Impact of various lipophilic substituents on ruthenium(II), rhodium(III) and iridium(III) salicylaldimine-based complexes: synthesis, in vitro cytotoxicity studies and DNA interactions

Article abstract of DOI:10.1007/s00775-018-1567-3

Abstract: A series of bidentate salicylaldimine ligands was prepared and reacted with either [RuCl(μ-Cl)(p-cymene)]₂, [RhCl(μ-Cl)(Cp*)]₂ or [IrCl(μ-Cl)(Cp*)]₂. All of the compounds were characterised using an array of spectroscopic and analytical techniques, namely, nuclear magnetic resonance (NMR) spectroscopy, infrared (IR) spectroscopy and mass spectrometry. Single crystal X-ray diffraction (XRD) was used to confirm the bidentate coordination mode of the salicylaldimine ligand to the metal centre. The platinum group metal (PGM) complexes were screened against the MCF7 breast cancer cell line. The ruthenium and iridium salicylaldimine complexes showed comparable or greater cytotoxicity than cisplatin against the MCF7 cancer cells, as well as greater cytotoxicity than their rhodium counterparts. Three of the salicylaldimine complexes showed potent activity in the range 18–21?μM. Two of these complexes had a greater affinity for cancerous cells than for CHO non-cancerous cells (SI > 4). Preliminary mechanistic studies suggest that the ruthenium complexes undergo solvation prior to 5′-GMP binding, whereas the iridium complexes were inert to the solvation process. Graphical abstract: [Figure not available: see fulltext.].

Full text of DOI:10.1007/s00775-018-1567-3

JBIC Journal of Biological Inorganic Chemistry
https://doi.org/10.1007/s00775-018-1567-3
ORIGINAL PAPER
Impact of various lipophilic substituents on ruthenium(II), rhodium(III)
and iridium(III) salicylaldimine‑based complexes: synthesis, in vitro
cytotoxicity studies and DNA interactions
Irwin Cassells1 · Tameryn Stringer1 · Alan T. Hutton1 · Sharon Prince2 · Gregory S. Smith1
Received: 14 March 2018 / Accepted: 28 April 2018
© SBIC 2018
Abstract
A series of bidentate salicylaldimine ligands was prepared and reacted with either [RuCl(µ-Cl)(p-cymene)]2, [RhCl(µ-Cl)
(Cp*)]2 or [IrCl(µ-Cl)(Cp*)]2. All of the compounds were characterised using an array of spectroscopic and analytical
techniques, namely, nuclear magnetic resonance (NMR) spectroscopy, infrared (IR) spectroscopy and mass spectrometry.
Single crystal X-ray difraction (XRD) was used to confrm the bidentate coordination mode of the salicylaldimine ligand
to the metal centre. The platinum group metal (PGM) complexes were screened against the MCF7 breast cancer cell line.
The ruthenium and iridium salicylaldimine complexes showed comparable or greater cytotoxicity than cisplatin against the
MCF7 cancer cells, as well as greater cytotoxicity than their rhodium counterparts. Three of the salicylaldimine complexes
showed potent activity in the range 18–21 µM. Two of these complexes had a greater afnity for cancerous cells than for
CHO non-cancerous cells (SI>4). Preliminarymechanistic studies suggest that the ruthenium complexes undergo solvation
prior to 5′-GMP binding, whereas the iridium complexes were inert to the solvation process.
Crystallographic data: CCDC 1572708 (11), CCDC 1572709 (12),
CCDC 1572710 (13).
* Gregory S. Smith
Gregory.Smith@uct.ac.za
1
Department of Chemistry, University of Cape Town,
Rondebosch 7701, South Africa
2
Department of Human Biology, University of Cape Town,
Medical School, Observatory 7925, South Africa
Vol.:(0123456789)
1 3

JBIC Journal of Biological Inorganic Chemistry
Graphical abstract
Keywords Salicylaldimine · Ruthenium–arene · Rhodium · Iridium · Anticancer activity
organometallic complexes through the splitting of the metal
dimers using N,N-, N,O- or O,O-bidentate ligands, where
complexes containing N,O-chelates were found to exhibit
the greatest activity [8, 9].
Introduction
Since the discovery of the anticancer properties of cispl-
atin by Rosenberg [1, 2], metal-based treatments have been
of interest to researchers. Cisplatin is a frequently used
metal-based drug against ovarian and testicular cancers
[3]. Considering the high levels of toxicity and unfavour-
able side efects of cisplatin, a range of biologically active
organometallic complexes, such as the half-sandwich piano
stool Ru(II) complexes [4, 5], has been investigated to treat
various cancers with reduced side efects. These platinum
group metal (PGM) complexes have shown exceptional
promise, with the development of a few complexes, such as
[(η6-p-cymene)Ru(en)Cl]PF6 and [(η6-p-cymene)Ru(PTA)
Cl2] (where en = ethylenediamine and PTA = 1,3,5-triaza-
7-phosphaadamantane), continuing to the pre-clinical stage
[6, 7]. Of these, piano stool structures with either an η6-p-
cymene or η5-pentamethylcyclopentadienyl (Cp*) ligand, in
general, have shown the most promise. The dimeric precur-
sors ([MCl2(p-cymene or Cp*)]2, M=platinum group metal)
of these complexes have been shown to be biologically
inactive, strongly suggesting that the activity of these com-
plexes is ligand dependent [8, 9]. Recently, McGowan and
co-workers have observed an increase in activity of these
Salicylaldimine ligands (N,O-chelates) are versatile and
easily functionalised building blocks for small molecules
and larger macromolecular metal complexes and selected
examples currently used in a wide range of biological
Fig. 1 Selected salicylaldimine structures that have shown promise in
biological applications [13–15, 19, 20]
1 3

JBIC Journal of Biological Inorganic Chemistry
applications are shown in Fig. 1 [10–18]. Salicylaldimine
compounds enhance the distribution of the metal into cell
organelles and interrupt DNA processes, increasing the
activity of the complex [16, 17]. Various salicylaldimine
PGM complexes have been reported to show moderate or
low activity against human ovarian cancer cell lines, A2780
and A2780cis [13, 19]. However, in other cell lines, namely
MCF7 (human breast carcinoma) and A549 (lung cancer),
similar complexes show comparable activity to cisplatin
[15, 20]. Three lipophilic groups of interest for this study
have been the ferrocenyl, trifuoromethyl and trimethylsi-
lane groups, which have been reported to show biological
activity [18, 21–23]. Ferrocenyl groups have the ability to
promote single electron redox transfers and have shown high
cytotoxicity in various invasive cancer cell lines in vitro [24,
25]. Fluorine groups are resistant to chemical degradation
[21], favourable for enzyme inhibition [21, 26], and have
shown inhibition of tumour growth at third- and fourth-
line cancers [27]. In a recent study, an increased activity
in chloroquine-sensitive (NF54) and chloroquine-resistant
(Dd2) Plasmodium falciparum strains of the malaria parasite
was discovered for organosilane complexes in comparison to
carbon analogues, further supporting its relevance in drug
discovery [22, 23].
elemental analyser. Electron impact (EI) mass spectrometry
was carried out on a JEOL GCmate II mass spectrometer
(1–10). Electrospray ionisation (ESI) mass spectrometry
was carried out on a Waters API Quattro instrument in the
positive-ion mode for complexes 11–16. Melting points were
determined using a Büchi melting point B-540 apparatus and
are uncorrected. Dimeric precursors [28, 29], compounds
1–3 [19, 30] and 5–10 [18, 19] were synthesised using pub-
lished methods.
Synthesis of (trimethylsilyl)propylsalicylaldimine
ligand (4)
Salicylaldehyde (0.437 g, 3.58 mmol) and 3-aminopropyl-
trimethylsilane (0.548 g, 4.17 mmol) were stirred in diethyl
ether under an inert atmosphere for 16 h. The resulting yel-
low solution was then washed with water (3 × 15 mL) to
remove the excess amine. The organic fractions were col-
lected, dried over anhydrous magnesium sulfate, fltered and
the solvent removed to yield a yellow oil (4). Yield: 0.783 g
(93%). 1H NMR (400 MHz, CDCl3): δH (ppm)=0.02 (9H,
s, CH3), 0.53–0.59 (2H, m, CH2), 1.66–1.76 (2H, m, CH2),
3
3
3.60 (2H, t, J = 7.4 Hz, CH2), 6.89 (1H, t, J = 7.4 Hz,
Ar–H), 6.99 (1H, d, 3J=8.3 Hz, Ar–H), 7.26–7.35 (2H, m,
Ar–H), 8.35 (1H, s, HC=N), 13.70 (1H, s, OH). 13C{1H}
NMR (100 MHz, CDCl3): δC (ppm)=−1.79 (CH3), 14.14
(CH2), 25.51 (CH2), 62.68 (CH2), 117.03 (Ar), 118.30 (Ar),
118.81 (Ar), 131.04 (Ar), 131.97 (Ar), 161.47 (Ar), 164.42
(Cimine). EI-MS m/z 235.1259 (100%, [M]+). Analysis Calc.
for C13H21NOSi: C, 66.33; H, 8.99; N, 5.95. Found: C,
66.31; H, 9.20; N, 5.96.
The combination of salicylaldimine ligands with PGM
dimeric precursors is thus a prolifc area of research and a
popular strategy in drug design, and we and other research-
ers have previously reported on this strategy [13, 18, 19, 22,
23]. This study details the synthesis and characterisation
of trifuoromethane- and organosilane-derivatised salicy-
laldimine complexes, as well as ferrocenyl- and propyl-
salicylaldimine complexes, with a particular focus on the
infuence of these lipophilic substituents on the biological
activity of salicylaldimine-containing metal complexes.
General procedure for the synthesis of complexes
11–13
Ligand 3 and triethylamine were stirred together in dichlo-
romethane for 60 min. To the yellow solution, [RuCl(µ-Cl)
(p-cymene)]2, [RhCl(µ-Cl)(Cp*)]2 or [IrCl(µ-Cl)(Cp*)]2
was added and stirred for an additional 18 h. The resulting
solution was washed with water (3×15 mL) in a separating
funnel, and the organic fractions collected and dried using
anhydrous magnesium sulfate. After removal of the drying
agent by fltration, the solvent was removed and the residue
taken up in a minimal amount of dichloromethane, to which
diethyl ether was added to precipitate the desired complex.
Experimental
Materials and methods
Reagents were purchased from either Sigma-Aldrich or
Merck and used as received. Solvents (from KIMIX) were
dried using molecular sieves. Nuclear magnetic reso-
nance (NMR) spectra were recorded on a Bruker Topspin
GmbH (1H: 400.22 MHz, 13C{1H}: 100.65 MHz, 19F{1H}:
376.58 MHz, 31P{1H}: 162.01 MHz) or a Varian Mercury
XR300 (1H: 300.08 MHz) spectrometer, with a Bruker Bio-
spin GmbH casing and sample injector at 30 °C. Chemical
shifts were reported using tetramethylsilane (TMS) as the
internal standard. Infrared (IR) absorptions were measured
on a Perkin-Elmer Spectrum One FT-IR spectrometer using
attenuated total refection (ATR) in the solid state. Elemen-
tal analysis was carried out using a Fisons EA 110 CHNS
Data for ruthenium–CF3 complex (11)
Compound 3 (0.0875 g, 0.330 mmol) and triethylamine
(0.1 mL, 0.717 mmol) were reacted with [RuCl(µ-Cl)
(p-cymene)]2 (0.101 g, 0.164 mmol). Complex 11 was
isolated as a brown powder. Yield: 0.0833 g (47%). Melt-
ing point: 226 °C dec. with melting. 1H NMR (400 MHz,
1 3

JBIC Journal of Biological Inorganic Chemistry
3
CDCl3): δH (ppm) = 1.14 (3H, d, J = 6.8 Hz, CH3), 1.19
(ppm)=8.43 (CH3), 85.81 (Cp*), 115.21 (Ar), 120.09 (Ar),
123.53 (Ar), 123.80 (q, 1JC–F =272.6 Hz, CF3), 125.45 (Ar),
125.57 (q, 2JC–F =3.7 Hz, Ar), 129.77 (q, 2JC–F =32.5 Hz,
Ar), 134.65 (Ar), 135.77 (Ar), 159.12 (Ar), 161.41
(Cimine), 164.34 (Ar). 19F{1H} NMR (376 MHz, CDCl3):
δF (ppm) = − 62.17 (s). IR (ATR): ν/cm−1 = 1615 (C=N).
ESI–MS (HR) m/z 592.1528 (80%, [M−Cl]+ requires
592.1439). Analysis Calc. for C24H24ClF3NOIr: C, 45.97;
H, 3.86; N, 2.23. Found: C, 45.33; H, 3.87; N, 1.95.
(3H, d, 3J=6.8 Hz, CH3), 2.13 (3H, s, CH3), 2.64 (1H, sept,
3J=6.8 Hz, CH), 4.25 (1H, d, 3J=5.4 Hz, Arpcy), 5.01 (1H,
d, 3J=5.4 Hz, Arpcy), 5.28 (1H, d, 3J=6.4 Hz, Arpcy), 5.36
(1H, d, 3J=6.4 Hz, Arpcy), 6.42–6.44 (1H, m, Ar–H), 6.95
3
4
(1H, dd, J = 7.8 Hz, J = 1.8 Hz, Ar–H), 6.99–7.01 (1H,
m, Ar–H), 7.22–7.27 (1H, m, Ar–H), 7.72–7.74 (3H, m,
Ar–H+HC=N), 7.80–7.82 (2H, m, Ar–H). 13C{1H} NMR
(100 MHz, CDCl3): δC (ppm)=18.47 (CH3), 21.68 (CH3),
22.66 (CH3), 30.41 (CH), 80.25 (Cpcy), 83.08 (Ccym), 83.25
(Cpcy), 86.59 (Cpcy), 98.31 (Cpcy), 101.74 (Cpcy), 114.50
(Ar), 117.95 (Ar), 122.91 (Ar), 123.85 (q, 1JC–F =270.1 Hz,
General procedure for the synthesis of complexes
14–16
3
CF3), 124.38 (Ar), 126.14 (q, JC–F = 2.9 Hz, Ar), 129.12
(q, 2JC–F =30.0 Hz, Ar), 135.46 (Ar), 136.08 (Ar), 160.83
(Ar), 164.6 (Cimine), 165.66 (Ar). 19F{1H} NMR (376 MHz,
CDCl3): δF (ppm) = − 62.21 (s). IR (ATR): ν/cm−1 =1615
(C=N). ESI–MS (HR) m/z 238.1773 (100%, [M+H+Na]2+
requires 238.2999). Analysis Calc. for C24H23ClF3NORu: C,
53.88; H, 4.33; N, 2.62. Found: C, 53.55; H, 4.14; N, 2.24.
[RuCl(µ-Cl)(p-cymene)]2, [RhCl(µ-Cl)(Cp*)]2 or [IrCl(µ-Cl)
(Cp*)]2 was added to a stirring solution of triethylamine
and 4 in diethyl ether and stirred for 18 h at room tempera-
ture. The excess dimer and triethylammonium chloride
that formed were fltered and a red or orange solution was
obtained. The solvent was removed in vacuo and the residue
washed with n-pentane to yield the desired complexes.
Data for rhodium–CF3 complex (12)
Data for ruthenium–silane complex (14)
Compound 3 (0.0866 g, 0.327 mmol) and triethylamine
(0.1 mL, 0.717 mmol) were reacted with [RhCl(µ-Cl)
(Cp*)]2 (0.101 g, 0.167 mmol). Complex 12 was isolated
as an orange powder. Yield: 0.132 g (75%). Melting point:
Compound 4 (0.0876 g, 0.372 mmol) and triethylamine
(0.1 mL, 0.717 mmol) were reacted with [RuCl(µ-Cl)
(p-cymene)]2 (0.137 g, 0.224 mmol). Complex 14 was iso-
lated as a dark brown crystalline powder. Yield: 0.0884 g
(46%). Melting point: 146 °C dec. with melting. 1H
NMR (400 MHz, CDCl3): δH (ppm) = 0.05 (9H, s, CH3),
0.44–0.54 (1H, m, CH2), 0.60–0.70 (1H, m, CH2), 1.15
1
241 °C dec. with melting. H NMR (400 MHz, CDCl3):
δH (ppm)=1.35 (15H, s, CH3), 6.43–6.47 (1H, m, Ar–H),
7.01–7.07 (2H, m, Ar–H), 7.24–7.28 (1H, m, Ar–H), 7.69
(2H, d, 3J=8.3 Hz, Ar–H), 7.94 (1H, s, HC=N), 8.01 (2H,
d, 3J=8.1 Hz, Ar–H). 13C{1H} NMR (100 MHz, CDCl3): δC
(ppm)=8.42 (CH3), 93.67 (d, JC–Rh =8.8 Hz, Cp*), 114.33
(Ar), 119.21 (Ar), 123.96 (q, 1JC–F =272.2 Hz, CF3), 124.35
3
3
(3H, d, J = 6.9 Hz, CH3), 1.25 (3H, d, J = 7.1 Hz, CH3),
1.84–1.96 (1H, m, CH2), 2.00–2.14 (1H, m, CH2), 2.22 (3H,
3
s, CH3), 2.79 (1H, sept, J = 7.0 Hz, CH), 3.94–4.04 (1H,
3
(Ar), 125.22 (Ar), 125.83 (q, JC–F = 3.7 Hz, Ar), 128.90
m, CH2), 4.20–4.29 (1H, m, CH2), 5.02 (1H, d, 3J=8.6 Hz,
Arpcy), 5.39 (3H, m, Arpcy), 6.38–6.42 (1H, m, Ar–H),
6.90–6.97 (2H, m, Ar–H), 7.13–7.19 (1H, m, Ar–H), 7.67
(1H, s, HC=N). 13C{1H} NMR (100 MHz, CDCl3): δC
(ppm) = − 1.70 (CH3), 14.13 (CH2), 18.49 (CH3), 21.61
(CH3), 22.70 (CH3), 25.74 (CH2), 30.46 (CH), 72.69 (CH2),
80.11 (Arpcy), 82.07 (Arpcy), 83.25 (Arpcy), 85.86 (Arpcy),
97.24 (Arpcy), 101.57 (Arpcy), 113.97 (Ar), 119.21 (Ar),
122.22 (Ar), 134.33 (Ar), 134.46 (Ar), 163.34 (Cimine),
164.92 (Ar). IR (ATR): ν/cm−1 = 1618 (C=N). ESI–MS
(HR) m/z 470.1460 (100%, [M−Cl]+ requires 470.1459).
Analysis Calc. for C23H34ClNORuSi·0.6H2O: C, 53.42; H,
6.89; N, 2.71. Found: C, 53.17; H, 6.81; N, 2.47.
(q, 2JC–F =32.5 Hz, Ar), 135.51 (Ar), 135.81 (Ar), 157.15
(Ar), 165.10 (Cimine), 166.80 (Ar). 19F{1H} NMR (376 MHz,
CDCl3): δF (ppm) = − 62.15 (s). IR (ATR): ν/cm−1 =1603
(C=N). ESI–MS (HR) m/z 502.0861 (100%, [M−Cl]+
requires 502.0865). Analysis Calc. for C24H24ClF3NORh: C,
53.60; H, 4.50; N, 2.60. Found: C, 53.33; H, 4.31; N, 2.48.
Data for iridium–CF3 complex (13)
Compound 3 (0.0685 g, 0.258 mmol) and triethylamine
(0.1 mL, 0.717 mmol) were reacted with [IrCl(µ-Cl)
(Cp*)]2 (0.0998 g, 0.125 mmol). Complex 13 was isolated
as an orange powder. Yield: 0.108 g (69%). Melting point:
1
243 °C dec. with melting. H NMR (400 MHz, CDCl3):
Data for rhodium–silane complex (15)
3
δH (ppm) = 1.33 (15H, s, CH3), 6.46 (1H, t, J = 7.2 Hz,
3
Ar–H), 6.97 (1H, d, J = 8.6 Hz, Ar–H), 7.11 (1H, d,
Compound 4 (0.0997 g, 0.423 mmol) and triethylamine
(0.1 mL, 0.717 mmol) were reacted with [RhCl(µ-Cl)
(Cp*)]2 (0.126 g, 0.204 mmol). Complex 15 was isolated
as a red powder. Yield: 0.101 g (49%). Melting point:
3
3J = 7.6 Hz, Ar–H), 7.37 (1H, t, J = 7.2 Hz, Ar–H), 7.67
(2H, d, 3J=8.1 Hz, Ar–H), 7.85 (2H, d, 3J=8.1 Hz, Ar–H),
8.01 (1H, s, HC=N). 13C{1H} NMR (100 MHz, CDCl3): δC
1 3

JBIC Journal of Biological Inorganic Chemistry
1
166 °C dec. with melting. H NMR (400 MHz, CDCl3):
images. The crystallographic data for complexes 11–13
δH (ppm) = − 0.01 (9H, s, CH3), 0.48–0.52 (2H, m, CH2),
1.53 (15H, s, CH3), 1.68–1.71 (1H, m, CH2), 2.15–2.25
(1H, m, CH2), 3.80–4.03 (2H, m, CH3), 6.41–6.45 (1H,
m, Ar–H), 6.92 (1H, d, 3J = 8.4 Hz, Ar–H), 7.01 (1H, dd,
3J=7.7 Hz, 4J=1.7 Hz, Ar–H), 7.13–1.17 (1H, m, Ar–H),
7.78 (1H, s, HC=N). 13C{1H} NMR (100 MHz, CDCl3):
δC (ppm) = − 1.72 (CH3), 8.61 (CH3), 14.02 (CH2), 25.28
are presented in Table 2.
Cytotoxicity studies
Cells (MCF7) were seeded in a 96-well plate (5000 cells
per well) and after 48 h treated with either vehicle (1.0 μM
in DMSO) or complexes 5–16 and cisplatin at 20 μM for
48 h. The impact of these complexes on cell viability of
these complexes was assessed using 3-(4,5-dimethylthia-
zol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay, as
described in the literature [35]. The absorbance at 550 nm
was determined for each well using a spectrophotometer
(RTQ2100C Microplate Reader, Ryto, China) and normal-
ised to the RPMI medium absorbance. This experiment was
performed three times in quadruplicate, and then the mean
cell viability determined using GraphPad Prism version 5.0.
For IC50 (concentration required for 50% viability) determi-
nation of 11, 13 and 14, cells were treated using concentra-
tions of 5–25 μM on three occasions in quadruplicate. For
the Chinese hamster ovarian (CHO) cells, 3000 cells per
well were plated and tested at concentrations of 0.001, 0.01,
0.1, 1, 10 and 100 μg/mL.
1
(CH2), 66.97 (CH2), 92.65 (d, JC–Rh = 8.1 Hz, Cp*),
114.05 (Ar), 121.92 (Ar), 123.70 (Ar), 133.71 (Ar),
133.90 (Ar), 162.31 (Cimine), 166.60 (Ar). IR (ATR):
ν/cm−1 = 1621 (C=N). ESI–MS (HR) m/z 472.1551
(100%, [M−Cl]+ requires 472.1543). Analysis Calc. for
C23H35ClNORhSi·0.2Et2O: C, 54.68; H, 7.13; N, 2.68.
Found: C, 55.32; H, 6.86; N, 2.39.
Data for iridium–silane complex (16)
Compound 4 (0.0337 g, 0.143 mmol) and triethylamine
(0.1 mL, 0.717 mmol) were reacted with [IrCl(µ-Cl)
(Cp*)]2 (0.0697 g, 0.0875 mmol). Complex 16 was iso-
lated as a fne yellow powder. Yield: 0.0478 g (56%).
Melting point: 158 °C dec. with melting. 1H NMR
(400 MHz, CDCl3): δH (ppm) = 0.03 (9H, s, CH3), 0.51
(2H, br m, CH2), 1.56 (15H, s, CH3), 1.70–1.93 (1H, br
m, CH2), 1.94–2.17 (1H, br m, CH2), 3.75–4.20 (2H, br
m, CH2), 6.40–6.44 (1H, m, Ar–H), 6.87–6.90 (1H, m,
NMR stability studies in DMSO and interactions
with DNA model 5′‑GMP
The stability of complexes 11 and 13 was investigated by
1H NMR experiments in which approximately 3.0 mg of
the complex was dissolved in 0.5 mL DMSO-d6, 0.5 mL of
50% H2O in DMSO-d6 (by volume) or 150 mM NaCl solu-
tion in 0.5 mL of 50% H2O in DMSO-d6 (by volume). The
NMR samples were heated and maintained at 37 °C between
sampling. 1H NMR spectra were collected at 0, 24 and 48 h
after the initial sample preparation. Similarly, approximately
3.0 mg of sodium 5′-guanosine monophosphate was dis-
solved in 0.25 mL H2O, and mixed with 3.0 mg of complex
in 0.25 mL DMSO-d6. The sample was monitored at 0, 24
and 48 h after preparation while heating to 37 °C between
NMR sampling.
3
4
Ar–H), 7.04 (1H, dd, J = 7.7 Hz, J = 1.8 Hz, Ar–H),
7.21–7.25 (1H, m, Ar–H), 7.75 (1H, s, HC=N). 13C{1H}
NMR (100 MHz, CDCl3): δC (ppm) = − 1.70 (CH3), 8.83
(CH3), 14.03 (CH2), 25.45 (CH2), 69.86 (CH2), 84.64
(Cp*), 114.79 (Ar), 121.88 (Ar), 122.71 (Ar), 133.15
(Ar), 134.00 (Ar), 159.72 (Cimine), 165.19 (Ar). IR (ATR):
ν/cm−1 = 1619 (C=N). ESI–MS (HR) m/z 562.2133
(100%, [M−Cl]+) requires 562.2117. Analysis Calc. for
C23H35ClIrNOSi·0.2Et2O: C, 46.70; H, 6.09; N, 2.29.
Found: C, 47.11; H, 5.86; N, 1.89.
X‑ray crystallography
Single crystal X-ray difraction data were collected on a
Bruker KAPPA APEX II DUO difractometer using graph-
ite-monochromated Mo-Kα radiation (λ = 0.71073Å).
The temperature was controlled using an Oxford Cryo-
stream cooling system (Oxford Cryostat). Data scaling and
absorption correction were performed using SAINT [31]
and SADABS [32]. Structures were solved by direct meth-
ods using SHELXS-97 and refned by full-matrix least-
squares methods based on F2 using SHELXL-2014 [33,
34] with X-Seed used as the graphical interface. X-Seed
and POV-Ray were used to prepare molecular graphic
Results and discussion
Synthesis and characterisation
A small series of four salicylaldimine ligands (Scheme 1)
was synthesised following published literature methods [19,
30]. Salicylaldehyde was reacted with either n-propylamine,
4-trifuoroaniline or 3-aminopropyltrimethylsilane via a
Schif-base condensation reaction to aford ligands 1, 3 and
4. The Schif-base products were isolated as either viscous
yellow oils (1 and 4) or an amorphous yellow powder (3) in
1 3

JBIC Journal of Biological Inorganic Chemistry
Scheme 1 Synthesis of
ligands 1–4. a n-Propylamine/
dichloromethane/18 h/r.t. b
(i) NH2NH2·xH2O/1 h/r.t, (ii)
ferrocenecarboxaldehyde/etha-
nol/18 h/r.t. c 4-(Trifuorome-
thyl)aniline/2 h/r.t. d 3-Ami-
nopropyltrimethylsilane/16 h/
Ar(g)/r.t
high yields (91–93%). Compound 2 was synthesised using
salicylaldehyde hydrazone and ferrocenecarboxaldehyde,
and was isolated as a deep red amorphous solid in high yield
(75%) [18].
observed for the carbons closest to the fuorine atoms. Cou-
1
2
3
pling constants for JC–F, JC–F and JC–F were observed
and found to be 272.0, 32.5 and 3.7 Hz, respectively, and
these signals were observed as quartets. A single peak was
observed in the 19F{1H} NMR spectrum of 3 at δF −62 ppm.
The IR spectra of compounds 1–4 displayed absorption
bands for the imine C=N between 1585 and 1618 cm−1,
providing further support that the proposed compounds had
been synthesised.
1
Analysis of the H NMR spectra of compounds 1–4
revealed the absence of the aldehyde resonance at δH
9.87 ppm (salicylaldehyde), confrming the successful syn-
thesis of the Schif-base ligands. The presence of a singlet in
the 1H NMR spectra of 1–4, in the region of δH 8.3–8.7 ppm,
confrmed the presence of the imine functionality. In the
13C{1H} NMR spectra of 1–4, signals for the C-OH car-
bon were observed at δC 161 ppm. In 3, C-F coupling was
A series of N,O-salicylaldimine Ru(II), Rh(III) and Ir(III)
complexes, 5–16, was synthesised using the aforemen-
tioned ligands via a bridge splitting reaction of [RuCl(µ-Cl)
Scheme 2 Synthesis of com-
plexes 5–16. a 1–4/[RuCl(µ-Cl)
(p-cymene)]2/Et3N/r.t. b 1–4/
[RhCl(µ-Cl)(Cp*)]2/Et3N/r.t. c
1–4/[IrCl(µ-Cl)(Cp*)]2/Et3N/r.t
1 3

JBIC Journal of Biological Inorganic Chemistry
Fig. 2 Molecular structures of 11, 12 and 13. Hydrogen atoms are omitted for clarity
(p-cymene)]2, [RhCl(µ-Cl)(Cp*)]2 or [IrCl(µ-Cl)(Cp*)]2
(Scheme 2). Complexes 5–7, 8–10, 11–13 and 14–16 were
synthesised by reacting 1–4 in the presence of triethyl-
amine to deprotonate the hydroxyl group and subsequent
reaction with the corresponding metal dimer to form the
desired (N,O)-complexes. Complexes 5–10 were synthesised
according to literature methods [18, 19]. Complexes 11–13
were synthesised using the method reported by Smith et al.
[19] and 14–16 were synthesised in diethyl ether. Complexes
were isolated as amorphous powders in moderate yields.
All spectroscopic and analytical data for 5–10 were in
accordance with the reported literature [18, 19]. In the 1H
NMR spectra of 11–13, an upfeld shift in the imine proton
signal was observed (δH 7.7–8.0 ppm) when compared to
the free ligand (δH 8.68 ppm). Minute shifts in the (trifuo-
romethyl)phenyl group suggest little to no interaction with
the p-cymene or Cp* ligands. In the 13C{1H} NMR spectra
of 11–13, a distinct downfeld shift of the carbon signals
adjacent to the oxygen from δC 151 to 160–161 ppm con-
frmed bonding of the oxygen atom to the metal centre.
The imine carbon resonance exhibited minor shifts from
δC 164 to 161–164 ppm. This was further confrmed using
19F{1H} NMR spectroscopy. As observed in 3, a single
resonance was observed at δF − 62 ppm in the spectra of
11–13, which supports the notion of a single fuorine spe-
cies. The 1H NMR spectra of 14–16 indicate the chirality
of the complexes, and a similar phenomenon is observed
in the spectra of complexes 5–7. Each CH2 proton of the
propyl chain was observed at two separate chemical shifts
and integrated for a solitary proton. An upfeld shift of the
imine signal from δH 8.35 to 7.6–7.6 ppm is indicative of
a change in the chemical environment of the imine sub-
structure. No hydroxyl signal was observed in the 1H NMR
spectra, strongly suggesting that the ligand was deproto-
nated in the synthesis process. As found for complexes
5–7, resonances for the imine and CO-M carbon atoms
in the 13C{1H} NMR spectra were observed to have an
upfeld or downfeld shift, respectively. In the IR spectra,
an overall trend indicating a shift to lower wavenumbers
of the C=N stretching vibration from the spectra of the
metal-free ligands (~ 1630 cm−1) to the PGM complexes
(~ 1615 cm−1) was observed, supporting the notion of
coordination to the N,O-chelate.
1 3

JBIC Journal of Biological Inorganic Chemistry
in relation to each other. A six-membered ring is formed
between the (N,O)-chelating ligand and the metal centre,
further supporting the proposed bidentate structure. Bond
distances between the metal and chlorido ligand are 20%
larger in comparison to the N,O-chelate and the arene/Cp*
group (2.0 Å for N,O compared to 2.4 Å for chlorido). Tor-
sion angles about the imine bond are observed to be 179°,
confrming a trans-confguration of the imine. Selected bond
lengths and angles are given in Table 1 and the crystallo-
graphic information is provided in Table 2.
X‑ray crystallography
Single crystal XRD is a useful technique to understand
the molecular structures of complexes. Single crystals of
11–13 were obtained by slow difusion of diethyl ether into
methanol and the structures of these complexes were elu-
cidated. These complexes crystallised in the P21/c space
group with four molecules present in each unit cell. The
geometry of the metal is consistent with the proposed piano
stool geometry, with the angles between the N, O and Cl
donor atoms ranging between 85° and 89°, indicative of a
pseudo-tetrahedral arrangement (Fig. 2). This is consistent
with the data obtained for complex 5, reported by Smith
et al. [19]. The two phenyl groups are observed to be ~90°
In vitro cytotoxicity studies
Complexes synthesised in this study were pre-screened for
cytotoxicity in the MCF7 carcinoma breast cancer cell line
at an initial dose concentration of 20 µM. This concentration
was chosen based on the National Cancer Institute (NCI,
USA) criteria which considers compounds with IC50 ≤20 µg/
mL as having better potential as anticancer drug leads [36].
Cisplatin was used as the positive control in this study. To
compare the cytotoxicity at the tested concentration, cell
viability was determined using the MTT assay [37, 38].
The anticancer data obtained for complexes 5–16 are
illustrated in Fig. 3. In general, the ruthenium (8, 11 and
14) and iridium complexes (10 and 13) showed greater activ-
ity than their rhodium (9, 12 and 15) counterparts. With
the exceptions of 8, 9 and 15, the complexes were found to
Table 1 Selected bond lengths and angles of complexes 11–13
Bond distance (Å)/angle (°)
Entity
11
12
13
2.095 (3)
2.4210 (9)
2.084 (2)
88.02 (10)
85.59 (7)
85.58 (7)
178.7 (3)
M–N
M–Cl
M–O
N–M–O
N–M–Cl
O–M–Cl
C=N–C–C
2.088 (2)
2.109 (2)
2.4288 (8)
2.0594 (18)
88.01 (7)
89.89 (9)
85.38 (6)
179.9 (3)
2.4258 (9)
2.068 (2)
88.42 (8)
85.18 (7)
88.89 (7)
179.9 (3)
Table 2 Selected crystallographic data and refnement for complexes 11–13
11
12
13
Empirical formula
Formula weight
Temperature (K)
Space group
Unit cell dimensions
a (Å)
b (Å)
c (Å)
α (°)
C24H23ClF3NORu
534.95
110
C24H24ClF3NORh
537.80
173
C24H24ClF3IrNO
627.11
173
P21/c
P21/c
P21/c
9.6891 (5)
13.8341 (7)
16.7075 (7)
90
11.8815 (14)
24.057 (3)
7.8436 (10)
90
7.7409 (6)
22.0339 (15)
12.9297 (9)
90
β (°)
96.279 (1)
90
2226.04 (19)
4
1.596
1080
103.124 (3)
90
2183.4 (5)
4
1.636
1088
95.103 (2)
90
2196.6 (3)
4
1.896
1216
γ (°)
V (Å3)
Z
Density (calc.) (g cm−3
F(000)
)
Crystal size (mm)
Theta min–max (°)
Data set
0.07×0.08×0.08
1.9, 27.9
−12:12; −18:18; −21:21
43,241
5332
−0.52, 0.46
0.13×0.14×0.15
1.7, 28.0
−15:15; −31:31; −10:9
26,557
5269
−0.55, 0.60
0.09×0.13×0.16
1.8, 28.0
−10:10; −29:29; 17:17
44,304
5291
−0.59, 0.96
Total refections
Unique refections
Min. and max. res. dens. (eÅ−3
)
1 3

JBIC Journal of Biological Inorganic Chemistry
Fig. 3 In vitro activity of com-
plexes 5–16 against the MCF7
breast cancer cell line at 20 µM
Table 3 In vitro activity of compounds 11, 13 and 14 against the
against non-cancerous Chinese hamster ovarian (CHO) cells.
Complexes 13 and 14 showed lower toxicity in the CHO cells
than in the MCF7 cell line, with IC50 values above 80 µM,
which is inactive according to the NCI [36]. Complex 11 was
found to be more active in the CHO cells (IC50 =5.67 µM)
than in the MCF7 cells (IC50 =16.83 µM). The selectivity
index (SI) of these complexes can give insight into which
cell line is more sensitive to the complexes. The selectivity
index shows that complex 11 has a greater selectivity for
non-tumorous cells (SI=0.34), which suggests that complex
11 may not be a suitable candidate for further testing as a
potential anticancer drug. Larger selectivity indices were
obtained for complexes 13 and 14 (SI>4), suggesting that
these complexes may be more selective towards cancerous
cells and may be potential candidates for further testing.
MCF7 breast cancer cell line
IC50 (MCF7) (μM)
IC50 (CHO) (μM)
SI (IC50
(CHO)/IC50
(MCF7))
11
13
14
17 1.2
21 1.3
19 1.2
6
0.6
0.3
4.6
4.3
96 6.0
84.0 17.7
be more active than the free metal dimers. This provides
evidence that the addition of the salicylaldimine group
increases the cytotoxicity of the metals tested. Four com-
plexes (10, 11, 13 and 14) displayed greater cytotoxicity than
cisplatin at 20 µM, against MCF7 cells, with the most active
complex being 11. Complexes bearing the trifuorometh-
ane (11–13) and trimethylsilane (14, 16) substructure were
more active than complexes with the propyl and ferrocenyl
(5–10) salicylaldimine ligand. Ligands 3 and 4 both have
higher ClogP values (4.42 and 3.89, respectively) than 1 and
2 (2.50 and 3.33 [39], respectively), suggesting that cytotox-
icity is linked to the degree of lipophilicity. Complexes 11,
13 and 14 showed promise in the pre-screen, and these com-
plexes were taken further to a multi-dose screen to determine
their IC50 values. The results are shown in Table 3. Complex
11 was found to be the most active, with an IC50 value of
16.83 µM. Changing from the ruthenium p-cymene (11) to
the iridium Cp* (13) moiety results in an IC50 of 20.75 µM.
This suggests that the metal fragment plays an important role
in the cytotoxicity exhibited in MCF7 cells. In changing the
ligand from trifuoromethane (11) to trimethylsilane (14) in
the ruthenium complexes, a slight drop in cytotoxicity was
also observed (16.83–19.48 µM, respectively). Compound
11 bears the ligand with the greatest ClogP value (4.42),
which could suggest greater accumulation in the cells and
therefore greater interaction with intracellular targets. To
evaluate their selectivity, the compounds were also tested
Stability in DMSO and Interactions with DNA model
5′‑GMP
Stability in media is an important feature to identify active
species, mechanism of action and potential drug candidates
[40]. The stability of complexes 11 and 13 was investigated
1
by H NMR spectroscopy in DMSO-d6 to simulate the
chemical environment prior to cell viability studies. The 1H
NMR spectrum of 11 shows that after 24 h, a new species
had formed (Fig. 4). Signals for the salicylaldimine group
were observed to double up, a further evidence that a new
species had formed. Splitting of the imine resonance of 11
was observed. No signifcant change from 24 to 48 h was
observed, suggesting that the formation of the new species
had reached equilibrium with 11 within 24 h. In contrast
to the ruthenium case, the iridium compound 13 showed
greater stability in DMSO-d6 under these conditions. The
stacked 1H NMR spectra of 13 in DMSO-d6 at 24 h intervals
are shown in Fig. 4. No additional signals were observed
within the 48 h testing period. Integration of the signals
after 48 h in DMSO-d6 for 13 is consistent with the reported
1 3

JBIC Journal of Biological Inorganic Chemistry
Since the biological assay medium contains a range of salts
and is mainly constituted of H2O, there is a high likelihood
that the active complex is diferent from the complex that
is initially isolated. Complex 13, though stable in DMSO,
was found to form a new species in the diferent mixtures of
DMSO and H2O. Cleavage of the imine bond was observed
when high concentrations of H2O were present, supported
by the appearance of the CHO signal in the 1H NMR spec-
tra. The Cp* signal was observed at 1.25 and 1.18 ppm for
the DMSO-d6:H2O and 150 mM NaCl(aq) samples, respec-
tively, which difers from the DMSO-d6 sample at 1.23 ppm.
Two signals are observed for the Cp* suggesting that the
active species may not be the complex isolated initially.
Guanosine 5′-monophosphate disodium (5′-GMP) is a
simplistic model used to study possible nucleotide interac-
tions with complexes. Complex 11, the most active complex,
was reacted with 5′-GMP in a DMSO-d6:H2O mixture, at
37 °C over 48 h, to understand the interactions with nucleo-
tides. Free GMP was studied under the same conditions and
was observed to have no change from the initial 1H NMR
spectrum. Ruthenium complexes have been reported to bind
to single-stranded DNA via nitrogen donor atoms from
guanosine [42–44]. Complex 11 was analysed for its GMP
binding afnity using 1H NMR spectroscopy (Fig. 5). A dis-
tinctive shift in the proton signals of the arene coordinated
to the metal centre indicated a change in the metal centre
environment. The aqua product of the complex observed in
the aqueous stability study was observed in the 11–GMP
mixture; however, this product is minor. A shift of the proton
resonance at 8.0 ppm strongly supports the coordination of
the complex at the N7 atom. This result is consistent with
the reported shift in the 5′-GMP proton signals for ruthe-
nium–GMP products [42]. This suggests why 11 is very
active against both cell lines tested, as this mechanism is not
specifc to cancerous cells only. The binding of 11 and GMP,
and the aqua species formed, suggests that an aqua species is
formed before 11 binds to the nucleotide. A similar mecha-
nism is observed for cisplatin [45], which could explain the
high levels of toxicity towards healthy cells.
Fig. 4 1H NMR spectral comparison of 11 (above) and 13 (below)
over a 48 h period in DMSO-d6; sampling taken at 24 h intervals
experimental data. The results obtained in this study suggest
that the ruthenium complex (11) undergoes solvation and the
resulting complex might be responsible for the cytotoxicity
observed in the MTT assay. In contrast to this, the iridium
complex (13) does not undergo solvation, which suggests
that this complex is likely to be the active species.
Conclusions
The behaviour of the aforementioned complexes was also
studied in partial aqueous solutions to mimic biological con-
ditions. NaCl was used to mimic the salt concentration out-
side the cells in the body [41]. Complex 11 was analysed
for its aqueous stability. A shift in the imine resonance was
observed when a high concentration of both H2O and NaCl is
present, suggesting there is a change in the species present in
solution. A downfeld shift, as well as a change in the split-
ting pattern of the p-cymene proton resonances, is observed
which strongly suggests substitution of the labile Cl ligand
with DMSO-d6 or H2O, as both are coordinating solvents.
Two new series of trifuoromethane (11–13) and organosi-
lane (14–16) complexes bearing Ru(II), Rh(III) or Ir(III)
metal centres, derived from two salicylaldimine ligands (3
and 4), were synthesised and characterised. Single crystals
of 11–13 were analysed by X-ray difraction and further con-
frmed the molecular structures of the complexes in the solid
state. Compounds were characterised using NMR and IR
spectroscopies, elemental analysis and EI or ESI mass spec-
trometry and correlate well with the proposed structures.
1 3

JBIC Journal of Biological Inorganic Chemistry
Fig. 5 1H NMR spectra of 11
and 5′-GMP after 48 h in a
DMSO-d6 and H2O mixture at
37 °C
The cytotoxicity of compounds 5–16, as well as the
dimeric metal precursors, was evaluated against the MCF7
breast carcinoma cell line. Complexes 11, 13 and 14 exhib-
ited the highest activity in this cell line, with IC50 val-
ues of 16.83, 20.75 and 19.48 µM observed, respectively.
The activity of these complexes can be linked to the lipo-
philicity of their ligands (the greater the ClogP value of
the ligand of the complex, the greater is the cytotoxicity
observed). The most active complexes were also evalu-
ated against the Chinese hamster ovarian (CHO) cell line
as a representative of normal, healthy cells. Complex 11
was highly active against the CHO cells (IC50 = 5.67 µM),
whilst the other two complexes (13 and 14) displayed
lower toxicity (IC50 > 80 µM).
References
1. Perone M (1884) Ann Chem Pharm 51:1–29
2. Rosenberg B, VanCamp L, Trosko JE, Mansour VH (1969)
Nature 222:385–386
3. Bhandari PR (2012) J Adv Pharm Technol Res 3:202–209
4. Hartinger CG, Metzler-Nolte N, Dyson PJ (2012) Organometal-
lics 31:5677–5685
5. Zhang P, Sadler PJ (2017) J Organomet Chem 839:5–14
6. Bergamo A, Masi A, Peacock A, Habtemariam A, Sadler PJ,
Sava G (2010) J Inorg Biochem 104:79–86
7. Berndsen RH, Weiss A, Abdul UK, Wong TJ, Meraldi P, Grif-
oen AW, Dyson PJ, Nowak-Sliwinska P (2017) Sci Rep 7:43005
8. Lucas SJ, Lord RM, Wilson RL, Phillips RM, Sridharan V,
McGowan PC (2012) Dalton Trans 41:13800–13802
9. Lucas SJ, Lord RM, Basri AM, Allison SJ, Phillips RM, Blacker
AJ, McGowan PC (2016) Dalton Trans 45:6812–6815
10. Yang J, Shi R, Zhou P, Qiu Q, Li H (2016) J Mol
Structure 1106:242–258
A possible mechanism of action for 11 and 13 was
1
investigated using H NMR spectroscopy in DMSO-d6,
11. Stringer T, Hendricks DT, Guzgay H, Smith GS (2012) Polyhe-
dron 31:486–493
DMSO:H2O and NaCl(aq). Doubling of signals in the
spectrum of complex 11 suggested low stability in DMSO-
d6; however, complex 13 was found to be inert. Aqua and
DMSO species of 11 and 13 were observed after 48 h in
DMSO-d6:H2O mixtures. Complex 11 showed evidence
of binding to 5′-GMP, suggesting a possible mechanism
of action similar to cisplatin; however, further studies are
needed to verify this.
12. Correia I, Roy S, Matos CP, Borovic S, Butenko N, Cavaco
I, Marques F, Lorenzo J, Rodríguez A, Moreno V, Pessoa JC
(2015) J Inorg Biochem 147:134–146
13. Govender P, Sudding LC, Clavel CM, Dyson PJ, Therrien B,
Smith GS (2013) Dalton Trans 42:1267–1277
14. Rahman FU, Ali A, Khan I, Guo R, Chen L, Wang H, Li ZT,
Lin Y, Zhang DW (2015) Polyhedron 100:264–270
15. Rahman FU, Ali A, Guo R, Wang WK, Wang H, Li ZT, Lin Y,
Zhang DW (2015) Dalton Trans 44:9872–9880
16. Sharma V, Beatty A, Wey SP, Dahlheimer J, Pica CM, Crank-
shaw CL, Bass L, Green MA, Welch MJ, Piwnica-Worms D
(2000) Chem Biol 7:335–343
Acknowledgements Financial support from the University of Cape
Town (UCT) and the National Research Foundation (NRF) of South
Africa and the Medical Research Council (MRC) of South Africa is
gratefully acknowledged. Dr Serah Kimani and Ms Sandra Jordaan are
thanked for their contributions.
1 3

JBIC Journal of Biological Inorganic Chemistry
17. Stringer T, Taylor D, Guzgay H, Shokar A, Au A, Smith PJ, Hen-
dricks DT, Land KM, Egan TJ, Smith GS (2015) Dalton Trans
44:14906–14917
31. SAINT Version 7.60a (2006) Bruker AXS Inc., Madison
32. Sheldrick GM (2004) SHELXS-97, SHELXL-2014 and SADABS
version 2.05, University of Göttingen, Germany
18. Nkoana W, Nyoni D, Chellan P, Stringer T, Taylor D, Smith PJ,
Hutton AT, Smith GS (2014) J Organomet Chem 752:67–75
19. Govender P, Renfrew AK, Clavel CM, Dyson PJ, Therrien B,
Smith GS (2011) Dalton Trans 40:1158–1167
33. Barbour LJ (2001) J Supramol Chem 1:189–191
34. Atwood JL, Barbour LJ (2003) Cryst Growth Des 3:3
35. Carmichael J, DeGraf WG, Gazdar AF, Minna JD, Mitchell JB
(1987) Cancer Res 47:943–946
20. Rahman FU, Ali A, Guo R, Zhang YC, Wang H, Li ZT, Zhang
DW (2015) Dalton Trans 44:2166–2175
36. Srisawat T, Chumkaew P, Heed-Chim W, Sukpondma Y, Kanok-
wiroon K (2013) Trop J Pharm Res 12:71–76
21. Isanbor C, O’Hagan D (2006) J Fluor Chem 127:303–319
22. Adams M, Barnard L, de Kock C, Smith PJ, Wiesner L, Chibale
K, Smith GS (2016) Dalton Trans 45:5514–5520
37. Mosmann T (1983) J Immunol Methods 65:55–63
38. Abate G, Asefa A, Selassie A, Goshu S, Fekade B, WoldeMeskal
D, Miörner H (2004) J Clin Microbiol 42:871–873
23. Adams M, de Kock C, Smith PJ, Land KM, Liu N, Hopper M,
Hsiao A, Burgoyne AR, Stringer T, Meyer M, Wiesner L, Chibale
K, Smith GS (2015) Dalton Trans 44:2456–2468
39. Stringer T, Taylor D, de Kock C, Guzgay H, Au A, An SH,
Sanchez B, O’Connor R, Patel N, Land KM, Smith PJ, Hendricks
DT, Egan TJ, Smith GS (2013) Eur J Med Chem 69:90–98
40. Shin YG, Bolton JL, van Breemen RB (2002) Comb Chem High
Throughput Screen 5:59–64
24. Kühnert J, Ecorchard P, Lang H (2008) Eur J Inorg Chem
2008:5125–5137
25. Ornelas C (2011) New J Chem 35:1973–1985
41. Prough DS, Bidani A (1999) Anesthesiology 90:1247–1249
42. Burgoyne AR, Kaschula CH, Parker MI, Smith GS (2016) Eur J
Inorg Chem 2016:1267–1273
26. Kirk KL (2006) J Fluor Chem 127:1013–1029
27. Temmink OH, Emura T, De Bruin M, Fukushima M, Peters GJ
(2007) Cancer Sci 98:779–789
43. Scolaro C, Chaplin AB, Hartinger CG, Bergamo A, Coc-
chietto M, Keppler BK, Sava G, Dyson PJ (2007) Dalton Trans
43:5065–5072
28. Bennett MA, Huang TN, Matheson TW, Smith AK, Ittel S, Nick-
erson W (2007) 16. (η6-Hexamethylbenzene)ruthenium com-
plexes. In: Inorganic Syntheses, Wiley, Hoboken, pp 74–78
29. White C, Yates A, Maitlis PM, Heinekey DM (2007) (η5-
Pentamethylcyclopentadienyl)rhodium and -iridium compounds.
In: Inorganic Syntheses, Wiley, Hoboken, pp 228–234
30. Maqeda L, Makhubela BCE, Smith GS (2015) Polyhedron
91:128–135
44. Dorcier A, Dyson PJ, Gossens C, Rothlisberger U, Scopelliti R,
Tavernelli I (2005) Organometallics 24:2114–2123
45. Reedijk I (2009) Eur J Inorg Chem 2009:1303–1312
1 3