mm, film thickness 25
lm, oven temperature 60 °C,
N2 carrier flow 114 l/min) and a flame ionization detector (FID,
250 °C detector temperature). After the thermostatization time a
sample was automatically drawn from the gas phase in the vial.
The product was analyzed by the GC retention time versus authen-
tic samples of cyclohexene and cyclohexane. Hydrogenation con-
version (%) was calculated from the obtained cyclohexane-to-
cyclohexene peak area and compared to a calibration curve from
different ratios of given cyclohexene/cyclohexane mixtures which
had been measured under identical headspace conditions.
[20] K. Kim, C. Lang, P.A. Kohl, J. Electrochem. Soc. 152 (2005) E9.
[21] E. Redel, R. Thomann, C. Janiak, Chem. Commun. (2008) 1789.
[22] J. Krämer, E. Redel, R. Thomann, C. Janiak, Organometallics 27 (2008) 1976.
[23] D.O. Silva, J.D. Scholten, M.A. Gelesky, S.R. Teixeira, A.C.B. Dos Santos, E.F.
Souza-Aguiar, J. Dupont, ChemSusChem 1 (2008) 291.
[24] A. Taubert, Z. Li, Dalton Trans. (2007) 723.
[25] J. Dupont, J. Brazil. Chem. Soc. 15 (2004) 341.
[26] C.S. Consorti, P.A.Z. Suarez, R.F. de Souza, R.A. Burrow, D.H. Farrar, A.J. Lough,
W. Loh, L.H.M. da Silva, J. Dupont, J. Phys. Chem. B 109 (2005) 4341.
[27] J. Dupont, P.A.Z. Suarez, R.F. de Souza, R.A. Burrow, J.P. Kintzinger, Chem. Eur. J.
6 (2000) 2377.
[28] E.J.W. Verwey, J.T.G. Overbeek, Theory of the Stability of Lyophobic Colloids,
2nd ed., Dover Publications Mineola, New York, 1999.
Acknowledgment
[29] B.W. Ninham, Adv. Colloid Int. Sci. 83 (1999) 1.
[30] G. Schmidt, Nanoparticles, Wiley-VCH, Weinheim, 2004.
[31] D. Astruc, F. Lu, J.R. Aranzaes, Angew. Chem., Int. Ed. 44 (2005) 7852.
[32] M. Antonietti, D. Kuang, B. Smarly, Y. Zhou, Angew. Chem., Int. Ed. 43 (2004)
4988.
[33] L.S. Ott, R.G. Finke, Coord. Chem. Rev. 251 (2007) 1075.
[34] A.P. Umpierre, G. Machado, G.H. Fechner, J. Morais, J. Dupont, Adv. Synth. Catal.
347 (2005) 1404.
[35] J.D. Scholten, G. Ebeling, J. Dupont, Dalton Trans. (2007) 5554.
[36] D. Astruc, Nanoparticles and Catalysis, Wiley, New York, 2007.
[37] A. Roucoux, J. Schulz, H. Patin, Adv. Synth. Catal. 345 (2003) 222.
[38] H.-G. Elias, Makromoleküle, Band 3, Industrielle Polymere und Synthesen,
Wiley-VCH, Weinheim, 2001.
[39] B.K. Hodnett, Heterogenous Catalytic Oxidations, Wiley, New York, 2000.
[40] J. Dupont, G.S. Fonseca, A.P. Umpierre, P.F.P. Fichtner, S.R. Texeira, J. Am. Chem.
Soc. 124 (2002) 4228.
We thank Dr. T.J.S. Schubert from IoLiTec (Ionic Liquids Tech-
nologies GmbH, Denzlingen, Germany) for donation of ionic
liquids.
Appendix A. Supplementary material
Supplementary material available: Statistical graphs for dy-
namic light scattering, additional TEM and TED pictures, X-ray
powder diffractograms, photographs of MNP/IL dispersions. Sup-
plementary data associated with this article can be found, in the
online version, at doi:10.1016/j.jorganchem.2008.09.050.
[41] L.M. Rossi, G. Machado, P.F.P. Fichtner, S.R. Teixeira, J. Dupont, Catal. Lett. 92
(2004) 149.
[42] J. Huang, T. Jiang, B. Han, H. Gao, Y. Chang, G. Zhao, W. Wu, Chem. Commun.
(2003) 1654.
References
[43] E.T. Silveira, A.P. Umpierre, L.M. Rossi, G. Machado, J. Morais, G.V. Soares, I.L.R.
Baumvol, S.R. Teixeira, P.F.P. Fichtner, J. Dupont, Chem. Eur. J. 10 (2004) 3734.
[44] D.J. Adams, P.J. Dyson, S.J. Travener, Chemistry in Alternative Reaction Media,
Wiley, New York, 2004.
[45] A.F. Holleman, N. Wiberg, Lehrbuch der Anorganischen Chemie, 102nd ed.,
Walter de Gruyter, Germany, Berlin, 2007.
[1] V.I. Pârvulescu, C. Hardacre, Chem. Rev. 107 (2007) 2665.
[2] R. Elghanian, J.J. Storhoff, R.C. Mucic, R.L. Letsinger, C.A. Mirkin, Science 277
(1997) 1078.
[3] J.A. Alonso, Chem. Rev. 100 (2000) 637;
H. Haberland, Clusters of Atoms and Molecules, Springer, Berlin, 1994.
[4] P. Braunstein, J. Rosé, in: P. Braunstein, L.A. Oro, P.R. Raithby (Eds.), Metal
Clusters in Chemistry, vol. 2, Wiley-VCH, Weinheim, 2001, pp. 616–677
(Chapter 2).
[5] A.H. Lu, E.L. Salabas, F. Schüth, Angew. Chem., Int. Ed. 46 (2007) 1222.
[6] A. Gedanken, Ultrason. Sonochem. 11 (2004) 47.
[46] STOE WINXPOW version 1.10, data base, STOE
Germany, 2002.
& Cie GmbH, Darmstadt,
[47] R. Allmann, Röntgenpulverdiffraktometrie, Springer, Berlin, 2005.
[48] A.N. Shipway, E. Katz, I. Willner, ChemPhysChem 1 (2000) 18.
[49] T. Cassagneau, J.H. Fendler, J. Phys. Chem. B 103 (1999) 1789.
[50] C.D. Keating, K.K. Kovaleski, M.J. Natan, J. Phys. Chem. B 102 (1998) 9404.
[51] E. Redel, J. Krämer, R. Thomann, C. Janiak, GIT Labor-Fachzeitschrift (04) (2008)
400.
[52] T.J. Gannon, G. Law, R.P. Watson, A.J. Carmichael, K.R. Seddon, Langmuir 15
(1999) 8429.
[53] G. Law, R.P. Watson, A.J. Carmichael, K.R. Seddon, Phys. Chem. Chem. Phys. 3
(2001) 2879.
[54] J.N.C. Lopes, M.F.C. Gomes, A.A.H. Padua, J. Phys. Chem. B 110 (2006) 16816.
[55] J.N.C. Lopes, A.A.H. Padua, J. Phys. Chem. B 110 (2006) 3330.
[56] I. Krossing, J.M. Slattery, Z. Phys. Chem. 220 (2006) 134.
[57] G. Ertl, H. Knözinger, J. Weitkamp, Handbook of Heterogonous Catalysis vols.
1–8, 2nd ed., Wiley-VCH, Weinheim, Germany, 2008.
[58] F. Endres, S.Z. El Abedin, Phys. Chem. Chem. Phys. 8 (2006) 2101.
[59] P. Wasserscheid, T. Welton, Ionic Liquids in Synthesis vol. 1, 2nd ed., Wiley-
VCH, Weinheim, 2007.
[60] R.P. Swatloski, J.D. Holbrey, R.D. Rogers, Green Chem. 5 (2003) 361.
[61] G.A. Baker, S.N. Baker, Aust. J. Chem. 58 (2005) 174.
[62] S. Özkar, R.G. Finke, J. Am. Chem. Soc. 127 (2005) 4800.
[63] H. Hachenberg, K. Beringer, Die Headspace-Gaschromatographie als Analysen-
und Meßmethode, Vieweg, Braunschweig/Wiesbaden, Germany, 1996, pp. 32–
35.
[7] C.N.R. Rao, S.R.C. Vivekchand, K. Biwas, A. Govindaraj, Dalton Trans. (2007)
3728.
[8] Y. Mastai, A. Gedanken, in: C.N.R. Rao, A. Müller, A.K. Cheetham (Eds.),
Chemistry of Nanomaterials, vol. 1, Wiley-VCH, Weinheim, 2004, p. 113ff.
[9] D. Mahajan, E.T. Papish, K. Pandya, Ultrason. Sonochem. 11 (2004) 385.
[10] J. Park, J. Joo, S.G. Kwon, Y. Jang, T. Hyeon, Angew. Chem., Int. Ed. 46 (2007)
4630.
[11] E. Redel, R. Thomann, C. Janiak, Inorg. Chem. 48 (2008) 14.
[12] T. Gutel, J. Garcia-Antón, K. Pelzer, K. Philippot, C.C. Santini, L.S. Ott, R.G. Finke,
Inorg. Chem. 45 (2006) 8382.
[13] G.S. Fonseca, A.P. Umpierre, P.F.P. Fichtner, S.R. Teixeira, J. Dupont, Chem. Eur.
J. 9 (2003) 3263.
[14] Z. Li, A. Friedrich, A. Taubert, J. Mater. Chem. 18 (2008) 1008.
[15] P. Migowski, G. Machado, L.M. Rossi, G. Machado, J. Morais, S.R. Teixeira,
M.C.M. Alves, A. Traverse, J. Dupont, Phys. Chem. Chem. Phys.
4814.
9 (2007)
[16] J.M. Zhu, Y.H. Shen, A.J. Xie, L.G. Qiu, Q. Zhang, X.Y. Zhang, J. Phys. Chem. C 111
(2007) 7629.
[17] M.A. Firestone, M.L. Dietz, S. Seifert, S. Trasobares, D.J. Miller, N.J. Zaluzec,
Small 1 (2005) 754.
[18] K. Peppler, M. Polleth, S. Meiss, M. Rohnke, J.Z. Janek, Phys. Chem. 220 (2006)
1507.