Haloaurate and halopalladate imidazolium salts: Structures, properties, and use as precursors for catalytic metal nanoparticles

Article abstract of DOI:10.1039/c2dt31984e

The synthesis and characterisation of a series of new gold- and palladium-containing symmetrical imidazolium salts are described which display significant cation-dependent effects determined by the structure of the alkyl chains of the imidazolium motifs. Whereas direct reduction of the Pd salts can produce stable nanoparticles (NPs) coated by imidazolium salts, the addition of strong base to the Pd or Au salts before reduction gives stable NPs, potentially pacified by N-heterocyclic carbene units. The possibility of NP surface protection by metal-carbon bonds in these systems is investigated by spectroscopic, synthetic, and catalytic investigations, providing support for the hypothesis. Significantly, the catalytic activity of the NPs is not inhibited by the continued presence of the ligands. The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c2dt31984e

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Haloaurate and halopalladate imidazolium salts:
structures, properties, and use as precursors for
Cite this: Dalton Trans., 2013, 42, 1385
catalytic metal nanoparticles†
Christopher J. Serpell,‡a James Cookson,b Amber L. Thompson,a
Christopher M. Brownb and Paul D. Beer*a
The synthesis and characterisation of a series of new gold- and palladium-containing symmetrical imida-
zolium salts are described which display significant cation-dependent effects determined by the structure
of the alkyl chains of the imidazolium motifs. Whereas direct reduction of the Pd salts can produce stable
nanoparticles (NPs) coated by imidazolium salts, the addition of strong base to the Pd or Au salts before
reduction gives stable NPs, potentially pacified by N-heterocyclic carbene units. The possibility of NP
surface protection by metal–carbon bonds in these systems is investigated by spectroscopic, synthetic,
and catalytic investigations, providing support for the hypothesis. Significantly, the catalytic activity of the
NPs is not inhibited by the continued presence of the ligands.
Received 29th August 2012,
Accepted 22nd October 2012
DOI: 10.1039/c2dt31984e
www.rsc.org/dalton
Ionic liquids (ILs) have been intensively studied as task-
specific solvents.1–4 Since the cation and anion can be varied
at will (subject to effects upon the melting point), an enor-
mous variety of solvents may be accessed with ease. Imidazo-
lium salts in particular have risen to the fore due to their
ease of synthesis. While most ILs designed as reaction media
−
make use of inert, non-coordinating ions such as BF4 and
PF6−, there is also substantial interest in the “added value”
obtained using active species, such as metallate anions.
Metal-containing ILs5 in particular have been studied for
a variety of motivations: primarily for catalysis,6–9 but also
metal deposition,10 electrochemistry,11 magnetism,12 and
photophysics.13,14
There is also considerable interest in the catalytic appli-
cations of metal nanoparticles (NPs) in imidazolium ionic
liquids,15–17 despite the fact that the nature of the interaction
between the ionic liquid and metal NPs remains controversial.
Early studies using XPS revealed interactions between the
Fig. 1 Possible stabilisation modes for nanoparticles in ionic liquids.
aChemistry Research Laboratory, Department of Chemistry, University of Oxford,
Mansfield Road, Oxford OX1 3TA, UK. E-mail: paul.beer@chem.ox.ac.uk;
Fax: +44 (0)1865 272609; Tel: +44 (0)1865 285142
bJohnson Matthey Technology Centre, Blount’s Court, Sonning Common, Reading,
RG4 9NH, UK. E-mail: cooksj@matthey.com; Fax: +44 (0)118 924 2338;
Tel: +44 (0)118 924 2061
surface and the anions of the solvent (even weakly coordinat-
ing species). These results suggested that the anions form a
coating over the surface (often positively charged), which is
further surrounded by a layer of cations, thus providing stabi-
lisation via the creation of Coulombic repulsions between NPs.
This is known as the DLVO (Derjaguin–Landau–Verwey–Over-
beek) model18 (Fig. 1a). In 2005 Finke reported hydrogen-deu-
terium exchange at all aromatic positions of the 1-butyl-3-
methylimidazolium cation catalysed by iridium NPs, and con-
cluded that N-heterocyclic carbene (NHC)-type stabilisation
†Electronic supplementary information (ESI) available: Further TEM images,
and crystal structures in .cif format. CCDC 898614–898618. For ESI and crystallo-
graphic data in CIF or other electronic format see DOI: 10.1039/c2dt31984e
‡Current address: Department of Chemistry, McGill University, Room 400,
801 Sherbrooke St. W., Montreal, Quebec, H3A 0B8, Canada, christopher.
serpell@mcgill.ca
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 1385–1393 | 1385

View Article Online
Paper
Dalton Transactions
occurs, at least transiently (Fig. 1b).19 However, in 2007,
Crystals suitable for single crystal X-ray diffraction struc-
Dupont disclosed results of surface-enhanced Raman scatter- tural analysis§ of 3a, 3c, 4a, and 4c were obtained by slow eva-
ing (SERS) studies on Au NPs which indicated that surface poration of chloroform solutions. The bis(hexyl) salt, 4b, was
stabilisation was occurring through parallel coordination by liquid but found to crystallise slowly over the course of two
the imidazolium cation (Fig. 1c).20 Even more recently, it has years and yielded small crystals which could be studied using
been claimed that the true stabilisers are actually methyl imi- synchrotron radiation.
dazole impurities present in the ionic liquid,21 whereas on the
The crystal structure of 3a (Fig. 2) consists of alternating
other hand the direct synthesis of NHC-AuNPs has been layers of anions and cations, primarily linked by anion–π inter-
reported from molecular NHC precursors22 or via ligand dis- actions, but with some C–H⋯Cl hydrogen bonds. The two crys-
placement.23 The actual situation may well depend on the tallographically independent imidazolium species are totally
exact system in question.
disordered, each requiring modelling over two sites. This lack
Herein we describe the synthesis and characterisation of order in the solid state can be considered to be reciprocal to
of a series of Au- and Pd-containing imidazolium salts the higher order displayed by ionic liquids when compared to
and report
a
new preparative route to catalytically conventional solvents. The anion layers consist of AuCl4− units
active NPs, via reduction of the metallate imidazolium ionic in an edge-to-face arrangement, with the chloride of one
−
liquid.
approaching the metal atom of the next such that the AuCl4
units are perpendicular. This arrangement has also been
reported for similar systems.25
In 3c (Fig. 3) the cations and anions are now located in an
alternating array – a version of the sodium chloride structure
with non-isotropic, mutually canted ions. Such an arrange-
ment permits the optimisation of electrostatic forces, while
also providing room for anion–π interactions. Although it
could be suggested that this separation of the anions is
responsible for the yellow colouration of the material, and the
edge-to-face interaction in 3a causes the red colour, previously
Synthesis and characterisation of halometallate imidazolium
salts
Bis(N,N′-propyl)imidazolium bromide (2a) and bis(N,N′-hexyl)-
imidazolium bromide (2b) were prepared by sequential
alkylation (Scheme 1). Imidazole, sodium hydride and alkyl
bromides were used to give 1a and 1b; this was followed by
refluxing in acetone with the same alkyl bromide, to afford
room temperature ILs in good yields. Conversely, bis(N,N′-tert-
butyl)imidazolium chloride (2c), a waxy solid, was synthesised
via a three-component reaction,24 due to the poor reactivity of
the hindered tert-butyl bromide in the reactions described
above. Biphasic anion exchange of these imidazolium
halide salts 2a–c using either HAuCl4 or K2PdCl4 followed
by solvent evaporation gave the metallic imidazolium salts
(Scheme 1). In the case of the propyl species crystalline solids
were obtained: 3a was deep red (mp 96 °C), whereas 4a was
brown (mp 100 °C). Conversely, the di-hexyl salts manifested
initially as liquids: blood red (3b) and deep brown (4b). In
contrast to the first two chloroaurate salts, the bis(tert-butyl)
analogue 3c (crystalline solid, mp 132 °C) was bright
yellow. The chloropalladate 4c (crystalline solid, mp 204 °C)
followed the trend of the other palladium salts, being deep
brown.
−
reported AuCl4 imidazolium systems have been documented
to be yellow11,25 or orange,26 whether or not they contain this
perpendicular arrangement of anions.
The motif of alternating layers of anions and cations seen
in 3a was repeated in the crystal structure of 4a (Fig. 4),
although it revealed not the expected tetrachloropalladate
§Single crystal diffraction data were collected using a Nonius Kappa CCD or
using Beamline I1932 at Diamond Light Source and processed with DENZO-
SMN33 or CrystalClear34 including unit cell refinement and inter-frame
scaling. Structures were solved with SHELXS,35 SIR9236 or SuperFlip37 and
refined using SHELXTL35 or CRYSTALS38–40 as per the SI (CIF). Crystallographic
data (excluding structure factors) have been deposited with the Cambridge Crys-
tallographic Data Centre (CCDC 898614–898618) and can be obtained via www.
ccdc.cam.ac.uk/data _request/cif. Single crystal data 3a: C9H17AuCl4N2, Mr
492.01, 150 K, orthorhombic, Pccn, a = 19.7735(2), b = 9.3333(1), c = 17.0270(2) Å,
V = 3142.37(6) Å3, Data/restraints/parameters – 3539/54/99, Rint = 0.0242, R1
=
=
0.0550, wR2 = 0.1597 (I > 2σ(I)). 3c: C11H21AuCl4N2, Mr = 520.08, 150 K, ortho-
rhombic, Cmc21, a = 9.9772(2), b = 14.7125(3), c = 11.6742(3) Å, V = 1713.65(7) Å3,
Data/restraints/parameters – 1917/45/103, Rint = 0.060, R1 = 0.0284, wR2 = 0.0633
(I > 2σ(I)). 4a: C18H34Br4Cl2N4Pd2, Mr = 909.83, 150 K, monoclinic, P21/n, a =
7.8505(3), b = 17.5620(6), c = 10.8085(4) Å, β = 94.616(2)°, V = 1485.34(9) Å3, Data/
restraints/parameters – 3358/0/144, Rint = 0.0863, R1 = 0.0438, wR2 = 0.0772 (I >
ˉ
2σ(I)). 4b: C30H58Br1.78Cl2.22N4Pd, Mr = 802.16, 100 K, triclinic, P1, a = 7.682(4),
b = 9.105(6), c = 13.838(10) Å, α = 82.14(2)°, β = 78.97(2)°, γ = 68.59(2)°, V =
882.1(10) Å3, Data/restraints/parameters – 5425/0/186, Rint = 0.029, R1 = 0.0414,
wR2 = 0.0675 (I > 2σ(I)). These data were collected using synchrotron data and
the sample suffered radiation damage. Since the crystals took two years to grow,
the experiment could not easily be repeated; although the data are incomplete,
it is not expected to have a significant influence on the structure determination.
4c: C22H42Cl4N4Pd, Mr = 610.81, 150 K, monoclinic, C2/m, a = 17.3545(7), b =
9.4171(5), c = 10.5259(5) Å, β = 125.898(2)°, V = 1393.50(12) Å3, Data/restraints/
parameters – 1673/0/91, Rint = 0.031, R1 = 0.0344, wR2 = 0.0769 (I > 2σ(I)).
Scheme 1 Synthesis of chlorometallate imidazolium salts.
1386 | Dalton Trans., 2013, 42, 1385–1393
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Fig. 3 Two views of the crystal structure of 3c; showing pseudo-NaCl structure
packing in space-filling representation (top), and detail of anion–π interactions
(bottom). Thermal ellipsoids displayed at 50% probability.
Fig. 2 Two views of the crystal structure of 3a; showing packing in layers of
cations and anions (top), and detail of disorder, and interactions between the
ions (bottom). Thermal ellipsoids displayed at 50% probability.
the protons in the 4- and 5-positions on the imidazolium ring
and one of the halides, an effect that has been previously dis-
cussed,29,30 but rarely observed directly.
dianion, but a dinuclear bromide bridged Pd(II) species, termi-
nated with halides (chlorides and bromides occurring in a
50 : 50 ratio). The formation of this anion must have occurred
during the anion exchange process in which the labile tetra-
chloropalladate would have encountered aqueous bromide
anions. The Pd : Br ratio of 1 : 2 in the anion exchange reaction
mixture is preserved here, suggesting that the structure is
indeed representative of the bulk sample, rather than being
an example of selective crystallisation effects. Using mass
spectrometry to demonstrate whether these dimeric species
were the only ones present before crystallisation was
inconclusive.
The crystal structure of 4c (Fig. 6) also featured anion–π
interactions as a packing motif. That these occur significantly
in both of the tert-butyl imidazolium salts is probably due to
steric hindrance of the acidic protons. In this structure
2−
the PdCl4 anion is observed, without dimerisation, which is
typical for crystals with more compact cations.
Synthesis of nanoparticles via reduction of metallate ionic
liquids
The same partial halide exchange process also occurred for
the hexyl analogue, 4b (Fig. 5). Interestingly, although this sub-
Having prepared the imidazolium metallate precursors,
reductions were performed with the aim of producing imidazo-
lium and/or carbene stabilised nanoparticulate Au and Pd
materials (Scheme 2).
Firstly, direct borohydride reduction of the bis(hexyl) imida-
zolium metallate salts 3b and 4b was conducted in 1 : 1
dichloromethane–toluene in an ice bath. In the case of the Pd
salt 4b this gave stable black NPs 5b, whereas using 3b to
obtain 5a gave only agglomerated bulk metal. To ensure that
the stability of the Pd NPs compared with the Au system was
not due to the extra equivalent of stabiliser in the former case,
2−
stitution was seen in this work when attempting Br−/PdCl4
exchange, it did not occur with AuCl4−, despite having been
observed by others.27 In the case of 4b the metal anion
remained monomeric, which is surprising given that large
cations are known to promote dimerisation.28 A possible expla-
nation comes from the prevalence of hydrogen bonding in this
structure over anion–π interactions, which are complemented
by a more concentrated negative charge. Interestingly, in this
case weak chelating hydrogen bonds were observed between
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 1385–1393 | 1387

View Article Online
Paper
Dalton Transactions
Fig. 6 Two views of the crystal structure of 4c; showing packing in layers (top),
and detail of anion–π interactions (bottom). Thermal ellipsoids displayed at 50%
probability.
Fig. 4 Two views of the crystal structure of 4a; showing packing in layers of
cations and anions (top), and detail of disorder, H-bonding and anion–π inter-
actions (bottom). Thermal ellipsoids displayed at 50% probability.
Attempts were then made to promote NHC-stabilisation of
NPs by deprotonation of the imidazolium cation prior to
reduction. This was achieved by treating the imidazolium salt
with sodium hydride in 1 : 1 dichloromethane–toluene, before
adding sodium borohydride to reduce the metals. This gener-
ated stable Au and Pd NPs 6a and 6b respectively as purple
and black solutions.
UV-vis spectroscopy of Au NPs 6a revealed a strong plasmon
resonance band at 514 nm, while as expected none was
observed for the Pd NPs. Examination of the 1H NMR spec-
trum of 6a in CDCl3 (Fig. 7) revealed the disappearance of the
acidic imidazolium C2 proton resonance (a), while broadening
and splitting was observed for the other peaks (b–h), typical of
NP-coordinated ligands. 13C NMR also failed to locate a peak
for C2, which is thought to be due to anisotropy resulting from
the surface coordination. Conversely, for the Pd NPs produced
without deprotonation (5b) the C2 1H peak was observed at
8.10 ppm.
TEM analysis of the three samples (Fig. 8) showed that the
Pd NPs were small and monodisperse with sizes of 1.8 nm
( 0.4, n = 261) and 2.7 nm ( 0.9, n = 191) for 5b and 6b respect-
ively. The Au NPs 6a were significantly larger and more polydis-
perse at 16.6 nm ( 6.5, n = 133), a fact which points towards a
weaker stabilisation of the Au surface by the NHC ligands com-
pared to Pd. Indeed, the NHC capping of Pd NPs (7) also
resulted in greater NP diameters than was seen for apparent
Fig. 5 Crystal structure of 4b. Thermal ellipsoids displayed at 50% probability.
the Au reduction was attempted again in the presence of one
equivalent of bis(N,N′-hexyl)imidazolium chloride, but no
stable NPs were formed.
1388 | Dalton Trans., 2013, 42, 1385–1393
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Scheme 2 Synthesis of imidazolium and NHC stabilised NPs.
surfactant stabilisation (5b) suggesting again that while the
NPs were sufficiently stable to characterise, NHCs are compara-
tively poor pacifying agents for metal NPs.
These results are consistent with the NHC coordination
mode in 6a and 6b. The implication of this type of capping
ligands is reaffirmed by the fact that stable Au NPs could only
be generated when the imidazolium cation was deprotonated,
permitting coordination of the surface through M–C bonds.
Using the bis(propyl) or bis(tert-butyl) chlorometallate imi-
dazolium salts in either of these protocols failed to give stable
NPs. In the former case, this is probably due to the propyl
chains being insufficiently bulky to prevent close approach
and subsequent thermodynamically favourable coagulation of
NPs, whereas for the latter example, it could also be due to
additional methyl groups sterically preventing the formation
of strong C–M bonds to surfaces.
Fig. 7 1H NMR spectra of ionic liquid 3b and Au NPs 6a (CDCl3, 293 K).
Catalytic studies with palladium nanoparticles
In order to test the catalytic activity of the NP samples, and in
particular test the differences due to coordination mode, por-
tions of the Pd NPs 5b and 6b were adsorbed onto activated
carbon (Norit GSX) to give a 2.5 wt% loading with respect to
the metal. These samples were then characterised by ICP-ES
and TEM, confirming the level of metal present and checking
for any changes in structure due to the support. TEM analysis
of the samples on carbon gave new NP diameters – it was
found that the surfactant stabilised Pd NPs 5b had increased
in average diameter from 1.8 nm to 3.5 nm ( 1.4, n = 110),
while the NHC Pd NPs 6b remained almost exactly the same at
2.6 nm ( 1.2, n = 99). This suggests that the more well-defined,
bonded coordination of 6b promotes NP stability and rigidity,
while the fluid surfactant stabilisation permits partial
agglomeration.
In these cases the presence or absence of ligand could
prove vital – while surfactant stabilisation is not expected to
block the NP surface to catalytic substrates, a strong M–C NHC
bond could well impede activity. Portions of the samples were
therefore calcined in air at 450 °C to remove the ligands
Fig. 8 TEM images of 5b (left), 6a (right), and 6b (bottom).
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 1385–1393 | 1389

View Article Online
Paper
Dalton Transactions
(and synthetically facile) modifications of the imidazolium
side-chains.
Conclusions
We have reported the synthesis of a series of new gold- and
palladium-containing imidazolium salts, and obtained crystal
structures for those solid at room temperature. These struc-
tures reveal a broad spectrum of interactions which are highly
dependent upon the structure of the alkyl chains of the imida-
zolium motifs. The stability of Au and Pd NPs produced by
reduction of these species is also determined by the nature of
the alkyl side chains, as well as the metal, and the presence or
absence of base, giving surfactant or NHC stabilised NPs.
Importantly the Pd samples retain their catalytic activity for
the reduction of nitrobenzene to aniline despite the continued
presence of the ligands on the surface, a significant finding
for the future production of novel task-specific catalysts.
Fig. 9 Hydrogen uptake curves for the reduction of nitrobenzene to aniline
catalysed by Pd NPs 5b and 6b. Reaction conditions: 3 bar H2, 50 °C, 1 : 1000
catalyst to substrate molar ratio.
Table 1 Rates of hydrogen uptake in the reduction of nitrobenzene to aniline
catalysed by Pd NPs 5b and 6ba
Catalyst
Reaction rate (ml min−1
)
Experimental
Instrumentation
5b
11.7
10.9
11.0
8.3
5b (calcined)
6b
NMR spectra were recorded on a Varian Mercury VX300,
Varian Unity Plus 500, Bruker DPX200, or Bruker AVC500
spectrometer. All 13C spectra were proton decoupled. Mass
spectrometry was performed on a Bruker microTOF (ESI) or
a Waters Micromass MALDI micro MX (MALDI-TOF) mass
spectrometer. Melting points were recorded on a Gallenkamp
capillary melting point apparatus and are uncorrected. Micro-
wave reactions were carried out using a Biotage Initiator
2.0 microwave. UV-vis spectra were recorded on a PG Instru-
ments T60U spectrometer or Shimadzu UV-2401PC UV-Vis
photospectrometer. Samples were prepared in a 3 ml quartz
cuvette and concentrations were adjusted to suit the absorp-
tion range of the instrument. Thermogravimetric analysis was
performed using a PerkinElmer Diamond TG/DTA. Calcination
was performed using a Carbolite RHF 1600 furnace. ICP-ES
analyses were undertaken using a PerkinElmer Optima 3300RL
instrument. Samples were prepared by Na2O2 infusion in a
zirconium crucible and compared against standard gravimetri-
cally determined solutions. TEM was conducted using a
Tecnai F20 Transmission Electron Microscope with a voltage of
200 kV, and C2 aperture of 30 and 50 m, in bright field (BF)
and STEM (HAADF) mode, and performing in situ EDX analy-
sis. A small portion of the sample was crushed and then
dusted onto a holey carbon film on a TEM grid (Cu). Approxi-
mately 12 micrographs were recorded for each sample.
6b (calcined)
Pd/C standard
10.7
a Reaction conditions: 3 bar H2, 50 °C, 1 : 1000 catalyst to substrate
molar ratio, 5% estimated error.
(the optimal temperature was determined by thermogravi-
metric analysis). The reduction of nitrobenzene to aniline was
then performed using the prepared catalysts.
The tests revealed complete conversion of nitrobenzene to
aniline within half an hour by all the samples, as well as by a
commercial Pd/C standard (Fig. 9). The differences in the reac-
tion rates (Table 1) and hence catalytic potency of the samples
– which cannot be correlated with NP surface area – can be
used to derive information about the physicochemical environ-
ment at the catalytic sites. The imidazolium-protected NPs 5b
were the most active, which is in concord with the expected
fluidity of the ionic liquid coating permitting access of the
substrate to the catalytic surface. The carbene-protected NPs
6b were less active (on a par with the Pd/C standard), which
implies a slightly less accessible surface, though still retaining
a useful level of potency. This is consistent with the formation
of a stronger bond between the surface and the ligand. For
both 5b and 6b calcination to remove the ligands resulted in a
decrease in reaction rate, which is likely to be due to some
degree of sintering and particle growth, but more significantly
indicates that the imidazolium moieties do not inhibit the
catalytic activity of the nanoparticles in either protecting
mode, and may indeed promote it. We have previously demon-
strated the advantages of functional supramolecular ligands
for the formation of catalytically optimised metal NPs,31 and
the retained activity of the NPs discussed herein suggest that
further novel systems could be produced through judicious
Synthesis
Commercially available materials were used without any pre-
treatment unless otherwise stated. Dry solvents were obtained
by purging with N2 and then passing through an MBraun
MPSP-800 column. H2O was de-ionised and microfiltered
using a Milli-Q® Millipore machine.
1390 | Dalton Trans., 2013, 42, 1385–1393
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
1-PROPYL-1H-IMIDAZOLE (1A). Sodium hydride (60% dispersion then glyoxal (40% aq, 4.83 ml, 3.33 mmol). The reaction
in mineral oil, 0.15 g, 3.8 mmol) was suspended in 10 ml dry mixture was heated to 40 °C for 15 hours, after which the
THF under nitrogen at 0 °C. Imidazole (0.25 g, 3.67 mmol) was organic solvent was removed in vacuo. The resulting aqueous
dissolved in 5 ml dry THF and added dropwise. After gas evol- mixture was extracted with CH2Cl2 (2 × 20 ml) to remove side
ution stopped, 1-bromopropane (0.44 g, 3.67 mmol) in 5 ml products. The water was evaporated under vacuum and the
dry THF was also added dropwise. The resultant mixture was solid was extracted by trituration with CH2Cl2 (3 × 50 ml),
then stirred at room temperature overnight. After careful giving 2c as an off-white solid after solvent removal (4.44 g,
addition of a few drops of water to quench the hydride, the 100%). 1H NMR (300 MHz, CDCl3) δ (ppm) 10.65 (1H, s, ImH),
organic solvent was removed and diethyl ether (25 ml) was 7.37 (2H, s, ImH), 1.81 (18H, s, CH3); ESMS: m/z calc. for
added. This solution was washed three times with water [M − Cl]+ 181.1, found 181.1.
(25 ml), and the combined aqueous layers were re-extracted
1,3-DIPROPYL-1H-IMIDAZOL-3-IUM
TETRACHLOROAURATE
(3A). 2a
with diethyl ether (50 ml). The collected organic layers were (0.10 g, 0.42 mmol) was dissolved in CHCl3 (12 ml) and stirred
then dried over MgSO4 and filtered. Solvent removal gave 1a vigorously with HAuCl4 (0.15 g, 0.42 mmol) dissolved in water
1
as a pale yellow oil (0.40 g, 100%). H NMR (300 MHz, CDCl3) (12 ml). After approximately two hours the yellow colouration
δ (ppm) 7.45 (1H, s, ImH), 7.04 (1H, s, ImH), 6.89 (1H, s, ImH), of the aqueous layer had disappeared, to be replaced by a deep
3.88 (2H, t, 3J = 7.0 Hz, CH2), 1.79 (2H, tq, 3J = 7.0, 7.3 Hz, red colour in organic portion. The solvent was removed from
CH2), 0.91 (3H, t, 3J = 7.3 Hz, CH3); ESMS: m/z calc. for the organic layer, giving 3a as an orange-red crystalline solid
[M + H]+ 111.09, found 111.09.
(0.12 g, 84%). 1H NMR (300 MHz, acetone-d6) δ (ppm) 9.16
3
1-HEXYL-1H-IMIDAZOLE (1B). Prepared as for 1b using sodium (1H, s, ImH), 7.84 (2H, s, ImH), 4.37 (4H, t, J = 7.2 Hz, CH2),
hydride (60% dispersion in mineral oil, 3.15 g, 78.8 mmol), 1.99 (4H, tq, 3J = 7.2, 7.3 Hz, CH2), 0.97 (6H, t, 3J = 7.3 Hz,
imidazole (4.47 g, 65.7 mmol) and 1-bromohexane (10.85 g, CH3); 13C NMR (125 MHz, 10 : 1 CDCl3–CD3OD) δ 135.5, 123.1,
65.7 mmol) giving the product as a yellow oil (9.74 g, 97%). 52.4, 24.0, 11.2; HR-ESMS: m/z calc. for [M − AuCl4]+ 153.1392,
1H NMR (300 MHz, CDCl3) δ (ppm) 7.47 (1H, s, ImH), 7.06 found 153.1390; ESMS: m/z calc. for [M − Pr2Im]− 336.84,
(1H, s, ImH), 6.91 (1H, s, ImrH), 3.93 (2H, t, 3J = 7.2 Hz, found 336.82; mp 96 °C.
NCH2), 1.77, (2H, m, NCH2CH2), 1.29 (6H, m, –(CH2)3–), 0.88
(6H, t, 3J = 6.7 Hz, CH2CH3); ESMS: m/z calc. for [M + H]+ 3a, using 2b (0.15 g, 0.47 mmol) and HAuCl4 (0.16 g,
153.13, found 153.12.
0.47 mmol), giving 3b as a blood red oil (0.23 g, 85%). 1H
1,3-DIHEXYL-1H-IMIDAZOL-3-IUM TETRACHLOROAURATE (3B). As for
1,3-DIPROPYL-1H-IMIDAZOL-3-IUM BROMIDE (2A). 1a (2.00 g, NMR (300 MHz, CDCl3) δ (ppm) 9.22 (1H, s, ImH), 7.35 (2H, s,
18 mmol) and 1-bromopropane (5.50 g, 45 mmol) were ImH), 4.30 (4H, t, 3J = 7.5 Hz, CH2), 1.95 (4H, m, CH2), 1.36
refluxed in acetone (250 ml) overnight. The solvent was (12H, m, CH2) 0.90 (6H, t, 3J = 7.0 Hz, CH3); 13C NMR (75 MHz,
removed, and the resultant oil was washed with diethyl ether CDCl3) δ 134.9, 122.5, 50.7, 31.0, 30.1, 25.9, 22.4, 13.9; ESMS:
(3 × 25 ml) to extract any starting materials, leaving 2a as a m/z calc. for [M − AuCl4]+ 237.23, found 237.22, calc. for
pure yellow oil (2.07 g, 49%). 1H NMR (300 MHz, CDCl3) [M − Hex2Im]− 338.84, found 338.85.
3
δ (ppm) 10.74 (1H, s, ImH), 7.32 (2H, s, ImH), 4.35 (4H, t, J =
1,3-DI-TERT-BUTYL-1H-IMIDAZOL-3-IUM
TETRACHLOROAURATE
(3C).
3
7.4 Hz, CH2), 1.99 (4H, tq, J = 7.3, 7.4 Hz, CH2), 1.00 (6H, t, chloride 2c (0.50 g, 2.3 mmol) was dissolved in CHCl3 (25 ml)
3J = 7.4 Hz, CH3); ESMS: m/z calc. for [M − Br]+ 153.13, found and HAuCl4 (0.78 g, 2.3 mmol) dissolved in water (25 ml) was
153.10.
added, forming an emulsion. Both solvents were removed
1,3-DIHEXYL-1H-IMIDAZOL-3-IUM BROMIDE (2B). 1b (0.34 g, in vacuo, and the solid was extracted by trituration with CH2Cl2
1.88 mmol) and 1-bromohexane (0.37 g, 2.26 mmol) were (3 × 50 ml). Evaporation of the combined organic phases gave
heated to 150 °C in 3 ml acetone in a sealed vial under micro- 3c as a yellow crystalline solid (0.95 g, 79%). 1H NMR
wave irradiation for ten minutes. The solvent was removed and (300 MHz, CDCl3) δ (ppm) 9.23 (1H, s, ImH), 7.46 (2H, s, ImH),
the resultant oil stirred extracted with diethyl ether (3 × 20 ml) 1.79 (18H, s, CH3); 13C NMR (125 MHz, 9 : 1 CDCl3–CD3OD)
to remove excess 1-bromohexane. The remaining yellow oil δ 130.7, 120.1, 60.4, 29.5; ESMS: m/z calc. for [M − AuCl4]+
constituted the pure bromide salt (0.50 g, 80%). 1H NMR 181.1, found 181.1; mp 132 °C.
(300 MHz, CDCl3) δ (ppm) 10.36 (1H, s, ImH), 7.45 (2H, s,
1,3-DIPROPYL-1H-IMIDAZOL-3-IUM HALOPALLADATE (4A). 2a (0.20 g,
3
ImH), 4.27 (4H, t, J = 7.6 Hz, NCH2), 1.83, 4H, m, NCH2CH2), 0.84 mmol) was dissolved in CHCl3 (12 ml) and stirred
3
1.22 (12H, m, –(CH2)3–), 0.77 (6H, t, J = 6.5 Hz, CH2CH3); 13C vigorously with K2PdCl4 (0.13 g, 0.42 mmol) dissolved in water
NMR (75 MHz, CDCl3) δ 137.0, 121.9, 49.8, 30.9, 25.6, 22.2, (12 ml). After four hours the colour had transferred from
13.7; ESMS: m/z calc. for [M − Br]+ 237.23, found 237.21.
the aqueous layer to the organic portion, although some oiling
1,3-DI-TERT-BUTYL-1H-IMIDAZOL-3-IUM CHLORIDE (2C).24 Paraformal- out was observed, which was countered by addition of CH2Cl2
dehyde (1.00 g, 3.33 mmol) was suspended in toluene (20 ml), (5 ml). The layers were separated, and solvent removed
and tert-butylamine (2.44 g, 3.33 mmol) was added dropwise. from the organic fraction, giving 4a as a dark brown crystalline
1
The mixture was then heated with stirring until a clear solu- solid (0.22 g, 49%). H NMR (300 MHz, CD3OD) δ (ppm) 9.21
3
tion was obtained. The solution was cooled in an ice bath, and (1H, s, ImH), 7.77 (2H, s, ImH), 4.34 (4H, t, J = 7.2 Hz, CH2),
further tert-butylamine (2.44 g, 3.33 mmol) was added drop- 2.02 (4H, tq, 3J = 7.2, 7.3 Hz, CH2), 1.02 (6H, t, 3J = 7.3 Hz,
wise, followed by conc. HCl (aq) (3.28 ml, 3.33 mmol), and CH3); 13C NMR (75 MHz, CD3OD3) δ 128.3, 124.1, 52.8, 25.2,
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 1385–1393 | 1391

View Article Online
Paper
Dalton Transactions
11.2; ESMS: m/z calc. for [M − PdnXm]+ 153.13, found 153.10; making use of the ICP-ES data for bulk composition. The reac-
mp 100 °C. tions were run at 50 °C under 3 bar of H2. The composition of
1,3-DIHEXYL-1H-IMIDAZOL-3-IUM TETRAHALOPALLADATE (4B). As for 3a the organic phase was analysed by gas chromatography using
using 2b (0.70 g, 2.2 mmol) and K2PdCl4 (0.32 g, 1.1 mmol) a PerkinElmer Autosample XL Gas chromatograph fitted with
giving 4b as a deep brown oil, which crystallised over the a 30 m PE-5MS column. Quantification was achieved by com-
1
course of two years (0.62 g, 87%). H NMR (300 MHz, CDCl3) parison to known standard solutions. The GC was connected
3
δ (ppm) 10.08 (1H, s, ImH), 7.42 (2H, s, ImH), 4.55 (4H, t, J = to a PerkinElmer TurboMass mass spectrometer to allow for
7.5 Hz, CH2), 1.98 (4H, m, CH2), 1.32 (12H, m, CH2) 0.87 determination of unknown peaks.
(6H, t, 3J = 7.2 Hz, CH3); 13C NMR (75 MHz, CDCl3) δ 136.8,
122.1, 50.7, 31.1, 30.3, 26.0, 22.4, 14.0; ESMS: m/z calc. for
[M − PdnXm]+ 237.23, found 237.22; mp 34 °C.
1,3-DI-TERT-BUTYL-1H-IMIDAZOL-3-IUM TETRACHLOROPALLADATE (4C).
Acknowledgements
C. J. S. thanks the EPSRC and Johnson Matthey for a CASE stu-
dentship and the EPSRC for a PhD+ award. We are grateful to
Diamond Light Source for the award of beamtime on I19
(MT1858) and to the beamline scientists for help and support.
1,3-Di-tert-butyl-1H-imidazol-3-ium chloride 2c (1.00 g,
4.6 mmol) and K2PdCl4 (0.75 g, 2.3 mmol) were stirred for two
hours in water (50 ml). The water was removed in vacuo, and
CH2Cl2 (50 ml) was used to dissolve the residue, which was
then filtered through Celite. Rotary evaporation of the solvent
gave 4c as a dark brown crystalline solid (1.00 g, 36%).
1H NMR (300 MHz, CDCl3) δ (ppm) 9.84 (1H, s, ImH), 7.51
(2H, s, ImH), 1.81 (18H, s, CH3); 13C NMR (125 MHz, 9 : 1
CDCl3–CD3OD) δ 131.4, 120.2, 60.2, 29.4; ESMS: m/z calc. for
[M − PdCl4]+ 181.1, found 181.1; mp 204 °C.
PD NPS 5A. Bis-hexylimidazolium halopalladate 4b (0.84 g,
1.16 mmol) was dissolved in 1 : 1 dry CH2Cl2–toluene (50 ml),
and cooled in an ice bath. NaBH4 (0.22 g, 5.80 mmol) in water
(5 ml) was added dropwise, with vigorous stirring, and the solu-
tion was further stirred at room temperature for two hours.
The reaction mixture was washed with water (3 × 25 ml), fil-
tered and the solvent removed in vacuo giving a brown-black
oil (0.51 g). TEM mean diameter 1.8 nm ( 0.4, n = 261).
Notes and references
1 J. Dupont, Acc. Chem. Res., 2011, 44, 1223–1231.
2 R. Giernoth, Angew. Chem., Int. Ed., 2010, 49, 2834–2839.
3 D. Freudenmann, S. Wolf, M. Wolff and C. Feldmann,
Angew. Chem., Int. Ed., 2011, 50, 11050–11060.
4 S. Tang, G. A. Baker and H. Zhao, Chem. Soc. Rev., 2012, 41,
4030–4066.
5 I. J. B. Lin and C. S. Vasam, J. Organomet. Chem., 2005, 690,
3498–3512.
6 K. Bica and P. Gaertner, Eur. J. Org. Chem., 2008, 2008,
3453–3456.
AU NPS 6A. Sodium hydride (60% dispersion in mineral oil,
0.041 g, 1.04 mmol) was suspended in 1 : 1 dry CH2Cl2–
toluene (40 ml) cooled in an ice bath. Bis-hexylimidazolium
tetrachloroaurate 3b (0.40 g, 0.69 mmol) dissolved in dry
CH2Cl2 (10 ml) was added dropwise, and the reaction mixture
was stirred for 20 min, exhibiting a colour change from red to
yellow. NaBH4 (0.13 g, 3.47 mmol) in water (5 ml) was adminis-
7 J. E. L. Dullius, P. A. Z. Suarez, S. Einloft, R. F. de Souza,
J. Dupont, J. Fischer and A. De Cian, Organometallics, 1998,
17, 815–819.
8 M. H. Valkenberg, C. deCastro and W. F. Hölderich, Appl.
Catal., A, 2001, 215, 185–190.
9 J. E. Dengler, A. Doroodian and B. Rieger, J. Organomet.
Chem., 2011, 696, 3831–3835.
tered dropwise, maintaining the temperature of the reaction 10 N. R. Brooks, S. Schaltin, K. Van Hecke, L. Van Meervelt,
using the ice bath. The reaction was stirred at room temp-
erature for two hours, after which it was washed with water
K. Binnemans and J. Fransaer, Chem.–Eur. J., 2011, 17,
5054–5059.
(3 × 50 ml), the aqueous portions being discarded. After 11 B. Ballarin, L. Busetto, M. Cristina Cassani and C. Femoni,
filtration, rotary evaporation gave a 6a as dark purple oil Inorg. Chim. Acta, 2010, 363, 2055–2064.
(0.21 g). UV-vis (toluene) PRB 514 nm; TEM mean diameter 12 S. Hayashi, S. Saha and H. Hamaguchi, Magn. IEEE Trans.,
16.6 nm ( 6.5, n = 133). 2006, 42, 12–14.
PD NPS 6B. The reaction was performed in the same manner 13 P. Nockemann, B. Thijs, N. Postelmans, K. Van Hecke,
as the preparation of 6a, using NaH (60% dispersion in
mineral oil, 0.12 g, 2.90 mmol), 4b (0.70 g, 0.97 mmol), and
L. Van Meervelt and K. Binnemans, J. Am. Chem. Soc., 2006,
128, 13658–13659.
NaBH4 (0.18 g, 4.84 mmol), giving a black oil (0.35 g). TEM 14 Y. Yoshida, J. Fujii, G. Saito, T. Hiramatsu and N. Sato,
mean diameter 2.7 nm ( 0.9, n = 191).
J. Mater. Chem., 2006, 16.
15 P. Migowski and J. Dupont, Chem.–Eur. J., 2007, 13, 32–39.
16 C. W. Scheeren, G. Machado, J. Dupont, P. F. P. Fichtner
and S. r. R. Texeira, Inorg. Chem., 2003, 42, 4738–4742.
Catalytic tests
Catalytic hydrogenations were performed in a HEL ChemSCAN
reactor. Within each reaction cell 5.0 ml of a 0.5 mol dm−3 17 J. Huang, T. Jiang, B. Han, H. Gao, Y. Chang, G. Zhao and
solution of the substrate in ethanol with 1,4-dioxane (internal
W. Wu, Chem. Commun., 2003, 1654–1655.
standard for GC) was used. The carbon-supported catalyst was 18 S. Ãzkar and R. G. Finke, J. Am. Chem. Soc., 2002, 124,
added in the ratio 1 : 1000 molar ratio of Pd to substrate,
5796–5810.
1392 | Dalton Trans., 2013, 42, 1385–1393
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
19 L. S. Ott, M. L. Cline, M. Deetlefs, K. R. Seddon and 30 G. T. Spence, C. J. Serpell, J. Sardinha, P. J. Costa, V. Félix
R. G. Finke, J. Am. Chem. Soc., 2005, 127, 5758–5759. and P. D. Beer, Chem.–Eur. J., 2011, 17, 12955–12966.
20 H. S. Schrekker, M. A. Gelesky, M. P. Stracke, 31 C. J. Serpell, J. Cookson, D. Ozkaya and P. D. Beer, Nat.
C. M. L. Schrekker, G. Machado, S. R. Teixeira, J. C. Rubim
and J. Dupont, J. Colloid Interface Sci., 2007, 316, 189–195.
21 P. Dash and R. W. J. Scott, Chem. Commun., 2009, 812–814.
22 J. Vignolle and T. D. Tilley, Chem. Commun., 2009.
23 E. C. Hurst, K. Wilson, I. J. S. Fairlamb and V. Chechik,
New J. Chem., 2009, 33, 1837–1840.
Chem., 2011, 3, 478–483.
32 H. Nowell, S. A. Barnett, K. E. Christensen, S. J. Teat and
D. R. Allan, J. Synch. Rad., 2012, 19, 435–441.
33 Z. Otwinowski and W. Minor, Processing of X-ray Diffraction
Data Collected in Oscillation Mode, Academic Press, 1997.
34 CrystalClear-SM Expert 2.0 r4, Rigaku, 2009.
24 Y. Chu, H. Deng and J.-P. Cheng, J. Org. Chem., 2007, 72, 35 G. M. Sheldrick, Acta Crystallogr., Sect. A: Fundam. Crystal-
7790–7793. logr., 2008, 64, 600–601.
25 M. Hasan, I. V. Kozhevnikov, M. R. H. Siddiqui, A. Steiner 36 G. C. A. Altomare, C. Giacovazzo, A. Guagliardi,
and N. Winterton, Inorg. Chem., 1999, 38, 5637–5641.
26 E. R. Schreiter, J. E. Stevens, M. F. Ortwerth and
R. G. Freeman, Inorg. Chem., 1999, 38, 3935–3937.
27 M. Deetlefs, H. G. Raubenheimer and M. W. Esterhuysen,
Catal. Today, 2002, 72, 29–41.
M. C. Burla, G. Polidori and M. Camalli, J. Appl. Crystal-
logr., 1994, 27, 435.
37 L. Palatinus and G. Chapuis, J. Appl. Crystallogr., 1997, 40,
786–790.
38 P. W. Betteridge, J. R. Carruthers, R. I. Cooper, K. Prout
and D. J. Watkin, J. Appl. Crystallogr., 2003, 36, 1487.
28 F. A. Cotton, G. Wilkinson, C. A. Murillo and
M. Bochmann, in Advanced Inorganic Chemistry, Wiley- 39 R. I. Cooper, A. L. Thompson and D. J. Watkin, J. Appl.
Interscience, 1999, p. 1071. Crystallogr., 2010, 43, 1100–1107.
29 C. J. Serpell, N. L. Kilah, P. J. Costa, V. Félix and P. D. Beer, 40 A. L. Thompson and D. J. Watkin, J. Appl. Crystallogr., 2011,
Angew. Chem., Int. Ed., 2010, 49, 5322–5326.
44, 1017–1022.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 1385–1393 | 1393