A cytotoxic bis(carbene)gold(I) complex of ferrocenyl complexes: Synthesis and structural characterisation

Article abstract of DOI:10.1039/b713917a

The N-heterocyclic carbene (NHC) precursors 1-[(E)-2-butenyl]-3-(4- ferrocenylphenyl)imidazolium bromide (2) and 1-[(E)-2-butenyl]-3-(4- ferrocenylphenyl)imidazolium tetrafluoroborate (3) were derived from 1-(4-ferrocenylphenyl)imidazole. Ferrocenyl complex 3 reacts with Ag ₂O and chloro(dimethylsulfide)gold(i) in the presence of tetraethylammonium chloride to produce the mixed metal species bis{1-[(E)-2-butenyl]-3-(4-ferrocenylphenyl)-2H-imidazol-2-ylidene}gold(i) tetrafluoroborate (4). Single crystal X-ray structure analyses of 1, 3 and 4 indicate that the NCHN-hydrogen in 3 is hydrogen bonded to the BF ₄^- anion [C(H1)...F, 3.265(4) A], as is also reflected in the position of its ¹H NMR chemical shift. Cytotoxicity studies show that complex 4 is selective for cancer cells and active against the tumour cell lines Jurkat and MCF 7. The Royal Society of Chemistry and the Centre National de la Recherche Scientifique.

Full text of DOI:10.1039/b713917a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/njc | New Journal of Chemistry
A cytotoxic bis(carbene)gold(I) complex of ferrocenyl complexes:
synthesis and structural characterisationw
Ulrike E. I. Horvath,*a Gino Bentivoglio,b Michael Hummel,b
Herwig Schottenberger,b Klaus Wurst,b Margo J. Nell,c
Constance E. J. van Rensburg,c Stephanie Cronjea and Helgard G. Raubenheimer*a
Received (in Durham, UK) 11th September 2007, Accepted 19th November 2007
First published as an Advance Article on the web 12th December 2007
DOI: 10.1039/b713917a
The N-heterocyclic carbene (NHC) precursors 1-[(E)-2-butenyl]-3-(4-ferrocenylphenyl)imidazolium
bromide (2) and 1-[(E)-2-butenyl]-3-(4-ferrocenylphenyl)imidazolium tetrafluoroborate (3) were
derived from 1-(4-ferrocenylphenyl)imidazole. Ferrocenyl complex 3 reacts with Ag2O and
chloro(dimethylsulfide)gold(I) in the presence of tetraethylammonium chloride to produce the
mixed metal species bis{1-[(E)-2-butenyl]-3-(4-ferrocenylphenyl)-2H-imidazol-2-ylidene}gold(I)
tetrafluoroborate (4). Single crystal X-ray structure analyses of 1, 3 and 4 indicate that the
ꢀ
NCHN-hydrogen in 3 is hydrogen bonded to the BF4 anion [C(H1)ꢁ ꢁ ꢁF, 3.265(4) A], as is also
1
reflected in the position of its H NMR chemical shift. Cytotoxicity studies show that complex 4
is selective for cancer cells and active against the tumour cell lines Jurkat and MCF 7.
Our interest in investigating the possible uses of NHC
Introduction
ligands in the synthesis of new biologically-active gold(I)
compounds as alternatives to gold(I) phosphines was sparked
by recent articles by Baker et al.13 and Barnard et al.10 which
involved the synthesis of gold(I) complexes of NHCs as
potential anti-mitochondrial and thus anti-tumour agents.
Previously all gold(I) compounds with anti-tumour activity
contained phosphine ligands14 and although gold(I) phosphine
compounds are known for their anti-tumour potential it is also
known that the phosphine ligands on their own show anti-
tumour activity which increases when complexed to gold.
Although initial studies of the mechanism of anti-cancer
activity in gold complexes pointed towards DNA as the target,
subsequent studies of gold(I) carbene complexes with potential
anti-tumor activity showed that they rather target mitochon-
drial membranes. Similarly, Au(III) cyclometalated complexes
do not cross-link DNA like cisplatin, indicating a different
mechanism of action. The analogy of d8 square planar Au(III)
and Pt(II) complexes does not extend to their interaction with
cells.15 The influence of the phosphine ligand on the activity of
a gold(I) complex remains unexplored.
Three of the main approaches for the preparation of gold
carbene complexes include (i) nucleophilic addition of alcohols
or amines to gold-coordinated isocyanides,1 (ii) transfer from
Fischer-type alkoxy-organo or amino-organo carbene com-
plexes of group 6 metals to form gold(I) and gold(III) com-
plexes,2,3 and (iii) deprotonation or activation of
N-heterocyclic compounds (NHC) before coordination to a
gold center.4
The activation strategies used for the preparation of
NHC–metal complexes can be divided into (i) cleavage of
the CQC bond of electron-rich alkenes,5 (ii) generation of a
free carbene by deprotonation of the corresponding imidazo-
lium precursor with a strong base (for example with BuLi,
NaH or t-BuOK),6 (iii) transmetallation of a deprotonated
azole followed by protonation or alkylation of the gold azolyl
compounds7 [lithiated imidazoles mainly form bis(carbene)
complexes with gold(I)],8 (iv) in situ-deprotonation of an
imidazolium salt with a weak base (for example NEt3 or
OAcꢀ)9,10 and (v) transmetallation from a silver–NHC com-
plex prepared by the direct reaction of an imidazolium
precursor with Ag2O.11,12
A number of studies of gold(I) complexes linked to ferrocene
derivatives, with the common denominator being that they
contain gold(I) phosphines or P-substituted ferrocene ligands,
underline the fact that such compounds exhibit significant
cytotoxic effects.16–20 Further examples of phosphine-free,
ferrocene-containing complexes were described by Gimeno
et al.,21 who studied the gold(I) complexes [Fe(C5H4S2-
CNEt2)2(AuCl)2]22 and [AuX(pzCH2Fc)] (X = Cl or C6F5;
pzCH2Fc = ferrocenylmethylpyrazole),23 and one gold(III)
complex: [Au(C6F5)3(pzCH2Fc)].23
a Department of Chemistry and Polymer Science, University of
Stellenbosch, Private Bag X1, Matieland, 7602, South Africa.
E-mail: hgr@sun.ac.za. E-mail: uhorvath@gmail.com; Fax: +27 21
808 3849; Tel: +27 21 808 3850
b Institut fur Allgemeine, Anorganische und Theoretische Chemie,
¨
Leopold Franzens Universitat, Innrain 52a, A-6020 Innsbruck,
¨
Austria
c Department of Pharmacology, Faculty of Medicine, University of
Pretoria, P.O. Box 2034, Pretoria, 0001, South Africa
w Electronic supplementary information (ESI) available: Alternative
preparation method for compound 1 and IR (KBr) results for 2–4. See
DOI: 10.1039/b713917a
One of the most general strategies for introducing a
large diversity of arene-spacered functionalities onto ferrocene
is the Gomberg–Bachmann–Hey arylation, as previously
described for carboxylic acids,24–26 carboxylic acid esters,27
ꢂc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2008
New J. Chem., 2008, 32, 533–539 | 533

View Article Online
8.08 in 1 to d 11.16 in the cation 2, and then again upfield to d
9.21 in 3, when only the counterion is changed. A similar
shielding of the NCHN proton by the close proximity of the
anion of 1-[prop-2-ynyl]-3-vinyl-1H-imidazolium tetraphenyl-
borate to this proton was recently investigated by NMR
studies and a crystal structure determination.37 The crystal
structure determination of 3ꢀshows hydrogen bonding between
H1 and an F atom on BF4 (vide infra). The assignments of
the signals observed for 4 in CD2Cl2 were unambiguously
confirmed by gHSQC and 1H,13C gHMQC measurements. No
NCHN signal is present, indicating successful deprotonation
of the azole unit with Ag2O, and a sharp singlet at d 182.5 for
the carbene carbon confirms that metallation has occurred.
Similar chemical shifts have been reported for related NHC
complexes of gold.10 The signals for the NCHQCHN protons
1
appear in the H NMR spectra between d 7.62 and 7.14 for
1–4. The same carbon atoms resonate at d 132.3 and 129.9 in
the 13C NMR spectrum of 1, and are shielded by more than Dd
8 in the spectra of 2–4, when compared to 1.
Scheme 1 Preparation of compounds 1–4: (a) NaNO2, HBF4; (b)
ferrocene; (c) 1-bromo-2-butene; (d) NH4BF4; (e) Ag2O, [NEt4]Cl; (f)
Me2SAuCl.
The signals of the unsubstituted cyclopentadienyl rings in
1
2–4 are essentially unchanged at ca. d 4.06 in their H NMR
spectra and at ca. d 70 in their 13C NMR spectra. The protons
of the monosubstituted cyclopentadienyl ring appear at d 4.81
and 4.38 in 1, and at ca. d 4.7 and 4.4 in 2–4 as an unsymme-
trical pair of triplets, indicating an A2B2 system with 3J = 4J,
where protons H2A and H5A are deshielded by the aromatic
phenyl ring bearing the ferrocenyl substituent. The cyclopen-
tadienyl carbon connected to the phenyl substituent is signifi-
cantly deshielded at ca. d 83, while the other carbons in the
same ring are observed at ca. d 67. The butenyl CH2 protons
(H12) and carbons (C12) of 2 and 3 have the expected
chemical shifts.
nitrophenyls,28 halophenyls,24–25,29 mercaptophenyls30 and
terminal phenylacetylenes.31 The same concept can also be
applied to the conjugative attachment of imidazoles in order to
access electrophoric NHC complexes, which have received
attention as redox responsive model systems.32–36 Our main
target, 4, is a phosphine-free bis(carbene) complex of gold,
carrying an electrophoric, cytotoxic ferrocene moiety. A pre-
liminary study regarding the cytotoxic activity of 4 has been
undertaken, and the characterisation of the two precursors, 1
and 3, are also reported here.
Infrared spectroscopy
The IR spectra of all new compounds have very similar
ꢀ
Results and discussion
patterns, and the presence of the BF4 anion in 3 and 4 is
confirmed by very intense bands between 1032 and 1107 cmꢀ1
A Gomberg arylation, depicted in Scheme 1, using diazotation
and coupling with ferrocene afforded precursor 1. Alterna-
tively, an equimolar reaction of 4-(1H-imidazolyl)phenyldia-
zonium bis(hexafluorophosphate) with ferrocene, followed by
treatment with NaHCO3, also yielded 1.w Subsequent alkyla-
tion with 1-bromo-2-butene in CH2Cl2 yielded the hygroscopic
NHC precursor 1-[(E)-2-butenyl]-3-(4-ferrocenylphenyl)imi-
dazolium bromide (2), and after anion metathesis in acetone,
.
Mass spectrometry
The molecular ion of 1 (EI-MS) and the cations of 2–4 (FAB-
MS) were present in mass spectra of the compounds. The most
characteristic peak in the mass spectrum of 4 corresponds to
the [M ꢀ BF4]+ fragment (m/z = 961), and another peak
representing the [cation-Fc]+ fragment (Fc = ferrocenyl)
(m/z = 777) is also observed.
1-[(E)-2-butenyl]-3-(4-ferrocenylphenyl)imidazolium
tetra-
fluoroborate (3) was obtained. The bis(carbene) gold(I) com-
plex 4 was generated by reaction of 3 with Ag2O and
subsequent carbene ligand transfer to (CH3)2SAuCl. Attempts
to synthesise the bis(carbene) according to alternative method
(iii), mentioned above, failed. Carbene complex 4 is stable in
air and decomposes only slowly in solution when exposed
to air.
Crystallography
Compounds 1, 3 and 4 furnished crystals suitable for X-ray
diffraction studies. Table 1 summarises selected bond lengths
and angles. All the dihedral angles between the planes in 1–4
are reported in Table 2. The molecular structure of 1 is shown
in Fig. 1.z It crystallises in the monoclinic space group P21/c
with one molecule in the unit cell and consists of a ferrocenyl
group appended to a phenyl-imidazolium moiety in the para-
position. The Z5-C5H4 cyclopentadienyl ring of the ferrocene
moiety is influenced by the phenyl-imidazolium substituent,
Nuclear magnetic resonance
1
The H NMR spectra of the new precursor compounds 1–3
and the complex product 4 are very similar to those of the
corresponding free organic ligand. The most active proton in
the azole ring, H1 (Fig. 1), is extremely sensitive towards
changes in the ring, and its chemical shift changes from d
z For crystallographic data in CIF or other electronic format see DOI:
10.1039/b713917a
ꢂc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2008
534 | New J. Chem., 2008, 32, 533–539

View Article Online
Table 1 Selected bond angles (1) and distances (A) for compounds 1,
3 and 4
1
3
4
C1–N2
N2–C3
C3–C4
C4–N5
N5–C1
C9–C1A
N5–C6
N2–C12
Au1–C1
C1–Au1–C10
N2–C1–N5
1.330(2)
1.381(2)
1.330(2)
1.3703(2)
1.371(2)
1.473(2)
1.425(2)
1.327(4)
1.376(4)
1.343(4)
1.390(4)
1.332(4)
1.477(4)
1.446(4)
1.484(4)
1.33(1)
1.39(1)
1.35(1)
1.39(1)
1.337(9)
1.48(2)
1.42(1)
1.45(1)
2.021(8)
179.997(1)
106.2(7)
109.4(2)
108.4(3)
Fig. 2 Packing diagram of 1 in the unit cell, shown along the c-axis;
hydrogen atoms are omitted for clarity.
Table 2 Dihedral angles (1) between planes for compounds 1, 3 and 4
ꢀ
is practically unchanged (difference only 1.21). The BF4
counterion is sandwiched between butenyl chains of neigh-
1
3
4
A and B
B and C
A and C
12.6(1)
24.34(9)
13.3(1)
13.82(4)
20.35(6)
6.53(7)
2.3(7)
50.2(3)
52.3(4)
bouring molecules. Lattice organisation is governed by hydro-
ꢀ
gen bonding of the BF4 counterion with the H atoms of
molecules in different layers. Such contacts, listed in Table 3,
involve the heterocyclic rings (C1–H1), the C6H4 group
(C7–H7), the butenyl group (C12–H12B), as well as the
Planes defined by atoms: A: C1A–C2A–C3A–C4A–C5A; B:
C6–C7–C8–C9–C10–C11; C: C1–N2–C3–C4–N5.
NCHQC(H)N fragment (C4–H4) and
(C11–H11) of neighbouring molecules.
a
C6H4 unit
leading to
a
distorted angle of 106.69(13)1 for
C2A–C1A–C5A. The phenyl ring is twisted 12.6(1)1 away
from the least-square plane of the cyclopentadienyl ring
defined by C1A–C2A–C3A–C4A–C5A and 24.34(9)1 from
the azole unit defined by C1–N2–C3–C4–N5. Furthermore,
the plane of the azole unit is rotated by 13.3(1)1 with respect to
the Z5-C5H4 plane. The heterocyclic group experiences a 1 : 1
occupation disorder at two positions, N2QC3b and
C3QN2b. The molecular packing of 1 along the crystallo-
graphic c-axis is shown in Fig. 2. In broad descriptive terms,
when viewed down the c-axis, the molecules in the unit cell are
arranged in such a way that every second molecule is turned by
901. This leads to an infinite chain, with the different layers
packed in opposite directions and no intermolecular bonding
between adjacent chains.
The molecular structure of 4, shown in Fig. 4, features
crystallographic inversion symmetry at the gold centre.z Com-
pound 4 crystallises in the trigonal space group R-3. The
asymmetric unit contains the gold complex, a highly disor-
ꢀ
dered BF4 anion and a solvent molecule, CH2Cl2. The
structure reveals the linear coordination of the central gold
atom to the two carbene carbons, with a C1–Au–C10 angle of
1801. The Au–Ccarbene bond lengths of 2.021(8) A are similar
to reported values in other essentially comparable compounds,
such as bis[m-[2,6-pyridinediylbis[methylene(3-methyl-1H-imi-
dazol-1-yl-2(3H)-ylidene)]]]digold dibromide [Au–Ccarbene
2.024(8) and 2.017(8) A].10
=
As in 3, both cyclopentadienyl rings are almost parallel,
with a torsion angle of 2.91 for C1A–center–center–C6A. The
ferrocene unit is not changed by carbene formation. However,
large differences are found for the dihedral angles between the
planes of the different connected rings. The Z5-C5H4 ring and
the phenyl ring are now nearly parallel, but the azole unit,
defined by C1–N2–C3–C4–N5, is twisted by 52.3(4)1 from the
Z5-C5H4 ring. This large tilt results from thꢀe linear orientation
of the C–Au–C arrangement. The BF4 counterions are
Crystals of 3 crystallise in the monoclinic space group P21/n,
and Fig. 3 illustrates this alkylated product of 1 in the form of
its BF4ꢀ salt.z The two cyclopentadienyl rings are now planar
and parallel, with a torsion angle of 0.21 for C1A–center–
center–C6A. The whole ferrocene unit remains unaffected by
the alkylation. The angle between the plane of the Z5-C5H4
ring and the plane of the phenyl moiety, when compared to 1,
Fig. 3 The molecular structure of 3, showing the C–Hꢁ ꢁ ꢁF hydrogen
interaction. The thermal ellipsoids are scaled to the 50% probability
level.
Fig. 1 Molecular structure of 1 showing the numbering scheme. The
thermal ellipsoids are depicted at 50% probability.
ꢂc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2008
New J. Chem., 2008, 32, 533–539 | 535

View Article Online
Table 3 C–Hꢁ ꢁ ꢁF hydrogen geometry (A, 1) in compound 3
Table 4 Cytotoxicity expressed as IC50 values (mM)a
o(C–Hꢁ ꢁ ꢁF) d(C–H)
d(Hꢁ ꢁ ꢁF) d(Cꢁ ꢁ ꢁF)
HeLa
CoLo
Jurcat
C–Hꢁ ꢁ ꢁF
C1–H1ꢁ ꢁ ꢁF1a
C1–H1ꢁ ꢁ ꢁF2a
C4–H4ꢁ ꢁ ꢁF3b
C7–H7ꢁ ꢁ ꢁF2a
C11–H11ꢁ ꢁ ꢁF3b
C12–H12Bꢁ ꢁ ꢁF1a
136
164
170
139
152
154
0.95
0.95
0.95
0.95
0.95
0.99
2.51
2.32
2.17
2.52
2.45
2.38
3.265(4)
3.247(4)
3.110(4)
3.297(4)
3.314(4)
3.297(4)
4
Cisplatin
0.572 ꢃ 0.027
1.007 ꢃ 0.081
0.407 ꢃ 0.043
0.253 ꢃ 0.031
0.783 ꢃ 0.054
0.638
a
Tumour specificity of 4 = 6.976.
tive to compound 4 than the cancer cells, indicating selectivity
for the cancer cells. The tumour specificity was calculated as
the average of the IC50 values for primary cultures divided by
the average of the IC50 values for the cancer cells.
Symmetry codes:a ꢀ x + 1, ꢀ y + 1, ꢀ z + 1. x + 21, ꢀy + 12, z ꢀ .
b
1
2
Conclusions
We have synthesised, in a series of steps, four new complexes
containing ferrocenyl ligands, two of them NHC precursors.
The gold(I)bis(carbene) complex was finally prepared by the
transfer of the carbene via the corresponding silver(I) carbene
complex, representing the first heterobimetallic gold(I) NHC
complex bearing a conjugatively-attached electrophoric ferro-
cenyl moiety. As a result of linear gold(I)bis(carbene) forma-
tion, large differences in the dihedral angles between the planes
of the NHC ligand were observed. Hydrogen bonding
(C–Hꢁ ꢁ ꢁF), involving the BF4ꢀ, counterion determines the
crystal lattice organisation in 3 and 4. The phosphine-free
gold(I) complex 4 is tumor specific and achieves growth
inhibition at lower concentrations than the well-known cispla-
tin for two of the three tested cancer cell lines. It is possible
that the anti-tumour activity of the gold(I)carbene complex is
enhanced by the presence of the ferrocene fragment, and this
possibility should be investigated in more detail.
Fig. 4 The molecular structure of 4 showing the numbering scheme.
1
2
2
The symmetry code for atom C10 is (ꢀx + , ꢀy + , ꢀz + ). Only
3
3
3
ꢀ
crystallographically unique atoms are labelled; the BF4 anion and a
CH2Cl2 molecule are omitted for clarity. The thermal ellipsoids are
depicted at 40% probability.
involved in hydrogen bonding and the boron atoms are
located on a 3-fold rotation axis (site occupancy factor 1/3)
with well ordered positions of the fluorine atoms and on the
center of a 3-fold rotoinversion axis (site occupation 1/6)
leading to positional disorder of fluorine atoms. The lattice
organisation is dominated by C–Hꢁ ꢁ ꢁF interactions between
the anion, the complex itself and the CH2Cl2 solvent
molecule, with relevant separations ranging from 3.030(19)
to 3.326(12) A (Table 5).
Experimental
Methods
Reactions were carried out under argon using standard
Schlenk and vacuum line techniques. All solvents were dried
according to standard literature methods under a nitrogen
atmosphere.39 Melting points were determined on a Stuart
Scientific melting point apparatus SMP3 and are uncorrected.
Elemental analyses were carried out by the Department of
Chemistry, University of Witwatersrand, South Africa. Mass
spectra were recorded on a AMD 604 high resolution mass
spectrometer (EI-MS, 70 eV) or on a VG 70SEQ (FAB, 70 eV)
instrument (nitrobenzyl alcohol matrices). The infrared spec-
tra were recorded on a Nicolet Avatar 330 FT-IR with an
ATR (attenuated total reflection) accessory (Smart Performer)
or on a Nicolet 5700 FT-IR with an ATR diamond window.
The 1H and 13C NMR spectra were recorded on a Varian 300
FT or a Varian Unity INOVA 400 MHz spectrometer (d
reported relative to the solvent resonance). Chloro(dimethyl-
sulfide)gold(I)40 was prepared via the addition of dimethylsul-
fide to a tetrachloroauric acid41 solution. All the other starting
materials were commercially available and were used without
further purification.
Cytotoxicity studies
Cytotoxicity assays were performed to establish the sensitivity
of cancer cell lines and normal cell cultures to compound 4,
and are given in Table 4. Complex 4 was screened against the
human HeLa cervix epithelioid carcinoma cell line, the CoLo
320 DM, a human colon adenocarcinoma line, the Jurkat
leukaemia cell line and the MCF-7 breast cancer cell line. A
known concentration of cells was exposed to different con-
centrations of complex 4 on a 96-well tissue culture plate and
incubated for a period of time. Drug-free solvent controls were
included. Also reported are the IC50 data determined under
the same conditions for cisplatin, a well-established anti-
cancer drug frequently used as a standard in drug performance
comparisons. The viability of cells was determined by the
MTT method, as described by van Rensburg et al.38 Jurkat
cells were the most sensitive, while the CoLo cells were the
least sensitive to the carbene complex. Excluding CoLo, the
derived values for the IC50, representing the mean concentra-
tion (mM) of experimental compound 4, requiring a 50% cell
growth inhibition to be achieved, were considerably lower
than those for cisplatin. Human lymphocytes were less sensi-
1-(4-Ferrocenylphenyl)imidazole (1). A solution of sodium
nitrite (0.44 g, 6.3 mmol) in H2O (10 cm3) was added dropwise
ꢂc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2008
536 | New J. Chem., 2008, 32, 533–539

View Article Online
to an ice cold solution of 4-(1H-imidazol-1-yl)aniline (1.0 g,
6.3 mmol) and aqueous tetrafluoroboric acid (50%) (2.0 cm3,
0.031 mol) in H2O (20 cm3), whereupon a yellowish precipitate
was observed. After complete addition of the sodium nitrite
solution, the mixture was stirred for another 5 min, followed
by the dropwise addition of a solution of ferrocene (1.18 g, 6.3
mmol) in a mixture of CH3CN and CH2Cl2 (1 : 2) (15 cm3),
resulting in a green-brown suspension. The cooling bath was
removed and the mixture stirred for 1 h. Subsequently, H2O
(200 cm3) was added and the reaction mixture extracted with
CH2Cl2 (70 cm3). The red organic layer was washed with a
diluted Na2CO3 solution, causing a deepening of the red
colour. After drying with Na2SO4, the solvent was removed
in vacuo, resulting in a brown residue, which was washed with
n-hexane (3 ꢄ 30 cm3) to remove residual ferrocene. The
residue was then dissolved in CH2Cl2 (2 cm3) and chromato-
graphed on a column charged with silica gel 100 (70 g), using
diethyl ether as the eluent. The first yellow fraction was
discarded, and the second dark orange fraction was collected
and dried in vacuo, to yield compound 1 (0.45 g, 22%) as a red-
brown, microcrystalline material.w Mp 160–162 1C. Found C,
69.26; H, 4.77; N, 8.14. C19H16FeN2 requires C, 69.53; H, 4.91;
N, 8.54%. nmax (ATR)/cmꢀ1: 3118w, 3095w, 1599w, 1530s,
1508vs, 1480s, 1299s, 1257m, 1102m, 1056s, 1032m, 1004m,
962m, 900m, 838m, 808vs, 739vs, 691m, 658vs, 530s, 500s,
479s and 454s. dH (300 MHz, acetone-d6): 8.08 (1H, s, H1),
7.73 (2H, m, C6H4), 7.58 (3H, m, C6H4, H4), 7.14 (1H, s, H3),
4.81 (2H, t, J3 = J4 = 1.9 Hz, Z5-C5H4), 4.38 (2H, t, J3 = J4
= 1.9 Hz, Z5-C5H4) and 4.06 (5H, s, Z5-C5H5). dC (75 MHz,
acetone-d6): 139.8 (C6), 136.5 (C7, C11), 133.8 (C1), 132.3
(C4), 129.9 (C3), 128.3 (C9), 121.9 (C8, C10), 85.1 (Z5-
C5H4-ipso), 70.5 (Z5-C5H5) and 67.4 (Z5-C5H4). m/z (EI-
MS): 328 (M+, 100%), 262 (M+ ꢀ imd [Himd = imidazole],
17%) and 207 (M+ ꢀ Z5 ꢀ C5H5Fe, 12%).
1-[(E)-2-Butenyl]-3-(4-ferrocenylphenyl)imidazolium tetra-
fluoroborate (3). Compound 2 (0.22 g, 0.47 mmol) was reacted
with ammonium tetrafluoroborate (74 mg, 0.71 mmol) in
acetone (30 cm3). The mixture was stirred for 24 h at room
temperature. After filtration through MgSO4 and solvent
removal, the residue was dried in vacuo to yield compound 3
(0.21 g, 99%) as an orange, microcrystalline material. Mp
82 1C (dec). Found C, 58.85; H, 4.81; N, 5.69. C23H23BF4FeN2
requires 58.76; H, 4.93; N, 5.96%. nmax (ATR)/cmꢀ1: 3171w,
3111vw, 2941vw, 1556m, 1531w, 1454w, 1105m, 1092m,
1049vs, 1030vs, 1011s, 982s, 887m, 825s, 811s, 734m and
690m.w dH (400 MHz, CD2Cl2): 9.21 (1H, s, H1), 7.67 (2H,
d, J3 8.2 Hz, H8, H10), 7.62 (1H, bs, H4), 7.52 (2H, m, H7,
H11), 7.47 (1H, m, H3), 6.10 (1H, m, H13), 5.75 (1H, m, H14),
4.96 (2H, dd, J3 = 7.3 Hz, J4 = 6.7 Hz, H12), 4.73 (2H, m, Z5-
C5H4), 4.42 (2H, m, Z5-C5H4), 4.07 (5H, s, Z5-C5H5) and 1.84
(3H, m, H15). dC (101 MHz, CD2Cl2): 143.3 (C6), 136.5 (C7,
C11), 134.3 (C1), 132.1 (C13), 127.8 (C9), 122.9 (C14), 122.5
(C4), 122.4 (C8, C10), 121.7 (C3), 82.9 (C5H4-ipso), 70.3 (Z5-
C5H5), 67.2 (Z5-C5H4), 49.7 (C12) and 17.9 (C15). m/z (FAB):
383 (M+ ꢀ BF4, 80%), 328 (M+ ꢀ BF4 ꢀ C4H7, 15%), 262
(M+ ꢀ BF4 ꢀ imd, 8%) and 199 (M+ ꢀ BF4 ꢀ Fc, 19%).
Bis{1-[(E)-2-butenyl]-3-(4-ferrocenylphenyl)-2H-imidazol-2-
ylidene}gold(I) tetrafluoroborate (4). Complex 3 (0.11 g, 0.24
mmol), tetraethylammonium chloride (40 mg, 0.24 mmol) and
Ag2O (61 mg, 0.26 mmol) were reacted in a mixture of CH2Cl2
and MeOH (1 : 1) (30 cm3). The mixture was stirred for 5 h at
room temperature, followed by the addition of chloro(di-
methylsulfide)gold(I) (42 mg, 0.14 mmol) and further stirring
for 4 h at room temperature. After filtration through Celite
and solvent removal, the residue was treated with CH2Cl2 (20
cm3), and the extract filtered through MgSO4/Celite. The
solution was reduced to dryness in vacuo yielding complex 4
(0.17 g, 68%) as an orange, microcrystalline material. Mp
58–65 1C (dec). Found C, 52.54; H, 4.33; N, 5.23.
C46H44AuBF4Fe2N4 requires C, 52.70; H, 4.23; N, 5.34%.
nmax (ATR)/cmꢀ1: 3136vw, 2964vw, 1527m, 1458w, 1423w,
1410w, 1259w, 1188w, 1095m, 1080s, 1047vs, 1030vs, 968m,
887m, 818s, 798s, 743m and 694m.w dH (400 MHz; CD2Cl2):
7.54 (4H, m, H8, H10), 7.47 (4H, m, H7, H11), 7.34 (2H, m, J3
= 1.9 Hz, H4), 7.22 (2H, m, J3 = 1.9 Hz, H3), 5.80 (2H, m,
H13), 5.63 (2H, m, H14), 4.77 (4H, m, H12), 4.69 (4H, m, Z5-
C5H4), 4.43 (4H, m, Z5-C5H4), 4.06 (10H, s, Z5-C5H5) and 1.74
(6H, m, H15). dC(101 MHz, CD2Cl2): 182.5 (CQAu), 141.7
(C6), 136.9 (C7, C11), 132.8 (C13), 127.0 (C9), 125.1 (C8, C10),
125.0 (C14), 122.7 (C4), 121.9 (C3), 83.4 (C5H4-ipso), 70.1
(Z5-C5H5), 67.1 (Z5-C5H4), 48.3 (C12) and 17.9 (C15). m/z
(FAB): 961 (M+ ꢀ BF4, 18%) and 777 (M+ ꢀ BF4 ꢀ Fc, 5%).
1-[(E)-2-Butenyl]-3-(4-ferrocenylphenyl)imidazolium bromide
(2). Complex 1 (0.26 g, 0.79 mmol) was reacted with 1-bromo-
2-butene (85%) (0.44 cm3, 4.3 mmol) in CH2Cl2 (25 cm3). The
mixture was stirred for 41 h at room temperature. After
solvent removal, the mixture was treated with Et2O (1 ꢄ 20
cm3), filtered, washed with Et2O (3 ꢄ 30 cm3) and dried in
vacuo to yield compound 2 (0.24 g, 66%) as an orange,
microcrystalline material. Mp 75 1C (dec.). nmax (ATR)/
cmꢀ1: 3088w, 3064w, 2918w, 2856w, 1606w, 1549s, 1525m,
1435m, 1410m, 1371w, 1282w, 1198s, 1105s, 1066s, 972s,
887m, 843s, 824vs, 808s, 737s and 689s.w dH (400 MHz,
CD2Cl2): 11.16 (1H, s, H1), 7.70 (4H, m, C6H4), 7.61 (1H,
m, H4), 7.42 (1H, m, H3), 6.09 (1H, m, H13), 5.76 (1H, m,
H14), 5.22 (2H, dd, J3 = 7.6 Hz, J4 = 7.0 Hz, H12), 4.72 (2H,
t, J3 = J4 = 1.9 Hz, Z5-C5H4), 4.41 (2H, t, J3 = J4 = 1.9 Hz,
Z5-C5H4), 4.06 (5H, s, Z5-C5H5) and 1.85 (3H, dd, J3 = 7.1
Hz, J4 = 6.5 Hz, H15). dC (101 MHz, CD2Cl2): 142.9 (C6),
135.9 (C7, C11), 133.9 (C1), 132.3 (C13), 127.8 (C9), 123.0
(C14), 122.2 (C4), 122.0 (C8, C10), 120.4 (C3), 82.9 (Z5-
C5H4-ipso), 70.1 (m, Z5-C5H5), 67.1 (Z5-C5H4), 46.8 (C12)
and 17.9 (C13). m/z (FAB): 383 (M+ ꢀ Br, 71%), 328
(M+ ꢀ Br ꢀ C4H7, 12%), 262 (M+ ꢀ Br ꢀ imd, 5%) and
199 (M+ ꢀ Br ꢀ Fc [Fc = ferrocene], 11%).
Crystal structure determination
Single crystals of 1 suitable for X-ray analysis, were obtained
by slow evaporation from a CH2Cl2 solution at room tem-
perature, and for 3 and 4 by slow crystallisation from CH2Cl2
solutions layered with n-hexane at ꢀ20 1C. Data and para-
meters of the crystal structure determinations are shown in
Table 5.z For compound 1, a Nonius Kappa CCD diffracto-
meter equipped with graphite-monochromated Mo-Ka radia-
tion (l = 0.71073 A) and a nominal crystal to area detector
ꢂc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2008
New J. Chem., 2008, 32, 533–539 | 537

View Article Online
Table 5 Crystal data and structure refinements for 1, 3 and 4
Compound
1
3
4
Empirical formula
Formula weight
Crystal habit
Crystal colour
Crystal size/mm
Crystal system
Space group
a/A
C19H16FeN2
328.19
Prism
Orange
C23H23BF4FeN2
470.09
Cubes
Red
C48H48AuBCl4F4Fe2N4
703.42
Prism
Orange
0.35 ꢄ 0.2 ꢄ 0.2
0.25 ꢄ 0.23 ꢄ 0.20
0.20 ꢄ 0.20 ꢄ 0.15
Monoclinic
P21/c
Monoclinic
P21/n
Trigonal
R-3
15.0730(4)
7.7163(2)
12.8686(2)
90
9.891(2)
18.733(3)
11.461(2)
90
22.070(2)
22.070(2)
25.746(6)
90
b/A
c/A
a/1
b/1
104.784(2)
90
1447.17(6)
4
1.506
99.699(3)
90
2093.3(6)
4
1.492
90
120
10 860(2)
9
1.676
g/1
V/A3
Z
Dcalc/mg cmꢀ3
Adsorption coefficient/mmꢀ1
F(000)
1.039
680
0.768
968
3.900
5436
y (min ꢀ max)/1
Index ranges, hkl
Temperature/K
Reflections collected
Independent reflections
Data/restraints/parameters
Final R indices [I 4 2s(I)]a
R indices (all data)a
Goodness-of-fit on F2
Largest peak/e Aꢀ3
CCDC number
a
2.80–26.00
ꢀ18–0, ꢀ9–9, ꢀ15–15
2.11–25.68
ꢀ12–10, ꢀ19–22, ꢀ13–11
2.93–25.69
ꢀ23–26, ꢀ23–26, ꢀ28–31
233(2)
9126
100(2)
11 371
100(2)
16 730
2837 [Rint = 0.0182]
2837/0/199
R1 = 0.0234, wR2 = 0.0627
R1 = 0.0263, wR2 = 0.0646
1.033
3958 [Rint = 0.0505]
3958/0/281
R1 = 0.0494, wR2 = 0.1075
R1 = 0.0790, wR2 = 0.1193
1.039
4472 [Rint = 0.0654]
4472/0/298
R1 = 0.0521, wR2 = 0.1261
R1 = 0.0929, wR2 = 0.1411
0.922
2.508
668308
2
0.222
668306
0.534
668307
The function minimized was S[w(Fo ꢀ Fc2)2], with the weight defined as wꢀ1 = [s2(Fo2) + (xP)2 + yP] and P = (Fo + 2Fc)/3.
2
distance of 36 mm was used. Intensities were integrated using
DENZO and scaled with SCALEPACK.42 Several scans in f
and o directions were made to increase the number of re-
dundant reflections, which were averaged in the refinement
cycles. This procedure replaces, in a good approximation, an
empirical absorption correction. For compounds 3 and 4, a
Bruker SMART Apex diffractometer43–46 with graphite
monochromated Mo-Ka radiation (l = 0.71073 A) was used
for data collection. Intensities were measured using the o-scan
mode and corrected for Lorentz and polarisation effects. All
structures were solved by direct methods (1 and 4) or Patter-
son synthesis (3) and refined by full matrix least-squares on F2
using the SHELXL-97 program package and X-Seed.47–50 All
non-hydrogen atoms were refined with anisotropic displace-
ment parameters. All hydrogen atoms were placed in calcu-
lated positions and refined using a riding model. Figures were
created using POV-RAY 3.5.
3 R. Aumann and E. O. Fischer, Chem. Ber., 1981, 114,
1853.
4 W. A. Herrmann, Angew. Chem., Int. Ed., 2002, 41, 1291.
5 B. Cetinkaya, P. Dixneuf and M. F. Lappert, J. Chem. Soc., Dalton
Trans., 1974, 1827.
6 C. Bohler, D. Stein, N. Donati and H. Grutzmacher, New J.
¨
Chem., 2002, 26, 1291.
¨
7 H. G. Raubenheimer and S. Cronje, J. Organomet. Chem., 2001,
617–618, 170.
8 H. G. Raubenheimer, L. Lindeque and S. Cronje, J. Organomet.
Chem., 1996, 511, 177.
9 M. Poyatos, M. Sanau and E. Peris, Inorg. Chem., 2003, 42,
2572.
10 P. J. Barnard, M. V. Baker, S. J. Berners-Price, B. W. Skelton and
A. H. White, Dalton Trans., 2004, 1038.
11 H. M. J. Wang and I. L. B. Lin, Organometallics, 1998, 17,
972.
12 H. M. J. Wang, C. S. Vasam, T. Y. R. Tsai, S.-H. Chen, A. H. H.
Chang and I. J. B. Lin, Organometallics, 2005, 24, 486.
13 M. V. Baker, P. J. Barnard, S. J. Berners-Price, S. K. Brayshaw, J.
L. Hickey, B. W. Skelton and A. H. White, Dalton Trans., 2006,
3708.
14 E. R. T. Tiekink, Crit. Rev. Oncol. Hematol., 2002, 42, 225.
15 S. P. Fricker, Dalton Trans., 2007, 4903.
16 D. T. Hill, G. R. Girard, F. L. McCabe, R. K. Johnson, P. D.
Stupik, J. H. Zhang, W. M. Reiff and D. S. Eggleston, Inorg.
Chem., 1989, 28, 3529.
17 A. Houlton, R. M. G. Roberts, J. Silver and R. V. Parish, J.
Organomet. Chem., 1991, 418, 269.
18 M. Viotte, B. Gautheron, M. M. Kubicki, Y. Mugnier and R. V.
Parish, Inorg. Chem., 1995, 34, 3465.
Acknowledgements
We thank the Claude Harris Leon Foundation for financial
support (UEIH). The group at the University of Pretoria is
grateful for grants from the Cancer Association of South
Africa (CANSA) and the National Research Foundation,
South Africa (NRF).
19 M. Viotte, B. Gautheron, I. E. Nifant’ev and L. G. Kuz’mina,
Inorg. Chim. Acta, 1996, 253, 71.
20 M. Viotte, B. Gautheron, M. M. Kubicki, I. E. Nifant’ev and S. P.
Fricker, Met.-Based Drugs, 1995, 2, 311.
References
21 M. C. Gimeno, A. Laguna, C. Sarroca and P. G. Jones, Inorg.
Chem., 1993, 32, 5926.
22 M. C. Gimeno and A. Laguna, Gold Bull., 1999, 32, 90.
1 F. Bonati and G. Minghetti, Synth. React. Inorg. Met.-Org. Chem.,
1971, 1, 299.
2 G. Minghetti and F. Bonati, J. Organomet. Chem., 1974, 73, C43.
ꢂc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2008
538 | New J. Chem., 2008, 32, 533–539

View Article Online
23 E. M. Barranco, M. C. Gimeno, A. Laguna and M. D. Villacampa,
Inorg. Chim. Acta, 2005, 358, 4177.
24 G. D. Broadhead and P. L. Pauson, J. Chem. Soc., 1955,
367.
25 G. Huang, B. Li, W. Liu, L. Shi and Y. Ma, J. Chem. Res., Synop.,
2000, 491 and references cited therein.
26 K.-Q. Zhao, P. Hu and H.-B. Xu, Molecules, 2001, 6, M246
(http://www.mdpi.org/molbank/m0246.htm).
27 D. Savage, G. Malone, S. R. Alley, J. F. Gallagher, A. Goel, P. N.
37 H. Schottenberger, K. Wurst, U. E. I. Horvath, S. Cronje, J.
Lukasser, J. Polin, J. M. McKenzie and H. G. Raubenheimer,
Dalton Trans., 2003, 4275.
38 C. E. J. van Rensburg, R. Anderson, G. Joone, M. S. Myer and J.
F. O’Sullivan, Anti-Cancer Drugs, 1997, 8, 708.
39 R. J. Errington, Advanced Practical Inorganic and Metalorganic
Chemistry, Chapman & Hall, London, 1997, p. 296.
40 M.-C. Brandys, M. C. Jennings and R. J. Puddephatt, J. Chem.
Soc., Dalton Trans., 2000, 24, 4601.
Kelly, H. Muller-Bunz and P. T. M. Kenny, J. Organomet. Chem.,
¨
2006, 691, 463.
41 A. Haas, J. Helmbrecht and U. Niemann, in Handbuch der
Praparativen Anorganischen Chemie, ed. G. Brauer, Ferdinand
¨
28 P. Hu, K.-Q. Zhao and H.-B. Xu, Molecules, 2001, 6, M249
(http://www.mdpi.org/molbank/m0249.htm).
29 A. Ambroise, R. W. Wagner, P. Dharma Rao, J. A. Riggs, P.
Hascoat, J. R. Diers, J. Seth, R. K. Lammi, D. F. Bocian, D.
Holten and J. S. Lindsey, Chem. Mater., 2001, 13, 1023.
30 D. Li, Surf. Sci., 2003, 522, 105.
31 H. Schottenberger, J. Lukassser, E. Reichel, A. G. Muller, G.
¨
Steiner, H. Kopacka, K. Wurst, K. H. Ongania and K. Kirchner,
J. Organomet. Chem., 2001, 637–639, 558.
Enke Verlag, Stuttgart, 1978, p. 1014.
42 Z. Otwinowski and W. Minor, Methods Enzymol., 1997,
307.
43 SMART data collection software (version 5.629), Bruker AXS Inc.,
Madison, WI, USA, 2003.
44 SAINT data reduction software (version 6.45), Bruker AXS Inc.,
Madison, WI, USA, 2003.
45 R. H. Blessing, Acta Crystallogr., Sect. A: Found. Crystallogr.,
1995, 51, 33.
32 V. N. Babin, Y. A. Belousov, V. V. Gumenyuk, R. B. Materikova,
R. M. Salimov and N. S. Kochetkova, J. Organomet. Chem., 1981,
214, C13.
33 V. I. Boev and A. V. Dombrovskii, J. Organomet. Chem., 2001,
637–639, 558.
46 SADABS (version 2.05), Bruker AXS Inc., Madison, WI, USA,
2002.
47 G. M. Sheldrick, Program package SHELXTL V.5.1,
Bruker Analytical X-Ray Instruments Inc., Madison, WI, USA,
1997.
34 H. Seo, B. Y. Kim, J. H. Lee, H.-J. Park, S. U. Son and Y. K.
Chung, Organometallics, 2003, 22, 4783.
35 A. Bertogg, F. Camponovo and A. Togni, Eur. J. Inorg. Chem.,
2005, 2, 347.
36 Y. Yuan, G. Raabe and C. Bolm, J. Organomet. Chem., 2005, 690,
5747 and references cited therein.
48 G. M. Sheldrick, SHELXS-97, Program for the solution of crystal
structures, University of Gottingen, Germany, 1997; G. M. Shel-
¨
drick, SHELXL-97, Program for the refinement of crystal struc-
tures, University of Gottingen, Germany, 1997.
¨
49 L. J. Barbour, J. Supramol. Chem., 2001, 1, 189.
50 J. L. Atwood and L. J. Barbour, Cryst. Growth Des., 2003, 3, 3.
ꢂc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2008
New J. Chem., 2008, 32, 533–539 | 539