Towards aqueous chiral heptadentate lanthanide complexes as selective shift and relaxation agents for MRS

Article abstract of DOI:10.1039/b706801h

Magnetic resonance spectroscopy (MRS) is of prime importance in diagnostics and offers a means of analyzing, in vivo, the chemical content of living tissue, as a non-invasive alternative to biopsy. Several heptadentate, lanthanide complexes have been synt

Full text of DOI:10.1039/b706801h

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Towards aqueous chiral heptadentate lanthanide complexes as selective shift
and relaxation agents for MRS†
Rachel S. Dickins* and Alessandra Badari
Received 8th May 2007, Accepted 1st June 2007
First published as an Advance Article on the web 4th July 2007
DOI: 10.1039/b706801h
Magnetic resonance spectroscopy (MRS) is of prime importance in diagnostics and offers a means of
analyzing, in vivo, the chemical content of living tissue, as a non-invasive alternative to biopsy. Several
heptadentate, lanthanide complexes have been synthesized and their potential to act as shift and
relaxation agents in MRS (for lactate, in particular) has been assessed through 1H NMR analysis. The
binding affinity and enantiopurity of the complexes have been modulated by systematic variation of the
lanthanide ion and ligand structure, in particular the peripheral electrostatic charge of the complex
(cationic versus neutral) and the local charge and steric demand at the metal centre.
The unique paramagnetic properties of the lanthanides have
Introduction
been exploited in NMR spectroscopy for many years.7,8 In order
Magnetic resonance spectroscopy (MRS) offers a means of ana-
to improve the diagnostic potential of MRS it would be highly
lyzing the chemical content of living tissue and may be applied to
advantageous to utilize the enhanced shifting and relaxation prop-
diagnostics as a non-invasive alternative to biopsy.1 For example,
erties of paramagnetic shift reagents, designed to bind strongly
MRS has been widely applied in the laboratory in the study of
and selectively to bioactive species of interest, to effect spectral
tumour biochemistry and physiology as the spectral metabolite
resolution and allow rapid acquisition times to be implemented,
signatures obtained from diseased tissue are very distinct and
giving increased signal intensity. Cationic, lanthanide complexes
based on L1 have two bound water molecules which may be readily
may provide information on tumour metabolism, tumour type
and degree of malignancy.2,3 Prognostic information on the likely
response to treatment and drug sensitivity is also available.4 31P
MRS offers potential information on phospholipid turnover, cell
proliferation and elevated levels of the phosphomonoesters and
phosphodiesters which represent a general feature of most tumour
cells and tissues and are associated with rapid tissue growth or cell
membrane synthesis.5 1H MRS also allows observation of several
key metabolites associated with malignancy (e.g. lactate, choline,
phosphocholine) and therefore provides a wealth of information.6
displaced by a variety of monodentate and chelating anions.9,10
1
Anion binding has been effectively signaled through H NMR
as the observed lanthanide induced shift of the macrocyclic ring
protons of [Ln·L1]3+ may be directly correlated to the nature of the
donor atom in the axial position: the more polarisable donors
giving the smallest shifts.11,12 Furthermore, anion substituents
lying within the ‘McConnell’ cone are also susceptible to a large
lanthanide induced shift. For example, [Yb·L1]3+ was found to
shift the CH and CH3 resonances of (S)-lactate, a key metabolite
to monitor, to +58 ppm and +20 ppm respectively (cf . +4.1 ppm
and +1.3 ppm in the free form respectively) well out of the range
of any interferent biological signals.13 However, the binding of
lactate (and other bioactive species) to [Ln·L1]3+ was found to
be in slow exchange on the NMR timescale, with resonances
corresponding to both bound and unbound species resulting
with sub-stoichiometric amounts present, the complex therefore
behaving as a chiral derivatising agent rather than a chiral shift
reagent.
In order to achieve the fast exchange conditions required for
potential use as shift reagents in MRS, the affinity and selectivity
of such macrocyclic lanthanide complexes for bioactive anions has
been modulated by systematic variation of the lanthanide ion and
ligand structure, in particular the peripheral electrostatic charge of
the complex (cationic versus neutral) and the local charge and steric
demand at the metal centre. Accordingly, we have synthesized
lanthanide complexes of L2–L7 and investigated their potential
to act as shift reagents for bioactive species. Herein, we report
our studies on lactate, the end product of the anaerobic glycolysis
pathway, the detection of which is of prime importance in the
diagnosis, grading and therapeutic monitoring of disease such as
cancer, stroke and heart disease.
1
Although H MRS is fifteen times more sensitive than 31P MRS
there have been a limited number of studies reported due to intense
background signals from water (55 M) and lipids (1 mM) which
mask the resonances of interest.
In vivo MRS suffers from some major limitations: long acqui-
sition times are required as S : N (signal-to-noise) is often very
poor, resolution is low as resonances are broad and often overlap,
presenting problems in analysis and quantification of spectra.
MRS is therefore widely applied in the laboratory but has yet
to find a routine clinical role. However, in vivo MRS is vital as
in vitro models experience different oxygenation, nutrient supplies,
exposure to hormone and immunological factors and can result
in degradation of some metabolites (e.g. high energy phosphates).
Since MRS can be performed on patients using a slightly modified
hospital MRI scanner, imaging and spectroscopic studies can be
combined to first locate then analyse a tumour non-invasively.
Hence there is great interest in putting the unique capabilities of
MRS to work in diagnostics.
Department of Chemistry, University of Durham, South Road, Durham, UK
DH1 3LE. E-mail: R.S.Dickins@durham.ac.uk
† Electronic supplementary information (ESI) available: Experimental
details – Ligand and complex syntheses. See DOI: 10.1039/b706801h
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 3661–3668 | 3661
©

View Article Online
anticlockwise (K) (NCCO torsion angle positive = D; negative =
K) and the macrocyclic ring can adopt two conformations (NCCN
torsion angle positive = dddd; negative = kkkk), giving rise to
four possible isomers in solution (Scheme 2). The presence of
the remote chiral centre at the amide functionality renders each of
the four isomers diastereomeric, yielding two square antiprismatic
(SAP) isomers, K(dddd) and D(kkkk), and two twisted square
antiprismatic isomers (TSAP), D(dddd) and K(kkkk). Variable
temperature NMR suggests that [Yb·L1(D2O)2]3+(CF3SO3−)3 ex-
ists in the two SAP forms which are in fast exchange at room
temperature.12 Indeed, upon addition of a-amino acids, resonances
corresponding to the two SAP amino acid ternary complexes are
observed at 298 K.12 The presence of the two isomers clearly
complicates spectral analysis therefore the design of an MRS agent
existing as a single isomer in solution is highly desirable.
Synthesis
The synthesis of the trisubstituted ligands, L2–L5, was attempted
by direct alkylation of tetraazacyclododecane with the appropriate
chloroamide. A mixture of mono, di and trisubstituted products
resulted. Although the trialkylated product was isolated via
exhaustive extraction of the organic phase with water or by column
chromatography, yields were poor. Therefore an alternative route
involving the selective protection of a single macrocyclic ring ni-
trogen, followed by trialkylation and subsequent deprotection was
employed (Scheme 1). Reaction of 1,4,7,10-tetraazacyclododecane
molybdenum tricarbonyl14 with benzyl bromide (dry DMF, 60 ◦C,
K2CO3), followed by decomplexation of the molybdenum moi-
ety in aqueous acid, yielded the monobenzylated macrocycle.
Subsequent reaction with three equivalents of the appropriate
chloroamide (dry DMF, Ar, 60 ◦C, K2CO3) afforded the tetra-
alkylated ligands (e.g. L6) after purification on neutral alumina.
The trialkylated ligands L2 and L4 were isolated following reduc-
tive deprotection of the N-benzyl group (H2, 45 psi, Pd(OH)2/C,
acidic EtOH, RT). Complexation was achieved by refluxing the
ligand in dry acetonitrile, with the appropriate lanthanide triflu-
oromethanesulfonate salt. Subsequent basic hydrolysis (0.02 M
NaOH) of [Ln·L2]3+, [Ln·L4]3+ and [Ln·L6]3+, followed by cation
exchange chromatography, yielded the neutral [Ln·L3], [Ln·L5] and
[Ln·L7] complexes respectively.
Scheme 2 Stereoisomerism associated with lanthanide complexes of L1.
Ligand variation
In devising enantiopure complexes, conformationally rigid on the
NMR timescale, pendent arms based on amino acids were chosen
as they offer simple routes to a variety of both cationic (as ester)
or neutral complexes (as acid, via subsequent hydrolysis of the
corresponding ester). Primarily, Yb was the lanthanide of choice
as it exhibits a large lanthanide induced shift with minimal line
broadening characteristics.
1
The H NMR spectra of the diaqua complexes, [Yb·L2]3+ and
[Yb·L3] (10 mmol, D2O–CD3OD, 200 MHz, 295 K), showed the
presence of two isomers in solution at room temperature, with
shifts indicative of SAP geometries [K(dddd) and D(kkkk)].11,15 The
presence of the two species coupled with poor water solubility
renders complexes of L2 and L3 unsuitable for use as shift reagents
in MRS.
In the search for water-soluble, enantiopure lanthanide com-
1
plexes, [Ln·L4]3+ and [Ln·L5] were synthesized. H NMR stud-
ies of [Yb·L4]3+ (10 mmol, D2O–CD3OD, 200 MHz, variable
temperature) revealed the presence of a single major species in
solution (Fig. 1a). The eight distinct resonances to high frequency
(+104 → +14 ppm) were attributed to the four most shifted axial
(᭹) and equatorial (x) protons of the macrocyclic ring. On addition
of 0.25 equivalents of (S)-lactate sodium salt, the appearance of an
additional set of axial and equatorial ring resonances were evident,
corresponding to the lactate bound species in slow exchange on the
NMR timescale with the free form. After addition of 1 equivalent
of (S)-lactate, only the bound form was observed (Fig. 1b) and an
additional singlet resonance at +26 ppm was evident, integrating
to three protons, which was subsequently attributed to the methyl
resonance of the bound (S)-lactate. The expected methyl doublet
Scheme 1 Synthetic procedure for the ligands.
Stereochemical and NMR analysis
Lanthanide complexes of L1 exhibit two independent features
of chirality: the pendent arms may be rotated clockwise (D) or
3662 | Dalton Trans., 2007, 3661–3668
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
solely one isomer (SAP) was observed in a mixture of D2O
and CD3OD. Lactate binding studies revealed slow exchange
conditions between the free and bound forms and upon addition
of racemic lactate, enantioselectivity in the binding to [Yb·L6]3+
was observed (cf . lack of enantioselectivity reported for the NH
analogue, [Yb·L4]3+), with the sterically demanding benzyl group
favouring binding to (S)-lactate (1.4 : 1, S : R).
1
The H NMR spectrum of [Yb·L7] (10 mmol, D2O–CD3OD,
200 MHz, 295 K) (Fig. 3) showed the presence of two isomers in
solution, in slow exchange on the NMR timescale. Interestingly,
the major species occurred with shifts indicative of a twisted square
antiprismatic (TSAP) conformation while the minor species
appeared to adopt the square antiprismatic (SAP) geometry (3 : 1,
TSAP : SAP). A change in the isomer ratio was observed in a range
of solvents, with polar aprotic solvents favouring purely the twisted
form. The N-benzylated [Yb·L7] appears to have a reduced affinity
towards lactate compared to the NH analogue [Yb·L5]. In fact, in
the presence of an excess of lactate (2 equivalents), resonances
corresponding to both the free SAP and TSAP forms (along
with the bound SAP form) were observed, with the fully bound
SAP form only being observed upon addition of 5 equivalents of
lactate. The addition of an excess of racemic lactate to [Yb·L7],
revealed preferential binding to (S)-lactate (2 : 1, S : R), (cf . lack
of enantioselectivity reported for the NH analogue, [Yb·L5]).
Fig. 1 Partial 1H NMR spectra of a) (SSS)-[Yb.L4]3+ [(᭹) axial and
equatorial (x) ring protons] and in the presence of b) 1 equiv. of (S)-lactate
[(*) (S)-lactate CH3]; c) 2 equiv. racemic lactate [(#) (R)-lactate CH3]
(200 MHz, D2O–MeOD, 295 K).
was not resolved due to the broadening effect of the paramagnetic
ion. The exact position of the lactate CH resonance could not
be defined due to inconclusive two dimensional spectra obtained
for this system,16 although an extra singlet resonance observed
in the high frequency region of the spectrum could be accounted
for. A lack of enantioselectivity in adduct formation was apparent
following addition of 2 equivalents of racemic lactate to [Yb·L4]3+
and a 1 : 1 mixture of diastereomeric complexes and was observed
1
by H NMR (Fig. 1c). The methyl lactate resonance was clearly
resolved for the (S) (+26 ppm) and (R) (+16 ppm) diastereoisomers
(DDd ∼ 9 ppm) and experienced a large lanthanide induced shift
(+1.3 ppm in the free form), indicative of bidentate chelation with
the hydroxyl group axially disposed.11,12
1H NMR investigation of a 10 mM solution of the water soluble
[Yb·L5] revealed the presence of many broadened peaks. However,
upon dilution to a 1 mM complex solution a single species was
evident, characterized in the high frequency range by the eight
singlet resonances of the four most shifted axial and equatorial ring
protons, consistent with the formation of dimers at higher concen-
tration, which was also confirmed by ES-MS [1735 (10%, M2Na+)].
The complex was found to behave in an analogous manner to
[Yb·L4]3+ upon addition of lactate; the bound and free forms in
slow exchange. The lactate CH3 resonances are clearly resolved
for the diastereomeric (R) and (S) adducts (DDd ∼ 10 ppm)
and exhibit large lanthanide induced shifts (+20 ppm and +30 ppm
for (R) and (S) respectively).
In an attempt to observe the effect of N-alkylation on the
anion binding properties of the lanthanide complexes, [Ln·L6]3+
and [Ln·L7] were prepared. 1H NMR of [Yb·L6]3+ (10 mmol,
CD3OD, 200 MHz, 295 K) (Fig. 2) revealed the presence of
mainly one isomer (square antiprism) in solution along with
traces of a minor species (< 8%) to lower frequency, with shifts
indicative of a twisted square antiprism conformation.15 However,
Fig. 3 Partial 1H NMR spectrum of [Yb·L7] showing the presence of
both the SAP and TSAP (ꢀ) isomers (200 MHz, D2O–CD3OD, 295 K).
Binding affinities
The relative binding affinities of [Eu·L4]3+, [Eu·L5] and [Eu·L7] for
(S)-lactate were measured by plotting the changes in intensity of
the DJ = 2 transition (614 nm) in the europium emission spectra as
a function of added anion concentration [1 mM complex (Eu·L4
and Eu·L5) or 0.25 mM (Eu·L7), in 0.1 M MOPS buffer, pH 7.4,
295 K]. The resulting binding isotherms were fitted to a 1 : 1
binding model by least-squares iterative analysis. A representative
binding isotherm is reported in Fig. 4. As expected, the neutral
[Eu·L5] exhibited a reduced binding affinity in comparison to the
cationic analogue [Eu·L4]3+ (log K = 2.58 and 3.20 respectively)
reflecting the increased electrostatic repulsion between the anion
and the negatively charged carboxylate groups of the pendent
arms. The apparent binding constant was found to increase by
an order of magnitude for the N-alkylated derivative, [Eu·L7]
(log K = 4.24) compared to the NH derivative [Eu·L5]. This
reflects the lesser degree of hydration around the anion binding
site for the more hydrophobic [Eu·L7] and the greater overall free
energy change accompanying the anion binding. This is consistent
with hydration state studies performed on the analogous terbium
complexes following analysis of the rate constants for radiative
5
decay of the terbium D4 excited state in H2O and D2O, using
established methodology.17 Such studies revealed the presence of
Fig. 2 Partial 1H NMR spectrum of [Yb·L6]3+ showing traces of the minor
TSAP species (*) (200 MHz, CD3OD, 295 K).
two bound water molecules for [Tb·L5] and only one for the
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 3661–3668 | 3663
©

View Article Online
occur for the less charge dense ions (e.g. Pr, Nd). Indeed, the use
of [Pr·DO3A] as a shift reagent for differentiating between intra
1
and extracellular lactate by H NMR has recently been reported
(DDd ∼ 0.2 ppm).18 However, in addition to displaying a high
rate of exchange between free and bound forms, the lanthanide
complex must also exhibit an optimal value for Dd/Dm1/2 (the
ratio of the chemical shift (Dd) to the linewidth at half height
maximum of the shifted resonance (Dm1/2)) in order to induce a
large paramagnetic shift in the resonance of interest to a spectral
region where background signals do not interfere along with
minimal line broadening effects.
1H NMR studies of the diaqua [Ln·L4]3+ and [Ln·L5] complexes
revealed the presence of a single species in solution for each
of the lanthanide ions. Complexes of the lighter lanthanides
(e.g. Ce3+–Eu3+) appeared to be characterized by relatively small
paramagnetic shifts, compared to the heavier ions (e.g. Tb3+–Yb3+)
(Tables 1 and 2). The size and direction of the paramagnetic
shift is determined by the Bleaney constant (Cj) and is unique
for each lanthanide ion.19 Therefore it is possible to divide the
lanthanide complexes into two groups: those with negative Cj
values (Ce3+, Pr3+, Nd3+, Sm3+, Tb3+, Dy3+, and Ho3+) shift the most
shifted axial proton of the macrocyclic ring to lower frequency
whereas those with positive Cj values (Eu3+, Er3+, Tm3+ and
Yb3+) shift to higher frequency. For example, among the cationic
[Ln·L4]3+ systems (Table 1), [Nd·L4]3+ and [Eu·L4]3+ (with Cj = −4.2
Fig. 4 Variation of europium emission intensity (614 nm) for [Eu·L4]3+
as a function of added (S)-lactate (1 mM complex, pH 7.4, 0.1 M MOPS,
295 K), showing the fit to the experimental data points.
sterically more demanding and hydrophobic [Tb·L7] analogue.
The apparent lack of consistency in binding affinity with the results
obtained by 1H NMR investigation for [Yb·L7], which appeared to
exhibit a weaker affinity for lactate compared to the NH derivative
[Yb·L5], can be explained by the different geometries adopted by
the complexes in solution. [Yb·L7] was found to exist as two species
in solution: the major TSAP isomer (3 : 1 TSAP : SAP) did not
bind to lactate, whereas the minor SAP species exhibited lactate
binding. The 1H NMR spectrum of [Eu·L7] however, showed the
presence of a single isomer in solution, with shifts indicative of
the SAP conformation. The SAP form appears to favour lactate
binding, thus explaining the increase in affinity towards the anion
observed in the Eu3+ binding experiment.
Table 2 1H NMR investigation of the diaqua neutral complexes [Ln·L5]],
reporting the chemical shift of the most shifted axial proton of the ring
(DdHax), the spectral range, the half peak height line width (m1/2) and the
ratio of DdHax/m1/2 (200 MHz or 65 MHz*, CD3OD, 295 K).
[Ln·L5] Ln =
dHax (ppm)
Range (ppm)
m1/2/Hz
DdHax/m1/2
Lanthanide variation
Ce
Pr
−15
−33
−17
—
+19 → −15
101
92
75
—
53
722
301
876
370
704
330
0.15
0.36
0.23
—
0.49
0.46
1.27
0.22
0.33
0.41
0.32
Although fast exchange conditions were not achieved by variation
of the ligand structure, complexes of L4 and L5 appear to exist as
a single isomer in solution, as confirmed by VT 1H NMR, and are
readily water soluble. Therefore in order to control the size and
direction of the paramagnetic shift and to achieve fast exchange
conditions, the lanthanide ion was varied for complexes based
on the enantiopure ligands, L4 and L5, and the effect on anion
affinity investigated. Lewis acidity increases across the lanthanide
series, therefore fast exchange conditions may be envisaged to
+34 → −33
Nd
Sm
Eu
Tb*
Dy*
Ho
Er
+17 → −17
+11 → −1.0
+26 → −15
+26
−333
−381
−195
+123
+290
+106
+250 → −333
+250 → −381
+142 → −195
+123 → −65
+290 → −389
+106 → −121
Tm*
Yb
Table 1 1H NMR investigation of the diaqua cationic complexes [Ln·L4]3+, reporting the pseudocontact (Cj)19 constant, the chemical shift of most
shifted axial proton of the ring (dHax), the spectral range, the half peak height line width (m1/2), and the ratio of DdHax/m1/2 (200 MHz or 65 MHz*,
CD3OD, 295 K).
[Ln·L4]3+ Ln =
Cj19
dHax (ppm)
Range (ppm)
m1/2/Hz
DdHax/m1/2
Ce
Pr
−6.3
−11.0
−4.2
−0.7
+4.0
−86
−100
−39
+33
−12
−30
−16
—
+15→ −12
+25→ −30
+16→ −16
+7→ 0.0
223
187
273
—
205
418
365
596
425
480
209
0.05
0.16
0.06
—
0.12
0.77
1.04
0.21
0.28
0.61
0.57
Nd
Sm
Eu
Tb*
Dy*
Ho
Er
+25
+25→ −19
+81 → −320
+300 → −379
+191→ −125
+118→ −75
+293→ −254
+120→ −91
−320
−379
−125
+118
+293
+120
Tm*
Yb
+53
+22
3664 | Dalton Trans., 2007, 3661–3668
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
and +4.0, respectively) exhibit a much narrower spectral range
(+16 → −16 ppm and +25 → −19 ppm respectively) compared to
[Tb·L4]3+ (Cj = −86) and [Tm·L4]3+ (Cj = +53) (+81 → −320 ppm
and +293 → −254 ppm respectively). In order to assess the relative
magnitude of the dipolar broadening of the nuclear resonances,
the line width at half peak height (m1/2) has been measured for the
most shifted axial proton of the ring in each of the cationic and
neutral complexes. The results, reported in Tables 1 and 2, reveal
that complexes of the Ln3+ ions belonging to the second part of the
series (Tb3+ to Tm3+), though better shifters, are also more efficient
dipolar broadeners than complexes of the early lanthanides. The
best compromise between line broadening and amplitude of the
shift (Dd/Dm1/2) appears to be exhibited by the dysprosium and
terbium complexes.
Conclusions
In the search for potential NMR shift reagents for bioactive
anions, a variety of heptadentate Ln3+ complexes have been syn-
thesised and assessed. In order to achieve the required conditions
of fast exchange on the NMR timescale, the binding affinity and
selectivity for anions has been modulated by variation of the Ln3+
ion and the ligand structure. The cationic [Ln·L4]3+ and neutral
[Ln·L5] complexes, were found to exist solely as a single species in
aqueous solution and displayed good water solubility rendering
them ideal candidates for assessment as potential MRS agents for
bioactive species.
1
In the H NMR studies for the detection of lactate, it appears
that the later lanthanides (Tb, Dy, Tm, Yb) may exhibit an optimal
Dd/Dm1/2 ratio, achieving maximal shifting properties coupled with
minimal line broadening. However, the high Lewis acidity of these
ions prevents the binding of anions under fast exchange conditions.
The reduced Lewis acidity of the earlier lanthanides (e.g. [Pr·L4]3+)
may possibly enable fast exchange conditions to be met but due to
the relatively large degree of line broadening and the small spectral
range observed for such complexes, spectral analysis is difficult.
Therefore the search goes on for a lanthanide complex capable
of exhibiting fast exchange conditions coupled with an optimal
Dd/Dm1/2 ratio.
Although complexes of L4 and L5 are not suitable as shift
and relaxation agents for MRS (slow exchange on the NMR
timescale), the existence of a single isomer in aqueous solution
(cf . two isomers previously reported for complexes of [Ln·L1]3+)12
coupled with the ability to induce a large lanthanide induced shift,
renders such complexes highly desirable as effective NMR probes
for a number of bioactive species. Furthermore, by suitable choice
of lanthanide ion luminescent (Eu, Tb) and chiroptical (Yb, Eu)
probes may also be envisaged.
A systematic study was then undertaken to assess whether anion
binding under fast exchange conditions could be achieved by
variation of the Ln3+ ion. Aliquots of a solution of (S)-lactate
sodium salt (175 mM, D2O, complex to lactate ratio ranging from
4 : 1 to 1 : 2) were added successively to a solution of [Ln·L4]3+ and
1
[Ln·L5] (10 mM, D2O) and the H NMR spectrum run for each
increment. Upon addition of 0.25 equivalents of (S)-lactate, slow
exchange on the NMR timescale was observed between the lactate-
bound and the free form for most of the cationic and neutral
complexes. The lactate CH3 did experience a large lanthanide
induced shift, the size and direction of which was highly dependant
upon the Ln3+ ion (e.g. dCH3 = −42 ppm for [Ho·L5]; +39 ppm
for [Er·L5]). A lack of enantioselectivity in the binding of racemic
lactate was observed for the cationic [Eu·L4]3+, [Tm·L4]3+, [Yb·L4]3+
and neutral [Tb·L5], [Dy·L5], [Ho·L5], [Er·L5], [Tm·L5], [Yb·L5]
complexes, and a 1 : 1 mixture of the diastereomeric R and S
adducts were apparent. Broad and overlapping proton resonances
for the remaining Ln complexes prevented the identification of
1
the diastereomeric peaks. Although a change in the H NMR
spectrum was observed upon addition of lactate to the cationic
[Pr·L4]3+ complex, the resonances experienced severe broadening
and the condition of fast or slow exchange could not be fully
defined. Therefore 0.1 equivalents of [Pr·L4]3+ were added to
a solution of (S)-lactate (10 mmol, D2O) and the 1H NMR
spectrum was run. The (S)-lactate CH resonance (+4.1 in the
free form) disappeared upon the addition of [Pr·L4]3+, consistent
with either a severe paramagnetic broadening of the resonance or
the occurrence of fast exchange conditions between the free and
bound forms. The lactate CH3 resonance, however, experienced
a slight shift (+0.2 ppm) and broadened significantly. Analogous
1H NMR studies were undertaken with other Ln complexes and
although a severe broadening of both the (S)-lactate CH and CH3
resonances were apparent, in all the cationic [Ln·L4]3+ and neutral
[Ln·L5] complexes, slow exchange conditions were observed and
no significant shift or disappearance of the lactate peaks were
observed. Therefore, fast exchange conditions may possibly have
been achieved for the [Pr·L4]3+ complex. However, due to the
severe line broadening exhibited by the complex, coupled with the
small spectral range, the actual position of the shifted lactate CH
peak could not be assigned and the [Pr·L4]3+ complex is therefore
unsuitable as a shift reagent for lactate. With the analogous neutral
[Pr·L5] complex it appears that the reduced electrostatic charge
coupled with the low charge density of the Pr3+ ion decreases the
affinity towards the lactate anion sufficiently, and no significant
binding was observed.
Experimental
General procedures and characterization techniques
Reactions requiring anhydrous or inert conditions were carried
out using Schlenk-line techniques under an atmosphere of dry
argon. Anhydrous solvents when required were freshly distilled
over the appropriate drying agent. Water was purified by the
‘PuriteSTILLplus’ system.
Thin-layer chromatography was carried out on neutral alumina
plates (Merck Art 5550) and visualised under UV (254 nm)
or by staining with iodine. Preparative column chromatography
was carried out using neutral alumina (Merck Aluminium Oxide
90, activity II-III, 70–230 mesh), pre-soaked in ethyl acetate.
Cation and anion exchange chromatography were performed using
Dowex 50 W strong ion exchange resin and Dowex 1-X8(Cl),
respectively, pre-treated with hydrochloric acid (3 M).
Infra-red spectra were recorded on a Graseby-Specac “Golden
Gate” Diamond ATR accessory spectrometer. Melting points were
recorded using a Ko¨fler block and are uncorrected.
1H and C NMR spectra were recorded on a Varian Mercury-
200 (1H at 199.975 MHz, 13C at 50.289 MHz), Varian Unity 300
(1H at 299.908 MHz, 13C at 75.412 MHz) or a Varian Inova
500 (1H at 499.7 MHz, 13C at 125.67 MHz) spectrometer. Two-
dimensional spectra were carried out on a Varian Inova 500
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 3661–3668 | 3665
©

View Article Online
spectrometer. Spectra were referenced internally relative to tert-
butanol for paramagnetic complexes or to the residual protio-
solvent resonances which are reported relative to TMS. All
chemical shifts are given in ppm and coupling constants are in
1731 (COO), 1650 (CONH); dH (200 MHz, CDCl3) 7.30–7.10 (5H,
m, Ar), 7.00 (1H, br s, NH), 4.84 (1H, m, CH), 4.17 (2H, q, 3J =
7.2, CH2O), 4.00 (2H, s, CH2Cl), 3.14 (2H, d, 3J = 5.8, CH2Ph),
1.24 (3H, t, 3J = 7, CH3); dC (50.29 MHz, CDCl3) 171.2 (CO2Et),
166.4 (CONH), 136.0 (ipso-Ar), 129.5 (m-Ar), 128.7 (o-Ar), 127.3
(p-Ar), 61.8 (OCH2), 53.9 (CH), 42.6 (CH2Cl), 37.8 (CH2Ph), 14.2
(CH3); m/z (ES+) 292 (100%, MNa+), 561 (90%, M2Na+), 270
(20%, MH+).
1
Hz. H NMR studies on the lactate bound ternary complexes
were obtained by adding successive 0.01 ml aliquots of a solution
of the appropriate sodium lactate salt (175 mmol, D2O, complex
to lactate ratio ranging from 4 : 1 to 1 : 2) to a solution of the
complex (10 mmol, 0.70 ml, D2O or CD3OD–D2O).
Electrospray mass spectra were recorded on a VG Platform II
(Fisons instrument), operating in positive or negative ion mode
as stated, with methanol as the carrier solvent. Accurate mass
spectra were recorded at the EPSRC Mass Spectrometry Service
at the University of Wales, Swansea.
Corrected emission spectra were recorded on a S.A. Fluorolog
3, operating with DataMax software using quartz fluorescence
cuvettes of path length 1 cm, a 420 nm cutoff filter and an excitation
wavelength of 397 nm (Eu) for direct excitation or 254 nm for
sensitized emission. Eu3+ spectra were recorded between 570 nm
and 730 nm, at 0.5 nm increments with 1 s integration time.
Excitation and emission slits were 10 and 1 nm respectively. Data
analysis was performed using an iterative least-squares fitting
procedure, assuming a 1 : 1 binding stoichiometry, operating in
Microsoft Excel. Statistical errors were estimated to be <10%,
and experimental errors were within 20%.
Excited state lifetimes were measured on a Perkin-Elmer LS55
spectrofluorimeter using FL WinLab Software. Reported decay
rates are the average of at least 3 separate measurements calculated
by monitoring the emission intensity, at 545 nm for Tb3+, after at
least 20 different delay times, covering 2 or more lifetimes. The
gate time was 0.1 ms and the excitation and emission slits were
set to 15 nm bandpass. The phosphorescence decay curves were
fitted by an equation of the form I(t) = I(0)exp(−t/s) using a
curve fitting program, operating in Microsoft Excel, where I(0)
is the initial intensity at t = 0 and I(t) is the intensity after
time t and s is the phosphorescence lifetime. Lifetimes for the
lactate bound adducts were obtained by adding 10 equivalents
of the sodium salt, as a solid, to a solution of the complex
(0.5 ml, 2.5 mmol for direct excitation; 0.25 mM for indirect
excitation).
2-Chloro-N-[(S)-(1-methoxycarbonyl-3-methyl)butyl] ethanamide
The title compound was prepared following a method similar
to that above, using (S)-leucinemethylester hydrochloride (3 g,
16.5 mmol) and triethylamine (5.52 ml, 39.6 mmol) in dry diethyl
ether (100 ml) and reacting with a solution of chloroacetylchloride
(1.58 ml, 19.8 mmol) in dry diethyl ether (60 ml). A brown
oil resulted (3.21 g, 88%); mmax (film)/cm−1 3304 (NH), 1745
(COO), 1680 (CONH); dH (200 MHz, CDCl3) 6.90 (1H, br s,
NH), 4.68–4.56 (1H, br m, CHNH), 4.04 (2H, s, ClCH2CO),
3.72 (3H, s, OCH3), 1.69–1.50 (3H, br m, CH2CH), 0.92 (6H, d,
3J = 5.8, CH3); dC (50.29 MHz, CDCl3) 172.8 (CO2Me), 166.6
(CONH), 52.2 (OCH3), 51.0 (CHNH), 42.4 (ClCH2CO), 40.7
(CH2CH), 24.7 (CH2CH), 22.7 (CH3), 21.6 (CH3); m/z (ES+) 244
(100%, MNa+) (Found: MNa+, 244.0739. C9H16NO3NaCl requires
MNa+, 244.0716).
General procedure for the preparation of tetraalkylated ligands.
1-Benzyl-4,7,10-tris-[(S)-{(1-ethoxycarbonyl-2-phenyl)ethyl}-
carbamoylmethyl]-1,4,7,10-tetraazacyclododecane. 2-Chloro-N-
[(S)-(1-ethoxycarbonyl-2-phenyl)ethyl] ethanamide (2.68 g,
9.96 mmol) in dry N,N-dimethylformamide (5 ml) was added to
a stirred solution of 1-benzyl-1,4,7,10-tetraazacyclododecane14
(0.872 g, 3.32 mmol) and fine mesh anhydrous potassium
carbonate (1.37 g, 9.96 mmol) in dry N,N-dimethylformamide
(25 ml). The reaction mixture was heated at 60 ◦C under an argon
atmosphere for 36 h. The solvent was distilled off under vacuum
and the resulting brown oil was extracted into dichloromethane
(40 ml), washed with Purite water (3 × 40 ml), brine (40 ml), dried
(K2CO3) and concentrated to dryness to give a yellow brown oil.
The mixture was purified by alumina column chromatography
(gradient elution from dichloromethane to 0.5% methanol–
dichloromethane) and the product was isolated as a pale yellow
oil (1.54 g, 48%); Rf (Al2O3; 10% CH3OH-CH2Cl2; I2 and uv
detection) 0.5; dH (200 MHz, CDCl3), 7.8 (1H, br s, NHCO), 7.5
(2H, br s, NHCO), 7.3–7.0 (20H, m, Ar), 4.7 (3H, m, CH), 4.01
(6H, m, OCH2), 3.8–2.2 (30H, m, ring-CH2, CH2CO, CH2Ph,
NCH2Ph), 1.14 (9H, m, CH3); dC (50.29 MHz, CDCl3) 171.8
(CO2Et), 171 (CONH), 138.1 (ipso-Ar), 136.6 (ipso-Ar), 129.6,
129.3, 128.5, 127.1 (C–Ar), 61.5 (OCH2), 59.4–59.1 (br, NCH2Ph,
NCH2CO), 53.9–52.0 (br, CH, ring-CH2), 37.0 (CCH2Ph), 14.3
(CH3); m/z (ES+) 985 (100%, MNa+), 963 (45%, MH+), 501 (5%,
MCa2+).
Ligand synthesis
2-Chloro-N-[(S)-(1- ethoxycarbonyl-2-phenyl)ethyl] ethanamide
(S)-Phenylalaninethylester hydrochloride (3 g, 13.1 mmol) was
dissolved in dry diethyl ether (100 ml) under argon and triethy-
lamine (4.37 ml, 31.3 mmol) was added. The solution was cooled to
−20 ◦C (acetone, CO2(s) bath) and chloroacetylchloride (1.25 ml,
15.7 mmol) in dry diethyl ether (60 ml) was added dropwise,
maintaining the temperature at −20 ◦C. The reaction mixture
was allowed to warm to room temperature and stirred for 1 h. The
resulting white precipitate was dissolved in water (60 ml) and the
organic layer washed with hydrochloric acid (0.1 M, 80 ml), water
(3 × 80 ml), dried (K2CO3) and the solvent was removed in vacuo to
give a pale yellow solid. The product was recrystallised from ether
by adding hexane and was isolated as white needles (2.47 g, 70%),
mp 38–40 ◦C; (Found: C, 57.7; H, 5.98; N, 5.12. C13H16NO3Cl
requires C, 57.9; H, 5.98; N, 5.16%); mmax (solid)/cm−1 3310 (NH),
1-Benzyl-4,7,10-tris-[(S)-{(1-methoxycarbonyl-3-methyl)butyl}-
carbamoylmethyl]-1,4,7,10-tetraazacyclododecane L6 see ESI†
3666 | Dalton Trans., 2007, 3661–3668
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
General procedure for the preparation of trialkylated ligands.
mp > 250 ◦C; mmax (solid)/cm−1 1619 (br, CONH, COO); dH
(200 MHz, CD3OD) partial assignment 90.0 (1H, br s, ring-Hax),
64.0 (1H, br s, ring-Hax), 52.0 (1H, br s, ring-Hax), 27.0 (1H,
br s, ring-Hax), −22.9, −25.1, −38.6, −46.7, −47.8, −49.9, −60.2,
−82.3; m/z (ES+) after addition of CF3COOH: 959 (100%, MH+)
(Found: MH+, 959.3137. C41H51N7O9Yb requires MH+ 951.3131).
1,4,7-Tris-[(S)-{(1-ethoxycarbonyl-2-phenyl)ethyl}carbamoyl-
methyl]-1,4,7,10-tetraazacyclododecane L2
A hydrochloric acid solution (1 M, 1 ml) and a catalytic amount
of palladium hydroxide on carbon were added to 1-benzyl-4,7,10-
tris-[(S)-{(1-ethoxycarbonyl-2-phenyl)ethyl}carbamoylmethyl]-
1,4,7,10-tetraazacyclododecane (0.473 g, 0.49 mmol) in ethanol
(25 ml) and the mixture was treated with hydrogen (45 psi) at room
temperature for 48 h. The reaction mixture was filtered through
Celite and the solvent was removed under vacuum. The residue was
taken into dichloromethane (30 ml) and washed with a solution
of sodium bicarbonate (30 ml), brine (1 × 30 ml), dried (K2CO3)
and concentrated to dryness to give the product as a pale yellow
glassy solid (0.31 g, 71%), mp 84–88 ◦C; mmax (solid)/cm−1 1734
(COO), 1656 (CONH); dH (200 MHz, CDCl3) 7.6 (2H, d, 3J = 8,
NHCO), 7.4 (1H, d, 3J = 8, NHCO), 7.30–7.00 (15H, m, Ar), 4.80
(3H, m, CH), 4.12 (6H, m, OCH2), 3.45–2.20 (29H, m, ring-CH2,
ring-NH, CH2Ph, CH2CO), 1.12 (9H, m, CH3); dC (50.29 MHz,
CDCl3) 172.1 (CO2Et), 171.4 (CONH), 136.8 (ipso-Ar), 129.6
(m-Ar), 128.6 (o-Ar), 127 (p-Ar), 61.5 (OCH2), 59.7 (CH2CO),
53.9–53.0 (ring-CH2, CH), 47.3 (ring-CH2), 37.8 (CH2Ph), 14.3
(CH3); m/z (ES+) 456 (100%, MCa 2+) (A satisfactory accurate
mass spectrum could not be obtained for this product).
Yb·L4. The title compound was prepared following a method
similar for Yb·L2, using ytterbium(III) triflate (0.44 g, 0.72 mmol)
in dry acetonitrile (3 ml) and L4 (0.512 g, 0.72 mmol) in dry
acetonitrile (4 ml). A yellow solid resulted (0.70 g, 75%), mp
142–144 ◦C; mmax (solid)/cm−1 1742 (COO), 1627 (CONH); dH
(200 MHz, CD3OD) partial assignment 119.5 (1H, br s, ring-Hax),
76.4 (1H, br s, ring-Hax), 67.8 (1H, br, ring-Hax), 53.9 (1H, br, ring-
Hax), 43.0 (1H, br s, ring-Heq), 35.8 (1H, br s, ring-Heq), 31.9 (1H,
br s, ring-Heq), 27.0 (1H, br, ring-Heq), 12.1, −17.5, −19.1, −29.7,
−33.8, −36.3, −41.1, −42.8, −48.5, −78.1, −91.0; m/z (ES+) 450
(100%, M2+) (Found: M2+, 450.2109. C35H64N7O9Yb requires M2+
450.2077).
Yb·L5. The title compound was prepared following a method
similar to that for Yb·L3, using Yb·L4 (0.25 g, 0.18 mmol),
dissolved in the minimum amount of methanol and an aqueous
sodium hydroxide solution (0.02 M, 37 ml). A yellow solid resulted
(0.080 g, 50%), mp > 250 ◦C; mmax (solid)/cm−1 1625 (br, CONH,
COO); dH (200 MHz, CD3OD) partial assignment 105.6 (1H,
br s, ring-Hax), 99.1 (1H, br s, ring-Hax), 79.0 (1H, br s, ring-
Hax), 75.0 (1H, br s, ring-Hax), 43.4 (1H, br s, ring-Heq), 27.0 (3H,
br s, ring-Heq), 21.7, 18.7, 10.4, 8.15, −17.0, −22.8, −24.3, −33.0,
−36.3, −46.7, −51.9, −54.3, −58.5, −61.7, −121.3; m/z (ES+)
after addition of CF3COOH: 429 (100%, [M + 2H+]2+) (Found:
MH+, 857.3681. C32H57N7O9Yb requires MH+ 857.3607).
1,4,7-Tris-[(S)-{(1-methoxycarbonyl-3-methyl)butyl}carbamoyl-
methyl]-1,4,7,10-tetraazacyclododecane L4 see ESI†.
Complex synthesis
Full experimental details on all of the Ln complexes of L2–L7 are
available via ESI†.
Yb·L6. The title compound was prepared following a method
similar to that for Yb·L2, using ytterbium(III) triflate (0.099 g,
0.16 mmol) in dry acetonitrile (1 ml) and L6 (0.131 g, 0.16 mmol)
in dry acetonitrile (4 ml). A yellow solid resulted (0.200 g, 87%),
mp > 140 ◦C (dec.); mmax (solid)/cm−1 1740 (COO), 1622 (CONH);
dH (200 MHz, CD3OD) partial assignment 140.0 (1H, br s, ring-
Hax), 134.0 (2H, br s, ring-Hax), 112.3 (1H, br s, ring-Hax), 34.8
(2H, br s, ring-Heq), 29.0 (1H, br s, ring-Heq), 24.65 (1H, br s,
ring-Heq), 23.8, 19.0, −9.63, −10.5, −15.7, −16.9, −28.9, −51.1,
−60.5, −66.3, −69.1, −75.4, −79.6, −86.1, −89.69; m/z (ES+) 495
(100%, M2+) (Found: M2+, 495.2322. C42H70N7O9Yb requires M2+,
495.2312).
Yb·L2. Ytterbium(III) triflate (0.099 g, 0.16 mmol) in dry
acetonitrile (1 ml) was added to a solution of L2 (0.135 g,
0.16 mmol) in dry acetonitrile (4 ml) and the mixture was
heated to reflux under argon overnight. The reaction solution
was dropped onto stirring diethyl ether (50 ml) and the resulting
white precipitate was collected by centrifugation and filtration.
The solid was redissolved in the minimum amount of acetonitrile
and the precipitation repeated once more. A white solid resulted
(0.134 g, 56%), mp 142–144 ◦C; mmax (solid)/cm−1 1736 (COO),
1625 (CONH); dH (200 MHz, CD3OD) partial assignment: major
isomer 92.0 (1H, br s, ring-Hax), 69.0 (1H, br s, ring-Hax), 56.0 (1H,
br s, ring-Hax), 43.0 (1H, br s, ring-Hax), 39.1 (1H, br s, ring-Heq),
25.2 (1H, br s, ring-Heq), 22.3 (1H, br s, ring-Heq), 15.0 (1H, br s,
ring-Heq); minor isomer 96.0 (1H, br s, ring-Hax), 67.0 (1H, br s,
ring-Hax), 59.0 (1H, br s, ring-Hax), 43.0 (1H, br s, ring-Hax), 41.0
(1H, br s, ring-Heq), 27.0 (1H, br s, ring-Heq), 23.0 (1H, br s, ring-
Heq), 19.0 (1H, br s, ring-Heq); m/z (ES+) 522 (100%, M2+), 349
(5%, M3+) (Found: M2+, 522.2077. C47H64N7O9Yb requires M2+,
522.2077).
Yb·L7. The title compound was prepared following a method
similar to that for Yb·L3, using Yb·L6 (0.100 g, 0.069 mmol),
dissolved in the minimum amount of methanol and an aqueous
sodium hydroxide solution (0.02 M, 13.9 ml). A yellow solid
◦
resulted (0.035g, 40%), mp > 250 C; mmax (solid)/cm−1 1623 (br,
CONH, COO); dH (200 MHz, CD3OD) partial assignment major
isomer 107.1 (1H, br s, ring-Hax), 102.7 (1H, br s, ring-Hax), 83.2
(1H, br s, ring-Hax), 61.74 (1H, br s, ring-Hax), 59.3, 51.0, 42.0,
38.6, 37.0, 34.0, 28.3, 19.1, 17.1, −10.89, −12.06, −24.67, −28.95,
−30.0, −43.0, −44.19, −46.3, −47.3, −55.8, −66.6, −72.1, −83.2,
−98.6, −102.0, −105.3, −131.4; d minor isomer 71.0 (1H, br s,
ring-Hax), 35.68 (1H, br s, ring-Hax), 20.32 (1H, br s, ring-Hax),
14.80 (1H, br s, ring-Hax), −4.86, −6.58; m/z (ES+) after addition
of CF3COOH: 474 (100%, [M + 2H+]2+) (Found: MH+, 947.4077.
C39H63N7O9Yb requires MH+ 947.4077).
Yb·L3. Yb·L2 (0.100 g, 0.067 mmol) was dissolved in the
minimum amount of methanol and treated with an aqueous
sodium hydroxide solution (0.02 M, 13.4 ml). The solution was
brought to pH 6, reduced to small volume and loaded onto a
cationic exchange column (Dowex, 50 WH+), eluting with water
and then aqueous ammonia solution (6%). The ammonia layer
was dried giving the product as a yellow solid (0.035 g, 54%),
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 3661–3668 | 3667
©

View Article Online
8 D. Parker, R. S. Dickins, H. Puschmann, C. Crossland and J. A. K.
Howard, Chem. Rev., 2002, 102, 1977.
Acknowledgements
9 J. I. Bruce, R. S. Dickins, L. J. Govenlock, T. Gunnlaugsson,
S. Lopinski, M. P. Lowe, D. Parker, R. D. Peacock, J. J. B.
Perry, S. Aime and M. Botta, J. Am. Chem. Soc., 2000, 122,
9674.
We thank EPSRC and the Royal Society for a Dorothy Hodgkin
Fellowship (RSD) for support.
10 R. S. Dickins, S. Aime, A. S. Batsanov, A. Beeby, M. Botta, J. I.
Bruce, J. A. K. Howard, C. S. Love, D. Parker, R. D. Peacock and
H. Puschmann, J. Am. Chem. Soc., 2002, 124, 12697.
11 R. S. Dickins, D. Parker, J. I. Bruce and D. Tozer, Dalton Trans., 2003,
1264.
12 R. S. Dickins, A. S. Batsanov, D. Parker, H. Puschmann, S. Salamano
and J. A. K. Howard, Dalton Trans., 2004, 70.
13 R. S. Dickins, C. S. Love and H. Puschmann, Chem. Commun., 2001,
2308.
14 K. P. Pulukkody, T. J. Norman, D. Parker, L. Royle and C. J. Broan,
J. Chem. Soc., Perkin Trans. 2, 1993, 605.
References
1 M. Stubbs, Acta Oncol., 1999, 38, 845; F. Podo, O. Henriksen,
W. M. M. J. Bovee, M. O. Leach, D. Leibfritz and J. D. Certaines,
Magn. Reson. Imaging, 1998, 16, 1085; J. Star-Lack, D. Spielman, E.
Adalsteinsson, J. Kurhanewicz and D. Vigneron, J. Magn. Reson., 1998,
133, 243.
2 S. J. Vaidya, G. S. Payne, M. O. Leach and C. R. Pinkerton,
Eur. J. Cancer, 2003, 39, 728; J.-K. Kim, S.-H. Park, H. M. Lee,
Y.-H. Lee, N.-K. Sung, D.-S. Chung and O.-D. Kim, Breast, 2003,
12, 179.
15 S. Aime, A. Barge, J. I. Bruce, M. Botta, J. A. K. Howard, J. M. Moloney,
D. Parker, A. De Sousa and M. Woods, J. Am. Chem. Soc., 1999, 121,
5672; S. Amin, J. R. Morrow, C. H. Lake and M. R. Churchill, Angew.
Chem., Int. Ed. Engl., 1994, 33, 773; L. L. Chappell, D. A. Voss, W.
deW. Horrocks and J. R. Morrow, Inorg. Chem., 1998, 37, 3989; L.
Alderighi, A. Bianchi, L. Calabi, P. Dapporto, C. Giorgi, P. Losi, L.
Pateari, P. Paoli, P. Rossi, B. Valtancoli and M. Virtuani, Eur. J. Inorg.
Chem., 1998, 1581.
3 M. O. Leach, Anticancer Res., 1996, 16, 1503; M. Stubbs and J. R.
Griffiths, Br. J. Cancer, 1999, 80, 86; A. P. Pathak, B. Gimi, K. Glunde,
E. Acherstaff, D. Artemov and Z. M. Bhujwalla, Methods Enzymol.,
2004, 386, 3.
4 S. P. Robinson, S. J. Barton, P. M. J. McSheehy and J. R. Griffiths,
Br. J. Radiol., 1997, 70, S60; M. O. Leach, Phys. Medica, 2001, 17, 50.
5 M. Stubbs and J. R. Griffiths, Br. J. Cancer, 1999, 80, 86; F. Podo,
NMR Biomed., 1999, 12, 413; M. O. Leach, M. Verrill, J. Glaholm,
T. A. D. Smith, D. J. Collins, G. S. Payne, J. C. Sharp, S. M. Rohen,
V. R. McCready, T. J. Powles and I. E. Smith, NMR Biomed., 1998, 11,
314.
16 Due to the broadening effect of the paramagnetic ion, not all
of the crosspeaks in the 1H–1H COSY experiment could be
identified.
17 A. Beeby, I. M. Clarkson, R. S. Dickins, S. Faulkner, D. Parker, L.
Royle, A. S. de Sousa, J. A. G. Williams and M. Woods, J. Chem. Soc.,
Perkin Trans. 2, 1999, 493.
18 S. Aime, M. Botta, V. Mainero and E. Terreno, Magn. Reson. Med.,
2002, 47, 10.
19 B. J. Bleaney, J. Magn. Reson., 1972, 8, 91.
6 L. C. Costello, R. B. Franklin and P. Narayan, Prostate, 1999, 38, 237;
C. E. Mountford, S. Doran, C. L. Lean and P. Russell, Chem. Rev.,
2004, 104, 3677; E. E. Kim and E. F. Jackson, in Molecular Imaging in
Oncology, Springer, New York, 1999.
7 I. Bertini and C. H. Luchinat, Coord. Chem. Rev., 1996, 150, 1; T. G.
Wenzel, in NMR Shift Reagents, CRC Press, Boca Raton, 1987.
3668 | Dalton Trans., 2007, 3661–3668
This journal is
The Royal Society of Chemistry 2007
©