Acetylacetonate bis(thiosemicarbazone) complexes of copper and nickel: Towards new copper radiopharmaceuticals

Article abstract of DOI:10.1039/b406429a

A series of copper(II) and nickel(II) 1,3-bis(thiosemicarbazonato) complexes have been synthesised by the reaction of the metal acetates with pyrazoline proligands. In each case the complexes have an overall neutral charge with a dianionic ligand. The cop

Full text of DOI:10.1039/b406429a

View Article Online / Journal Homepage / Table of Contents for this issue
F U L L P A P E R
Acetylacetonate bis(thiosemicarbazone) complexes of copper and
nickel: towards new copper radiopharmaceuticals
a
b
b
c
Andrew R. Cowley, Jonathan R. Dilworth,* Paul S. Donnelly, Antony D. Gee and
b
Julia M. Heslop
a
Chemical Crystallogaphy, Laboratory and University of Oxford, 12 Mansfield Road, Oxford,
UK OX1 3TA
b
Chemistry Research Laboratory, University of Oxford, 12 Mansfield Road, Oxford,
UK OX1 3TA. E-mail: jon.dilworth@chem.ox.ac.uk
Clinical Research Unit, GlaxoSmithKline, Translational Medicine and Technologies,
c
ACCI and Addenbrookes Hospital, Hills Road, Cambridge, UK CB2 2GG
Received 29thꢀAir llꢁ0hcceꢀ9edpl9tꢂ3ꢃe llꢁ
FiAs9ꢀ3bristedasaꢃhdvaꢃcehA9icreꢄꢃ9teꢅeb ꢃdꢂ3rꢆ llꢁ
A series of copper(II) and nickel(II) 1,3-bis(thiosemicarbazonato) complexes have been synthesised by the reaction of the
metal acetates with pyrazoline proligands. In each case the complexes have an overall neutral charge with a dianionic
ligand. The copper 1,3-bis(4-methyl-3-thiosemicarbazonato complex has been characterised by X-ray crystallography,
2 2
which shows the copper is in an essentially square-planar symmetric N S environment. The nickel 1,3-bis(4-methyl-3-
thiosemicarbazonato) and nickel 1,3-bis(4-phenyl-3-thiosemicarbazonato) complexes have been characterised by X-ray
crystallography and show that in these cases the nickel is in a distorted square-planar environment, but the bonding mode
of the ligands is unusual; the nickel binds to one of the aza-methinic nitrogen atoms and one hydrazinic nitrogen, creating
one five-membered N–N–C–S–Ni chelate ring and one four-membered N–C–S–Ni chelate ring. Interestingly, the X-ray
structure of the ethyl analogue {1,3-bis(4-ethyl-3-thiosemicarbazonato)nickel(II)} shows that in this case the nickel is
symmetrically coordinated in the usual manner. The nickel complexes are diamagnetic and the different coordination
modes are confirmed in solution by NMR spectroscopy. The complexes are susceptible to oxidation in air and a nickel
complex, in which the central methylene carbon has been oxidised, has been characterised by X-ray crystallography and
NMR spectroscopy. Electrochemical measurements show that the copper complexes undergo a reversible one-electron
reduction at biologically accessible potentials.
Introduction
Thiosemicarbazones are of great pharmacological interest since
they show a broad spectrum of biological activity including
antitumour, antibacterial, antimalarial and antifungal activities.1
Moreover, metal complexes of thiosemicarbazones often
display enhanced activities when compared to the uncomplexed
thiosemicarbazones. Copper complexes of bis(thiosemicarbazones)
have been investigated for use as anti-cancer chemotherapeutic
2
,3
4
agents and as superoxide dismutase-like radical scavengers. It is,
however, their use as delivery agents for radioactive copper in new
copper based radiopharmaceuticals and the hypoxic selectivity of
certain copper bis(thiosemicarbazonato) complexes that has created
Fig. 1 Bis(thiosemicarbazone) ligands and abbreviations.
5
–11
protonated Cu(I) species, which is trapped inside the cell.10,11 In
normal oxic cells, this Cu(I) complex is rapidly oxidised back to the
original neutral Cu(II) complex, which quickly washes out of the
cell. In the case of the non-selective copper complex [Cu(GTS)],
which shows high cell uptake independent of oxygen concentra-
tion, it is believed that once the Cu(II) complex enters the cell, it
is reduced to Cu(I) and then dissociates from the ligand. The ‘free’
copper is then sequestered by intra-cellular ligands and irreversibly
trapped inside the cell.15
much recent interest.
Hypoxia (low oxygen concentrations) is
often associated with tumours and heart disease and can affect
the outcome of anti-cancer treatments. This means it is of great
clinical utility to develop systems that allow the ‘non-invasive’
imaging of hypoxia with a suitable radiopharmaceutical.12 Copper
radionuclides are of considerable interest in the development of new
radiopharmaceuticals as they include isotopes that have both PET
1
3
imaging and therapeutic potential.
The cell uptake and retention characteristics of copper
bis(thiosemicarbazonato) complexes derived from 1,2-diones is
dependent on the nature of the ‘backbone’ substituents on the
The marked dependence of the nature of the backbone on the
specificity of copper complexes of bis(thiosemicarbazones) led us
to investigate further bis(thiosemicarbazone) proligands with ex-
tended, three-carbon backbones as ligands for the development of
new copper radiopharmaceuticals (Fig. 2). It was hoped that these
larger ligands would better accommodate both Cu(I) and Cu(II). A
cationic nickel complex of R = H has been structurally character-
ised16 and the copper complex of L has been recently reported but
6
64
ligand. The radioactive copper complex [ Cu(ATSM)] (ligand
structures; Fig. 1) shows rapid uptake and washout in normal tis-
sues but is selectively trapped in hypoxic heart and tumour tissue;
1
0
it is therefore referred to as ‘hypoxia selective’. Other complexes,
such as [Cu(GTS)] and [Cu(PTSM)] (Fig. 1), are non-selectively
trapped in all cells, which leads to alternative uses. For example,
1
1
7
1
[
Cu(PTSM)] has shown great potential as a non-selective blood
was not structurally characterised. Whilst complexes of L with
1
4
18,19
perfusion tracer.
technetium have been investigated as potential imaging agents,
20
In the case of the hypoxia-selective complex [Cu(ATSM)], the
trapping in hypoxic cells is believed to occur via reduction of the
copper(II) complex by intracellular reducing agents to give a stable
there is some doubt as to the nature of the species formed. To our
knowledge nobody has attempted to study the radiocopper com-
plexes of these proligands.
2
4 0 4
D a l t o n T r a n s . , 2 0 0 4 , 2 4 0 4 – 2 4 1 2
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4

View Article Online
Fig. 2 Bis(thiosemicarbazone) proligands and abbreviations.
Fig. 4 An ORTEP representation of ethylthiosemicarbazido–pyrazoline
2
(H
2
L ).
This paper describes the synthesis and full characterisation
of a series of (bis)thiosemicarbazone proligands derived from
amide H atoms link the molecules to form dimers (N(4)N(2)′
.194(2), N(5)S(2)′ 3.3365(18) Å, a symmetry operator generat-
3
1
,3-diketones, their copper complexes, including a study of their
ing acceptor atoms −x, 1 − y, 1 − z). The remaining NH groups form
longer NS hydrogen bonds that link the dimers to form infinite
chains running parallel to the crystallographic c axis (N(6)S(1)″
redox properties, and an evaluation of their potential to act as
radiopharmaceuticals. In an attempt to characterise further the
coordination behaviour of these ligands, we have also investigated
their nickel complexes.
3
.5427(18) Å, a symmetry operator generating acceptor −x,
1
− y, −z).
Whilst this method worked well for the synthesis of the methyl
Results and discussion
1
2
and ethyl analogues (H
2
L and H
2
L ) in their pyrazoline form, prob-
Synthesis
3
lems were encountered in synthesising the phenyl analogue (H
2
L ).
In this case, the elemental analysis of the isolated bulk material was
not consistent with H L in the pyrazoline form, although it was of
2
sufficient purity to make metal complexes that gave satisfactory
The synthesis of bis(thiosemicarbazones) by the reaction of thios-
emicarbazides with 2,4-pentanedione is complicated by cyclisation
reactions, which result in the isolated product being in the pyrazoline
form. Fortunately, it has been shown recently that these pyrazo-
lines undergo ring opening back to a bis(thiosemicarbazone) in the
presence of certain transition metal ions, resulting in the formation
of the bis(thiosemicarbazonato) metal complex.17 It has been high-
lighted recently that this has led to confusion in the literature, and
3
2
0
mass spectra and, in the case of the nickel complex crystals suit-
3
able for X-ray crystallography. An attempt to purify crude H
2
L by
recrystallisation from ethanol gave colourless crystals that were
shown by X-ray crystallography, mass spectroscopy and NMR to
be N,N-diphenyl[1,3,4]thiadiazole-2,5-diamine (2) (Fig. 5), which
proved to be a major contaminant in the product. The synthesis of
this compound has been reported directly from 4-phenylthiosemi-
some misinterpretation of the nature of materials isolated from the
reaction of thiosemicarbazides directly with 2,4-pentanedione.20
22
3
2
carbazide in the presence of sulfuric acid. Pure H L in the pyrazo-
The pyrazoline proligands were readily synthesised and isolated
by literature procedures, which involve the reaction in the presence
of acid of two equivalents of thiosemicarbazide with 1,2-bis(1-
methyl-3-oxobutylideneamino)ethane (1), which is isolated from
line form could be obtained in moderate yield (58%) by performing
the synthesis in the absence of acid.
2
1
the reaction of 2,5-pentanedione with 1,2-ethanediamine (Fig. 3).
2
2
An X-ray crystal structure of the molecule (H L ) isolated from the
reaction of 4-ethyl-3-thiosemicarbazide with 1 is shown in Fig. 4
and confirms the molecule is in the pyrazoline form.
Fig. 5 An ORTEP representation of N,N-diphenyl[1,3,4]thiadiazole-2,5-
diamine (2).
The copper complexes of the three-carbon backbone ligands were
prepared by reaction of the pyrazoline proligands with Cu(CH
3 2 2
CO )
in methanol under an atmosphere of nitrogen and copper complexes
1
3
of L –L were isolated in this manner. The complexes precipitated
from the reaction mixture and could be readily isolated by filtra-
tion. An ORTEP representation of the complex obtained by the
1
reaction of H
2
L with copper acetate is shown in Fig. 6 (crystal-
lographic data appears in Table 1 and selected bond lengths for the
complexes are given in Table 2). The neutral complex contains a
square-planar copper atom coordinated to each of the sulfur and
azomethinic nitrogen atoms. The Cu–S bond distances of 2.2469(5)
and 2.2335(5) Å are very similar to copper(II) thiolate bond dis-
tances found in “simple” N
radiopharmaceutical [Cu(ATSM)] (Fig. 1).
2 2
S chelates and the hypoxia selective
11,23
The Cu–N bond
distances of 1.9737(16) and 1.9893(16) Å are also similar to the
Cu–N bond distances in [Cu(ATSM)]. Although the bond distances
are similar to [Cu(ATSM)], there are significant differences in the
bond angles around the copper, which are closer to the ideal for a
1
square-planar system in the complex with the larger ligand, [Cu(L )].
Fig. 3 Reaction scheme for the synthesis of [M(AATS)] complexes.
The S–Cu–N bond angles are 86.79(5) and 85.60(5)°, whereas in
[Cu(ATSM)] the S–Cu–N bond angles average is 85.17°. Bigger
differences are evident in the N–Cu–N angle, which is 96.10(7)° in
The structure is similar to the methyl analogue which has been
1
7,20
reported recently.
tramolecular hydrogen bond to one of the S atoms (N(3)S(1)
.1715(19) Å). The second hydrazine H atom and one of the thio-
One of the hydrazine H atoms forms an in-
1
[Cu(L )] compared with an average of 80.60° in [Cu(ATSM)], and
1
3
the S–Cu–S angle, which is 91.728(19)° in [Cu(L )] compared with
D a l t o n T r a n s . , 2 0 0 4 , 2 4 0 4 – 2 4 1 2
2 4 0 5

View Article Online
Table 1 Crystallographic data
L2
[Cu(L )]
1
[Ni(L )]
1
[Ni(L )]
3
[Ni(L O)]
2
Crystal identification
2
H
2
Chemical formula
C
14
H
12
N
4
S
C
11
H
22
N
6
S
2
C
10
H
6
20CuN OS
2
C
9
H
16
N
6
NiS
2
C
19
H
20
N
6
NiS
2
11 18 6 2
C H N NiOS
M
268.34
Monoclinic
Pc
11.0953(2)
7.0982(2)
8.5325(2)
90
104.4051(9)
90
302.46
Triclinic
P1
8.2326(3)
367.97
Triclinic
P1
331.10
Orthorhombic
Pcba
8.2450(2)
17.2004(2)
19.5299(3)
90
455.24
Monoclinic
P2 /c
7.6882(2)
20.5594(5)
12.7497(4)
90
105.6694(11)
90
1940.38(9)
4
373.14
Monoclinic
P2 /c
20.5059(4)
8.1622(2)
19.1350(4)
90
104.6903(9)
90
3097.99(12)
8
Crystal system
Space group
a/Å
b/Å
c/Å
1
1
7.6332(2)
8.2419(2)
12.8765(3)
72.6838(9)
85.4540(9)
85.3228(8)
769.51(3)
2
9.8339(3)
10.0725(4)
106.4508(17)
95.7000(17)
91.7108(13)
776.77(5)
2


/°
/°
90
90
/°
3
V/Å
Z
650.83(3)
2
2769.68(9)
8
R
int
0.055
0.043
0.029
0.032
0.047
0.050
wR
0.0480
−0.29, 0.45
0.0374
−0.24, 0.19
0.0351
−0.36, 0.34
0.0316
−0.32, 0.36
0.0310
−0.35, 0.25
0.0335
−0.40, 0.41

max/min/e Å−3
Table 2 Selected bond distances (Å) in the complexes
1
1
3
2
[
Cu(L )]
[Ni(L )]
[Ni(L )]
[Ni(L O)]
Cu(1)–N(3)
Cu(1)–N(6)
Cu(1)–S(1)
Cu(1)–S(2)
1.9737(16)
1.9893(16)
2.2469(5)
2.2335(5)
Ni(1)–N(2)
Ni(1)–N(4)
Ni(1)–S(1)
Ni(1)–S(2)
1.9026(15)
1.9065(15)
2.1675(5)
2.2279(5)
Ni(1)–N(2)
Ni(1)–N(4)
Ni(1)–S(1)
Ni(1)–S(2)
1.8999(19)
Ni(1)–N(2)
Ni(1)–N(3)
Ni(1)–S(1)
Ni(1)–S(2)
1.905(2)
1.904(3)
2.1505(9)
2.1512(8)
1.908(2)
2.1464(7)
2.2210(7)
23
an average of 109.22° in [Cu(ATSM)]. This suggests that the larger
1
ligand, L , is in general a better fit for copper(II) than the smaller
ATSM ligand. The thiosemicarbazonato ligand is dianionic and all
the non-hydrogen atoms are approximately coplanar (rms deviation
from best plane 0.09 Å). This results in considerable strain at the
methylene C atom, as shown by the abnormally large value of the
C(3)–C(9)–C(4) angle (126.15(16)°). The bond distances within
the ligand show extensive conjugation within the ligand but the
resonance form depicted in Fig. 3, in which the ligand is dianionic,
predominates. It is interesting to note that this resonance form is
2
Fig. 7 Diagram of the hydrogen bonding array in [Cu(L )].
2
preferred to one in which the central CH group is deprotonated,
resulting in a conjugated ligand with an overall monoanionic charge.
The C–S bond lengths of 1.737(2) and 1.751(2) Å are similar to the
analogous bond distances in [Cu(ATSM)] and consistent with a bond
order greater than 1 but less than 1.5. There are also similarities be-
structure of the complex reveals an extensive hydrogen bonding
array (Fig. 7).
The nature of the hydrogen bonding interaction within these com-
pounds could be of some relevance to the biological activity through
possible interactions with nucleoside bases and peptide bonds.
One of the NH groups forms a hydrogen bond to the O atom of a
nearby solvent (N(1)O(1)′ 2.824(2) Å, symmetry operator 1 − x,
1
tween [Cu(ATSM)] and [Cu(L )] in the carbon to nitrogen bond
distances in the SN–CN fragments and the N–CN fragments. In
1
[
Cu(L )] the SN–CN bond distances are 1.320(3) and 1.310(3) Å;
the N–CN bond distances are 1.293(2) and 1.292(3) Å. These
2
− y, −z). The second NH group forms a hydrogen bond to a S
distances are indicative of a bond order greater than 1.5 but less
atom of a second molecule of complex (N(4)S(2)″ 3.3956(19) Å,
symmetry operator −x, −y, 1 − z). The solvent OH forms a hydro-
gen bond to a N atom in the same asymmetric unit (O(1)N(2)
1
than 2. The greatest difference between [Cu(ATSM)] and [Cu(L )]
arises in the N–N bond distance, which is significantly shorter in
[
1
Cu(ATSM)] (average value 1.369 Å) than it is in 1 (1.397(2) and
.402(2) Å). This relatively long value is indicative of a reduc-
2.819(2) Å). These hydrogen bonds link the complexes and solvents
23
to form infinite chains.
The reaction of Ni(CH CO ) with the proligands in their
3 2 2
tion in the delocalisation of electron density. Perhaps this explains,
1
at least in part, why we have found complexes of the type [Cu(L )]
pyrazoline form also resulted in ring opening of the proligand to
more susceptible to protonation at the hydrazinic nitrogen (N5, N2)
than complexes of the ATSM type. This could be of importance for
any biological selectivity.
1
3
allow the isolation of the nickel complexes of L –L . In the case
1
3
of L and L green and green–brown solids were isolated from the
reaction mixture respectively, but interestingly, the product from the
As is the case for the two-carbon backbone copper(II)
2
reaction of L was pink. The ES-MS all show peaks corresponding
2
3
bis(thiosemicarbazone) radiopharmaceuticals,
the crystal
1–3
+
to [Ni(L ) + H ].
1
The nickel complex of L was isolated as crystals that exhibit
pronounced pleichroism; when viewed along the b axis they appear
very dark green, but when viewed along the c axis they appear
red–brown. A representation of the X-ray structure is shown in
Fig. 8.
Like the copper complex, the nickel complex is essentially square
planar with N S coordination, but somewhat surprisingly the bond-
2 2
ing mode of the ligands is quite different, with the nickel binding
to one azamethinic nitrogen and one hydrazinic nitrogen, creating
one five-membered N–Ni–S chelate ring and one four-membered
N–Ni–S chelate ring. Whilst this binding mode is known for thios-
24
emicarbazones, it is rare. The presence of the four-membered che-
late ring does introduce a significant amount of distortion and the
bond angles around the nickel are much further away from the ideal
1
Fig. 6 An ORTEP representation of [Cu(L )].
2
4 0 6
D a l t o n T r a n s . , 2 0 0 4 , 2 4 0 4 – 2 4 1 2

View Article Online
3
Fig. 9 An ORTEP representation of [Ni(L )].
6
the d -DMSO signal. There are also two resonances corresponding
to the imino carbons ( 151.1 and 157.8 ppm) and the C–S groups
( 173.3 and 176.2 ppm).
In the case of the N-phenyl substituted ligand, L , the H NMR
spectrum also shows the complex is doubly deprotonated at the
hydrazinic nitrogens and that the nickel is non-symmetrically bound.
1
Fig. 8 An ORTEP representation of [Ni(L )].
3
1
2
value than in a similar nickel complex (where R = NH ), which is
symmetrically bound with two five-membered N–Ni–S chelate
1
6
3
There are two signals for the CH groups ( 2.09 and 2.33 ppm), a
series of overlapping multiplets corresponding to the aromatic
rings. The N4–Ni–S2 bond angle in the four-membered chelate is
3.94(5)°, whilst the N2–Ni–S1 bond angle of the five-membered
ring is 86.98(5)°. Despite these distortions, the geometry of the
nickel atom is still approximately planar with the rms deviation of
the Ni, S and coordinated nitrogen atoms from the best plane being
7
resonances ( 6.90–7.58 ppm) and two NH resonances ( 9.42 and
1
0.18 ppm). The 13C NMR spectrum has two methyl resonances
(
1
 23.0 and 23.6 ppm) and a series of aromatic resonances (
18.3–141.0 ppm). There are two resonances corresponding to the
0
.054 Å. The Ni–S bond lengths are similar to other compounds
2
5
imino carbons ( 153.8 and 162.3 ppm) and the C–S groups ( 169.4
and 173.2 ppm). These spectra confirm that in these cases the solid
state structures correspond with those in solution.
with N–S thiolate coordination. Once again, the ligand is doubly
deprotonated, but, in this case, much less planar than the copper
analogue.
The conformation of the ligand generates a ‘bowl-like’ shape
2
see Fig. 8) with the CH of the backbone folded below the plane
Interestingly, the H and 13C NMR spectra of the pink solid,
1
2
+
[
Ni(L )] , indicate that in this case the nickel is symmetrically
(
1
coordinated. The H NMR spectrum has one signal corresponding
to the CH groups ( 2.11 ppm) and two multiplets corresponding
to the terminal NHCH CH groups ( 3.12 and 1.03 ppm,
respectively). The NH resonance is split into a multiplet by the
neighbouring CH CH groups and occurs at  6.94 ppm. The
C NMR spectrum has a single resonance corresponding to the
backbone’ methyl groups ( 20.4 ppm) and a signal corresponding
to the terminal methyl groups of the ethylene moieties ( 14.4 ppm).
A HSQC experiment has shown that the NHCH signal is partially
obscured by the d -DMSO signal. There are single resonances
containing the nickel. This conformation is perhaps facilitated by
the non-symmetrical coordination that reduces the strain within the
3
2
3
C–CH
pears as if the ligand consists of two nearly planar sections, which
are joined at the ‘backbone’ CH with a C2–C5–C7 bond angle of
09.46(14)°. The bond lengths within the ligand still indicate ex-
2
–C backbone when compared to the copper complex. It ap-
2
3
2
1
3
1
‘
tensive delocalisation throughout the ligand but the resonance form
depicted in Fig. 3 still predominates.
2
Crystals of the nickel complex of the phenyl substituted ligand,
L , suitable for X-ray studies, were grown by slow evaporation of
3
6
corresponding to the imino carbons ( 156.9 ppm) and the C–S
groups ( 171.0 ppm).
a dichloromethane solution of the complex. A representation of the
structureisdepictedinFig. 9. Onceagain, themetalatomcoordinates
to the bis(thiosemicarbazone) ligand in a non-symmetrical fashion,
with an approximately square-planar nickel and a 4–7–5 chelate
system. The Ni–S and Ni–N bond lengths are similar to the previous
Oxidation of the complexes
Recently Durrant and co-workers reported that the ligand in
1
example, [Ni(L )], as are the bond angles. The ligand is doubly
1
[
Cu(L )] is susceptible to oxidation in air. They suggested that
deprotonated with a similar conformation to the methyl analogue,
although in this case there is slightly less deviation from planarity
the oxidation involved the addition of a single oxygen atom to
the methylene backbone with a concomitant loss of two hydrogen
at the C–CH
2
–C ‘backbone’ of the ligand, with a C2–C3–C4 bond
17
atoms (Fig. 10).
angle of 114.2(2)°. Once again, the ‘bowl-like’ conformation of the
ligand reduces the strain within the backbone.
NMR studies
8
The d square-planar nickel complexes are all diamagnetic and
give well resolved NMR spectra, which confirm the nature of the
1
1
+
binding. The H NMR spectrum of [Ni(L )] shows signals for the
CH groups ( 1.91 and 2.16 ppm) and a multiplet corresponding to
the two NHCH resonances ( 2.63 ppm). The two NH resonances
are split into multiplets by the neighbouring CH groups and are
clearly distinct, occurring at  6.85 and 8.00 ppm, respectively. The
Fig. 10 Oxidative formation of complexes [M(L1–3O)].
3
3
1
The electronic spectra of [Cu(L )] prepared under nitrogen has a
3
weak intensity d–d band at 778 nm. There is also a weak intensity
band at 555 nm and medium to high intensity bands at 417, 318 and
293 nm which are assigned to ligand to metal charge transfer and
intraligand transitions. On exposure to air, the spectrum gradually
changes over time, but rapid oxidation, as previously reported,17
was not observed. After the sample had been exposed to air for 15 h
13
C NMR spectrum is also consistent with non-symmetric binding,
with two methyl resonances corresponding to the ‘backbone’ CH
groups ( 21.5 and 22.9 ppm) and two resonances corresponding to
the terminal CH groups ( 27.3 and 30.8 ppm).AHSQC experiment
has shown that the central methylene signal is partially obscured by
3
3
D a l t o n T r a n s . , 2 0 0 4 , 2 4 0 4 – 2 4 1 2
2 4 0 7

View Article Online
the electronic spectrum showed peaks that have been attributed to
the oxidised complex (483 and 378 nm) but also retains a band at
ligand is doubly deprotonated with approximately planar geometry
with the exception of its ethyl groups.
4
17 nm which has been attributed to the original ‘non-oxidised’
One of the molecules forms hydrogen bonds with both NH
groups (N(5)O(2) 3.141(5) Å, N(6)O(1)′ 2.938(3) Å, sym-
metry operator of acceptor x, 1/2 − y, 1/2 + z) (Fig. 12). The second
molecule does not act as a hydrogen-bond donor. The hydrogen
bonds link the complexes to form infinite chains running parallel to
the crystallographic c axis.
1
+ 17
product, [Cu(L )] . This suggests that the after a sample of the
compound has been exposed to air for 15 h a mixture of the oxidised
and non-oxidised copper complexes is present.
1
If the copper acetate is reacted with the proligand H
2
L in
ethanol, but this time in an aerobic environment (rather than under
nitrogen), a brown solid is obtained. The electronic spectrum of
this material suggests that it contains a mixture of [Cu(L )] and its
1
oxidised product. The ES-MS of this material also shows a peak
1
+
at m/z = 335, corresponding to [Cu(L ) + H ] (copper in the +III
2
6
oxidation state), and a peak at m/z = 350, which corresponds to
the oxidised product. HPLC analysis (C18 reverse phase column
with ammonium acetate–acetonitrile gradient elution) of the solid
dissolved in DMSO revealed two peaks with elution times of 15.3
and 16.0 min. The two peaks presumably correspond to non-
oxidised and oxidised product.
1
The crystals used for the X-ray studies of [Cu(L )] (Fig. 6) were
grown under an atmosphere of nitrogen. The restrictions of the
square-planar coordination about the copper results in considerable
strain at the methylene C atom, as shown by the abnormally large
value of the C(3)–C(9)–C(4) angle (126.15(16)°). The oxidation of
the ligand may actually be a consequence of the ligand trying to
reduce this strain, and may involve deprotonation of the methylene
2
Fig. 12 Diagram of the hydrogen bonding array in [Ni(L O)].
NMR analysis of the crystals used for the X-ray studies
confirmed that the oxidised product could be identified by its NMR
spectra with loss of the peaks attributed to the methylene ‘backbone’
2 3
carbon followed by electrophilic attack. Addition of Na CO to a
1
1
solution, which contained a mixture of [Cu(L )] and the oxidised
complex, resulted in increasing the intensity of a band at 363 nm in
the UV/vis spectrum which is attributed to the oxidised complex.
in the H NMR and an additional carbonyl peak at  181.9 ppm in
the 13C NMR spectrum.
Copper(II) and zinc(II) complexes of -diketiminate ligands,
derived from reacting an amine with acetylacetonante, have been
demonstrated to undergo a similar oxidation to give ketone diimine
derivative under aerobic conditions. In this case the oxidation is
assumed to occur via a series of steps, which are facilitated by the
presence of the metal ion, following addition of dioxygen to the
deprotonated methylene carbon. Presumably a similar mechanism
can be invoked for the oxidation of the metal complexes of
bis(thiosemicarbazones) ligands encountered in this work.
1
The X-ray structures of the green/brown crystals of [Ni(L )] and
3
[
Ni(L )], that display the non-symmetric coordination environment
of the metal, show that the ligand is in a ‘bowl-like shape’ with
reduced strain at the methylene carbon. Perhaps this explains why
these nickel complexes are more resistant to oxidation, the crystals
3
28
of [Ni(L )] were grown via evaporation of a dichloromethane
solution under an aerial environment.
As mentioned previously, the nickel complex of the ethyl ligand
2
L H
2
was isolated as a pink solid in which the nickel was shown
2
by NMR to be symmetrically coordinated, [Ni(L )]. An attempt to
grow crystals suitable for X-ray studies from a methanolic solution
exposed to air resulted in the isolation of dark brown crystals which
Electrochemistry
Structure–activityrelationshipsofcopper(II)(bis)thiosemicarbazone
radiopharmaceuticals derived from 1,2-diones show a correlation
with the reduction potential for the Cu(II)/Cu(I) couple and hy-
poxic cell selectivity. The hypoxia selective radiopharmaceutical,
Cu(ATSM)], undergoes a reversible reduction at E1/2 = −0.62 V in
were shown by X-ray crystallography to be the oxidised product,
2
[
Ni(L O)] (Fig. 11).
29
[
DMF at a glassy carbon working electrode. The cyclic voltammo-
1
grams of [Cu(L )] in DMF with a glassy carbon electrode exhibits a
quasi-reversible reduction at E1/2 = −0.37 V, which was assigned to a
Cu(II)/Cu(I) process, but also a pre-absorption wave. There was also
a broad irreversible oxidation at higher potentials (Eox = 0.76 V),
which is presumably due to a ligand based oxidation.
The cyclic voltammetry measurements of the copper complexes,
which were prepared in air, were sometimes complicated by the fact
that the samples contained mixtures of both the oxidised copper
complex and that of the original ‘non-oxidised’ complex. In these
cases two independent reversible reductions could be seen in the
cyclic voltammogram. Addition of a small amount of sodium car-
bonate to these analyte solutions to complete the oxidation resulted
in diminution of one of the processes, to give a single reversible
process, which was attributed to a reversible one- electron reduc-
tion which we attribute to a Cu(II)/Cu(I) process (Fig. 13). The
2
Fig. 11 An ORTEP representation of [Ni(L O)].
1
1
‘
oxidised’ copper complex of L , [Cu(L O)], has an E1/2 = −0.30
In this case, the ligand has indeed been oxidised at the methylene
V (vs. SCE) with an anodic to cathodic peak separation of 102 mV
carbon, to give a ketone (bis)thiosemicarbazone.Although there has
been some earlier suggestions of this oxidative degradation,1 this
is the first structurally characterised example. The asymmetric unit
contains two distinct molecules, neither of which has any crystal-
lographic symmetry. The complex is essentially square planar with
the ligand coordinating in a symmetric manner. Despite the different
coordination mode and oxidised ligand, the Ni–N and Ni–S bond
lengths and angles are similar to the two previous examples. The
(under the same conditions ferrocene had a peak separation of 102
mV); the oxidised complex of L , [Cu(L O)], has an E1/2 = −0.34
7,27
2
2
V with a peak separation of 96 mV. The copper complex of the
3
phenyl substituted ligand, [Cu(L O)] exhibits a dramatically dif-
ferent Cu(II)/Cu(I) reduction potential; the reduction occurs at a
potential about 250 mV more positive, (E1/2 = −0.06 V vs SCE),
than the alkyl substituted analogues. It is worth mentioning that the
alkyl substituted compounds give red/brown solutions in the DMF
2
4 0 8
D a l t o n T r a n s . , 2 0 0 4 , 2 4 0 4 – 2 4 1 2

View Article Online
used for electrochemical measurements, whereas the solution of the
phenyl derivative, [Cu(L O)], is dark green. This could suggest that
the large shift in the reduction potential is related to some degree of
protonation of the complex and that the phenyl substituent dramati-
contain a pyrazoline ring. Reaction with either copper or nickel ac-
etate opens the ring to allow the formation of metal complexes. In all
the cases reported here, where the metal acetate is used to form the
complex, the ligand deprotonates twice to give a dinegative ligand,
which is best described as being in the form depicted in Fig. 3. In a
3
a
cally effects the pK of the complex. The cyclic voltammograms
1
–3
of the ‘oxidised’ complexes, [Cu(L O)], all show an irreversible
oxidation at higher potentials, which is presumably due to a further
ligand based oxidation, which obscures any possible Cu(II)/Cu(III)
redox processes.
related example, in which Ni(NO
ion, the X-ray structure shows that the ligand is monoanionic, and
deprotonation occurs at the central CH
group.16 The ligands show
3 2
) was used as the source of metal
2
some flexibility in their binding characteristics, as shown by the
nickel complexes that have four-, five- and seven-membered chelate
rings and the complexes that display the more expected symmetric
5–6–5 membered chelate ring system. The symmetrically coordi-
nated copper complex reveals considerable strain at the backbone
methylene carbon and the complexes are sensitive to oxidative deg-
radation to give keto-bis(thiosemicarbazonato) metal complexes, as
shown by the structural characterisation of an oxidised nickel com-
plex. This oxidation is accelerated in basic solution. To prevent this
oxidation it is essential to prepare the compounds with the rigorous
exclusion of air. If the copper complexes are prepared in air, then a
mixture of the oxidised and non oxidised products is obtained. The
copper complexes show reversible one-electron reductions, which
are attributed to a Cu(II)/Cu(I) process.
The presence of oxo and non-oxo complexes could explain the
1
Fig. 13 Cyclic voltammogram of [Cu(L O)].
presence of two different compounds, as determined by radio-
2
HPLC, in experiments that have attempted to radiolabel H
2
L with
Since hypoxic cells are known to be mildly acidic, the measure-
ments were repeated after addition of some aqueous hydrochloric
acid to the sample solutions. The resultant voltammograms showed
loss of reversibility of the Cu(II/I) couple. This suggests that in the
acidic environment of hypoxic cells reduction is likely to be coupled
99m
18,19
Tc. All of these complicating factors need to be considered, in
these deceptively simple systems, if they are to be used as ligands in
radiopharmaceutical preparations. Despite these complications, the
copper compounds exhibit good stability in human serum and our
1
64
2
preliminary radiolabelling experiments with H L and Cu show
3
0
to protonation as has been suggested for [Cu(ATSM)].
that, under certain conditions, a single copper complex can be syn-
thesised with high radiochemical purity that exhibited high levels
Stability studies in human serum
33
of cell uptake. The complexes decribed here offer the potential
to act as more lipophilic variants, with different electrochemistry,
and consequently cell uptake characteristics, to the well known
copper bis(thiosemicarbazone) radiopharmceuticals derived from
If these complexes are to be used as copper radiopharmaceuticals,
it is essential that they are sufficiently stable with respect to loss of
metal ion from the chelate in human serum. The perfusion tracer
1
,2-diones.
[
Cu(PTSM)] (Fig. 1) was found to be a promising tracer in animals,
6
2
but problems were encountered in the use of [ Cu(PTSM)] at high
flow rates in humans. This was attributed to interspecies variability
in the binding of [Cu(PTSM)] to serum albumin. It was shown that
Experimental
General
[
Cu(PTSM)] had stronger interaction with human serum albumin
Elemental and ICP-AES analyses were carried out by the
microanalysis service of the department. NMR spectra were
recorded on a Varian Mercury VX300 spectrometer using the
deuterated solvent signal as an internal reference. Mass spectra were
recorded on a Micromass LCT Time of Flight Mass Spectrometer
using the electrospray ionisation technique. UV/VIS spectra
were recorded on a Cintra 10 UV/VIS spectrometer (solution
concentration 2.5 × 10 M). Cyclic voltammograms were recorded
on a CH instruments Electrochemical Analyser using a platinum
working electrode, a platinum counter electrode and a silver
pseudo reference electrode. Ferrocene was used as an internal
reference which was taken as having an E1/2 = 0.53 V in DMF vs.
SCE. All reagents and other solvents were obtained from standard
commercial sources and were used as received.
than it did with dog serum albumin, but that the chelate remains
3
1
intact in its interaction with the albumin.
1
[
Cu(L O)] was incubated in human serum at 37 °C for 6 h.
Copper(II) acetate and the copper complex of histidine, [Cu(hist) ],
were also incubated as controls. [Cu(hist) ] is known to form a
well defined complex with human serum albumin, the primary
2
2
3
2
component of human serum. The mixtures were then applied to
a Sephadex G-25 PD-10 size exclusion column, which had been
−
5
−
1
2 4
previously equilibrated with 0.01 mol L KH PO . The column was
eluted with phosphate buffer and 2 mLfractions were collected. The
high molecular weight proteins rapidly eluted (second fraction) and
were identified by monitoring their UV absorbance at 280 nm. The
protein fractions were then analysed for copper content by ICP-
AES. The protein fractions collected from the human serum that had
2
been incubated with copper(II) acetate and [Cu(hist) ] contained 19
Syntheses
and 48% of the total copper concentration added, respectively. In
comparison, the protein fraction that was collected from the sample
1–3
Synthesis of the ligands 1 and H
described by Durrant and co-workers
2
L
was based on a procedure
1
that had been incubated with [Cu(L O)] analysed as containing
21
1
negligible amounts of copper. Separation of [Cu(L O)], from the
high molecular weight proteins could be observed visually with the
dark brown complex eluting down the column slowly (3rd 2 mL
fraction). This suggests [Cu(L O)] does not relinquish its copper
to high molecular weight proteins or bind to human serum albumin
after being incubated in human serum at 37 °C for 6 h.
1
. Pentane-2,4-dione (20 g, 19.98 mmol) was added to ethylene-
diamine (6 g, 99.83 mmol) in water (150 mL). After 6 h standing, a
bulky white solid settled out of solution and was recrystallised from
water to give 1 as white feathery needles (12.76 g, 56.89 mmol,
1
5
7%).
Concluding remarks
Synthesis of proligands
1
2
The above studies show that the synthesis of bis(thiosemi-
carbazone) proligands from acetylacetonate is complicated by ring
closure reactions, that result in the formation of compounds that
H L . 1 (3.15 g, 14.04 mmol) and 4-methyl-3-thiosemicarbazide
(2.97 g, 28.24 mmol) were added to ethanol (150 mL) and concen-
trated sulfuric acid (5 drops) was added. The mixture was heated
D a l t o n T r a n s . , 2 0 0 4 , 2 4 0 4 – 2 4 1 2
2 4 0 9

View Article Online
at reflux under an atmosphere of nitrogen for 3 h and allowed to
cool slowly to room temperature. A white precipitate formed which
(Found C, 36.1; H, 5.6; N, 22.6. Calc. for C11
H, 5.5; N, 23.1%). MS: m/z 376 = {[Cu(L )] − H }.
H
20
+
N
6
S
2
Cu: C, 36.3;
2
was collected by filtration and washed with diethyl ether to give
1
3
L3
H
2
L as a white powder (2.43 g, 8.85 mmol, 62%) (Found: C, 39.3;
[Cu(L )]. Method 1. As per general procedure except with H
2
H, 6.4; N, 29.9. Calc. for C
9
H
18
N
6
S
2
C, 39.4; H, 6.6; N, 30.6%).
):  1.73, 3H, s, 5-CH ; 1.95, 3H, s, 3-CH
; 2.85, 3H, d, CH N; 2.89, 3H, d, CH N; 6.40, 1H,
s, 5-NH; 7.75, 1H, s, NHNH; 8.19, 1H, m, NHCH
. 13C NMR: 
6.90, 3-CH ; 24.0, 5-CH ; 31.0, NHCH ; 31.50, NHCH ; 47.60,
CH ; 85.0, C ; 155.0, C ; 165.61, CS; 183.85, CS. MS: m/z
(0.069 g, 0.17 mmol) and copper acetate (0.034 g, 0.17 mmol).
[Cu(L )] was isolated as a green/brown powder (0.056 g,
0.12 mmol, 71%). Elemental analysis results were not consistent
1
3
H NMR (DMSO-d
.79, 2H, d, CH
6
3
3
;
2
2
3
3
3
with the calculated formula C19
20 6 2
H N S Cu due to some N,N-
1
3
3
3
3
diphenyl[1,3,4]thiadiazole-2,5-diamine contamination. MS: m/z
3
+
2
5
3
459 = {[Cu(L )] + H }.
1
+
2
2
73 = [H L − H ].
3
2
Method 2. As per general procedure except with H L (from
2
ligand method 2) (0.40 g, 1.00 mmol) and copper acetate (0.20 g,
H
2
L . As per general procedure above except with 1 (2.19 g,
3
1
.00 mmol). [Cu(L )] was isolated as a dark brown powder
9
H
(
7
3
4
2
2
1
.76 mmol) and 4-ethyl-3-thiosemicarbazide (2.34 g, 19.63 mmol).
2
(0.35 g, 0.76 mmol, 76%) (Found C, 47.8; H, 3.8; N, 17.4. Calc.
2
L was isolated as a white powder (1.32 g, 4.37 mmol, 44%)
Found: C, 43.6; H, 7.3; N, 27.7. Calc. for C11 : C, 43.7; H,
):  1.05, 6H, m, CH CH ; 1.74,
; 2.81, 2H, AB quartet, CH ; 3.44,
; 6.25, 1H, s, 5-NH; 7.71, 1H, NHNH; 8.15, 8.25,
for C19
458 = {[Cu(L )] − H }.
H
20
N
6
S
2
3
Cu. H
2
O: C, 47.9; H, 4.2; N, 17.6%). MS: m/z
22 6 2
H N S
+
1
.3; N, 27.8%). H NMR (DMSO-d
H, s, 5-CH ; 1.95, 3H, s, 3-CH
H, m, CH CH
H, t, NHCH
4.2, 5-CH
74.4, 182.4, CS. MS: m/z 302 = H
6
2
3
3
3
2
1
1
[
Ni(L )]. As per general procedure except with H
2
L (0.30 g,
2
3
1
1
3
1.09 mmol) and nickel acetate (0.27 g, 1.09 mmol). [Ni(L )] was
2
CH
3
.
C NMR:  15.1, 15.2, CH
2
CH
3
; 16.5, 3-CH
; 155.0, C
3
;
;
isolated as dark-green prismatic crystals (0.13 g, 0.40 mmol, 37%)
(Found C, 28.2; H, 5.2; N, 22.2. Calc. for C H N S Ni·3H O: C,
9 16 6 2 2
3
; 38.4, 38.8, CH
2
CH ; 47.7, CH ; 85.1, C
3
2
5
3
2
2
L . Crystals suitable for single
1
2
2
8.1; H, 5.8; N, 21.8%). H NMR (DMSO-d
.16, 3H, s, CH ; 2.63, 6H, m, NHCH ; 3.28, 2H, s, CH
; 27.3, 30.8,
(obscured by solvent peak); 151.1, 157.8, CN;
6
):  1.91, 3H, s, CH
3
;
crystal X-ray crystallography were grown by slow evaporation of an
EtOH–diethyl ether solution.
3
3
2
; 6.85, H,
m, NH; 8.0, H, m, NH. 13C NMR:  21.5, 22.9, CH
3
3
NHCH
1
3 2
; 39.5, CH
H
2
L . Method 1. As per general procedure except with 1
1
+
73.3, 176.2, C–S. MS: m/z 331 = {[Cu(L )] + H }. Crystals suit-
(
2
6
2.39 g, 10.65 mmol) and 4-phenyl-3-thiosemicarbazide (3.56 g,
3
able for single crystal X-ray crystallography were isolated from the
reaction mixture.
1.29 mmol). H
.10 mmol, 57%). Elemental analysis results were not consistent
due to some N,N-diphenyl-
1,3,4]thiadiazole-2,5-diamine being formed as a by-product.
The product obtained was of sufficient purity to synthesise metal
2
L was isolated as a white powder (2.43 g,
22 6 2
with the calculated formula C19H N S
2
2
[
Ni(L )]. As per general procedure except with H
2
L (1.15 g,
[
2
0
.50 mmol) and nickel acetate (0.12 g, 0.48 mmol).[Ni(L )] was
3
+
isolated as a pink microcrystalline solid (0.11 g, 0.30 mmol,
2%) (Found C, 36.7; H, 5.5; N, 23.2. Calc. for C11 Ni:
):  1.03, 6H, m,
CH ; 4.02, 2H, s,
C NMR:  14.4, CH CH ; 20.4,
; 156.9, CN; 171.0, C–S. MS: m/z
2
complexes. MS: m/z 399 = [H L − H ].
6
20 6 2
H N S
1
C, 36.8; H, 5.6; N, 23.4%). H NMR (DMSO-d
6
1
N,N-Diphenyl[1,3,4]thiadiazole-2,5-diamine. H NMR (DMSO-
6
):  6.94, 2H, m, Ar; 7.31, 4H, m, Ar; 7.57, 4H, m, Ar; 9.91, H, s,
NH. C NMR:  116.8, Ar; 121.0, Ar, 129.0, Ar; 141.1, Ar; 155.7,
CH
CH
CH
2
2
3
CH
; 6.94, 2H, m, NHCH
; 39.5, CH CH ; 46.2, CH
2
3
; 2.11, 6H, s, CH
3
; 3.12, 4H, m, CH
2
3
d
1
3
2
CH
3
.
2
3
1
3
2
3
+
CN. MS: m/z 267 = [product − H ].
2
+
3
59 = {[Ni(L )] + H }. Crystals suitable for single crystal X-ray
crystallography were grown by slow evaporation of a methanol
solution. These were analysed as the oxidised complex [Ni(L O)].
Method 2. As per general procedure except with 1 (1.80 g,
2
8
.02 mmol) and 4-phenyl-3-thiosemicarbazide (2.68 g, 16.0 mmol)
3
without adding sulfuric acid. H
2
L was isolated as a white powder
3
3
[
2
Ni(L )]. Method 1. As per general procedure except with H L
(
C
d
1.87 g, 4.68 mmol, 58%) (Found C, 56.2; H, 5.4; N, 20.9. Calc. for
3
(
1.06 g, 2.66 mmol) and nickel acetate (0.66 g, 2.66 mmol). [Ni(L )]
1
19
H
22
N
6
S
2
·0.5H
):  1.92, 3H, s, 5-CH
CH ; 6.67, 1H, s, 5-NH; 7.11–7.60, 10H, m,Ar; 8.93, 1H, s, NHNH;
2
O: C, 56.0; H, 5.7; N, 20.6%). H NMR (DMSO-
was isolated as a green/brown powder (1.07 g, 2.35 mmol, 88%).
Elemental analysis results were not consistent with the calculated
6
3
; 2.06, 3H, s, 3-CH ; 3.06, 2H, AB quartet,
3
2
formula C19
zole-2,5-diamine contamination. H NMR (DMSO-d
s, CH ; 2.33, 3H, s, CH ; 3.54, 2H, s, CH ; 6.90–7.58, 10H, m, Ar;
.42, H, s, NH; 10.18, H, s, NH. 13C NMR:  23.0, CH
; 23.6, CH
(obscured by solvent peak); 118.3–141.0, Ar; 153.8,
20 6 2
H N S Ni due to some N,N-diphenyl[1,3,4]thiadia-
1
3
9
2
1
.74, 1H, s, NHPh; 9.94, 1H, s, NHPh. C NMR:  15.8, 3-CH
3.7, 5-CH ; 47.3, CH ; 84.6, C ; 124.7–139.0,Ar; 155.2, C ; 173.2,
81.0, CS.
3
;
1
6
):  2.09, 3H,
3
2
5
3
3
3
2
9
3
3
3
;
9.5, CH
2
2
+
Synthesis of complexes
162.3, CN; 169.4, 173.2, C–S. MS: m/z 455 = {[Ni(L )] + H }.
Crystals suitable for single crystal X-ray crystallography were
grown by slow evaporation of a dichloromethane solution.
1
1
[
Cu(L )]. H
2
L (0.60 g, 2.19 mmol) and copper acetate (0.44 g,
2
.20 mmol) were added to methanol (20 mL). The mixture was
heated at reflux for 2 h under an atmosphere of nitrogen and allowed
to cool slowly to room temperature. Small dark brown crystals and
a brown powder formed which were collected by filtration, washed
with diethyl ether and dried in vacuo to give [Cu(L )] as dark brown
crystals and a brown powder (2.23 g, 8.14 mmol, 74%) (Found C,
3
Method 2. As per general procedure except with H
2
L (from
ligand method 2) (0.40 g, 1.00 mmol) and nickel acetate (0.25 g,
.00 mmol). [Ni(L )] was isolated as a light brown powder (0.37 g,
.81 mmol, 81%) (Found C, 48.7; H, 4.5; N, 17.4. Calc. for
19 20 6 2 2
C H N S Ni·H O: C, 48.4; H, 4.3; N, 17.8%). NMR results were
3
1
0
1
3
2
4
2
7
1.3; H, 4.5; N, 24.1. Calc. for C
9
H
14
−1
N
6
S
2
OCu: C, 30.9; H, 4.0; N,
3
+
similar to method 1. MS: m/z 455 = {[Ni(L )] + H }.
−
1
4.0%). UV-VIS: /nm (/M cm ): 293 (17200), 318 (14800),
17 (6600), 555 (1200), 778 (1400). After 15 h exposure to air:
83 (19200), 314 (16200), 378 (8000), 417 (6240), 483 (2200),
Serum stability studies
1
+
78 (400). MS: m/z 335 = {[Cu(L )] + H }. Crystals suitable for
2 2
Stock solutions of blank samples Cu(OAc) and Cu(hist) and the
single crystal X-ray crystallography were isolated from the reaction
mixture.
1
−2
−1
test sample [Cu(L O)] (5.1 × 10 mol L ) were made up in DMSO
solution. The stock solutions (0.125 mL) were added to 2.5 mL of
human serum and the samples were incubated in a thermostated oil
bath at 37 °C for 6 h before being chromatographed. The PD-10
(Sephadex G-25) size exclusion columns were equilibrated with
2
L2
[
Cu(L )]. As per general procedure above except with H
2
(
[
0.40 g, 1.32 mmol) and copper acetate (0.26 g, 1.30 mmol).
Cu(L )] was isolated as a brown powder (0.14 g, 0.38 mmol, 29%)
2
2 4
25 mL of 0.01 M KH PO buffer prior to use. The columns were
2
4 1 0
D a l t o n T r a n s . , 2 0 0 4 , 2 4 0 4 – 2 4 1 2

View Article Online
eluted with 0.01 mol L−1 KH
PO
. Three fractions (ca. 2 mL each)
each cycle of refinement. A three-term Chebychev polynomial
weighting scheme was applied. Refinement converged satisfacto-
rily to give R = 0.0266, wR = 0.0316.
2
4
were collected and the protein containing fractions were identified
by their absorbance at 280 nm. The fractions were diluted to 25 mL
with phosphate buffer solution and the copper concentrations were
determined by ICP-AES. The total copper concentration in the
diluted fractions was expected to be 16 ppm. The first and second
fractions were found to be the major protein-containing fractions
by their high intensity absorption in the UV spectrum at 280 nm.
The copper concentrations measured in the fractions collected were:
3
[Ni(L )]. Examination of the systematic absences of the intensity
datashowedthespacegrouptobe P2 /c. Coordinatesandanisotropic
1
thermal parameters of all non-hydrogen atoms were refined. The NH
hydrogen atoms were located in a difference Fourier map and their
coordinates and isotropic thermal parameters subsequently refined.
Other hydrogen atoms were positioned geometrically after each
cycle of refinement.Athree-term Chebychev polynomial weighting
scheme was applied. Refinement converged satisfactorily to give
R = 0.0291, wR = 0.0310.
2 2
Cu(OAc) 1: 2.4 ppm, 2: 3.1 ppm; Cu(hist) 1: 2.5 ppm, 2: 7.7 ppm
1
CuL O 1: 0.12 ppm, 2: 0.09 ppm (background level).
X-Ray crystallography
Crystals were mounted on a glass fibre using perfluoropolyether oil
and cooled rapidly to 150 K in a stream of cold N using an Oxford
2
2
[
Ni(L O)]. Examination of the systematic absences of the
1
intensity data showed the space group to be P2 /c. Coordinates and
Cryosystems CRYOSTREAM unit. Diffraction data were measured
using an Enraf-Nonius KappaCCD diffractometer (graphite-mono-
chromated Mo-K radiation,  = 0.71073 Å). Intensity data were
anisotropic thermal parameters of all non-hydrogen atoms were
refined. The thermal parameters of the C atoms of one of the ethyl
groups were seen to be excessively large and anisotropic, indicative
of disorder. This was modelled as disorder over two orientations. The
coordinates, anisotropic thermal parameters and site occupancies of
the disordered C atoms were refined. Geometric restraints were
applied: the N–C bonds were restrained to be 1.47(2) Å long; the
C–C bonds to 1.54(2) Å and the N–C–C angles to 110(2)°. Similarity
restraints have been applied to the components of the atomic
displacement parameters of directly bonded N and C atoms parallel
to their bonds. The NH hydrogen atoms were located in a difference
Fourier map and their coordinates and isotropic thermal parameters
subsequently refined. Other hydrogen atoms were positioned
geometrically after each cycle of refinement. The difference Fourier
map suggested that the ‘backbone’ methyl groups were disordered
and their hydrogen atoms were placed at positions corresponding to
the idealised geometry of the predominant orientation. A three-term
Chebychev polynomial weighting scheme was applied. Refinement
converged satisfactorily to give R = 0.0334, wR = 0.0335.
3
4
processed using the DENZO-SMN package.
The structures were solved using the direct-methods program
3
5
SIR92, which located all non-hydrogen atoms. Subsequent full-
matrix least-squares refinement was carried out using the CRYS-
3
6
TALS program suite.
CCDC reference numbers 237483–237488.
See http://www.rsc.org/suppdata/dt/b4/b406429a/ for crystal-
lographic data in CIF or other electronic format.
2
H
2
L . The structure was solved in the space group P1 using the
35
direct-methods program SIR92, which located all non-hydrogen
atoms. Coordinates and anisotropic thermal parameters of all non-hy-
drogen atoms were refined. The NH hydrogen atoms were located in
a difference Fourier map and their coordinates and isotropic thermal
parameters were subsequently refined. Other hydrogen atoms were
positioned geometrically after each cycle of refinement.Athree-term
Chebychev polynomial weighting scheme was applied. Refinement
converged satisfactorily to give R = 0.0340, wR = 0.0374.
Acknowledgements
1
[
Cu(L )]. The structure was solved in the space group P1 using
We thank EPSRC for funding a postdoctoral fellowship (P.S.D)
and GSK for financial support and funding a D. Phil scholarship
(J. M. H.).
3
5
the direct-methods program SIR92, which located all non-hydro-
gen atoms. Coordinates and anisotropic thermal parameters of all
non-hydrogen atoms were refined. The NH and OH hydrogen atoms
were located in a difference Fourier map and their coordinates and
isotropic thermal parameters subsequently refined. Other hydrogen
atoms were positioned geometrically after each cycle of refine-
ment. A three-term Chebychev polynomial weighting scheme was
applied. Refinement converged satisfactorily to give R = 0.0277,
wR = 0.0351.
References
1
D. X. West, S. B. Padhye and P. B. Sonawane, Struct. Bonding (Berlin),
991, 76, 1.
1
2 D. H. Petering, Bioinorg. Chem., 1972, 1, 255.
3 D. X. West, A. E. Liberta, S. B. Padhye, R. C. Chikate, P. B. Sonawane,
A. S. Kumbhar and R. G. Yerande, Coord. Chem. Rev., 1993, 123, 49.
4
K. Wada, Y. Fujibayashi and A. Yokoyama, Arch. Biochem. Biophys.,
994, 310, 1.
1
2
. Examination of the systematic absences of the intensity data
5
Y. Fujibayashi, H. Taniuchi, Y. Yonekura, H. Ohtani, J. Konishi and
showed the space group to be either Pc or P2/c. The structure
could not be solved in the latter space group but was solved in the
space group Pc using the direct-methods program SIR92,35 which
located all non-hydrogen atoms. Examination of the resulting model
showed no additional symmetry elements to be present. Coordi-
nates and anisotropic thermal parameters of all non-hydrogen atoms
were refined. The NH hydrogen atoms were located in a difference
Fourier map and their coordinates and isotropic thermal parameters
were subsequently refined. Other hydrogen atoms were positioned
geometrically after each cycle of refinement. The relative contri-
bution of the two twin components to overlapping reflections was
included as a parameter in the refinement. A four-term Chebychev
polynomial weighting scheme was applied. Refinement converged
satisfactorily to give R = 0.413, wR = 0.0480.
A. Yokoyama, J. Nucl. Med., 1997, 38, 1155.
6 J. L. J. Dearling, J. S. Lewis, D. W. McCarthy, M. J. Welch and P. J.
Blower, Chem. Commun., 1998, 2531.
Y. Fujibayahsi, C. S. Cutler, C. J. Anderson, D. W. McCarthy, L. A.
7
8
Jones, T. Sharp, Y. Yonekura and M. J. Welch, Nucl. Med. Biol., 1999,
2
6, 117.
J. S. Lewis, R. Laforest, T. L. Buettner, S.-K. Song, Y. Fujibayashi,
J. M. Connet and M. J. Welch, Proc. Natl. Acad. Sci. USA, 2001, 98,
1206.
9 J. S. Lewis, T. L. Sharp, R. Laforest, Y. Fujibayashi and M. J. Welch,
J. Nucl. Med., 2001, 42, 655.
0 J. L. J. Dearling, J. S. Lewis, G. D. Mullen, M. J. Welch and P. J.
Blower, J. Biol. Inorg. Chem., 2002, 7, 249.
1 A. R. Cowley, J. R. Dilworth, P. S. Donnelly, E. Labisbal and A. Sousa,
1
1
J. Am. Chem. Soc., 2002, 124, 5270.
12 J. R. Ballinger, Semin. Nucl. Med., 2001, XXXI, 1.
1
3 P. J. Blower, J. S. Lewis and J. Zweit, Nucl. Med. Biol, 1996, 23, 957.
14 M. A. Green, D. L. Klippenstein and J. R. Tennison, J. Nucl. Med.,
988, 29, 1549.
5 C. J. Anderson and M. J. Welch, Chem. Rev., 1999, 99, 2219.
6 V. M. Leovac, V. I. Cešljevic, G. Argay, A. Kálmán and B. Ribár,
J. Coord. Chem., 1995, 34, 357.
1
[
Ni(L )]. Examination of the systematic absences of the intensity
1
data showed the space group to be Pbca. Coordinates and aniso-
tropic thermal parameters of all non-hydrogen atoms were refined.
The NH hydrogen atoms were located in a difference Fourier map
and their coordinates and isotropic thermal parameters subsequently
refined. Other hydrogen atoms were positioned geometrically after
1
1
17 S. C. Davies, M. C. Durrant, D. L. Hughes, A. Pezeshk and R. L.
Richards, J. Chem. Res. (S), 2001, 100.
D a l t o n T r a n s . , 2 0 0 4 , 2 4 0 4 – 2 4 1 2
2 4 1 1

View Article Online
1
1
2
8 Y. Arano, M. Yabuki, T. Yahata, K. Horiuchi and A. Yokoyama, Chem.
Pharm. Bull., 1990, 38, 3099.
9 Y. Arano, M. Yabuki, A. G. Jones and A. Yokoyama, Chem. Pharm.
Bull., 1991, 39, 104.
27 V. B. Arion, N. V. Gerbeleu and K. M. Indrichan, Russ. J. Inorg. Chem.,
1985, 30, 70.
28 S. Yokota, Y. Tachi and S. Itoh, Inorg. Chem., 2002, 41, 1342.
29 R. I. Maurer, P. J. Blower, J. R. Dilworth, C. A. Reynolds, Y. Zheng and
G. D. Mullen, J. Med. Chem., 2002, 45, 1420.
0 D. H. Hunter, C. McRoberts and J. J. Vittal, Can. J. Chem., 1998, 76,
5
22.
30 A. R. Cowley, J. R. Dilworth, P. S. Donnelly, E. Labisbal and A. Sousa,
J. Am. Chem. Soc., 2002, 124, 5270.
31 C. J. Mathias, S. R. Bergmann and M. A. Green, J. Nucl. Med., 1995,
36, 1451.
32 J. Masuoka and P. Saltman, J. Biol. Chem., 1994, 41, 25557.
33 J. R. Ballinger, J. R. Dilworth, P. S. Donnelly and J. M. Heslop, 2003,
unpublished results.
34 Z. Otwinowski and W. Minor, Methods Enzymol., 1997, 276, 307.
35 A. Altomare, G. Cascarano, G. Giacovazzo,A. Guagliardi, M. C. Burla,
G. Polidori and M. Camalli, J. Appl. Crystallogr., 1994, 27, 435.
36 D. J. Watkin, C. K. Prout, J. R. Carruthers, P. W. Betteridge and R. I.
Cooper, in CRYSTALS issue 11, Oxford, UK, 2001.
2
2
2
1 C. N. O’Callaghan and D. Twomey, J. Chem. Soc., 1967, 22, 2400.
2 B. Stanovnik and M. Tišler, J. Org. Chem., 1961, 26, 5200.
3 P. J. Blower, T. C. Castle, A. R. Cowley, J. R. Dilworth, P. S. Donnelly,
E. Labisbal, F. E. Sowrey, S. J. Teat and M. J. Went, Dalton Trans.,
2
003, 4416.
2
2
2
4 E. Bermejo, A. Castiñeiras, L. J. Ackerman, M. D. Owens and D. X.
West, Z. Anorg. Allg. Chem., 2001, 627, 1966.
5 G. J. Colpas, M. Kumar, R. O. Day and M. J. Maroney, Inorg. Chem.,
1
990, 4779.
6 M. E. Padilla-Tosta, D. Fox, M. G. B. Drew and P. D. Beer, Angew.
Chem., 2001, 40, 4235.
2
4 1 2
D a l t o n T r a n s . , 2 0 0 4 , 2 4 0 4 – 2 4 1 2