Neutral mononuclear and dinuclear complexes of gold(I) featuring azole ligands: Synthesis, structure and cytotoxicity

Article abstract of DOI:10.1016/j.poly.2011.12.026

The aim of this work was to extend the small library of existing compounds of gold(I) that contain pentafluorophenyl and thiazole ligands and to determine the biological activity of such compounds against a well-known cancer cell line. Seven mono and dinuclear new compounds that contain imidazole, thiazole and tetrazole rings were isolated and characterised. Three of the crystal structures determined indicated Au...Au interactions between constituent complex molecules. A simple adduct, [Au(C₆F₅)(N-methyl-imidazole)] , exhibits a relatively high anti-tumour specificity indicating the viability of selected phosphine-free gold complexes as cytotoxic agents.

Full text of DOI:10.1016/j.poly.2011.12.026

Polyhedron 34 (2012) 188–197
Contents lists available at SciVerse ScienceDirect
Polyhedron
journal homepage: www.elsevier.com/locate/poly
Neutral mononuclear and dinuclear complexes of gold(I) featuring azole ligands:
Synthesis, structure and cytotoxicity
William F. Gabrielli a, Stefan D. Nogai a, Margo Nell b, Stephanie Cronje a,c, Helgard G. Raubenheimer a,
⇑
a Department of Chemistry and Polymer Science, University of Stellenbosch, Private Bag X1, Matieland, 7602 Stellenbosch, South Africa
b Department of Pharmacology, Prinshof Campus, University of Pretoria, P.O. Box 2034, Pretoria 0001, South Africa
c Institut für Anorganische und Analytische Chemie, Goethe-Universität Frankfurt, Max-von-Laue-Strasse 7, D-60348 Frankfurt am Main, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
The aim of this work was to extend the small library of existing compounds of gold(I) that contain pen-
tafluorophenyl and thiazole ligands and to determine the biological activity of such compounds against a
well-known cancer cell line. Seven mono and dinuclear new compounds that contain imidazole, thiazole
and tetrazole rings were isolated and characterised. Three of the crystal structures determined indicated
Au. . .Au interactions between constituent complex molecules. A simple adduct, [Au(C6F5)(N-methyl-
imidazole)], exhibits a relatively high anti-tumour specificity indicating the viability of selected phos-
phine-free gold complexes as cytotoxic agents.
Received 19 November 2011
Accepted 24 December 2011
Available online 4 January 2012
Keywords:
Azole
Gold complexes
Pentafluorophenyl
Cytotoxicity
HeLa cells
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction
recently unambiguously characterised neutral, linear, N-bonded
gold(I) complexes of two substituted thiazole ligands with penta-
The number of gold(I) complexes that contain nitrogen donors
have steadily grown in recent years [1]. This class of complexes
was recently expanded by description of the synthesis and charac-
terisation of neutral azol-2-ylideneamine complexes containing
C6F5, PPh3 and (1,3-di-tert-butylimidazol-2-ylidene) as ancillary li-
gands [2]. A large number of neutral complexes based on azoles
contain azolate ligands and they have been discussed in a previous
article [3]. Phosphine complexes of gold(I) linked to neutral azoles
are mostly of the cationic form [AuL(PPh3)]+ (L = N-bonded azole)
[4,5], although dicationic complexes with bidentate phosphine li-
fluorophenyl as ancillary ligand.
The anti-proliferative, i.e. anti-viral, anti-bacterial, anti-fungal,
anti-microbial, anti-cancer, anti-inflammatory and anti-infective,
activity of gold(I) compounds and their clinical use in the treat-
ment of severe of rheumatoid arthritis has provided a powerful
incentive for the continued pursuit of the bonding of biologically
active molecules to gold(I) and the investigation of their biological
effects [11–13]. The coordination of biologically active molecules
to metal centres has afforded compounds with amplified activity
or therapeutic effect [14].
The development and investigation of the mechanisms of cyto-
toxicity and anti-tumour activity of gold(I) phosphine and carbene
compounds has provided a new momentum in the search for new
generations of chemotherapeutic agents [15,16].
Combination of nucleoside analogues with gold(I) might be
beneficial in eliminating resistance and lowering toxicity. Azole-
platinum compounds with comparable in vitro anticancer activity
to cisplatin have shown improved cytotoxic activity in cell lines
resistant to cisplatin [17]. A phosphine-free ‘complex of a complex’
consisting of a ferrocenyl moiety conjugatively part of an NHC li-
gand which is bound to gold(I), is tumour specific against the HeLa
and Jurkat cancer cell lines. The question regarding possible antitu-
mour activity enhancement by the presence of the ferrocenyl
group remains unresolved, and needs to be further addressed
[18]. Exocyclic imine complexation of azol-2-ylideneamine ligands
with (Ph3P)Au+ increases their antitumour as well as antimalarial
activity [2].
gands, [Au2L2(
l
-P^P)]2+ (P^P = ddpm, dppe, dppp) [6,7], are also
known. Leonesi et al. [8] have prepared the first neutral azole com-
plexes of gold(I) by substituting Me2S from [AuCl(Me2S)] with
imidazole and substituted imidazoles. No definite structure have
been assigned to the insoluble products but gold coordination
numbers varying between two and three are proposed for the
intermolecularly-associated compounds.
A number of neutral
azole complexes with empirical formula [AuX(L)] (L = substituted
´
imidazole or oxazole; X = Cl or Br), have been prepared by Matovic
et al. [9] who, on the basis of vibrational spectra analyses, propose
rather unlikely non-linear, halogen-bridged, dimeric structures in
the solid state, which have not been confirmed by crystal and
molecular structure determination. Cronje et al. [10], have more
⇑
Corresponding author.
E-mail address: hgr@sun.ac.za (H.G. Raubenheimer).
0277-5387/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2011.12.026

W.F. Gabrielli et al. / Polyhedron 34 (2012) 188–197
189
Here we report the results of an investigation to prepare and
characterise neutral pentafluorophenyl complexes of gold(I) that
contain substituted neutral imidazoles, a bidentate ligand with thi-
azole units and, uniquely, a substituted tetrazole positioned in the
second coordination position of the central metal(I) ion. Crystal
structure determinations were carried out as well as a preliminary
assay of the cytotoxicity of some of the new compounds to estab-
lish the biological activity of such phosphine-free complexes.
Complex 5, although stable at room temperature and under in-
ert conditions for prolonged periods, is somewhat more air and
moisture sensitive than 1–3.
The microcrystalline complex 6 is insoluble in most polar and
apolar organic solvents, and since characterisation by NMR was
impossible, a molecular structure analysis by X-ray diffraction
was crucial. In order to obtain crystals suitable for a crystal struc-
ture determination, we slowed down the reaction and thus the
immediate formation of the microcrystalline product by working
at lower temperature.
2. Results and discussion
Whereas imine-coordination with imidazole substrates led to
immediate product formation, the related tetrazole coordination
required prolonged reaction periods and were incomplete. In this
manner compound 7 was obtained as a colourless solid that is both
air and moisture sensitive. Recrystallisation from dichloromethane
layered with n-pentane at ꢀ20 °C afforded colourless needles of
the targeted complex and co-crystallised free ligand. These crystals
were subjected to a crystal and molecular structure determination.
Subsequent washing of the crystalline product with diethyl ether
and stripping of solvent under vacuum, afforded 7 in pure form
for further analysis and biological screening.
2.1. Preparative studies
The neutral ligands I–V shown in Scheme 1 were coordinated to
the C6F5Au-unit by simple substitution of labile tht (tht = tetrahy-
drothiophene) [19]. Compound II, that functions as a ligand, is a
gold complex of derivatised benzyl imidazole and was prepared
according to a method used by Burini et al. [20] while substituting
[AuCl(Me2S)] rather than [AuCl(tht)] that was used by them as
starting material.
The complexes 1–5 were finally obtained as analytically pure
crystals from solutions of dichloromethane (1–4) or tetrahydrofu-
ran (5) layered with hexane at ꢀ20 °C.
2.2. Characterisation
X-ray quality crystals of 1–3 grew successfully, and the
resultant structural determination by means of X-ray diffraction
confirmed the molecular structures of these compounds. It is note-
worthy that no homoleptically rearranged by-products of 1–3 in
various solvents were identified during analysis in solution or in
the solids.
The 1H and 13C NMR spectra for the compounds 1–3 and 5 pro-
vide limited information on the bonding in these complexes. The
resonances of the atoms in the complexes are shifted consistently
downfield (DdH 0.3 and DdC ca. 4.0) from their positions in the
spectra of the free ligands. The most notable changes occur in
the proton resonances of the acidic CH’s (C2), which show a more
significant and consistent downfield shift of
Dd 1.0 upon coordina-
C6F5Au
tion. High resolution 13C and 19F NMR spectra of these complexes
confirm the presence of C6F5Au moieties in the noted distinctive
C–F and F–F coupling patterns of the signals. Since no crystals of
suitable dimension for crystal structure determination of 5 could
be grown, we relied on elemental analysis and the pronounced
R1
R2
N
R1
R2
N
[A u(C6F5)(tht)]
N
R3
N
R3
R1=R2=H; R3 = Me
I
1
2
3
R1=R2=H; R3 = CH2Ph
R1R2 = cyclo-C4H4; R3 = Me
(Dd 52) upfield shift for the gold-coordinated nitrogen atoms
(N3) of the two imidazole rings compared to their signals in the
free ligand, as evidence for dinuclear gold complex formation.
The chemical shifts of the dinuclear complex 4 can be compared
to those for the parent complex II with all the spectra measured in
CD2Cl2. The broadened, incompletely resolved 1H NMR signal for
the CH2 unit can be ascribed to H–P coupling across (alkylated)
N1, in contrast to the sharp singlet found for the same proton in
II. The two distinct sets of doublets of doublets – both integrating
for single protons and representing two chemically non-equivalent
nuclei, are assigned to protons H4 and H5. Apart from the 3JHH cou-
C6F5Au
AuCl
Ph
Ph
AuCl
N
N
N
[Au( C6F5)(tht)]
P
P
Ph
Ph
N
CH2Ph
CH2Ph
II
4
C6F5Au
AuC6F5
4
N
N
N
N
N
N
N
N
pling of 7.2 and 7.8 Hz, long range JHP coupling of 1.6 and 1.9 Hz
2 [Au(C6F5)(tht)]
across the nitrogen atom was noted, for H4 and H5, respectively.
Similar coupling patterns cannot be observed in complex II due
to overlap with the aromatic signals. No conclusion could thus be
drawn on whether complexation at N3, favours longer range cou-
pling of the allylic protons to the phosphorus atom or not. The
13C NMR spectrum of 4 shows distinctive C–P coupling patterns
for the chemically non-equivalent phenyl carbons and diagnostic
C2 carbon. The azole carbon in the 2-position appears as a doublet,
(CH2
)
(CH2)4
III 4
5
C6F5Au
N
N
S
N
S
N
S
[Au(C6F5)(tht)]
S
with an insignificant upfield shift (Dd 1.3) compared to the corre-
IV
6
sponding complex–ligand II. The characteristic and complicated C–
F coupling patterns were partly (because of overlap with aromatic
signals) assigned to the pentafluorophenyl moiety. The most
apparent difference is noted in the 31P NMR singlet resonance in
4 which experiences a downfield shift of 6.54 ppm (to d 17.24) rel-
ative to complex II.
C6F5 Au
N
N
N
N
N
N
[Au (C6F5)(tht)]
N
N
CH2Ph
CH2Ph
A pronounced downfield shift (Dd 49) of the resonance for coor-
V
7
dinated N4 in the 15N NMR [1H,15N gHMQC] spectrum of 7 com-
Scheme 1.
pared to that of free 1-benzyltetrazole, provides unambiguous

190
W.F. Gabrielli et al. / Polyhedron 34 (2012) 188–197
evidence for the coordination of a gold(I) centre to this nitrogen
atom.
Molecular ions of 1 (m/z 446), 2 (m/z 522), 3 (m/z 496), 6 (m/z
588) and 7 (m/z 524) are observed in the electron impact mass
spectra, whereas the molecular ion peaks of 4 (m/z 938; 35Cl) and
5 (m/z 918) were only obtained by the softer FAB ionisation. Diag-
nostic fragmentations for 4 include the loss of AuCl (m/z 706) and
loss of C6F5Au (m/z 574) from the molecular ion.
A conspicuous omission from the fragmentation pattern of the
imine complex 7 is the initial loss of dinitrogen that is characteris-
tic in the ionisation of free 1- and 5-substituted tetrazoles,
1-benzyltetrazol-2-yl(diphenyl)phosphines and gold(I) complexes
thereof, various tetrazol-5-ylidenegold(I) complexes and trialkyl-
phosphine(tetrazol-1-yl)gold(I) complexes [21]. The N4-imine
coordination to gold(I) appears to stabilise the ring against the
fragmentation prevalent in this class of compounds.
2.3. X-ray structure determination
The crystal and molecular structures of compounds were deter-
mined by single crystal X-ray diffraction. Their molecular struc-
tures and selected bond lengths and bond angles are depicted in
Figs. 1–6. Although compounds 4 and 6 that differ from the rest
in having somewhat more complicated ligands or structures exhi-
bit seemingly somewhat longer Au–N separations and respectively
somewhat shorter and longer Au–C distances, these elongations
and shortenings remain within 4r.
The neutral molecules in complex 1 are paired in the crystal lat-
tice by weak hydrogen bonds involving the acidic azole hydrogen
and a phenyl fluorine atom. It is striking that no intermolecular
aurophilic interactions occur in this rather simple gold(I) complex
derived from non-bulky ligands especially considering the remark-
able stability of the complex in crystalline form.
The molecular structure of 2 (Fig. 2) can be described as two
discrete linear two-coordinate molecules, which assemble in the
unit cell as dimeric units. Weak aurophilic interactions between
the gold centres [Au(1)–Au(2) 3.2237(3) Å] are attained by a stag-
gered conformation about the metallic bond plane. This places the
planes through the pentafluorophenyl substituents parallel to one
another, whilst the planes through the imidazole rings appear at an
angle.
Fig. 2. Molecular structure of 2 showing the numbering scheme. Bond lengths:
Au(1)–C(31) 1.996(5) Å, Au(1)–N(11) 2.049(4) Å, Au(1)–Au(2) 3.2237(3) Å; bond
angles: C(31)–Au(1)–N(11) 179.0(2)°, C(41)–Au(2)–N(21) 177.6(6)°.
proach each other in a head to toe fashion. In addition, the C6F5
and imine substituents are co-planar and as a result of weak
p-
stacking interaction between pentafluorophenyl and the 5-mem-
bered ring of benzimidazole (distance between centroids
3.484 Å) of the paired units, allow efficient packing in the crystal
lattice.
The molecular structure of complex 4 is given in Fig. 4. The col-
ourless complex crystallises together with one mole quantity of
ꢀ
dichloromethane in the triclinic space group P1.
The neutral binuclear structure shows the two metal centres
bridged by a BzimPPh2 (Bzim = benzimidazole) ligand, with Au(1)
coordinated to the phosphorus atom and Au(2) to the imine-N
The linear complex 3, depicted in Fig. 3, is involved in auro-
philic interaction [Au. . .Au at 3.2896(3) Å] with a neighbouring
molecule across an inversion centre, and the two complexes ap-
Fig. 1. Molecular structure of 1 showing the numbering scheme with 50% probability ellipsoids. Bond lengths: Au(1)–C(11) 2.001(4) Å, Au(1)–N(1) 2.044(3) Å; bond angle:
C(11)–Au(1)–N(1) 177.1(1)°.

W.F. Gabrielli et al. / Polyhedron 34 (2012) 188–197
191
Fig. 3. Molecular structure of 3 showing the numbering scheme. Bond lengths: Au(1)–C(21) 2.008(4) Å, Au(1)–N(11) 2.046(3) Å, Au(1)–Au(1) 3.2896(3) Å; bond angles:
C(21)–Au(1)–N(11) 175.6(1)°, C(21)–Au(1)–Au(1) 98.1(1)°.
Fig. 4. Molecular structure of 4 showing the numbering scheme; phenyl hydrogens and unlinked solvent (dichloromethane) omitted. Bond lengths: Au(1)–P(1) 2.228(2),
Au(1)–Cl(1) 2.290(1) Å, Au(2)–C(41) 1.986(7) Å, Au(2)–N(11) 2.060(5) Å, Au(1)–Au(1) 3.0079(5) Å; bond angles: P(1)–Au(1)–Cl(1) 174.15(6)°, C(41)–Au(2)–N(11) 179.0(2)°.
Fig. 5. Molecular structure of 6 showing the numbering scheme. Bond lengths: Au(2)–C(31) 2.014(7) Å, Au(2)–N(11) 2.061(6) Å; bond angle: C(31)–Au(2)–N(11) 178.4(2)°.

192
W.F. Gabrielli et al. / Polyhedron 34 (2012) 188–197
Fig. 6. Molecular structure of 7ꢁbenzyltetrazole showing the numbering scheme and the presence of a free ligand in the unit cell. Bond lengths: Au(1)–C(31) 2.004(6) Å,
Au(1)–N(11) 2.050(5) Å; bond angle: C(31)–Au(1)–N(11) 177.5(2)°.
(N11) on the imidazole ring. Compared to the linear C(41)–Au(2)–
N(II) angle, some distortion from linearity occurs in P1–Au(1)–
Cl(1) [174.15(6)°].
ing of a C6F5Au unit, coordinated to a neutral imine within tetra-
zole while creating an almost linear geometry about the gold(I)
atom. Of the three unsubstituted nitrogen atoms, only N(14) (N4
in the NMR discussion) is involved in coordination to the gold cen-
tre. It is worth mentioning that the Au(1)–N(11) [2.050(5) Å], adja-
cent N(11)–N(12) [1.368(7) Å] and C(10)–N(11) [1.310(7) Å] bond
distances in the tetrazole imine complex are consistent with Au–
The relatively short intramolecular aurophilic separation of
3.0079(5) Å is similar to the intramolecular metal contacts in the
related dibridged compound [(AuCl)2{(Ph2P)2CH2}2] [2.962(1) Å]
[22] and in the tetranuclear complex, [Au{S@CNC(CH3)@CHS}]4
[3.02(4) Å] [10]. The two centres of coordination, Cl–Au–P and C–
Au–N, are crossed, with the resultant plane of the five membered
metallocyclic ring twisted by about 30° from planarity, [torsion an-
gles Au(1)–P(1)–C(11)–N(11) and Au(1)–Au(2)–N(11)–C(11) at
35.7(6)° and ꢀ26.6(5)°, respectively]. In contrast, the cationic com-
plex [Au2(imPPh2)2]2+ (im = imidazole) shows almost perfect pla-
narity [23]. Dichloromethane solvent molecules are dispersed
along each of the regularly packed rows of molecules.
N
[2.043(5) Å], N–N [1.336(6) Å] and C–N [1.306(7) Å] bond
lengths in the related triphenylphosphine(tetrazol-1-yl)gold(I)
amine complex (the only entry of a ring substituted tetrazole com-
plex of gold in the Cambridge Crystallographic database) [24]. The
free ligand guest in the unit cell associates with the gold(I) com-
plex through weak fluorine and hydrogen interactions. Apart from
additional fluorine–hydrogen interaction,
p-stacking interactions
of the phenyl rings govern the lattice organisation.
ꢀ
Compound 6 (Fig. 5) crystallises in the triclinic space group P1
as yellowish needles. The planes of the two thiazole rings that con-
stitute the ligand are almost co-planar [torsion angles N(11)–
C(11)–C(21)–N(21) and S(11)–C(11)–C(21)–N(21) at 171.5(7) and
ꢀ8.6(9), respectively]. The bond lengths in the coordinated ring
[N(11)–C(11) 1.329(9) Å and N(11)–C(13) 1.408(9) Å] remain lar-
gely unchanged compared with bond lengths in the uncoordinated
ring [N(21)–C(21) 1.321(9) Å and N(21)–C(23) 1.375(9)Å].
Possible electron charge delocalisation from N(21) effected by
gold(I) coordination to N(11) that could rationalise the mono-sub-
stitution, remains unsubstantiated, owing to typical single bond
character of C(21)–C(11) [1.447(10) Å]. Molecules in the crystal
lattice are organised in closely knitted sheets in the bc plane that
stack upon one another along the a-axis by translation. Molecules
arrange across the gold centres, with C6F5 moieties in a trans con-
figuration about the coordinative axis of adjacent molecules. The
structure is devoid of sulfur–Au(I) or aurophilic interactions. Inter-
2.4. Cytotoxicity studies
Cytotoxicity assays are performed to establish the sensitivity of
cancer cell lines and normal cell cultures to selected compounds. In
a standard procedure, a known concentration of cells is exposed to
different concentrations of the experimental drug in a 96 well tis-
sue culture plate and incubated for a selected period of time. Incu-
bation periods can range from three to 7 days. The results obtained
in such experiments enable the calculation of the concentration of
the experimental compound that inhibits 50% of growth (IC50). This
method is applicable to both cancer cells and lymphocytes. From
such results a coarse tumour specificity can be calculated, i.e.
(mean IC50 of healthy lymphocytes)/(mean IC50 of the cancer cells).
The in vitro cytotoxic potency of altogether five compounds as
well as cisplatin serving as benchmark were evaluated by determin-
ing their activity using cervical carcinoma cells (HeLa; ATCC CCL-2)
as the target cell line. Compounds 1 and 7 are neutral azolyl (penta-
fluorophenyl)gold(I), complexes whereas the structures of VI [25],
VII (see Section 4 for preparative details) and VIII [21c] that all con-
tain a cationic gold phosphine unit attached to different N-heterocy-
clic ring systems, are shown in Scheme 2. To evaluate whether any
cytotoxicity observed against HeLa cells could not be tumour spe-
cific, additional dose–response assays were performed to derive
molecular interactions are limited to p-stacking between the thia-
zole and pentafluorophenyl ring (distance between centroids
3.554 Å), which brings the non-interactive gold(I) centres in the
closest possible separation at 3.679 Å.
The molecular structure of 7 (Fig. 6) that co-crystallises with a
molecule of free ligand (1-benzyltetrazole) in the monoclinic space
group P21/c, can best be described as a discrete monomer consist-

W.F. Gabrielli et al. / Polyhedron 34 (2012) 188–197
193
SO3CF3
4. Experimental
Me
N
N
N
N
N
C
N
N
N
N
N
N
4.1. Materials and physical measurements
H
H
AuPPh3
Reactions were carried out under argon using standard Schlenk
and vacuum-line techniques. Tetrahydrofuran (thf), n-hexane, n-
pentane and diethyl ether were distilled under N2 from sodium
benzophenone ketyl, dichloromethane and N,N0-dimethylformam-
ide (DMF) from CaH2, ethanol and methanol from magnesium. 1-
Methylimidazole, 1-benzylimidazole, sodium azide and ammo-
nium chloride were purchased from Aldrich. Butyllithium (1.6 M
solution n-hexane), sodium hydride, dimethyldisulfide, triethyl
orthoformate and glacial acetic acid were purchased from Merck.
Literature methods were followed to prepare 1-benzyltetrazole
[27], 1-methylbenzimidazole [28] 1,4-bis(imidazol-1-yl)butane
[29], 1,2-di(tetrazol-2-yl)ethane [30], 2,20-bis(4,5-dimethylthia-
zole) [31], [AuCl(tht)] [32], [Au(C6F5)(tht)] [33], and chloro[(1-ben-
zylimidazol-2-yl)diphenyl]phosphinegold(I) [30].
AuPPh3
AuPMe3
CH2Ph
VI
VII
VIII
Scheme 2.
Table 1
Cytotoxicity (IC50a,b) of cisplatin and gold complexes against HeLa cells and human
lymphocytes.
Compound HeLa
Human
lymphocytes
(resting)
Human lymphocytes (PHA
stimulated)
1
7
VI
VII
0.55 (0.09) 8.37 (0.96)
1.39 (0.16)
1.27 (0.72) 2.58 (0.01)
8.18 (1.75)
–
–
Melting points were determined on a Stuart SMP3 apparatus
and are uncorrected. Mass spectra were recorded on an AMD 604
(EI, 70 eV), VG Quattro (ESI, 70 eV methanol, acetonitrile) or VG
70 SEQ (FAB, 70 eV) instruments. In the EI and FAB MS results
the isotopic distribution patterns confirmed the theoretical distri-
bution. All NMR spectra were recorded on a Varian 300 FT or INO-
VA 600 MHz spectrometer (1H NMR at 300/600 MHz, 13C{1H} NMR
at 75/150 MHz, 31P{1H} NMR at 121/243 MHz, 15N NMR at
60.8 MHz and 19F NMR at 376 MHz). Chemical shift (d is reported
relative to solvent resonance or external reference of 85% H3PO4
2.45 (1.29)
2.33 (0.14)
1.67 (0.20)
–
–
–
–
VIII
Cisplatin
0.45 (0.12) 48.71 (1.82)
2.88 (1.17)
a
Mean drug concentration
cell death under chosen conditions.
lmol/l (standard deviation in brackets) causing 50%
b
Average of not less than three independent experiments.
IC50 values against normal, resting and phytohaemagglutinin(PHA)-
stimulated human lymphocytes for the two compounds, 1 and VI,
that exhibited the lowest IC50s against HeLa cells. The results for
the cytotoxicity assays are reported in Table 1.
(
31P), NH3NO2 (15N) and CFCl3 (19F). Infrared spectra were recorded
on a Thermo Nicolet Avatar 330FT-IR with Smart OMNI ATR (atten-
uated total reflectance) sampler. Elemental analysis was carried
out at the School of Chemistry, University of the Witwatersrand.
For elemental analysis, products were evacuated under high vac-
uum for 10 h.
From the results it is clear that the relatively simple neutral
complex 1 performs much better against cervical carcinoma cells
(HeLa; ATCC CCL-2) than the phosphine-free tetrazolate complex
7 or the phosphine complexes VII and VIII. The phosphine (tetraz-
olate) complex VI probably has a similar potency to the latter three
but, unfortunately, its results show a larger variability. The imidaz-
olate(pentafluorophenyl)gold(I) complex, 1, effects a tumour spec-
ificity [IC50(av. lymph)/IC50(HeLa)] of 15 compared to 2.0 for VI and
57 for cisplatin. We established, however [3b], that all these mono-
nuclear compounds exhibit much lower cytotoxicity than dinucle-
ar azolate-type gold compounds of diphosphines with carbon
chain lengths of 5 and 6 between the two phosphorous atoms –
even when normalised with respect to the mole quantities gold
and phosphorous that they contain. It could, in the light of the
present results, prove worthwhile to investigate related dinuclear
complexes with neutral azoles as ligands, particularly phosphine-
free analogues. The carbene ligands developed by Baker and co-
workers [26] could be ideally suited for such a purpose.
4.2. Preparation of 1-methylimidazole(pentafluorophenyl)gold(I) (1)
The addition of
a solution of 1-methylimidazole (0.17 g,
2.1 mmol) in diethyl ether (20 ml) to a solution of freshly prepared
[Au(C6F5)(tht)] (0.95 g, 2.1 mmol) in diethyl ether (20 ml) pro-
duced a cloudy solution. The mixture was stirred for 1.5 h at room
temperature upon which the solution was reduced to dryness in
vacuo. The white residue was extracted with diethyl ether, filtered
through anhydrous MgSO4 and the filtrate concentrated in vacuo to
yield a white microcrystalline material. Recrystallisation of the
white solid by layering a dichloromethane solution with n-hexane
produced colourless needles, 1 (0.70 g) at ꢀ20 °C.
Yield: 75%. Mp 136 °C. EI-MS, m/z (relative intensity) for
C
10H6AuF5N2: 446 (M+, 9), 364 (AuC6F5+, 2), 334 (C12F10+, 100),
279 (M+ꢀC6F5, 5), 265 (C11F7+, 36), 82 (M+ꢀAuC6F5, 28). 1H NMR
(300 MHz, CD2Cl2, dH) 3.77 (s, 3H, NCH3), 7.13 (m, 1H, CHN1),
7.15 (m, 1H, CHN3), 7.78 (s, 1H, NC(H)N). 13C NMR (75 MHz,
CD2Cl2, dC) 35.1 (s, NCH3), 117.9 (tm, JCF = 56.4 Hz, ipso-C6F5),
121.7 (s, CHN1), 129.5 (s, CHN3), 137.1 (dm, JCF = 249.9 Hz, meta-
C6F5), 138.8 (dm, JCF = 245.3 Hz, para-C6F5), 139.4 (s, NC(H)N),
149.9 (ddm, JCF = 228.0 Hz, JCF = 24.2 Hz, ortho-C6F5). 15N NMR
(61 MHz, CD2Cl2, dN) ꢀ217.7 (s, N1), ꢀ166.3 (s, N3). Anal. Calc.
for C10H6AuF5N2 (446.13): C, 26.92; H, 1.36; N, 6.28. Found: C,
26.58; H, 1.41; N, 6.22%.
3. Conclusion
The small number of well-characterised neutral gold complexes
that contain imine-bonded azoles, has been extended by the prep-
aration and crystal structure determination of a number of very
stable complexes based on substituted imidazole, thiazole and tet-
razole. The simplest complex now prepared, [Au(C6F5)(N-methyl-
imidazole)], shows no aurophilic interaction in the solid state
whereas three other complexes crystallizes with one or other form
gold–gold interaction present. Our investigation clearly indicates,
firstly, that phosphine auxiliary ligands in certain gold(I) com-
plexes can be replaced by the pentafluorophenyl anion without
loss of cytotoxicity and, secondly, that neutral azoles should be in-
cluded in the list of viable ligands for biological activity studies.
4.3. Preparation of 1-benzylimidazole(pentafluorophenyl)gold(I) (2)
This product was prepared in a similar fashion as 1 from
1-benzylimidazole (0.04 g, 0.3 mmol) and [Au(C6F5)(tht)] (0.12 g,

194
W.F. Gabrielli et al. / Polyhedron 34 (2012) 188–197
Table 2
Crystallographic data.
1
2
Empirical formula
Mr
T (K)
C
10H6F5N2Au
C16H10AuF5N2
522.23
100(2)
446.13
100(2)
k (Å)
0.71073
0.71073
Crystal system
Space group
a (Å)
triclinic
P1
4.7900(3)
triclinic
P1
8.1714(8)
ꢀ
ꢀ
b (Å)
c (Å)
10.3738(6)
11.0755(7)
88.0840(10)
86.0360(10)
82.6480(10)
544.35(6)
13.4065(13)
14.5984(13)
77.470(2)
86.845(2)
79.489(2)
a
(°)
b (°)
c
(°)
V (Å3)
1534.8(3)
Z
2
4
Dcalc (g cmꢀ3
)
2.722
13.563
2.260
9.640
Absorption coefficient (
Absorption correction
F(000)
l
, mmꢀ1
)
semi-empirical from equivalents
408
semi-empirical from equivalents
976
Crystal size (mm)
h-range for data collection (°)
0.30 ꢂ 0.20 ꢂ 0.20
0.30 ꢂ 0.20 ꢂ 0.15
1.84–28.24
1.89–25.68
Index range
ꢀ6 6 h 6 6
ꢀ9 6 h 6 9
ꢀ13 6 k 6 13
ꢀ14 6 l 6 14
6256
ꢀ16 6 k 6 15
ꢀ10 6 l 6 17
8529
Number of reflections collected
Number of independent reflections
Maximum and minimum transmission
Refinement method
2494 (Rint = 0.0235)
0.2355 and 0.1318
Full-matrix least-squares on F2
2494/0/165
5710 (Rint = 0.0176)
0.3258 and 0.2133
Full-matrix least-squares on F2
5710/0/433
1.025
Data/restraints/parameters
Gof on F2
1.056
Final R-indices [I > 2
r
> (I)]
R1 = 0.0194
wR2 = 0.0444
R1 = 0.0209
wR2 = 0.0451
1.487 and ꢀ0.566
a = 0.0221/b = 0.2729
R1 = 0.0263
wR2 = 0.0635
R1 = 0.0308
wR2 = 0.0656
1.952 and ꢀ0.641
a = 0.0384
R indices (all data)
Largest difference in peak and hole (e Aꢀ3
Weighing scheme
)
0.27 mmol). Colourless crystals of 2 (0.11 g) were obtained, from a
dichloromethane solution layered with n-hexane, at ꢀ20 °C.
Yield: 75%. Mp: 120–122 °C. EI-MS, m/z (relative intensity) for
(dm, JCF = 242.8 Hz, meta-C6F5), 140.4 (dm, JCF = 244.5 Hz,
para-C6F5), 142.4 (s, CHN3), 148.6 (s, NC(H)N), 151.7 (ddm,
JCF = 227.6 Hz, JCF = 24.1 Hz, ortho-C6F5). 19F NMR [376 MHz,
(CD3)2CO, dF] ꢀ115.8 (m, 2F, ortho-F), ꢀ157.1 (t, JFF = 19.3 Hz, 2F,
meta-F), ꢀ161.3 (m, F, para-C6F5). Anal. Calc. for C14H8AuF5N2
(496.19): C, 33.89; H, 1.63; N, 5.65. Found: C, 33.60; H, 1.51; N,
5.82%.
C
16H10AuF5N2: 522 (M+, 25), 364 (AuC6F5+, 6), 334 (C12F10+, 62),
355 (M+ꢀC6F5, 2), 265 (C11F7 22), 158 (M+ꢀAuC6F5, 100). 1H
+
,
NMR [600 MHz, (CD3)2CO, dH] 5.47 (bs, 2H, CH2Ph), 7.29 (m, 1H,
CHN3), 7.36–7.45 (m, 5H, –CH2C6H5), 7.60 (m, 1H, CHN1), 8.46
(m, 1H, NC(H)N). 31C NMR [150 MHz, (CD3)2CO, dC] 53.4 (s,
NCH2), 119.2 (tm, JCF = 54.1 Hz, ipso-C6F5), 122.9 (s, CHN1), 129.4
(s, ipso-C6H5CH2), 129.9 (s, meta-C6H5CH2), 130.4 (s, para-
C6H5CH2), 130.7 (s, CHN3), 130.9 (s, ortho-C6H5CH2), 135.3 (dm,
JCF = 251.4 Hz, meta-C6F5), 139.4 (dm, JCF = 243.6 Hz, para-C6F5),
141.5 (s, NC(H)N), 150.3 (ddm, JCF = 227.2 Hz, JCF = 24.2 Hz, ortho-
C6F5). 19F NMR [376 MHz, (CD3)2CO, dF] ꢀ115.7 (m, 2F, ortho-F),
ꢀ161.7 (t, JFF = 19.5 Hz, 2F, meta-F), ꢀ164.3 (m, F, para-F). Anal.
Calc. for C16H10AuF5N2 (522.23): C, 36.80; H, 1.93; N, 5.36. Found:
C, 36.91; H, 1.90; N, 5.71%.
4.5. Preparation of chloro[1-benzyl-N-
pentafluorophenylgold(I)imidazol-2-yl](diphenyl)phosphinegold(I) (4)
A
solution of chloro[(1-benzylimidazol-2-yl)diphenyl]phos-
phinegold(I) (0.13 g, 0.22 mmol) in diethyl ether/thf (10:1, 20 ml)
was treated with a solution of freshly prepared [Au(C6F5)(tht)]
(0.10 g, 0.22 mmol) in diethyl ether (20 ml). The clear solution
was stirred at room for 48 h temperature. The resultant cloudy
solution was reduced to dryness under vacuum. The off-white oily
residue was repeatedly washed with n-hexane to afford a white so-
lid. Recrystallisation by layering a dichloromethane solution with
n-hexane afforded colourless, crystalline material, 4 (0.15 g,) at
ꢀ20 °C.
4.4. Preparation of 1-methylbenzimidazole(pentafluorophenyl)gold(I)
(3)
This product was prepared in a similar manner as 1 from 1-
methylbenzimidazole (0.04 g, 0.3 mmol) and [Au(C6F5)(tht)]
(0.12 g, 0.27 mmol). Colourless crystals of 3 (0.09 g) were obtained,
from a dichloromethane solution layered with n-hexane, at ꢀ20 °C.
Yield: 71%. Mp: 233 °C (decomp.). EI-MS, m/z (relative intensity)
Yield: 73%. Mp: 117 °C (decomp.). EI-MS, m/z (relative intensity)
for C28H19N2F5ClPAu2: 938 (M35Cl+, 100), 741 (M+ꢀAu, 1), 706
(M+ꢀAuCl, 16), 574 (M+ꢀAuC6F5, 12), 342 (BzimPPh2+, 70). 1H
NMR (300 MHz, CD2Cl2, dH) 5.30 (bs, 2H, CH2Ph), 6.83 (dd, 1H,
3J = 7.2 Hz, 4JHP = 1.6 Hz, CHN3), 7.23 (dd, 1H, 3J = 7.8 Hz,
4JHP = 1.9 Hz, CHN1), 7.18–7.44 (m, 5H, –CH2C6H5), 7.42–7.72 (m,
10H, P(C6H5)2). 31C NMR (75 MHz, CD2Cl2, dC) 53.3 (s, CH2Ph),
for C14H8AuF5N2: 496 (M+, 2), 334 (C12F10+, 100), 265 (C11F7 55),
+
,
132 (M+ꢀAuC6F5, 100). 1H NMR [300 MHz, (CD3)2CO, dH] 4.15 (s,
1
3H, NCH3), 7.53 (m, 2H, ortho-C6H4), 7.96 (m, 1H, meta-C6H4) and
7.83 (m, 1H, meta-C6H4), 8.82 (s, 1H, NC(H)N). 31C NMR [75 MHz,
(CD3)2CO, dC] 33.6 (s, NCH3), 113.8/120.1 (s, ortho-C6H4), 126.4/
127.2 (s, meta-C6H4), 117.7 (m, ipso-C6F5), 135.6 (s, CHN1), 138.7
117.7 (m, ipso-C6F5), 124.8 (d, JCP = 65.3 Hz, ipso-C6H5P), 127.0–
135.0 (m, para-C6F5), 127.0–135.0 (m, meta-C6F5), 127.7
(s, CHN1), 128.1 (s, ipso-C6H5CH2), 129.4 (s, m-C6H5CH2), 129.8
3
(s, CHN3), 130.5 (d, JCP = 12.7 Hz, meta-P(C6H5)2), 133.0 (s,

W.F. Gabrielli et al. / Polyhedron 34 (2012) 188–197
195
Table 3
Crystallographic data.
3
4ꢁCH2Cl2
Empirical formula
Mr
C14H8AuF5N2
496.19
C43H23F5N2Cl2PAu2
1158.47
T (K)
100(2)
100(2)
k (Å)
0.71073
0.71073
Crystal system
Space group
a (Å)
monoclinic
P21/c
11.0548(11)
8.1204(8)
triclinic
P1
10.505(2)
12.697(2)
ꢀ
b (Å)
c (Å)
15.1295(16)
90
13.241(2)
113.914(3)
a
(°)
b (°)
101.383(2)
90
1331.5(2)
103.872(2)
90.991(3)
1554.2(4)
c
(°)
V (Å3)
Z
4
2
Dcalc (g cmꢀ3
)
2.475
11.104
2.187
9.791
Absorption coefficient (
Absorption correction
F(000)
l
, mmꢀ1
)
semi-empirical from equivalents
920
semi-empirical from equivalents (SADABS
956
)
Crystal size (mm)
h-range for data collection (°)
Index range
0.20 ꢂ 0.20 ꢂ 0.10
1.88–26.37
ꢀ11 6 h 6 13
ꢀ10 6 k 6 9
ꢀ18 6 l 6 15
7396
0.20 ꢂ 0.10 ꢂ 0.10
1.75–26.73
ꢀ7 6 h 6 13
ꢀ13 6 k 6 16
ꢀ16 6 l 6 16
9244
Number of reflections collected
Number of independent reflections
Maximum and minimum transmission
Refinement method
2707 (Rint = 0.0220)
0.4031 and 0.2039
Full-matrix least-squares on F2
2707/0/200
1.059
6419 (Rint = 0.0146)
0.3749 and 0.2408
Full-matrix least-squares on F2
6419/0/379
Data/restraints/parameters
Gof on F2
1.104
Final R-indices [I > 2
r
> (I)]
R1 = 0.0221
wR2 = 0.0520
R1 = 0.0244
wR2 = 0.0531
1.557 and ꢀ1.283
a = 0.0312/b = 0.9283
R1 = 0.0350
wR2 = 0.0837
R1 = 0.0400
wR2 = 0.0860
2.140 and ꢀ1.464
a = 0.0360/b = 11.0208
R indices (all data)
Largest difference in peak and hole (e Aꢀ3
Weighing scheme
)
1
para-P(C6H5)2), 133.4 (d, JCP = 63.1 Hz, NC(P)N), 133.4 (s, para-
F, para-F). Anal. Calc. for C22H14N4F10Au2 (918.30): C, 28.78; H,
1.54; N, 6.10. Found: C, 28.86; H, 1.51; N, 6.15%.
2
C6H5CH2), 134.8 (d, JCP = 14.8 Hz, ortho-P(C6H5)2), 141.2 (s, ortho-
C6H5CH2), 149.6 (ddm, JCF = 231.0, JCF = 28.0 Hz, ortho-C6F5). 31P
NMR (121 MHz, CD2Cl2, dP) 17.24 (s). 19F NMR (376 MHz, CD2Cl2,
dF) ꢀ114.7 (m, 2F, ortho-F), ꢀ161.4 (t, J = 21.3 Hz, F, para-F) and
4.7. Preparation of [2,2-bis(4,5-dimethylthiazol-2-
yl)](pentafluorophenyl)gold(I) (6)
ꢀ162.7 (m, 2F, meta-F). Anal. Calc. for
C28H19N2F5ClPAu2
(938.82): C, 35.82; H, 2.04; N, 2.98. Found: C, 36.10; H, 2.22; N,
3.14%.
The addition of a solution of 2,20-bis(4,5-dimethylthiazole)
(0.03 g, 0.20 mmol) in diethyl ether (10 ml) to [Au(C6F5)(tht)]
(0.14 g, 0.30 mmol) in diethyl ether (10 ml), produced an immedi-
ate thick precipitate. The resultant suspension was stirred at room
temperature for 1 h, filtered and washed sequentially with diethyl
ether (10 ml) and dichloromethane (10 ml), to afford a highly
insoluble off-white microcrystalline material of 6 (0.08 g). Repeat-
ing the reaction with equimolar amounts of [Au(C6F5)(tht)] and the
bisthiazole effected the same result. Crystals of suitable dimension
were obtained by layering a solution of [Au(C6F5)(tht)] (0.02 mg,
0.03 mmol) in thf (1 ml) in a crystallisation tube with diethyl ether
(1 ml) followed by a solution of bisthiazole (0.01 g, 0.03 mmol) in
n-pentane (1 ml). Leaving the reaction at ꢀ20 °C, the first crystal-
line product formed within hours at the contact point between
the reagent solutions.
4.6. Preparation of [1,4-bis(imidazol-1-
yl)butane]bis[(pentafluorophenyl)gold(I)] (5)
This product was prepared in a similar fashion as 1 from 1,4-
bis(imidazol-1-yl)butane (0.05 g, 0.3 mmol) in methanol (10 ml)
and [Au(C6F5)(tht)] (0.23 g, 0.50 mmol) in thf/methanol (10:1,
10 ml). Colourless microcrystalline material of 5 (0.19 g) was ob-
tained, from a thf solution layered with n-hexane, at ꢀ20 °C.
Yield: 83%. Mp: 148 °C (decomp.). EI-MS, m/z (relative intensity)
for
C
22H14N4F10Au2: 918 (M+, 2), 751 (M+ꢀC6F5, 13), 555
(M+ꢀAuC6F5, 12), 387 (M+ꢀAu(C6F5)2, 12), 191 (M+ꢀ[Au(C6F5)]2,
15). 1H NMR [600 MHz, (CD3)2SO), dH] 1.75 (q, 4H, 3J = 2.8 Hz,
NCH2CH2), 4.12 (t, 4H, 3J = 5.0 Hz, NCH2), 7.29 (dd, 2H, 3J = 2.8 Hz,
4J = 0.4 Hz, CHN3), 7.64 (dd, 2H, 3J = 3.0 Hz, 4J = 0.3 Hz, CHN1),
8.48 (dd, 2H, 3J = 2.6 Hz, 4J = 0.2 Hz, CHN3). 31C NMR [150 MHz,
(CD3)2SO), dC] 26.7 (s, NCH2CH2), 46.7 (s, NCH2), 118.8 (t,
JCF = 53.2 Hz, ipso-C6F5), 121.0 (s, CHN1), 128.2 (s, CHN3), 136.0
(dm, JCF = 249.8 Hz, para-C6F5), 137.4 (dm, JCF = 243.9 Hz, meta-
C6F5), 139.7 (s, NC(H)N), 148.6 (ddm, JCF = 228.6 Hz, JCF = 23.8 Hz,
ortho-C6F5). 15N NMR [61 MHz, (CD3)2SO), dN] ꢀ168.4 (s, N3),
ꢀ200.7 (s, N1). 19F NMR [376 MHz, (CD3)2SO), dF] ꢀ116.2
(m, 2F, ortho-F), ꢀ161.3 (t, JFF = 21.2 Hz, 2F, meta-F), ꢀ164.0 (m,
Yield: 93%. Mp: 182 °C (decomp.). EI-MS, m/z (relative intensity)
+
for
C
16H12AuF5N2S2: 588 (M+, 21), 334 (C12F10
,
37), 265
(C12F10 ꢀCF3, 16), 224 (M+ꢀAuC6F5, 100). Anal. Calc. for
+
C
16H12AuF5N2S2 (588.36): C, 32.66; H, 2.06; N, 4.76. Found: C,
32.78; H, 2.09; N, 4.91%.
4.8. Preparation of 1-benzyltetrazole(pentafluorophenyl)gold(I) (7)
This product was prepared in a similar manner as 1 from
1-benzyltetrazole (0.39 g, 2.4 mmol) and [Au(C6F5)(tht)] (1.1 g,

196
W.F. Gabrielli et al. / Polyhedron 34 (2012) 188–197
Table 4
Crystallographic data.
6
7ꢁC8H8N4
Empirical formula
Mr
C16H12F5N2S2Au
588.36
C22H16F5N8Au
684.38
T (K)
100(2)
100(2)
k (Å)
0.71073
0.71073
Crystal system
Space group
a (Å)
triclinic
P1
6.875(4)
11.536(6)
monoclinic
P21/c
18.4500(13)
5.0734(4)
ꢀ
b (Å)
c (Å)
11.747(6)
112.809(7)
94.191(8)
99.105(8)
838.7(8)
25.1579(17)
90
107.5310(10)
90
a
(°)
b (°)
c
(°)
V (Å3)
2245.5(3)
Z
2
4
Dcalc (g cmꢀ3
)
2.330
9.074
2.024
6.623
Absorption coefficient (
Absorption correction
F(000)
l
, mmꢀ1
)
semi-empirical from equivalents
556
semi-empirical from equivalents
1312
Crystal size (mm)
h-range for data collection (°)
Index range
0.30 ꢂ 0.10 ꢂ 0.05
1.90–25.68
ꢀ8 6 h 6 8
0.20 ꢂ 0.10 ꢂ 0.10
1.70–26.73
ꢀ23 6 h 6 22
ꢀ6 6 k 6 6
ꢀ26 6 l 6 31
12484
ꢀ7 6 k 6 14
ꢀ14 6 l 6 14
4412
Number of reflections collected
Number of independent reflections
Maximum and minimum transmission
Refinement method
3081 (Rint = 0.0277)
0.6597 and 0.2428
Full-matrix least-squares on F2
3081/0/239
1.064
4727 (Rint = 0.0404)
0.5572 and 0.4477
Full-matrix least-squares on F2
4727/0/325
1.057
Data/restraints/parameters
Gof on F2
Final R-indices [I > 2
r
> (I)]
R1 = 0.0355
wR2 = 0.0909
R1 = 0.0393
wR2 = 0.0931
3.027 and ꢀ2.290
a = 0.0586
R1 = 0.0391
wR2 = 0.0851
R1 = 0.0502
wR2 = 0.0894
2.977 and ꢀ1.524
a = 0.0490
R indices (all data)
Largest difference in peak and hole (e Aꢀ3
Weighing scheme
)
2.4 mmol). An n-hexane extract of the dried product residue, was
recrystallised and afforded colourless crystals of 7 (0.79 g), from
a dichloromethane solution layered with diethyl ether, at ꢀ20 °C.
Yield: 62%. Mp: 121 °C (decomp.). EI-MS, m/z (relative intensity)
for C14H8AuF5N4: 524 (M+, 3), 364 (AuC6F5+, 3), 334 (C12F10+, 100),
4.10. X-ray crystal structure determinations
Specimens of suitable quality and size of 1, 2, 3, 4ꢁCH2Cl2, 6 and
7ꢁC8H8N4 were mounted on the ends of glass fibres in inert oil and
used for intensity data collection on a Bruker SMART Apex CCD dif-
265 (C11F7 40). 1H NMR [300 MHz, (CD3)2CO, dH] 5.80 (s, 2H,
fractometer [34], employing graphite-monochromated Mo Ka
+
,
CH2Ph), 7.40–7.46 (m, 5H, C6H5), 9.24 (s, 1H, NC(H)N). 13C NMR
radiation (k = 0.71073 Å). Data reduction was carried out using
the SAINT [35] suite of programs and multi-scan absorption correc-
tions were performed with SADABS [36,37].
[75 MHz, (CD3)2CO, dC] 53.2 (s, NCH2), 118.7 (tm, JCF = 56.8 Hz,
ipso-C6F5), 130.5 (s, meta-C6H5CH2), 130.8 (s, para-C6H5CH2),
131.2 (s, ortho-C6H5CH2), 137.2 (s, ipso-C6H5CH2), 141.3 (dm,
JCF = 250.7 Hz, para-C6F5), 139.9 (dm, JCF = 245.4 Hz, meta-C6F5),
145.6 (s, NC(H)N), 150.1 (ddm, JCF = 223.9 Hz, JCF = 26.9 Hz, ortho-
C6F5). 15N NMR [61 MHz, (CD3)2CO, dN] ꢀ9.0 (s, N3), ꢀ37.0 (s, N4),
ꢀ50.2 (s, N2), ꢀ141.2 (s, N1). Anal. Calc. for C14H8AuF5N4 (524.20):
C, 32.08; H, 1.54; N, 10.69. Found: C, 32.24; H, 1.51; N, 10.78%.
The structures were solved by a combination of direct methods
(SHELXS-97) and difference-Fourier syntheses and refined by full
matrix least-squares calculations on F2 (SHELXL-97) [38] within the
X-seedenvironment[39]. The thermalmotionwas treatedanisotrop-
ically for all non-hydrogen atoms. All hydrogen atoms were calcu-
lated in ideal positions and refined using a riding model. Further
details regarding the data collection and structure refinement for
compounds 1, 2, 3, 4ꢁCH2Cl2, 6 and 7ꢁC8H8N4 are listed in Tables 2–
4. Figures were generated with X-seed [39] and POV Ray for
Windows, with the displacement ellipsoids at 50% probability level.
4.9. Preparation of (trimethylphosphine)(tetrazol-1-yl)gold(I) (VII)
This product was prepared in a similar manner as VI [25] from
[AuCl(PMe3)] (0.56 g, 1.8 mmol) and 1-H tetrazole (0.13 g,
1.8 mmol) in acetone (60 ml), treated with 1.0 M aqueous NaOH
(1.8 ml, 1.8 mmol). Colourless microcrystalline material of VII
(0.46 g) was obtained, from a methanol solution layered with n-
pentane, at ꢀ20 °C.
4.11. Biological assays
4.11.1. Formulation of drugs
Cisplatin was included as a control for comparison. Ten millimo-
lar stock solutions of the experimental compounds in DMSO were
prepared (concentration 60.1%). Further dilutions were made in
the appropriate tissue culture medium which was supplemented
with 10% heat- inactivated, foetal calf serum (FCS) just before use.
Yield: 74%. Mp: 154 °C (decomp.). FAB-MS, m/z (relative inten-
sity) for C4H10AuN4P: 349 (Au(PMe3)2+, 15), 342 (M+, 72), 314
(M+ꢀN2, 4), 273 ({M+ꢀCHN4, 100). 1H NMR [300 MHz, (CD3)2SO,
dH] 0.87 (d, 9H, 1J = 12.0 Hz, CH3), 7.95 (s, 1H, CH). 13C NMR
[75 MHz, (CD3)2SO, dC] 5.4 (d, 1J = 42.1 Hz, CH3), 140.7 (bs, CH).
31P NMR [61 MHz, (CD3)2SO, dP] ꢀ10.1 (s). Anal. Calc. for
C4H10AuN4P (342.09): C, 14.04; H, 2.95; N, 16.38. Found: C, C,
14.31; H, 2.72; N, 16.10%.
4.11.2. Cell lines and culture conditions
The human cervical carcinoma (HeLa, ATCC No. CCL2) cell line –
adherent epithelial cells maintained in Eagles minimum essential

W.F. Gabrielli et al. / Polyhedron 34 (2012) 188–197
197
[2] J. Coetzee, S. Cronje, L. Dobrzan´ ska, H.G. Raubenheimer, M.J. Nell, G. Jooné, H.
Hoppe, Dalton Trans. 40 (2011) 1471.
[3] (a) U.E.I. Horvath, S. Cronje, J.M. McKenzie, L.J. Barbour, H.G. Raubenheimer, Z.
Naturforsch. 59b (2004) 1605. and references therein;
(b) Also: U.E.I. Horvath, H.G. Raubenheimer, unpublished results.
[4] J.R. Lechat, R.H. De Almeida-Santos, G. Banditelli, F. Bonati, Cryst. Struct.
Commun. 11 (1982) 471.
[5] F. Bonati, A. Burini, B.R. Pietroni, E. Georgini, J. Organomet. Chem. 344 (1988)
119.
[6] F. Bachechi, A. Burini, R. Galassi, B.R. Pietroni, M. Severini, J. Organomet. Chem.
575 (1999) 269.
medium (EMEM) containing 2 mM
essential amino acids, 1.0 mM sodium pyruvate and 5% bovine
FCS – were used.
L-glutamine, 0.1 mM non-
To determine the tumour specificity of 1, VI and cisplatin the
non-cancerous cell culture, primary human lymphocytes, isolated
as described by Anderson et al. [40], from heparinised human
blood, obtained from healthy volunteers, was used. Cell suspen-
sions were prepared using Histopaque-1077 (Sigma–Aldrich, St.
Louis, MO, USA) as described by Smit et al. [41]. The cells were
maintained in RPMI medium. The bovine FCS that was used to sup-
plement the growth media was heat inactivated at 56 °C for
30 min. All the cultures were cultivated in the presence of 1% pen-
icillin and streptomycin and were maintained at 37 °C with 5% CO2.
Cultures were sub-cultured as needed.
[7] F. Bachechi, A. Burini, R. Galassi, B.R. Pietroni, D. Tesei, Z. Kristallogr. 214
(1999) 497.
[8] D. Leonesi, A. Lorenzotti, A. Cingolani, F. Bonati, Gazz. Chim. Ital. 111 (1981)
483.
[9] D. Matovic´, D.J. Radanovic´, G. Ponticelli, P. Scano, I.A. Efimenko, Transition Met.
Chem. 19 (1994) 461.
[10] S. Cronje, H.G. Raubenheimer, H.S.C. Spies, C. Esterhuysen, H. Schmidbaur, A.
Schier, G.J. Kruger, Dalton Trans. 14 (2003) 2859.
[11] E.R.T. Tiekink, Inflammopharmacology 16 (2008) 138. and references therein.
[12] S. Nobili, E. Mini, I. Landini, C. Gabbiani, A. Casini, L. Messori, Med. Res. Rev. 30
(2010) 550. and references therein.
[13] E. Vergara, E. Cerrada, A. Casini, O. Zava, M. Laguna, P.J. Dyson, Organometallics
29 (2010) 2596. and references therein.
[14] E.R.T. Tiekink, Crit. Rev. Oncol./Hematol. 42 (2002) 225.
[15] P.J. Barnard, S.J. Berners-Price, Coord. Chem. Rev. 251 (2007) 1889. and
references therein.
[16] C.K. Mirabelli, D.T. Hill, L.F. Faucette, F.L. McCabe, G.R. Girard, D.B. Byran, B.M.
Sutton, J. O‘Leary Bartus, S.T. Crooke, R.K. Johnson, J. Med. Chem. 30 (1987)
2181.
[17] K. Spiegel, A. Magistrato, P. Carloni, J. Reedijk, M.L. Kelin, J. Phys. Chem. B 111
(2007) 11873.
[18] U.E.I. Horvath, G. Bentivoglio, M. Hummel, H. Schottenberger, K. Wurst, M.J.
Nell, C.E. Janse van Rensburg, S. Cronje, H.G. Raubenheimer, New J. Chem. 32
(2008) 533.
[19] A. Luquin, E. Cerrada, M. Laguna, in: F. Mohr (Ed.), Gold Chemistry:
Applications and Future Directions in the Life Sciences, Wiley-VCH Verlag,
Weinheim, Germany, 2009, p. 93;
(b) J. Coetzee, W.F. Gabrielli, K. Coetzee, O. Schuster, S.D. Nogai, S. Cronje, H.G.
Raubenheimer, Angew. Chem., Int. Ed. 46 (2007) 2497.
[20] A. Burini, B.R. Pietroni, R. Galassi, G. Valke, S. Cologero, Inorg. Chim. Acta 229
(1995) 299.
4.11.3. Cytotoxicity assays
Cytotoxicity assays were performed to establish the sensitivity
of cancer cell lines and normal cell cultures to the experimental
compounds. 5 ꢂ 102 cells/well were exposed to different concen-
trations (0.5–50 lM) of the complexes in 96-well tissue culture
plates and incubated in a 5% CO2 incubator for 7 days at 37 °C.
Drug-free solvent controls were included. The IC50 data for cis-
platin were determined under the same conditions. Dose–response
assays were performed to derive IC50 values against resting and
phytohaemagglutinin(PHA)-stimulated human lymphocytes in or-
der to determine whether cytotoxicity observed against HeLa cells
may not be tumour specific. The final concentration of PHA that
was added to the relevant wells was 5 lg/ml. Cancer cells were
seeded at 4–5 ꢂ 102/well and were incubated for 7 days. Lympho-
cytes were seeded at 2 ꢂ 105/well and incubated for 3 days. The
viability of cells was determined with an MTT assay [42] originally
described by Mosmann [43] with modifications by van Rensburg
[21] (a) E. Takach, R.H. Henry, D.W. Moore, W.M. Tolles, G.A. Gray, Magn. Res.
Chem. 24 (1986) 984;
et al. [44]. A volume of 20 ll MTT [3-(4,5-dimethylthiazol-2-yl)-
2,5-diphenyltetrazolium bromide] (Sigma Diagnostics Inc.) was
added to each well. MTT is a pale yellow substance that is metab-
olized to dark blue formazan crystals by unaffected metabolically
active cells, which is then quantified by means of spectrometry.
The same method was used for both cancer cells and normal cell
lines. A minimum of three independent experiments were per-
formed where the availability of the test compounds allowed. Data
were processed using a Wilcoxon signed-rank test at a confidence
interval of 95% and with GRAPHAD Prism 4 softwareÓ.
(b) M. Forkey, W.R. Carpenter, Org. Mass Spectrosc. 2 (1969) 433;
(c) W.F. Gabrielli, S.D. Nogai, J.M. McKenzie, S. Cronje, H.G. Raubenheimer,
New J. Chem. 33 (2009) 2208.
[22] (a) H. Schmidbaur, A. Wohlleben, F.E. Wagner, D.F. Van de Vondel, G.P. Van der
Kelen, Chem. Ber. 110 (1977) 2758;
(b) P. Pyykkö, Chem. Soc. Rev. 37 (2008) 1967;
(c) H. Schmidbaur, A. Schier, Chem. Soc. Rev. 41 (2012) 370.
[23] V.J. Catalano, S.J. Horner, Inorg. Chem. 42 (2003) 8430.
[24] K. Nomiya, R. Noguchi, M. Oda, Inorg. Chim. Acta 298 (2000) 24.
[25] R.L. Kieft, W.M. Peterson, G.L. Blundell, S. Horton, R.A. Henry, H.B. Jonassen,
Inorg. Chem. 15 (1976) 1721.
[26] P.J. Barnard, L.E. Wedlock, M.V. Baker, S.J. Berners-Price, D.A. Joyce, B.W.
Skelton, J.A. Steer, Angew. Chem., Int. Ed. 45 (2006) 5966. and references
therein.
Acknowledgements
[27] Y. Satoh, N. Marcopulos, Tetrahedron Lett. 36 (1995) 1759.
[28] L.J. Mathias, D. Burkett, Tetrahedron Lett. 49 (1979) 4709.
[29] J.-F. Ma, J. Yang, G.-L. Zheng, L. Li, J.-F. Liu, Inorg. Chem. 42 (2003) 7531.
[30] R. Bronisz, Inorg. Chim. Acta 340 (2002) 215.
[31] R. Mostaghim, Y.A. Beni, Indian J. Chem. 40B (2001) 498.
[32] R. Usòn, A. Laguna, in: R.B. Lang, J.J. Eisch (Eds.), Organometallic Synthesis, vol.
3, Elsevier, Amsterdam, 1986, p. 325.
We gratefully thank the Alexander von Humboldt Stiftung
(H.G.R., S.C., S.N.) and the NRF (National Research Foundation,
South Africa) for financial support and Harmony Gold Mining Co.
Ltd. (WFG), through Mintek (Project Autek), for the generous loan
of gold.
[33] R. Usòn, A. Laguna, in: R.B. Lang, J.J. Eisch (Eds.), Organometallic Synthesis, vol.
3, Elsevier, Amsterdam, 1986, p. 326.
[34] SMART Data Collection Software, Version 5.629, Bruker AXS Inc., Madison, WI,
2003.
Appendix A. Supplementary data
[35] SAINT, Data Reduction Software, Version 6.45, Bruker AXS Inc., Madison, WI,
2003.
[36] R.H. Blessing, Acta Crystallogr., Sect. A 51 (1995) 33.
[37] SADABS, Version 2.05, Bruker AXS Inc., Madison, WI, 2002.
[38] G.M. Shelrick, SHELX-97. Program for Crystal Structure Analysis, University of
Göttingen, Germany, 1997.
CCDC 847311, 847312, 847313, 847314, 847315 contain the
supplementary crystallographic data for 1, 2, 3, 4ꢁCH2Cl2, 6 and
7ꢁC8H8N4, respectively. These data can be obtained free of charge
via http://www.ccdc.cam.ac.uk/conts/retrieving.html, or from the
Cambridge Crystallographic Data Centre, 12 Union Road, Cam-
bridge CB2 1EZ, UK; fax: (+44) 1223-336-033; or e-mail:
deposit@ccdc.cam.ac.uk.
[39] L.J. Barbour, J. Supramol. Chem. 1 (2001) 189.
[40] R. Anderson, M.J. Smit, C.E. Janse van Rensburg, Mol. Pharmacol. 44 (1993) 605.
[41] T. Smit, J.R. Snyman, E.W. Neuse, L. Bohm, C.E. Janse van Rensburg, Anti-Cancer
Drug 16 (2005) 501.
[42] G.D. Hoke, G.F. Rush, G.E. Bossard, J.V. McArdle, B.D. Jensen, C.K. Mirabelli, J.
Biol. Chem. 263 (1988) 11203.
[43] T. Mosmann, J. Immunol. Meth. 65 (1983) 55.
References
[44] C.E. Janse van Rensburg, R. Anderson, G. Joone, M.S. Myer, J.F. O’Sullivan, Anti-
Cancer Drug 8 (1997) 708.
[1] M.C. Gimero, A. Laguna, in: J.A. McCleverty, T.J. Meyer, D.E. Fenton (Eds.),
Comprehensive Coordination Chemistry II, vol. 6, Elsevier, Amsterdam, 2004, p.
911.